Identification and Characterization of Two Novel Syncytin-Like Retroviral Envelope Genes, Captured for a Possible role in the Atypical Structure of the Hyena Placenta and in the Emergence of the Non-Mammalian Mabuya Lizard Placenta a Mathis Funk

To cite this version:

Mathis Funk. Identification and Characterization of Two Novel Syncytin-Like Retroviral Envelope Genes, Captured for a Possible role in the Atypical Structure of the Hyena Placenta and in the Emergence of the Non-Mammalian Mabuya Lizard Placenta a. Virology. Université Paris Saclay (COmUE), 2018. English. ￿NNT : 2018SACLS106￿. ￿tel-02377630￿

HAL Id: tel-02377630 https://tel.archives-ouvertes.fr/tel-02377630 Submitted on 24 Nov 2019

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés.

Identification and characterization of 106 S two novel syncytin-like retroviral

SACL envelope genes, captured for a 8 possible role in the atypical structure : 201 : of the hyena placenta and in the NNT emergence of the non-mammalian Mabuya lizard placenta

Thèse de doctorat de l'Université Paris-Saclay préparée à l'UMR 9196, Gustave Roussy

École doctorale n°582 cancérologie: biologie, médecine, santé (CBMS) Spécialité de doctorat: aspects moléculaires et cellulaires de la biologie

Thèse présentée et soutenue à Villejuif, le 23 mai 2018, par

Mathis Funk

Composition du Jury :

Uriel Hazan Professeur des université, ENS Paris-Saclay (– UMR 8113) Président Jean-Luc Battini Directeur de recherche, IRIM (– UMR 9004) Rapporteur Olivier Schwartz Directeur de recherche, Institut Pasteur (– UMR 3569) Rapporteur Pascale Chavatte-Palmer Directrice de recherche, INRA (– UMR 1198) Examinatrice François Mallet Directeur de recherche, bioMérieux (– EA 7426) Examinateur Thierry Heidmann Directeur de recherche, CNRS (– UMR 9196) Directeur de thèse

Acknowledgments

I would first like to thank the members of the jury for taking the time to read the present manuscript, which turned out a bit longer than I had planned. I would like to thank Uriel Hazan for accepting to be the president of this jury, book-ending his involvement in my studies. What had started at the ENS Cachan and continued during my Master’s degree at the Institut Pasteur, finally reaches its culmination with the present work, on a topic that Uriel suggested I look into. I would like to sincerely thank Jean-Luc Battini and Olivier Schwartz for their critical reading and evaluation of the present manuscript and their positive feedback. Finally, I want to thank Pascale Chavatte-Palmer and François Mallet, for taking the time to read my work and for accepting to be a part of the jury.

I would, of course, also like to thank Thierry Heidmann for giving me the opportunity to work in his laboratory for the past four years, for being always available when I was in need of a guiding hand and present when I needed some prodding. Thank you for not only helping me to get these results, but also for the countless hours we spent to turn them into actual scientific articles.

I am grateful to Anne Dupressoir and Gérard Pierron for spending time reading through my manuscript and their words of encouragement and advice. I also want to thank Christa Kuhn for her enthusiasm and encouragement, which played an important part in getting me to where I am today.

Je souhaite remercier l’intégralité de l’unité 9196, en particulier Marie et Anne pour avoir partagé leurs savoirs, leurs conseils ainsi que leurs bureaux, Cécile Vernochet pour son aide inestimable, Seila pour son soutien administratif et émotionnel, Grégoire pour son optimisme inébranlable et infectieux, Caroline, Cécile Lemaître et Anthony pour les bon moments passés ensemble, et ce-dernier tout particulièrement pour nos nombreuses discussions (profondes et/ou inutiles) et son énergie aussi inépuisable qu’insupportable. Un grand merci également à Guil- laume, ex-membre de l’unité, pour m’avoir encadré pendant mes débuts et avoir toujours été disponible pour mes questions.

J’aimerais remercier le gang des pasteuriens, premiers parmi eux mes ex-colocataires, Matthieu et Thomas, merci de m’avoir supporté pendant ces trois ans et quelques de vie com- mune au 20 avenue d’Ivry, qui s’est transformé en haut-lieu de la science, accueillant maints et maints congrès réunissant la fine fleur de la virologie et de l’immunologie parisienne. I would

iii like to thank Oksana, one of the kindest and most generous persons I have had the pleasure to know and someone you can always count on. Merci à Simon l’invisible, ton rétro nous a apporté de nombreuses heures de bonheur. Merci à Vincent le sale cylon pour les côtes de bœuf et à Essia pour ses couscous, toujours aussi bons que son humeur. Merci aux deux Juju pour la bonne ambiance, et je vous souhaite bien du courage, à Juju pour supporter Matthieu et à Juju pour finir ta thèse. Merci aussi à Anaïs, Laura, Cécile, Alix, Marion et tous les autres qui ont rendu la vie parisienne tellement plus agréable. J’aimerais également remercier mes nouveaux ex-colocataires, Raphaël, Ferretti et Viking, pour m’avoir permis de garder l’appartement quelques mois de plus, puis pour m’avoir accueilli sur notre canapé qui, entretemps, était devenu le leur.

On a somewhat stranger note I would like to thank all of the friends I have never met, but who have been there to play, talk and just hang out in the virtual parts of this world. Thank you to all sigmas, especially 48, Mori Kaal and koreamax, but also willemd, :goatdrügs:, Cal, and even Thuneral, Slap Chop and Koahi. A great thank you to the Aqueduct Acrobats who have been able to observe my slow and steady descent into placenta and retrovirus induced madness over the past few months, especially Topaz with his awful K/D, expensive haircuts and shared disgust of terrible book adaptations, mrcompson for our academia themed diversions and unmasking the terrible truth behind omics- conferences, and Malloot, everyone’s favorite tactical espionage action ground team leader and platforming liability. I also want to thank Norns and Tarth, our resident Americans, as well as Charisma and Waka.

Ich möchte natürlich auch meiner Familie danken, meiner Mutter, meiner Schwester und meinen beiden Brüdern, von deren Hilfe ich immer ausgehen konnte und deren Gesellschaft und Vertrauen mein Leben bereichert haben. Ich danke meiner Oma und meinem Opa für ihre Ermutigungen und Liebe. J’aimerais remercier Patrice, qui a été pour moi pendant ces 25 dernières années le meilleur père que j’aurais pu souhaiter, et qui m’a poussé à être curieux et intéressé en espérant de peut-être, un jour, être aussi instruit et intéressant que lui. Ich möchte hier auch ganz besonders meiner Tante Jutta danken, die innerhalb von ein paar Tagen die ganze Einführung durchgelesen und korrigiert hat.

Last but not least, I want to thank all of the people that I have shamefully forgotten to mention here, please forgive me for this oversight.

I am still unsure of how I made it this far, but one thing I know: I would not have made it without you. The words above pale in comparison to what I owe you all, I have written and rewritten them many times and I know I’ll never be able to express the feeling appropriately, so all I can do is apologize and say it once more, thank you all so very much.

iv Contents

Acknowledgments iii

Abbreviations xi

I Introduction 1

1 Placenta 5 1.1 Challenges of viviparity ...... 5 1.1.1 Maternofetal exchanges ...... 5 a. Nutrients 6 b. Oxygen 8 c. Waste 9 1.1.2 Placentation and the maternal immune system ...... 9 a. Major histocompatibility complex expression 9 b. Uterine natural killer cells 10 c. Adaptive immunity 11 1.1.3 Maternofetal communication ...... 11 a. Placental hormones 11 b. Extracellular vesicles 12 1.2 Mammalian eutherian placentas ...... 13 1.2.1 Origin and evolution of the mammalian placenta ...... 13 a. Monophyletic origin 13 b. Placental diversity 13 1.2.2 Placental types by maternofetal interface ...... 16 a. Epitheliochorial placentation 18 b. Synepitheliochorial placentation 20 c. Endotheliochorial placentation 22 d. Hemochorial placentation 22 1.3 Viviparity in non-mammalian species ...... 26 1.3.1 Prevalence of viviparity outside of mammals ...... 26

v a. Invertebrates 26 b. Non-mammalian vertebrates 27 1.3.2 Examples of non-mammalian placentas ...... 30 a. Non-vertebrates: the placenta of salps 30 b. Fish: placentation in sharks 31 c. Lizards: placentation in Mabuya and other Scincidae 32

2 Retroviruses 38 2.1 Exogenous Retroviruses ...... 39 2.1.1 Genomic orgarnisation ...... 39 a. Non-coding features 39 b. Gag 41 c. Pro and Pol 42 d. Env 44 e. Regulatory and accessory proteins 44 2.1.2 Retroviral phylogeny ...... 48 a. Subfamily Orthoretrovirinae 48 b. Subfamily Spumaretrovirinae 51 2.1.3 Viral cycle ...... 53 a. Viral entry 53 b. Reverse transcription 53 c. Genome integration 57 d. Protein expression 60 e. Assembly and budding 61 f. Maturation 64 2.2 Endogenous Retroviruses ...... 66 2.2.1 Endogenization of retroviral sequences ...... 67 a. Infection of germinal cells 67 b. Intra-genome spreading of ERVs 69 c. Groups of ERVs 71 2.2.2 Impact of retroviral integration on the host ...... 72 a. LTR integration can drive ectopic gene expression 72 b. Impact of retroviral gene acquisition on host physiology 76 2.2.3 Fate of ERVs and control by the host ...... 78 a. Epigenetic regulation of ERV expression 78 b. Accumulation of mutations and solo LTR formation 79 c. Exaptation and purifying selection 80

3 Syncytins 81 3.1 Retroviral envelopes ...... 81 vi 3.1.1 Functional domains ...... 81 a. Domains of the SU 81 b. Domains of the TM 83 3.1.2 Env induced membrane fusion ...... 86 a. Implication of the cognate receptor in fusion 86 b. Model of viral fusion protein class I induced membrane fusion 89 3.2 Syncytins in mammals ...... 90 3.2.1 Human syncytins: Syncytin-1 and Syncytin-2 ...... 90 a. Syncytin-1 92 b. Syncytin-2 94 3.2.2 Mouse syncytins: Syncytin-A and Syncytin-B ...... 95 a. Discovery 95 b. Expression in 2 segregated layers 95 c. Impact of Syncytin knockout on placental physiology 96 3.2.3 Other mammalian syncytins ...... 98 a. Rabbit: Syncytin-Ory1 98 b. Cat and dog: Syncytin-Car1 99 c. Old world monkey: EnvV2 99 d. Cattle and sheep: Syncytin-Rum1 and Fematrin-1 99 e. Squirrel and marmot: Syncytin-Mar1 100 f. Tenrec: Syncytin-Ten1 100 g. Short-tailed opossum: Syncytin-Opo1 100 3.2.4 Syncytin-like Env ...... 101 a. Conserved ERV Env with placental expression in human 101 b. Non-human syncytin-like genes 102 3.3 Syncytins and evolution ...... 103 3.3.1 Ancestral syncytin and syncytin replacement ...... 103 a. Example of EnvV2, a decaying syncytin in humans 104 b. Baton pass hypothesis 105 3.3.2 Syncytin diversity and placental diversity ...... 106 3.3.3 Other roles of syncytins ...... 108 a. Cell-cell fusion in muscles 109 b. Osteoclast fusion 109 c. Syncytins and pathologies 110

4 Objectives 111

II Results 113

vii 1 Implication of ERV env genes in non-mammalian Mabuya placen- tation 114 1.1 Summary ...... 114 1.2 A syncytin in a placental lizard ...... 117

2 Implication of ERV env genes in placental structural transitions in Hyaenidae 135 2.1 Summary ...... 135 2.2 Hyena-specific retroviral envelope gene capture and placenta evolution . . . 139

III Discussion 167

1 Two novel Env with a potential placental role and a new SLC1A5 Env168 1.1 Hyena-Env2 is a conserved syncytin-like envelope gene expressed during hyena placentation ...... 168 1.2 Hyena-Env3 is a member of the SLC1A5 interference group with a peculiar tropism ...... 169 1.3 Syncytin-Mab1: characterization of a non-mammalian syncytin and its re- ceptor ...... 170

2 Implication of syncytins and syncytin-likes in placental evolution 172 2.1 Retroviral insertions have been a constant driving force of vertebrate evolution172 2.2 Syncytin-Mab1 and convergent evolution of animal placentas ...... 174 2.2.1 Convergent use of ERV env genes in placentation ...... 174 2.2.2 Syncytin-Mab1 as the first example of a founding syncytin ...... 175 2.3 Syncytins and syncytin-likes and divergent evolution of the mammalian placenta ...... 175

3 Syncytin-likes and alternative envelope functions in the placenta 178 3.1 Growing evidence of conservation of non-fusogenic Env ...... 178 3.2 Alternative placental roles for Env ...... 178 3.2.1 Immunosuppression ...... 178 3.2.2 Differentiation and proliferation ...... 180 3.2.3 Invasion ...... 180 3.3 Syncytin-likes in epitheliochorial placentation ...... 183

4 Conclusion 185 viii IV Bibliography 187

V Résumé Français 223

List of Figures

1 Dynamics of placental glucose and amino-acid transport ...... 7 2 Timed tree of major mammalian clades...... 14 3 Diversity of eutherian placentas ...... 17 4 Types of maternofetal interface in eutherians ...... 19 5 Heterologous syncytium formation in ruminant placentas ...... 21 6 Remodeling of maternal blood vessels by extravillous trophoblast cells in humans 24 7 Matrotrophy in Animalia clades ...... 28 8 Mabuya placental structure ...... 35

9 Organization of retroviral virions and genomes ...... 40 10 Organization of the complex HIV-1 genome ...... 45 11 Phylogeny of retroviral genera based on RT or TM analysis ...... 49 12 Retroviral replication cycle ...... 54 13 Reverse transcription ...... 56 14 Retroviral genome integration ...... 59 15 Model of ESCRT mediated viral budding ...... 64 16 Retroviral maturation ...... 65 17 Retroviral endogenization ...... 68 18 ERV amplification and fate ...... 70 19 ERV family representation ...... 72 20 LTR insertional mutagenesis ...... 74

21 Detailed structure of Env ...... 82

22 Sequence of the ISD and CXnC1-2 motif and rationale of the mice immunosup- pression assay ...... 85 23 Retroviral Env mediated cell fusion ...... 86 24 Nature of retroviral receptors and SLC1A5 RBD ...... 88 25 Phylogeny of mammals with known syncytins positioned ...... 91 26 Impact of syncytin KO in mice ...... 97

ix 27 Compared evolutionary fates of HERV-W Env and EnvV2 in simians ...... 104 28 Phylogeny of retroviral syncytins among retroviral genera based on RT or TM analysis ...... 107 29 Differences in syncytin transcript initiation ...... 108

1 Phylogeny of mammals with emphasis on carnivorans ...... 143 2 Characterization of the identified Crocuta crocuta envelope candidates . . . . . 145 3 Entry date and conservation of env genes in Carnivora ...... 147 4 Alignment of Hyena-Env2 genes from all four extant hyena species ...... 148 5 Hyena-Env2 associated provirus and its genomic location ...... 150 6 The Hyena-Env2 associated virus is a gammaretrovirus ...... 151 7 Comparison of carnivoran and hyena placentas ...... 153 8 Expression pattern of syncytin-Car1 and Hyena-Env2 ...... 154 9 Fusogenicity assay for the three hyena envelope proteins ...... 155 10 Viral titers obtained in the pseudotyping assay using all three hyena envelopes . 156 11 Hyena-Env3 belongs to the SLC1A5 interference group of envelopes ...... 158

x Abbreviations

APC: antigen presenting cell

ALV: avian leukosis virus

BaEV: baboon endogenous virus

BLV: bovine leukemia virus

BNC: binucleate cell

CA: capsid protein

CCD: catalytic core domain

CT: cytotrophoblast

CTD: C-terminal domain

EIAV: equine infection anemia virus enJSRV: endogenous Jaagsiekte sheep retrovirus

Env: envelope

ER: endoplasmic reticulum

ERV: endogenous retrovirus

ESCRT: endosomal sorting complex required for transport

ETn: early transposon

EVT: extravillous trophoblast

FLV: feline leukemia virus

Gag: group-specific antigen

GaLV: gibbon ape leukemia virus

GCMa: chorion-specific transcription factor GCMa

xi HA: hemagglutinin

HBZ: HTLV-1 basic zipper factor

HEMO: human endogenous MER34 (medium re-iteration family 34) ORF

HERV: human endogenous retrovirus

HIV: human immunodeficiency virus

HLA: human leukocyte antigen hPL: human placental lactogen

HR: heptad repeat

HTLV: human T-lymphotropic virus

IAP: intra-cisternal A-type particle

IAPE: IAP-related retroviral element containing an env gene

ICTV: International Committee on Taxonomy of Viruses

IN: integrase

ITIM: immunoreceptor tyrosin-based inhibitory motif

JSRV: Jaagsiekte sheep retrovirus

KoRV: koala retrovirus

KRAB-ZFP: Krüppel associated box zinc finger proteins

LTR: long terminal repeat

Ly6E: lymphocyte antigen 6E

MA: matrix protein

MFSD2A: major facilitator superfamily domain containing 2 A

MHC: major histocompatibility complex

MHR: major homology region

MLV: murine leukemia virus

MMTV: mouse mammary tumor virus xii MPMV: Mason-Pfizer monkey virus

MPZL1: myelin protein zero-like 1

My: million years

Mya: million years ago

NAHR: non-allelic homologous recombination

NC: nucleocapsid protein

Nef: negative factor

NK: natural killer

NTD: N-terminal domain

ORF: open reading frame

PBS: primer binding site

PIC: pre-integration complex

Pol: polymerase

PPT: polypurine tract

Pro: protease

PRR: proline-rich region

PtdIns(4,5)P2: phosphatidylinositol-4,5-biphosphate

PTLV: primate T-lymphotropic virus

RDR: RD114 and D-type receptor

Rev: regulator of expression of viral proteins

Rex: Rev homolog

RSV: Rous sarcoma virus

RT: reverse transcriptase

RTC: reverse transcription complex

SFV: simian foamy virus

xiii SIV: simian immunodeficiency virus

SLC: solute carrier

ST: syncytiotrophoblast

SU : surface subunit

Tat : trans-activator

Tax : trans-activating regulatory protein

TM : transmembrane subunit

TSD: target site duplication uNK: uterine natural killer

UTR: untranslated region

Vif: viral infectivity factor

Vpr: viral protein r

Vpu: viral protein u

WDSV: walleye dermal sarcoma virus

xiv Part I

Introduction

1 2 The field of virology is a relatively new one: less than two centuries ago the notion of a pathogen smaller than even bacteria was proposed by Pasteur, but he lacked the means to prove its existence. Years after these speculations, Chamberland invented his famous filter, able to remove even bacteria from a solution passed through it. It was this discovery that enabled Beijerinck in 1898 to show that filtered sap from plants infected with what is now known as tobacco mosaic virus was able to infect other plants. He concluded from this that the causative pathogen was smaller than a bacteria and called it a virus. From there virology advanced in leaps and bounds and in the following decades a great number of viruses were identified and linked to diverse diseases. In 1911 Rous published his discovery of what would later become known as Rous sarcoma virus (RSV, Rous, 1911). While this was not the first retrovirus to be discovered, and while Rous himself did not know about the true nature of the infectious agent he described, this discovery would lead to the creation of both the fields of tumor viruses and retrovirology. In the 1960s, cancers caused by RSV in egg-laying hens were becoming an ever more important problem, leading to many new investigations into the characteristics and behavior of the virus. In 1961 Crawford & Crawford (1961) isolated the virus on a density gradient and determined that it had an RNA genome. Further research in the following years, making use of Mendelian genetics, both in live chicken and in infected cells, suggested that the viral genes could be stably inherited, which would imply that they had become part of the genome. This posed a major problem since at the time the "central dogma of biology" postulated that DNA could be transcribed into RNA, but that the opposite was impossible. How then could an RNA virus integrate its genes into a DNA genome? Nevertheless, in 1964 Temin introduced the provirus hypothesis, coining this term in the context of eukaryote viruses, and postulating that RSV could convert its RNA genome into a DNA genome which could then be inserted into the host genome (Temin, 1964). This hypothesis proved unpopular until the discovery 6 years later of reverse transciptase (RT, Baltimore, 1970; Temin & Mizutani, 1970), which finally delivered the key element: an enzyme able to convert RNA information into DNA information, and which led to the naming and creation of the Retroviridae family by the International Committee on Taxonomy of Viruses in 1975 (ICTV, 1975). Since viral replication of members of this family necessitates integration into the host

3 genome, endogenous retroviruses (ERV) were discovered in parallel with the family itself, first in chicken and mice and later in humans and other vertebrates. The identification of retroviral sequences that were an integral part of host genomes (sequencing of the human genome later revealed that ERVs represent 8% of the genome Lander et al., 2001) led to the search for the function of these elements, and a number of them have been identified as protecting their host from infection by exogenous retroviruses. A role not linked to viral infection, and probably the most remarkable example of ERV function, is the exaptation of syncytins in mammals. The first such gene was described in 2000 as participating in the fusion of trophoblastic cells in the human placenta (Blond et al., 2000; Mi et al., 2000). Since then numerous other syncytins have been identified in other mammalian placentas and their role in trophoblast fusion has been confirmed using KO mice. To delve further into these topics, I will first discuss the characteristics of mammalian and non-mammalian placentas, a pre-requisite for discussion of syncytin role and acquisition. I will then present retroviruses in greater detail and finally I will talk specifically about captured envelope genes, especially syncytins.

4 1. Placenta

The transition from oviparity (egg-laying) to viviparity (live-bearing) occured on numerous occa- sions during vertebrate evolution (Blackburn, 2015), most famously in Eutherians. In viviparous species, by definition, the embryo is not exposed to the external environment and the exchanges necessary for embryonic development must take place between the embryo and the mother. These exchanges are rendered possible by the placenta. The placenta is a highly variable organ, both in structure and functionality. Its simplest definition was given by Mossman in his seminal 1937 work on mammalian placentas: "an apposition or fusion of the fetal membranes to the uterine mucosa for physiological exchange" (Mossman, 1937). This definition encompasses two related notions: the placenta joins maternal and fetal tissues, and it forms an exchange interface between mother and fetus.

1.1 Challenges of viviparity

1.1.1 Maternofetal exchanges

The first major challenge facing viviparous species is for the mother to act as the go-between of the embryo and the exterior. This means that the mother must not only provide the substances required for embryo development but also that all embryonic waste products must be taken up by the mother and eliminated. The importance of these exchanges during development is very variable depending on the species (Blackburn, 2015). In some cases, the embryo obtains most nutrients from its own yolk and is reliant on the mother mostly for gas and water exchanges, in others the embryo also needs to acquire almost all nutrients necessary for development. These two extremes are called lecithotrophy and matrotrophy, respectively, and form the two ends of a continuum of reproductive strategies that differ by the degree of implication of the mother and the functional capacities of the placenta (Blackburn, 2015). Of note, while almost all oviparous species are by necessity lecithotrophic, the mammalian monotremes (platypus and echidnas) have a unique matrotrophic oviparous mode of reproduction in which females ovulate small eggs with a 3 mm diameter yolk which takes up maternal nutrients before oviposition, to reach a final diameter of 17 mm (Hughes, 1993). While this is mentioned here for completeness, "oviparity" will refer to the much more common lecithotrophic oviparous mode of reproduction from here on. Combining the two distinctions of oviparity/viviparity and lecithotrophy/matrotrophy allows classification of vertebrate reproductive strategies into four major categories, see Table 1.

5 1. Placenta

Table 1 – Classification of vertebrate reproductive modes. The examples provided for each reproductive type are not exhaustive. Adapted from Blackburn (2015).

oviparity viviparity 1.1 lecithotrophy lecitotrophic oviparity: lecitotrophic viviparity: all birds and crocodilians, many squamates, most squamates, teleosts, some teleosts, chondrichtyans some chondrichtyans matrotrophy matrotrophic oviparity: matrotrophic viviparity: only monotremes all eutherians and marsulpials, several teleosts and chondrichtyans, a few clades of squamates

Since most of our species of interest will be matrotrophic viviparous, the description of placental characteristics will focus on this subgroup, however even within this group placental structures show a great amount of variability (see 1.2.1.b.).

a. Nutrients

In matrotrophic placentation, nutrients that must be supplied to the embryo include carbohy- drates, most amino-acids, fatty acids and metal ions. In mammals, glucose is one of the most important fetal nutrients, providing much of the energy required during fetal growth (Wooding & Burton, 2008, chap. 2). In eutherians, maternal glycemia is higher than fetal glycemia, es- tablishing a concentration gradient that induces glucose transport from the mother to the fetus. In order to transfer the necessary amount of glucose from maternal to fetal blood, diffusion is facilitated using glucose transporters (glucose transporter type 1 in all mammalian placentas and additionally type 3 in some, like rat, horse and ewe, see Wooding & Burton, 2008, chap. 2). Depending on the species, 30-70% of the glucose provided by the mother is used in placental tissues to permit growth and functionality of the placenta itself and does not reach the embryo (Fig. 1, Wooding & Burton, 2008, chap. 2), though some of the products of placental glucose metabolism, like lactate, will enter the fetal circulation to serve as metabolites, lessening the impact of placental glucose consumption (Fowden, 2010). Indeed, lactate is another major actor of fetal metabolism (up to 25% of total fetal energy requirements, see Wooding & Burton, 2008, chap. 2). Similarly to glucose, lactate transport makes use of transporters for facilitated diffusion, but in this case most of the lactate is produced by the placenta itself and maternal transplacental provision of this metabolite seems absent (Fowden, 2010), leading to higher concentrations of lactate in fetal than in maternal blood. Contrarily to carbohydrates, amino acid concentrations are higher in fetal than in maternal blood in most species, indicating that there is a net uptake of amino acids by the fetus and that active transport mechanisms must be involved (Fig. 1, Vaughan & Fowden, 2016). Interestingly, placental amino acid concentrations are even more elevated than fetal concentrations and the placenta exchanges amino acids both with the maternal and fetal circulations. Most amino acids

6 ea hni aenlbod(odn utn 08 hp 2). in chap. 2008, concentration Burton, higher & a (Wooding establishing blood maternal in in participate than transporters fetal active since acids, especially essential fatty the fetus, chain like others and long For most. mother for concentrations separating fetal than membranes higher are the concentrations maternal across diffusion by supplied be 1). that (Fig. and passive fetus, simply and not in the mother is between in role role middleman lactate active its placental with an introduced The is been placenta 2016). had the Fowden, that that notion, & paragraph, the to previous (Vaughan reinforces glycine fetus mother the and the to methionine supply from acid glutamine, up amino takes more provides it than it fetus but consumer net the isoleucine, a and is and placenta leucine generation sheep valine, ATP the addition, of including In needs, 2016). Fowden, metabolic taken & instead (Vaughan placental are synthesis satisfy they protein net to null, The is circulation example. acids fetal for amino the sheep three these from in for glutamate be blood and maternal must serine from and uptake aspartate, fetus placental for more case or much is the placenta transit is acids, the amino this some by complicated, For 2016). synthesized Fowden, & be (Vaughan essential the mother for cannot the the by case provided definition, into the by cross notably is then that, this and acids gradient, tissues amino concentration placental their the to according into circulation transport fetal active by imported simply are example the on wedges. based black and as codes represented letter are one gradients as Concentration given sheep. are of metabolism Examples placental metabolism. in placental transport. involved the of amino-acids amino-acid implication and of (bottom) and acids amino glucose and (top) glucose of of transport Dynamics – 1 Figure amino acids maternal maternal en yrpoi yntr,lpd uha reftyais lcrladcoetrlcan cholesterol and glycerol acids, fatty free as such lipids nature, by hydrophobic Being glucose mother facilitated transport amino acidconcentration ,L Q,M,G V, L, I amino acids metabolism glycemia placenta placental placental placental glucose 30-70% D, S,E lactate ceai ersnaino placental of representation Schematic active transport ..Calne fviviparity of Challenges 1.1. amino acids glucose fetal fetal metabolism fetal fetal 7

1.1 1. Placenta

The final nutrient category discussed here concerns ions. For some like calcium, the transfer from maternal blood to the placenta can simply occur by facilitated diffusion through calcium channels. Once in the placenta, it is bound to calcium binding proteins, so as to isolate it from

1.1 sensitive cellular signaling mechanisms that rely on cytoplasmic Ca2+ concentration, and then transferred to the fetus using calcium ATPase pumps (Wooding & Burton, 2008, chap. 2). Iron transport is a bit more complex since iron ions are not present in their free form in the organism. Instead they are bound to transport proteins (uteroferrin in horse and pig or transferrin) or are acquired by breakdown of red blood cells. Uteroferrin and transferrin are secreted by maternal uterine glands and enter the placenta by endocytosis. Iron ion acquisition from red blood cells concerns mainly carnivorans and ruminants in which specialized regions of the placenta (called hemophagous zones) break down these cells after phagocytosis (Wooding & Burton, 2008, chap. 2). Of note, in some placentas (human for example) a similar endocytic method is used to transfer a few specific proteins from the mother to the fetus, such as immunoglobulin G (Wooding & Burton, 2008, chap. 2). The substances acquired directly from maternal blood are grouped under the term "hemotroph" while the substances secreted by the mother and taken up through endocytosis are termed "his- totroph".

b. Oxygen

As with many other organs and organisms, both the placenta and the fetus need a constant supply

of O2. Since biological membranes are highly permeable to gases like O2, there is no need for facilitated diffusion or active transport, instead diffusion occurs according to Fick’s law along the concentration gradient, leading to a lower oxygenation of fetal blood than maternal blood.

The two main factors influencing the efficiency of O2 transfer from the mother to the fetus are blood flow velocity and relative disposition of the blood vessels (see Fig. 3C Wooding & Burton, 2008, chap. 2). Interestingly both of these factors are very variable across mammalian

placentation, indicating that there is not necessarily a single efficient strategy for O2 transport since all mammals are able to produce viable offspring.

The placental O2 consumption is high, higher than that of the fetus when normalized by weight and on par with that of the fetal brain (Fowden, 2010). In sheep, about 75% of the

O2 consumed by the placenta is used to generate ATP using a variety of substrates, including glucose and amino acids, as discussed in 1.1.1a.. This ATP is then used primarily for synthesis of placental proteins and to fuel active transport against electrochemical gradients. 15-20% of

the O2 uptake of sheep placentas is used to produce intermediaries necessary for synthesis of

progesterone and other steroid hormones (Vaughan & Fowden, 2016). The proportion of O2 that actually reaches the fetal circulation is very dependent on the species and sometimes the gestational stage. In sheep it varies from 80% at the beginning to about 45% towards the end of gestation, while in horse and pig it remains at a constant 55% or 65% respectively (Vaughan & Fowden, 2016).

8 Ca hi ufc,hwvri h lcnams el ontepesayHA(pse al., cells. T the et cytotoxic to especially (Apps invisible and HLA attack cells maternal any placental from express them the protecting not renders system, effectively do immune adaptive expression cells MHC most of placenta absence the This in 2009). however and surface, -B HLA-A, their express usually at T cells -C and MHC and (HLA) the antigens B humans leukocyte In to human detected. called antigens is are element display molecules foreign a and if by responses immune recognized histocompatibility to leading are major lymphocytes, of proteins expression These the (MHC). on proteins relies complex system immune adaptive the mammals In expression complex histocompatibility Major a. 2006). Loke, & Moffett in (reviewed models rodent material) genetic in foreign failure present of pregnancy not absence in does the result fetus in the molecules, will which anti-inflammatory (in that of pregnancies aspect presence syngeneic even important the which be An to aspects. seems major here few detailed a such be on as not focus species, only considered will the discussion on following dependent and the highly placenta and by the complex between shielded are Interactions are system 2014). they immune Colucci, as maternal system, & Moffett immune in maternal the (reviewed of the tissues cells other with The placental tissues. interact in fetal not and seen maternal do between generally is contact itself of placenta as point fetus the the system, is again, it Here immune as 2014). role maternal crucial Colucci, a & the plays Moffett by in (reviewed tissues situations fetal allograft immunological equivalent of particular attack a of avoid establishment to the context necessitates exchanges contact transplacental for intimate necessary This absolutely 1.1.1). is (see In tissues system. maternal physical immune and close however, a viviparity fetal by In maternal between 9). separated chap. contact the are 2008, Burton, fetus of & Wooding and eyes in mother (reviewed shell since the egg new problem the barrier, in major a a to tissue pose rise foreign not give does a this and oviparity is mix fetus genes the maternal and genome, paternal hybrid the reproduction sexual system during immune Since maternal the and Placentation 1.1.2 in transport permit CO to fetal of placenta the form the enter the for in weight need circulation fetal the fetal directions. of illustrating the both exit kg 1978), to Meschia, per need & carbon kg (Battaglia per of urea carbon g and of 7.8 gas g late day 4.6 a during and each is sheep circulation that one In membranes. estimated as biological diffusion, is cross simple it to by able gestation, molecule circulation uncharged maternal small the a necessary other CO to also with the return case is the to it example able for but is being substrates, This necessary both byproducts. metabolic the the with of them that some provide and eliminate active to to metabolically necessary are fetus only the not and placenta is the it both that remember to important is It Waste c. ..Calne fviviparity of Challenges 1.1. 2 n urea, and 2 and 9

1.1 1. Placenta

The presence of mostly MHC null cells is also observed in rodent, pig, horse and ruminant placentas for example (reviewed in Wooding & Burton, 2008, chap. 9), and as such seems to be a common mechanism of fetal tolerance. 1.1 While this is true for the bulk of placental cells, some subsets still express MHC molecules, this is mostly the case for invasive cells. In horses for example, some cells that migrate into the maternal tissue do express MHC proteins, only becoming MHC null once they have reached their destination. In cows the so-called binucleate cells (BNC) fuse with maternal cells, giving rise to heterologous trinucleate syncytia that express MHC molecules but do not seem to present allogenic antigens (reviewed in Wooding & Burton, 2008, chap. 9). The human placenta also presents invasive cells that express MHC proteins: the extravillous trophoblasts (EVT). These cells separate from the rest of the placenta and invade maternal tissues where they induce a major remodeling of blood vessels in order to increase placental blood supply by the mother (see Fig. 6, reviewed in Wooding & Burton, 2008, chap. 8). At their surface the EVT cells express the unique combination of HLA-C, -E and -G (Apps et al., 2009). HLA-E and -G are non classical HLA molecules and display a very restricted expression profile, HLA-G being exclusively expressed in EVT (Apps et al., 2009). The presence of these HLA molecules is crucial for the regulation of maternal vascular remodeling, as they provide both activatory and inhibitory signals to uterine natural killer (uNK) cells (see below, Parham & Moffett 2013).

b. Uterine natural killer cells

During pregnancy in humans up to 70% of uterine lymphocytes are uNK lymphocytes and an equivalent population of NK has been described in rodents (reviewed in Wooding & Burton, 2008, chap. 9), suggesting that they play an important role in all highly invasive (hemochorial, see 1.2.2.d.) placentas. Despite their name and unlike systemic NK cells, uNK cells are only weakly cytotoxic in vitro and do not kill MHC-null placental cells (Ritson & Bulmer, 1989; King et al., 1989). Using NK specific receptors at their surface, uNK interact with the HLA expressed at the surface of EVTs, receiving both activating and inhibiting signals (reviewed in Parham & Moffett, 2013). When activated, uNK cells release cytokines and proteases that regulate EVT migration and assist in vascular remodeling (reviewed in Burton & Fowden, 2015). The dialog between EVT and uNK plays a critical role in pregnancy as it has been shown that a spectrum of pregnancy complications, referred to as the "great obstetrical syndromes" (preeclampsia, intrauterine growth restriction, pre-term labor and abruption placentae among others), are linked to defects in remodeling of the maternal circulation by EVT (reviewed in Brosens et al., 2011). While the non-classical HLA-E and -G show very little genetic variation, HLA-C presents several different epitopes that interact with distinct receptors on the uNK cells. Some combinations of placental HLA-C epitopes and maternal uNK receptors have been shown to increase the risk of pregnancy complications as they bias the EVT/uNK interaction, favoring uNK inhibition over activation (Hiby et al., 2010).

10 rvee nBro odn 05.A icse n111a,alntinsncsayfrpla- for necessary nutrients all 1.1.1.a., in discussed As 2015). en- Fowden, major & pregnancy Burton a during physiology in being maternal (reviewed on placenta effects the important have pathways, hormones hormonal whose organ classical docrine uses communication this of Part hormones Placental a. for conditions an correct establish as to acting as for growth. so fetal responsible side is maternal the placenta with the communicating fetus, and and intermediary mother the between link the Establishing communication Maternofetal 1.1.3 2017). al., et (Moffett antibodies EVT-specific of generation local substantial no is effect side a pregnancy. observed be during to this established it that context consider fact hormonal and success the particular reproductive the stress Th1 on of (2017) impact of any al. have to severity et seem Moffett the not 2016). does (reviewed pregnancy shift al., increases et during disorders Marder linked why 1999; Th2 Ostensen, and of in that 2017) while (like al., decreases infections disorders et viral autoimmune some Mertz linked to shifted susceptible This more in 2017). are reviewed al., women influenza, pregnant et why Moffett Th1 reason in pregnancy the (reviewed in during is generation) decreased balance that (antibody are seems infections) role responses It viral Th2 The clear. against of cells. effective less favor are T is and pregnancy effector cytotoxicity favor on suppress (which cells and responses helper interface T placental of the impact to and protective fetus-specific migrate pregnancy cells during T that show regulatory (2008) al. et the Tilburgs of immunity. adaptive rest the in functionality it APC APC, impact not maternal does the system. and and immune context maternal EVT placental expressing the HLA-G to the restricted between is contact direct a APC necessitates with cells well interferes T as on immunosuppressive binding MHC of II receptors this class induction that through inhibitory as peptides and to allogenic 2003) of binds presentation al., lowered cells, in et resulting (Shiroishi EVT maturation, affinity the human high of of very one surface with HLA-G, the APC that shown on been expressed has types it MHC cells, placental al., expressing et (Apps MHC system immune Regarding MHC- adaptive 2009). are the by cells tissues placental foreign most as above, recognition discussed direct preventing As null, antigen MHC. specialized II by class or using MHC I (APC) lymphocytes, class cells T using presenting or cells B generic by by either antigens presented of being recognition antigens the said on based is mammals in immunity Adaptive immunity Adaptive c. fnt,mtra el r aeyosre ertepaetltsus niaigta there that indicating tissues, placental the near observed rarely are cells B maternal note, Of in role important an play also cells T helper and regulatory cells, T effector from Apart nvitro in Rsihe l,20) ic hsmechanism this Since 2005). al., et (Ristich ..Calne fviviparity of Challenges 1.1. 11

1.1 1. Placenta

cental and fetal growth have to be supplied by the mother, who also must fulfill the needs of her own metabolism. This is particularly true for glucose, the main substrate of placental and fetal metabolism (reviewed in Wooding & Burton, 2008, chap. 2) and an important metabolite

1.1 in adults. In humans the main hormones of placental origin involved in these mechanisms are progesterone, human placental lactogen (hPL) and human placental growth hormone (hGH) (re- viewed in Freemark, 2006; Burton & Fowden, 2015). Expression and impact of these hormones on the maternal metabolism can be separated into two phases: early to mid-gestation and late gestation. During early to mid-gestation, secretion of progesterone and hPL by the placenta increase progressively (reviewed in Freemark, 2006; Burton & Fowden, 2015). Both of these hormones are appetite stimulants and have been associated with weight gain and fat deposition in mice and humans (Grueso et al., 2001; Freemark, 2006), leading to a disruption of energetic homeostasis and a buildup of reserves. In addition hPL participates in establishing leptin resistance during pregnancy, attenuating the negative feedback loop of fatty tissues on food intake normally mediated by this hormone (reviewed in Ladyman et al., 2010). Finally, hPL also stimulates beta-cell proliferation in the pancreas of the mother, leading to increased insulinaemia and thus fat deposition and energy storage (reviewed in Burton & Fowden, 2015). Taken together, these effects lead to a 10-15% increase in food intake and a 60% increase in insulin secretion during this first phase, which starts even before a significant fetal energy demand arises and corresponds to a pre-preemptive increase in energy storage (reviewed in Freemark, 2006). In late pregnancy there is a significant secretion of hGH by the syncytiotrophoblast of the placenta. This growth hormone is capable of inducing insulin resistance (Barbour et al., 2002) and the resulting 45-70 % decrease in maternal insulin sensitivity leads to breakdown of fat reserves and mobilization of glucose into the maternal bloodstream, from where it can then be absorbed by the placenta (see 1.1.1.a.). In addition to the roles in nutrient provision discussed above, fetal hormones also play an important role in the establishment and maintenance of pregnancy (and interrupting the menstrual cycle) and can be involved in the process leading to parturition at the end of gestation (reviewed in Wooding & Burton, 2008).

b. Extracellular vesicles

In addition to the soluble secreted factors described above, the placenta releases a great amount of extracellular vesicles (Salomon et al., 2014). These particles are lipid-bilayer particles that can contain proteins, lipids, RNAs and DNA, be classified as exosomes (40-120 nm) or microvesicles (0.2-1 µm), and participate in inter-cellular communication processes (reviewed in Adam et al., 2017). Salomon et al. (2014) show that the concentration of exosomes in plasma of pregnant woman is increased 50-100 fold when compared to non-pregnant women, depending on the stage of pregnancy. The role of these particles in maternofetal communication is still poorly understood, though it has been shown that placenta generated extracellular vesicles can interact

12 o ayadcercasfiain(i.3 eiwdi hvtePle ard,2016). allows Tarrade, system & single Chavatte-Palmer to in no reviewed tries as 3, one species, (Fig. classification different when clear of apparent and placentas easy immediately for the becomes classifying of This 2004; system Carter, 1). a & chap. establish Enders organ 2008, in the Burton, (reviewed into diversity evolved & inter-species since Wooding have of they amount origin, greatest shared the single exhibiting a have placentas eutherian While diversity Placental b. the during placenta primitive common in My. this observed 90 from specializations last evolved and have placenta adaptations placentas mammalian different eutherian the common day that a modern indicating of emergence 2012), al., after et long Reis Mya, (dos 90 about occurred radiation clade the diversification), this recent (relatively of evolution eutherian of accordance model In fuse" Mya. "long 175 the about with occurred have would eutherians and marsupials to common the origin branch this in 1987). involved al., least of et at (Lillegraven monophyletic similarity that is suggest the organ clearly the case of placentation placentation any origin eutherian of In in emergences 2015). tissues indendent Blackburn, and two 1987; processes that al., placentation excluded et marsupial entirely (Lillegraven and be occurred eutherian cannot of it comparison the though by placenta monotremes. traits, supported from mammalian separation is their the origin after clades, common of two the This emergence of ancestor are the common last marsupials that the and in suggests eutherians occurred placentas principle both complex Since parsimony and species. the mammalian periods extant placental, gestation the long of present most Eutherians regroup and 2). not eutherians the Fig. do iii) and (see and placenta oviparous placentalia short-lived are a or present that which monotremes marsurpials into the the ii) split placenta, i) be a placentation: present can to Mammalia regard class with estimed The be subgroups 1987). can three emergence al., of et time (Lillegraven the phylogeny and mammalian origin using monophyletic of probably very is placenta mammals in the Nevertheless, evolution. and origin in placental placentation of of studies record fossil complicating substantial species, no extinct is there tissue, soft of composed is placenta the Since origin Monophyletic placenta mammalian a. the of evolution and Origin 1.2.1 placentas eutherian Mammalian 1.2 al., et (Sabapatha effect immunosuppressive an have and system 2006). immune maternal the with h oslclbae ie reetbihdb o ese l 21)idctsta placental a that indicates (2012) al. et Reis dos by established tree timed calibrated fossil The ..Mmaineteinplacentas eutherian Mammalian 1.2. 13

1.2 1. Placenta

Afrotheria MAMMALIAN Xenartha PLACENTA eutherians Euarchontoglires

Laurasiatheria marsupials monotremes 200 150 100 50 0 Mya 1.2

Jurassic CretaceousTertiary

Figure 2 – Timed tree of major mammalian clades. Placental clades are indicated in red, branch length is proportional to time, see scale at the bottom (dos Reis et al., 2012). A red arrow marks the emergence of the mammalian placenta.

Classification by shape. The most straightforward way of classifying different eutherian placentas is to simply go by external appearance, but even here there are two possible approaches. The first is based on description of the distribution of areas of high maternofetal exchange at the surface of the placenta (Fig. 3A). These areas present folds or extruding placental villi that increase the exchange area and the distribution of such regions can follow four major patterns. In diffuse placentas (e.g. pig, horse, Friess et al., 1980; Samuel et al., 1974) there are no particular areas of the placenta concentrating villi or folds. Cotyledonary or placentomal placentas (e.g. ruminants, Björkman, 1969) present several spots of increased maternofetal exchange surface. The number of these so-called cotyledons ranges from 5 in deer to 150 in giraffe (reviewed in Wooding & Burton, 2008, chap. 1), illustrating the high amount of potential variability even within a given category. In zonary placentas (e.g. carnivorans, Leiser & Koob, 1993) the folds are grouped into a single band that circumvents the placenta. Finally, in discoid placentas the exchange zone takes the form of one (e.g. human, rodents) or two (e.g. rhesus monkey) discs at the surface of the placenta (reviewed in Chavatte-Palmer & Tarrade, 2016; Hafez, 2017). The second shape-based classification relies on the way the above described folds and villi interdigitate fetal and maternal tissues (Fig. 3B). The simplest structure is found in folded placentas (e.g. pig, Friess et al., 1980) where placental folds simply interlock with corresponding maternal endometrial folds. Carnivores present the slightly more complex lamellar form of placentation in which the folds are more extensive and branched (Leiser & Koob, 1993). Humans as well as ruminants and horses present villous placentas where placental tissues form tree-like structures with numerous primary villi that branch out into secondary and tertiary villi that fill corresponding maternal recesses (Enders, 1965; Björkman, 1969; Samuel et al., 1974). Finally the most complex form is found in labyrinthine placentas (e.g. rodents, Enders, 1965) in which the placental tissues are crossed by an elaborate assembly of maternal channels, forming a large interconnected mesh-like structure. We can already see that both types of classification establish

14 n lcna lo esl Fg C.Tedsoiino lo esl a ao maton impact major a has vessels blood of disposition The 3C). (Fig. vessels blood placental and (reviewed former the 2017). of deeper Hafez, tearing the in necessitates placentation tissues high-invasion placental in and while maternal the cleanly, between relationship separate to apposed they linked simply closely parturition are is during tissues placetal that this however and so maternal birth, the during placentation low-invasion lost in tissue one: maternal previous of amount the on based factors. other by Classification the and the degraded 1.2.2.d.). in largely (see blood culminates been maternal spectrum have in directly This tissues migrate bathes 1.2.2.a.). maternal placenta and also where epithelium see placentas their maternal 1973, hemochorial maintain al., the so-called et epithelia cross (Allen while cells endometrium invasion: placental the more binucleated into invasiveness, bit of represents of a subset This spectrum displays a a 1985). example exists integrity, Dantzer, for there 1980; on placenta there al., horse From et the placentation. (Friess of former type epithelia, invasive the placental least and of the maternal integrity of of apposition the loss from without stated mostly As results 4A). (Fig. placenta invasiveness pig of terms the in above, classified be also can structures placental definitive Wooding in (reviewed epithelium 1973). 8). al., uterine chap. et the 2008, (Allen crosses Burton, cells conceptus & epithelial entire uterine the of placentas body sepa- the human cells through In binucleated pass individual horse and placenta, in placenta horse found the the be from In can rate degrees. and different intrusion to is albeit implantation human, former. during the and invasion of integrity placental of of loss type without origin final fetal The (partial) in of cells resulting penetration by allows epithelium epithelium, process maternal maternal This the 1.2.2.b.). of the also of see 1981, cells al., invasion with et early (Wooding fuse syncytia of BNC heterologuous type third Fetal the fusion. of tissues: example an maternal Wooding are of in placentas (reviewed Ruminant example the 8). for beyond chap. mice 2008, cells and Burton, rats placental & in of observed penetration is This to off barrier. slough leading epithelial epithelium epithelium, maternal uterine fetal the the of cells with of apposition groups after or interme- simply cells an are individual describes which epithelia Displacement in fetal 1985). phenotype (Dantzer, diate and integrity maternal their degree where maintain lowest both pigs and The in apposed development. observed early is during interdigitation, invasion, placentas of eutherian of invasiveness of degrees invasiveness. by Classification development. they placental that the not is and classifications in structures these ruminants placental to definitive and the factor humans account recurrent with into and but take first important the Another system. in pigs second with the grouped being horse groups, different ti lopsil odsigihpaetswt ead oterltv ipsto fmaternal of disposition relative the to regards with placentas distinguish to possible also is It step, implantation early very the during invasiveness on based classification this to addition In clfe&Edr 17)dsigihbtenfu major four between distinguish (1975) Enders & Schlafke nte osbecasfiaino uhra lcna is placentas eutherian of classification possible Another ..Mmaineteinplacentas eutherian Mammalian 1.2. 15

1.2 1. Placenta

the efficiency of exchanges between them. In a concurrent system (parallel blood flows) for example, the theoretical maximal efficiency is 50% while a countercurrent system allows for a theoretical maximum of 100%. Perhaps unsurprisingly, no eutherian placenta employing a fully concurrent system has been described to this day. An example of a placenta presenting an almost perfect countercurrent system can be found in guinea pigs, while other species present intermediate arrangements like cross-current (e.g. cats) and villous (e.g. humans)(Fig. 3C, reviewed in Wooding & Burton, 2008; Hafez, 2017). 1.2

The final major classification is based on the histological nature of the maternofetal interface and the amount of layers that separate the maternal and fetal circulations (see Fig 4A). Wooding & Burton (2008, chap. 1) suggest that this way of categorizing placental structures is more relevant to the physiological functions of eutherian placentas and as such it will be presented in more detail than the other classifications.

1.2.2 Placental types by maternofetal interface

This classification was first proposed by Grosser (1909) and is based on his observation that while all fetal tissue layers (endothelium, connective, epithelium called trophoblast) were al- ways present in every eutherian placenta, the amount of maternal tissue layers (epithelium, connective, endothelium) was variable. Grosser proposed four categories of placenta: Placenta epitheliochorialis, where all maternal tissues are still present and the trophoblast is in contact with the maternal epithelium, Placenta syndesmochorialis, in which the maternal epithelium is absent and the trophoblast is in direct contact with connective tissue, Placenta endotheliochori- alis, in which both the maternal epithelium and connective tissue have been destroyed and the trophoblast is in direct contact with the maternal endothelium, and Placenta haemochorialis, in which all maternal tissues have disappeared and the trophoblast is in direct contact with maternal blood. Most of this classification proved remarkably accurate for being based only on light mi- croscopy, but the category of syndesmochorial placentation was based on a mistake. Grosser (1909) had concluded that in ruminant placentas the maternal epithelium was destroyed and replaced by the fetal trophoblast, but closer investigation with more elaborate methods revealed that both epithelia remain present until term. This led to ruminant placentas being described as epitheliochorial for a time, however as reviewed in Wooding (1992), ruminant placentas present fusion events between fetal and maternal cells, separating them from all other epitheliochorial placentas and justifying their belonging to a separate category: synepitheliochorial placentation. Grosser (1909) had also suggested that this classification might represent an evolutionary progression, that epitheliochorial placentas are the most basic and least efficient, having six cell layers separating maternal and fetal circulations, while hemochorial placentas are the most evolved and efficient, having only three layers in-between both circulations. This supposed

16 h ulttv ocnrto fa mgnr ouei niae nteshmswe osbeand possible when schemes the on transferred. the amount indicated indicates on thickness is solute, impact solute the major of imaginary pink. transfer a in an denote vessels has arrows of maternal vessels red, concentration in blood represented qualitative vascular fetal are by The vessels Orange and Classification blood Fetal established. (C) exchanges. maternal tissues. maternofetal be of of maternal can efficiency indicate disposition categories M relative tissues, major and The yellow four maternal while arrangement. and with tissues fetal variable interdigitate indicate is placentas F interdigitation all and the help While in of can indicated interdigitation. complexity regions distinct are by exchange the these Classification are maternofetal of for (B) there specialized distribution red. Areas diffuse) and systems dark categories. shape (except classification distinct The placentas into several . placentas by eutherian exchanges separate maternofetal illustrated many for as In specialized morphology. placentas regions eutherian gross by of Diversity Classification – 3 Figure C B A M F ..pge.g.cat e.g. pig ..pge.g.cattle e.g. pig ifs cotyledonary diffuse folded M F F 100 0 0 crosscurrent 0 100 100 ocretcountercurrent concurrent M M x ..nn e.g.guineapig e.g. none e.g. cat classification byvasculararrangement M F classification byinterdigitationtype classification bygrossmorphology lamellar y 50 50 x+y M M M F 0 100-x 100 e.g. human 100 ..Mmaineteinplacentas eutherian Mammalian 1.2. 99 e.g. cat zonary villous x F e.g. human villous F M e.g. guineapig 0 labyrinthine e.g. human 1 0 discoidal F (A) 17

1.2 1. Placenta

significant difference in efficiency based on the number of layers separating maternal and fetal blood has since been disproved as Grosser did not take into account the finer details of placental structure that can have major effects on the permeability of placental layers and the efficiency of exchanges (reviewed in Wooding & Burton, 2008; Chavatte-Palmer & Tarrade, 2016; Hafez, 2017). In the following parts the different types of placentation shall be described using representative species and with particular focus on the structure of the maternal interface, cell-cell fusion events and the invasiveness of fetal tissues. 1.2

a. Epitheliochorial placentation

In epitheliochorial placentation, all three layers of maternal tissue remain present and conserve their integrity during gestation (Fig. 4A). The maternal epithelium and trophoblast are apposed and their microvilli interdigitate. The trophoblast does not present cell-cell fusion events, re- maining in the form of individualized trophoblast cells called cytotrophoblasts (CT). Pig and horse placentas are the most studied representatives of this type of placentation.

Pig. At first glance, the pig placenta seems like the simplest eutherian placenta, it is lowly invasive, diffuse, folded and epitheliochorial, in other words an apposition of membranes that then fold to increase the available exchange surface (Friess et al., 1980; Dantzer, 1985). This results in the presence of the full complement of six layers separating maternal and fetal bloodstreams, hinting at low exchange efficiency. In reality this is not the case, while all layers remain present until term, their thickness is decreased considerably at the tip and sides of the folds, and the blood vessels indent the epithelial layers reducing the interhemal distance to as little as 2 µm, instead of 20-100 µm at the base of the fold (reviewed in Wooding & Burton, 2008, chap. 5). Comparing the weight of fetal mass supplied per gram of placental mass suggests that, according to this particular metric, even with six layers to cross the pig placenta is more efficient (9 g of fetus/g of placenta) than human placentas that present only three interhemal layers (6 g of fetus/g of placenta) (reviewed in Dantzer et al., 1988).

Horse. The horse placenta is structurally very similar to that of pigs, also resulting from a simple apposition of fetal trophoblast and maternal epithelium. Its general structure however, is villous rather than folded (Samuel et al., 1974) and its efficiency might be even higher (14 g of fetus supplied by 1 g of placenta, reviewed in Dantzer et al., 1988). A particularity of horse placentas is the presence of a subpopulation of binucleate invasive cells. These cells start as a so-called "chorionic girdle", encircling the fetus (Allen et al., 1973). The girdle cells are binuclear, express MHC I molecules (Donaldson et al., 1990) and detach from the conceptus to invade maternal tissues. They cross the uterine epithelium by traversing the body of epithelial cells, rather than by crossing at the tight inter-cellular junctions (Allen et al., 1973). Once on the other side they aggregate to constitute what is called endometrical cups, stop expressing MHC molecules and start secreting hormones that assist placental growth, before finally being

18 h atrtepaetltp sidctduigtesm oosa h rmsi A n h rsneor time 2012). presence to al., the et proportional and Reis is (dos (A) length tree Branch in respectively. the frames minus below or the scale plus the as For a in paraphyletically. colors by indicated line same distributed indicated as the A is are using cells tissues. types indicated fused fetal interface of is absence Placental type of invasiveness placental (B) the of side. latter fetal order the and in maternal arranged the eutherians, tree. delimits in species mam- found the eutherian in interfaces on interface maternofetal maternofetal repartition of their types four and the mals of representation Schematic – 4 Figure PLACENTA MAMMALIAN B A 200 cytotrophoblast (CT)cytotrophoblast uascCeaeu Tertiary Cretaceous Jurassic blood vesselblood blood vesselblood epithelium uterine 150 Eutheria epithelio- chorial heterologous syncytiumheterologous 100 Invasion of maternal tissues of maternal Invasion synepithelio- chorial A ceai ersnain ftefu ye of types four the of representations Schematic (A) 50 endothelio- syncytiotrophoblast (ST)syncytiotrophoblast chorial ..Mmaineteinplacentas eutherian Mammalian 1.2. 0 Mya haemo- Monotrema Marsupiala Afrotheria Xenarthra Strepsirrhini Haplorrhini Rodentia Lagomorpha Insectivora Carnivora Chiroptera Perissodactyla Suidae Ruminantia Cetacea chorial chorial

mother fetus fusion - - - + - + + + + + + + + 19

1.2 1. Placenta

destroyed by cells of the maternal immune system (Allen & Moor, 1972; Allen et al., 1973; Donaldson et al., 1990). The presence of this invasive subpopulation is especially striking since apart from it epitheliochorial placentas are non-invasive.

b. Synepitheliochorial placentation

Ruminant placentas are the only synepitheliochorial placentas described to this day. At first glance they are very similar to epitheliochorial ones, presenting all six layers at term. They 1.2 were in fact classified as such until Wooding (1992) established the synepitheliochorial category based on the presence of cell-cell fusion events at the maternofetal interface. In contrast to the epitheliochorial placentas described above, here binucleated trophoblast cells fuse with cells of the maternal epithelium, forming heterologous syncytia that directly connect the fetal and maternal sides (see Fig 4A, reviewed in Wooding (1992)). The degree of syncytialization of the interface depends on the species, it is for example low in cattle, but high in sheep, the two most studied synepitheliochorial placentas and thus the two common examples of each category.

Binucleate cells in synepitheliochorial placentation The appearance of BNC is a character- istic of synepitheliochorial placentation (reviewed in Wooding, 1992; Wooding & Burton, 2008, chap. 6). These cells are generated by division of trophoblastic nuclei without cytokinesis and represent about 15-20% of trophoblast cells (Wooding, 1983). Mature BNC are very large and present a well developed translational apparatus as well as numerous microvesicles containing dense granules (reviewed in Wooding, 1992; Wooding & Burton, 2008, chap. 6). As in horse placentas (see 1.2.2.a. above), the BNC are migratory and invade the uterine epithelium, however the employed mechanism is very different. BNC mature between two trophoblastic cells before extending a pseudopod towards the maternofetal interface. This pseudopod squeezes between the two neighboring trophoblast cells and separates their tight junctions, replacing them with its own (Morgan & Wooding, 1983). This maintains the integrity of the placental epithelium while allowing the BNC to contact the maternal epithelium. Once the pseudopod has made contact with the maternal epithelium, a number of specialized vesicles flock to its tip to extend the membrane surface area (Wooding & Burton, 2008). This process induces a break in the microvillous surface of the maternofetal interface since it generates a flat membranous area. The flat area of the BNC membrane bulges into the flattened membrane of the maternal epithelium and both membranes fuse, allowing the BNC cell to inject its contents into what was the maternal epithelial cell (Wooding et al., 1981). Following the fusion event the membrane of the BNC that remains between trophoblast cells gets resorbed by the trophoblasts, re-establishing a tight junction contact between the trophoblast cells (Morgan & Wooding, 1983). For a summary of these processes, see Fig. 5A. The fusion of the BNC with a maternal epithelial cell creates a heterologous syncytium in which nuclei and cell contents are both of placental and maternal origin. This also gives the contents of the BNC cell, and especially its dense vesicles, access to the basal lamina of the maternal epithelium and

20 fteBC h ycta lqelyras a napie irvlosbre oincrease to border microvillous amplified an secretions has for also circulation layer maternal plaque the syncytial to the access BNC, direct the more (Wooding, a of syncytium providing the the into to into fusing addition BNC fuse more In reach dozen that can 1984). a and cells about above, to only described corresponding as the nuclei, 25 cell, are to trinucleate up BNC a as that starts indicating plaque syncytial odd, Every always syncytium. show is (1984) Wooding number al., plaques, total et syncytial the Wooding the in 5B, present that (Fig. nuclei placenta of syncytia amount existing definitive the into counting the or cells By sheep of 1981). placental like interface new some of maternofetal In fusion the species. repeated at the through develop on depends plaques cells sycnytial fused large the goats, of fate the occur, BNC implying placentation synepitheliochorial in their Syncytia allows process fusion barrier. epithelial and maternal migration the bypassing particular circulation, very maternal the this to and close release 6) chap. Wooding in 2008, reviewed Burton, example, for & glycoproteins cells, associated hormones fused placental pregnancy the contain and granules of lactogen The membrane (placental 1990). al., basal et the (Wango circulation at maternal specifically the towards observed is vesicles these of exocytosis cells. trinuclear heterologous of formation existing the already for (2013) as an al. same et into the Cornelis is fused their from mechanism can Adapted secrete The junctions. plaques. BNC they adherent syncytial re-establishes placentation, where form and to epitheliochorial membrane membrane syncytium type basal BNC the sheep the of epithelial In to rest maternal transferred (B) the the are absorbs with trophoblast granules mixes The cytoplasm BNC contents. BNC the the and fusion, and Following cytoplasm process. BNC cell the fetal in own junctions between adherent their replacing observed with interface, maternofetal processes the towards projections fusion placentas. cytoplasmic extending the before ruminant of in representation epithelium Schematic maternal – 5 Figure B A cow, sheep A ea N auebtentotohbatcells trophoblast two between mature BNC Fetal (A) sheep hl nalrmnn lcna uinevents fusion placentas ruminant all in While tight junction uterine epithelium syncytial plaque(cow, sheep) trinucleate cell(cow, sheep) BNC granule binucleate cell(BNC) trophoblast ..Mmaineteinplacentas eutherian Mammalian 1.2. 21

1.2 1. Placenta

exchange surface and is thinner than the maternal epithelium, reducing interhemal distance to less than 1-2 µm (reviewed in Wooding & Burton, 2008, chap. 6). In cow and deer placentas the syncytial plaques are only present during implantation after which the uterine epithelium regrows rapidly (King et al., 1979; Wathes & Wooding, 1980), at any subsequent step and in the definitive placenta only transient trinucleate cells resulting from a single fusion event can be observed (Wooding & Wathes, 1980). The trinucleate cells then undergo apoptosis after having exocyted their granules, using a pathway that is also implicated 1.2 in human placentation (Ushizawa et al., 2006), which explains the absence of a more extensive syncytium.

c. Endotheliochorial placentation

As the name indicates, in endotheliochorial placentation the fetal tissue is in direct contact with maternal endothelium, the maternal epithelium and connective tissue having disappeared in the definitive placenta. The most well studied family presenting endotheliochorial placentation are the carnivorans (except for the hyena which is the only species of this family presenting a hemochorial placenta, Morton, 1957; Wynn & Amoroso, 1964; Oduor Okelo & Neaves, 1982), and more precisely dog and cat. In both of these species, the placenta is of the zonary type (see 1.2.1.b.). In the zone specialized for maternofetal exchanges, the outer trophoblast cells fuse together to form a large syncytium called the syncytiotrophoblast (ST) while the underlying layer remains individualized CT (Barrau et al., 1975; Leiser, 1979). As the ST comes into contact with the maternal epithelium it sends out projections that wedge between uterine cells and contain phagosomes (Barrau et al., 1975; Leiser, 1979). Both the uterine epithelium and the cells of the connective tissue are destroyed by the syncytial front which penetrates these layers and surrounds and isolates cells, detaching them from their basal membranes (Barrau et al., 1975; Leiser, 1979). At the tip of the fetal villi the maternal tissues show clear signs of breakdown, the debris being taken up by the ST (Barrau et al., 1975; Leiser, 1979). This way the ST rapidly establishes an intimate contact with the maternal blood vessel walls, which remain intact during pregnancy (Barrau et al., 1975; Leiser, 1979). The two layers are only separated by the so-called interstitial lamina, which is very close in both structure and composition of the basal lamina of maternal vessels. The lamina is of variable thickness and the fetal capillaries often indent the syncytiotrophoblast, reducing the interhemal distance locally (reviewed in Wooding & Burton, 2008, chap. 7). Outside of the specialized region of the zonary carnivoran placenta, the maternofetal interface is of the epitheliochorial type.

d. Hemochorial placentation

The hemochorial (or haemochorial) category contains the most invasive placentas. Placental invasion of maternal tissues is so strong that the definitive maternofetal interface does not

22 Trmiigidvda el.Bten1 n 9weso etto h ubro uliin nuclei of number the gestation of from of exponentially number weeks increases the 39 ST reducing and the syncytium, 13 Between the cells. maintained to CTs individual be remaining more to CT of needs to layer addition connect syncytial implies to the which projections Additionally expanded, to out 2008). and cuboidal send al., a and et from flattened (Jones switch are cells a CTs surface neighboring by the villus accompanied which in is in increase and hemomonochorial phenotype vast CT ’octopoid’ a the of an to rate to due growth be rise the to outpaces giving seems which change area thus This 2008). discontinuous, al., becomes et (Jones layer placenta CT the while m 12 1995), about of continous one area remains syncytium an the (covering expand layer villi the a overlying As layer 2008). ST al., et outer (Jones an layer During of CT example. consisting continuous for hemodichorial, placenta is human placenta definitive human the the in stages, early case the the is this term, at trophoblast of layer Hemomonochorial. in (reviewed cellular or syncytial be may 8). which chap. layers group, 2008, this cell Burton, of three & species However,Wooding and among one blood. trophoblast between maternal this have of in can structure directly it the bathing in variation is considerable trophoblast is the there tissue, maternal any contain ta.(92 ecieta h muto Tnce nrae hogotpenny( pregnancy throughout increases nuclei to CT 13 Simpson of week Accordingly, amount 6). the inflow chap. that the 2012, describe to al., (1992) due et al. Benirschke be et in must (reviewed increase layer fused this ST, the human into CTs the of in desribed been ever has activity mitotic nraigtevlm n lwn h eoiyo loflw n rvdn nuhblood enough providing both 1983), and bloodflow, al., et of Pijnenborg velocity 1967; the al., slowing et and (Brosens volume arteries the the sheath increasing of muscle mouth the of the breakdown the widening (Pijnenborg for the and lumen responsible are vessel invades populations blood two these the Together the invades 1980). al., second of et the more one while sub-populations, tissues, arteries two maternal the are around invade endometrium There to arteries. villi maternal the extravillous the from the specifically separate cells: invasive they of because population so-called another trophoblasts, presents also (Enders, placenta spaces inter-villous human filled The maternal blood breaches 1989). create and ST circulation the maternal endometrium, the the access the reach to in and capillaries embedded epithelium is the traverse conceptus to the conceptus Once whole endometrium. the allows CTs. and of cells supply epithelial constant uterine al. a by et enabled Mayhew is by ST human obtained the further were of CT), results growth the Similar in that pregnancy. 10% confirming during that ST, (1994), phenomenon rate the continous constant a in is a layer (90% at syncytial the pregnancy operates into cells during of constant incorporation the remains that syncytium. ST indicating the nuclei and of for proportion CT cells the that the of suggest supply (1992) in al. constant et a Simpson by provide supplied to numbers the able Interestingly thus are and capacities stem-cell h Ti nivsv ise uigipatto tsnsotpoetosbtenthe between projections out sends it implantation during tissue, invasive an is ST The 5 . 8 × 10 9 twe 9,idctn hta es oeclso hspplto uthave must population this of cells some least at that indicating 39), week at sternm niae,hmmncoilpaetscnanasingle a contain placentas hemomonochorial indicates, name their As 6 2 . 8 ttr,Gut&Okeod 96 utn&Jauniaux, & Burton 1986; Ockleford, & Gaunt term, at × 10 9 to 5 . 8 × 10 ..Mmaineteinplacentas eutherian Mammalian 1.2. 10 Smsne l,19) ic no Since 1992). al., et (Simpson 0 . 6 × 10 9 23 at

1.2 1. Placenta

ABC placental villus syncytiotrophoblast maternal blood endocascular EVT uNK interstitial EVT

1.2 endometrium myometrium arterial muscle sheath endothelium

Figure 6 – Remodeling of maternal blood vessels by extravillous trophoblast cells in humans. (A) Schematic representation of maternal uterine blood vessels before remodeling. The vessel is narrow and surrounded by a muscle sheath around its whole length. (B) Maternal blood vessel post-remodeling. The endovascular extravillous trophoblasts (EVT; green) have invaded the opening of the vessel, widening it to alter blood flow and speed. The interstitial EVT (blue) invade the maternal tissues to remove the arterial muscle sheath and reduce vasoreactivity. The EVT are assisted and regulated during this process by the maternal uNK cells (white). (C) Representation of the relative positions of fetus, placenta, maternal tissues and maternal blood vessels during human pregnancy. Adapted from Moffett & Loke (2006); Parham & Moffett (2013).

supply to the placenta while also avoiding excessive shear stress from blood velocity. Burton et al. (2009) use computer models to predict that the observed dilation of the mouth of maternal arteries from 0.4 mm to 2.5 mm of diameter slows the blood flow from 2.5 m/s to about 10 cm/s. This is important both to avoid damaging the fetal villi and to insure that the blood remains in the intervillous spaces for long enough for exchanges to take place. The degradation of the the arterial muscle sheath is probably also important to diminish vasoreactivity to maternal signals, which might decrease the blood supply and damage the placenta and fetus (Burton et al., 2009). The remodeling process also necessitates interaction with the maternal uNK cells, see 1.1.2.b.. In addition to the blood supply increase following these changes in maternal blood flow, the fetal blood vessels bulge into the syncytiotrophoblast towards the end of gestation (Burton & Tham, 1992), reducing the thickness of the layers separating both circulations to about 3 µm over 40% of the total placental surface (reviewed in Wooding & Burton, 2008; Benirschke et al., 2012).

Hemodichorial In the rabbit placenta, implantation occurs when the conceptus sends out syncytial knobs that fuse with the maternal epithelium (Larsen, 1963; Enders & Schlafke, 1971). This triggers a syncytialisation of the latter, forming large uterine syncytial plaques which subsequently are dislodged and phagocyted by the hybrid syncytium (Enders & Schlafke, 1971). This syncytium then sends out smaller processes that invade the basal lamina and reach

24 nogatclsb noioi,sgetn oeta norn oe si omnfrthese for common is as role, endocrine potential a suggesting differentiate blood endomitosis, layer by outer and cells the flow of giant blood suggests cells into slower and the blood of Interestingly maternal exchanges. regions maternofetal with establish enhancing contact might stasis, direct epithelium in fenestrated be outer confirmed to ST-I (later this layer allow that outer that the 2005) of al., cells al., et et the Coan Coan in by, 1965; perforations (Enders, II describes and also II I (1965) layers between spaces Enders to filled 2005). attached fluid loosely of formation only the is to and leading III, term and until layer I continuous trophoblast integrity and The cellular structural 1976). al., remains both et meanwhile, enhancing Metz 1965; junctions (Enders, gap layers the and between tight communication and of series the a in they by are observed ST-II human, connected been and intimately in ST-I ever 1975). having As Legrand, figures syncytial. inner & mitotic Hernandez-Verdun are two no 1974; (Hernandez-Verdun, ST-II) events, syncytium The fusion and ST-I side. cell-cell from called maternal result also the to found. III, seem is from and trophoblast starting (II trilaminar I-III trophoblast characteristic numbered of a layers are villi layers the of the surface convention the By At 2004). of al., unit et 267 per (Coan from area gestation during the dramatically 1909), increases blood Grosser, cm maternal 1892; with (Duval, contact in blood is maternal that volume with contact direct in are 2010). Cross, embryo & the Hu to in nutrients (reviewed and gestation gases early penetrate supplies that during that structure cells transient giant a invasive form and into tissues cells maternal trophoblast Implantation the some 1992). Given, of & differentiation cells Blankenship the uterine maternal 1991; triggers of the also Enders, action & with the (Welsh contact to trophoblast due establishes the seemingly than conceptus destroyed, rather is the and as off soon sloughs latter as the placentas epithelium, these In mice. and rat Hemotrichorial and (1926) species. Mossman hemochorial by other rabbit and in rabbit for described (1981) first al. as et countercurrent Pijnenborg arteries, by a the reviewed induces around forms that sheath cells This trophoblastic a by 1971). invasion of is drains formation al., there and term, et side to persist maternal vessels Carter the the if 1926; towards Even exchanger. down (Mossman, back veins flows it maternal where the on through from blood lacunae, supplying and the placenta of the side of length fetal full the across running gestation, during intact 1971). Schlafke, & trophoblast Enders of 1965; layers (Enders, two CTs which present of in are layer placenta a hemodichorial always a is establishing there thus ST, processes, maternal the these by underlying later During and 1963). capillaries Larsen, by 1926; supplied (Mossman, first blood, arteries intrasyncytial maternal to with rise filled gives be which will vacuolises that ST lacunae the established, is contact circulation Once maternal 1963). the (Larsen, with contact establishing and them breaching vessels, blood maternal the 2 h entv lcnata setbihdi h eodpr fgsainpeet il that villi presents gestation of part second the in established is that placenta definitive The remain arteries maternal major the placenta, human the for observed is what to Contrarily /cm 3 o49cm 489 to h eorcoiltp fpaetto sfudi oet,frexample for rodents, in found is placentation of type hemotrichorial The 2 /cm 3 niaigta ope ewr fvliadbodsae sformed is spaces blood and villi of network complex a that indicating , ..Mmaineteinplacentas eutherian Mammalian 1.2. 25

1.2 1. Placenta

cells in rodent placentas (Coan et al., 2005; Simmons et al., 2007). Of note, while the presence of three layers in these placentas might suggest a huge increase in thickness compared to the single layer present in hemomonochorial placentation, stereology studies reveal that the three layers are about 4 µm thick at term in mice (Coan et al., 2004), remarkably close to the 3 µm observed in human (see above). Similarly to what is observed in rabbits but not in humans, the major arterial channels persist in rodent placentas, but there is significant invasion of the vessels by giant trophoblastic cells that leads to the establishment of large blood sinuses and reduces vasoreactivity to maternal signals (Pijnenborg et al., 1981; Simmons et al., 2007). As in human, the participation of uNK cells in blood vessel remodeling in rodents has been established, see 1.1.b.. 1.3 1.3 Viviparity in non-mammalian species

Before going into further details on non-mammalian viviparity, a quick reminder about defini- tions seems to be in order. As defined in Blackburn (2015) and described in 1.1.1, viviparity is defined by giving birth to live young and regroups a spectrum of strategies for embryonic nutrition ranging from lecithotrophy (the embryo is dependent on the yolk, very few nutrients provided by the mother) to matrotrophy (the mother provides almost all nutrients necessary). Viviparity can be inferred from the size of ova or the presence of well-developed fetuses in the oviducts for example (Blackburn, 1993c), while matrotrophy can be inferred from diverse criteria such as an important increase in dry mass of the conceptus during gestation or chemical composition analysis of the conceptus at different states (an important increase in mass or spe- cific components indicates uptake of material by the fetus, Blackburn, 1994). It is important to distinguish viviparity and matrotrophy, the latter being a subset of the former that we will focus on in more detail, since mammals are extremely matrotrophic.

1.3.1 Prevalence of viviparity outside of mammals

a. Invertebrates

Contrary to popular belief, it has been known for a long time that viviparity, both lecithotrophic and matrotrophic, is far from restricted to mammals. Existence of both viviparous fishes and snakes was described by Aristotle (ca -350) in his pioneering "History of animals" and de- scriptions of live birth and placental structures in invertebrate species go back several hundred years. Quoy & Gaimard (1835) for example make references to the placenta of salps, as do Krohn (1846) and Huxley (1851) a few years later. A recent comprehensive review of available literature combined with their own data led Ostrovsky et al. (2015) to the conclusion that out of the 34 animal phyla, 21 present at least one matrotrophic viviparous species, and in 11 of the 21 matrotrophy was a widespread or common phenomenon (Fig. 7). While the authors con- sider this a conservative estimate due to lacking data for many species, it still represents about

26 rsns15idpnetoiiso iiaiy bu 5 falvrert mrecs ayof many emergences, vertebrate all of 75% about viviparity, of origins independent 115 presents Squamata. in Viviparity 2015). al., example et no Ostrovsky Interestingly, 2015; Mya). (Blackburn, 175 birds Reis at in mammals dos observed in been comparison, has viviparity For viviparity of of 1997). of origin fishes al., the comephorid place et the (2012) Hunt al. in 1981; et Mya Wourms, 5 1873; most (Dybowski, the about while Baikal occurred 2008), Lake al., viviparity example et of known (Long emergence oldest fishes placoderm described the now-extinct wildly, recent in levels Mya varies uneven 380 viviparity about the vertebrate at by set of impacted underes- being origin significant are a of numbers to date these leading The that potentially timation. clades, mind different in of kept knowledge be and group investigation origins to of the 29-31 has vertebrates with also close makes It come this 7). arthropodes (2015) though (Fig. matrotrophy, al. of et origins Ostrovsky most ones the to singular presenting According the 7). to addition (Fig. in origins mammals 2015) 32 in al., and et viviparity Ostrovsky of placentotrophy, (mostly origins matrotrophy 152 of distinguishes (2015) Blackburn animals, vertebrate In vertebrates Non-mammalian b. observed those to mostly correspond that clade. they noteworthy matrotrophic that best-studied also the and is mammals, emerge, it in to aspect, seem interesting sites an and number is modes immense clades dominant distant the these While in variation. observed intra-phyla strategies great of show others 2015). sites whereas and al., modes matrotrophy, et uniform very (Ostrovsky of present phyla cavities most some body that it report and cavities, (2015) system body al. or et genital Ostrovsky Additionally, tissues female other the in in occur place can takes it is commonly while matrotophy but of species, site the the on Similarly, dependent 2015). highly al., invertebrates et (Ostrovsky in different placentotrophy matrotrophy and at of histotrophy successively modes are dominant or The simultaneously 2015). occur (Blackburn, stages either developmental These can 2015). al., and et non-exclusive Ostrovsky are 2015; (Blackburn, categories which implicated in 5) placentotrophy is nutrients, 6) structure and said placental organs ingests specialized or tissues a embryo maternal the eats which embryo the in which histophagy in absorbs matrophagy 4) embryo mother, the which the in by histotrophy secreted 3) nutrients embryos, embryophagy sibling 2) yolks, cannibalizes sibling embryo assimilate the and which ingest in embryos the which in nutrient of oophagy animals, modes 1) diverse matrotrophic delivery, six all regroups example, between for itself, observed matrotrophy diversity term The the not. to or vertebrate comparison in pales it work, this of 2015). (Blackburn, higher be total to the assumed reproduction, safely can of viviparity modes of viviparous emergences of of subset number stunning a only a Since describes representing clades. viviparity related animals, distantly matrotrophic of of number impressive evolution an the across convergence evolutive in of example matrotrophy of origins independent 145 hl h osdrbedvriyi amla lcna a h ou ftepeiu part previous the of focus the was placentas mammalian in diversity considerable the While qaaa(cldrpie)icuelzrsadsae.Ti family This snakes. and lizards include reptiles) (scaled Squamata ..Vvprt nnnmmainspecies non-mammalian in Viviparity 1.3. 27

1.3 1. Placenta

matrotrophic Animalia clades Xenoturbellida Acoelomorpha 2 2/11 Chordata ** 37 221/1084 Hemichordata Echinodermata 19 20/175 Loricifera 1 1/3 Priapulida

1.3 Kinorhyncha Nematoda 10 21/267 Nematomorpha Tardigrada Onychophora 1 2/2 Arthropoda 29 83/3022 Chaetognatha Rotifera 1 1/35 Acanthocephala 1 3/22 Cycliophora 2 1/1 Kamptozoa 1 2/4 Bryozoa 23 53/189 Mollusca 13 13/660 Platyhelminthes 11 162/300 Gnathostomulidae Gastrotricha 1 1/17 Phoronida Brachiopoda Nemertea 5 5/42 matrotrophic Chordata clades Annelida 9 8/130 Sipunculida Mammalia 1 157 Orthonectida 1 2/2 Squamata 6 13 Dicyemida 1 3/3 Amphibia 9 9 Ctenophora 1 1/25 Teleosts 13 15

Cnidaria 4 4/288 vertebrates Placozoa ** Chondrichtyans 5 24 Porifera 4 25/146 Tunicata 3 3

Figure 7 – Matrotrophic viviparity is a widespread phenomemon among Animalia. Cladogram of Animalia with phyla containing matrotrophic species in red. The number after the taxon name represents the conservative estimate of origins of viviparity, the bar the number of families containing at least one matrotrophic species (truncated for Chordata and Platyhelminthes) and the number following the bar is the number of families containing at least one matrotrophic species compared to the total number of estimated families of the phylum. (inset) Detail of the origins and prevalence of matrotrophy in selected Chordata clades, using the same representation as the main cladogram. Adapted from Ostrovsky et al. (2015).

28 otnal vrtecus fgsain si h aei anis raogietefertilized the 1993). Gilmore, alongside 1992; or (Yano, Lamnoids, Pseudotriakidae in in case case the the is is as as ovum, gestation, of course released the in either are over oophagy that continually ova with unfertilized other (Wourms, associated by provided placenta often resources assimilates a is embryo the forming matrotrophy which tissues, Chondrichtyans, like maternal In some, to 2015). in apposed Blackburn, closely 1981; and are species embryo the different of in pattern surfaces This times tissue. ovarian multiple specialized by arisen released has are that nutrients ingests or absorbs embryo have clades viviparous of matrotrophy of 75% emergence the about 2015). to (Blackburn, but predisposed are viviparity, latter of the that origins suggesting matrotrophy, evolved of fishes cartilaginous 15% and about bony whereas represent 5% species, only only matrotrophic viviparity, contain of clades origins viviparous vertebrate these of of 75% represent Squamates While particularly fishes. be in to seems common matrotrophy fact In of 2015). emergence Blackburn, in present 1993; single Gilmore, clades case (Chondrichtyans, a following extant the several Both is even 2008). sometimes as the matrotrophy, al., viviparity, in of et origins as Long of well origin, number as 1 a 2015), (Placodermi, (Blackburn, fishes origins) armored 9 extinct least at (Chondrichtyans, fishes cartilaginous fishes. in Viviparity 1.3.2.c.). also see 2012, & Flemming, (Jerez & fetus Blackburn the 2007; by al., tissues et maternal Vieira of 2001; invasion zones Ramírez-Pinilla, cellular and specialized placentomes, layers, and cells areolae placentas syncytialized as the presenting such to often similar species, very mammalian structurally in and transfer observed nutrient for specialized highly is placenta named (previously genera African the and Mediterranean the lizards: neotropical the matrotrophic 1993b), four of contains clades that family six a (skinks), the Scincidae in of case the mostly Blackburn, is & This (Flemming 2015). matrotrophy Blackburn, of 2003; amount substantial a exhibiting there but clades mother, different the six by provided are generally are nutrients of amounts small only one, matrotrophic 1998). al., et Cruz la Mendez-de in oviparous (reviewed are identified been others have while vivipariy genus viviparous are the genus In a 1999). of (Blackburn, species occurred some have where emergences levels, many subgeneric Accordingly at 2015). 1999, (Blackburn, recent relatively them ntlot,ms artohcseisepo omo ittoh rhsohg hr the where histophagy or histotrophy of form a employ species matrotrophic most teleosts, In the than spectrum the of end lecithotrophic the to closer are species viviparous these of Most rcyei ivensii Trachylepis iiaiyhsbe bevdbt nbn Tlot,1 rgn)and origins) 13 (Teleosts, bony in both observed been has Viviparity Mabuya uei anchietae Eumecia Poeciliopsis eu Bakune l,18;Jrz&RmrzPnla 2001) Ramírez-Pinilla, & Jerez 1984; al., et (Blackburn genus Sceloporus lcbr lmig 00.I hs pce the species these In 2010). Flemming, & Blackburn , tlot,Rzike l,20)o carcharhiniforms or 2002) al., et Reznick (teleosts, Femn rnh 01 and 2001) Branch, & (Flemming o xml,a es oridpnetoiisof origins independent four least at example, for ..Vvprt nnnmmainspecies non-mammalian in Viviparity 1.3. hlie chalcides Chalcides Poeciliidae h absorptive the , uuaivensii Lubuya (Blackburn, 29

1.3 1. Placenta

1.3.2 Examples of non-mammalian placentas

In this section selected non-mammalian placentas will be briefly described and compared to mammalian examples. Higher emphasis will be put on the description of the Mabuya placenta as it is of particular relevance later.

a. Non-vertebrates: the placenta of salps

The most complex invertebrate placenta is found in Salpidae, tunicate chordates that live in the ocean (Bone et al., 1985; Ostrovsky et al., 2015). As was first pointed out by von Chamisso (1819), salps have a very particular mode of reproduction involving the alternation of two distinct generations: the asexual oozooid and the protogynous hermaphrodite (first female then 1.3 male) blastozooid. The oozooid results from sexual reproduction but reproduces asexually by growing a long stolon that differentiates into a chain of linked blastozooids (Bone et al., 1985). The chain then breaks into several pieces and the linked blastozooids swim around in smaller chains. As protogynous hermaphrodites, the blastozooid first develops a single egg and afterwards mature testis and spermatozoids, this leads to younger blastozooids being fertilized by older blastozooids and avoids auto-fertilization (Bone et al., 1985). The fertilized egg then grows within the blastozooid, establishing a round placenta, and develops into an oozooid which then gets released into the sea alongside the placenta. After separation, the blastozooid starts producing sperm to fecundate younger still female blastozooids and the released oozoid produces a stolon generating the next generation of blastozooids (Bone et al., 1985). In smaller salp species, this cycle might be as fast as 46 hours under favorable conditions (Heron, 1972), while in larger species it is thought to take 36 hours to grow the stolon and less than a week for gestation of the oozooid within the blastozooid (Bone et al., 1985). After fertilization of the egg in the blastozooid ovary, the conceptus breaks through the ovary wall into the jet chamber of the mother, where it implants and establishes the salp placenta (Sutton, 1960; Bone et al., 1985). The mature placenta is composed of two syncytial, highly interdigitated cell layers separating the maternal and fetal blood spaces, reminiscent of mammalian hemodichorial placentation except for the absence of fetal endothelium (see 1.2.2.d., Bone et al., 1985). A major difference is the origin of the cell layers, both syncytia are initially of maternal origin and can therefore not be termed syncytiotrophoblast. The inner (mother-facing) syncytial layer presents additional fusion events of maternal blood cells into the syncytium while fetal leukocytes seem to add to the outer (fetus-facing), resulting in a heterologous syncytium as is observed in some cases of mammalian placentation, even though in this case the fetal contribution does not implicate specialized placental trophoblasts (Bone et al., 1985). At term the thickness of each of the two layers is between 50 and 100 µm, the outer layer being the thicker of the two. This represents a significantly higher distance to cross when compared to mammals, where the interhemal distance can be as low as 2 µm in pigs, 3 µm in humans or 4 µm in mice (see 1.2.2). It is important to note however that the salp placenta is probably not

30 lcna hr ato h itrgopto group sister the of part shark placental Blackburn, 1981; (Wourms, appendiculae for than opposite is 2015). direction the case even exchanges, this maternofetal for in specialized secrete also though are that they such epithelium as maternal embryo, the the feed trophonemata to of (Wourms,histotrophe the projections species are with matrotrophic Trophonemata highly similarity 2015). presents Blackburn, structural that 1981; high Chondrichtyans in of a observed family present be another also rays, only of they not but can species, appendiculae shark these other Interestingly, 1993). (Wourms, exchanges gaseous particular probably exchanges, contact maternofetal in in direct implicated in be tissues to not placental nevertheless considered are seem they be projections cannot these they Since tissues, 1993). maternal (Wourms, with embryo connecting stalk the the to on originate placenta that the epithelium, thin a with projections well-vascularized long is it layers. that the proof maternal with definite intact contact completely no direct with is epitheliochorial, there in not but is structure, blood hemochorial maternal a suggesting that seems epithelium, interface opinion placental the the of of without side being maternal border (1993) the Wourms smooth of unclear, a an nature presents histological what by epithelium The to covered 1993). the contrarily (Wourms, and placentas, microvilli but vascularized mammalian distance, in highly interhemal seen is in commonly sphere decrease is a The allowing stalk. connected epithelium, is umbilical thin and the an tissues extremely maternal by by with embryo contact accompanied in the mm remains is that matrotrophy 1 to growth sphere the extreme mm of this of 12 parts a indicator months, extra-embryonic into develop strong the 6 egg implantation, a of After times, 1994). course 58000 Blackburn, mm the about 1993; 1 (Wourms, over of a mass mm from grows in 150 embryo increase of the important shark size this a In to 2006). of al., egg et 30% Haines about 1993; but the (Wourms, histotrophy/phagy placentas of or Most oophagy ( 2015). shark employ spadenose the Blackburn, like group sharks, to this according in (55% species percentage vertebrates highest matrotrophic second among the species with group viviparous vertebrate the of Chondrychtians, the of part are Sharks sharks in placentation Fish: thickness layer b. of impact the b.) which and for 1.1.1.a. (see transport, important active less or be might facilitated many distance), of diffusion forms in increase use is an by nutrients water impacted strongly diffusion which is simple law, through using Fick’s layers to jet-chamber placental according cross (which, maternal gases While the of 1985). transfer al., in et are (Bone embryo placenta constantly pumped the this by of accomplished attachment functions and major respiratory contain nutrients the not that does probable blood is salp It as pigments. functions, respiratory and exchanges gas in involved oercn td nteAlni hrns hr ( shark sharpnose Atlantic the on study recent more A appendiculae, so-called develops embryo shark spadenose the placenta, the to addition In cloo laticaudus Scoliodon Scoliodon netgtdteimn otx fshark of context immune the investigated ) ,aepaettohcaddvlpcomplex develop and placentotrophic are ), ..Vvprt nnnmmainspecies non-mammalian in Viviparity 1.3. hzpindnterranovae Rhizoprionodon sensu osa 13) but (1937), Mossman another , 31

1.3 1. Placenta

placentation and found cells that resemble mammalian uNK and are associated with maternal blood vessel (Haines et al., 2006). The presence of such cells with a potential similar role in maternal blood supply would be a striking example of evolutive convergence in an already very convergent context since the shark placenta has emerged independently of the mammalian one.

c. Lizards: placentation in Mabuya and other Scincidae

As stated above, while the squamates present an astonishing 115 independent origins of vivipar- ity, most of the viviparous species are lecithotrophic (Flemming & Blackburn, 2003; Blackburn, 2015). Of the six known examples of highly matrotrophic squamates, four belong to the Scinci- dae family of lizards (commonly called skinks) and at least two of these four are associated with invasive and highly complex placentas. 1.3

Viviparity and placentation in the Neotropical skinks Mabuya. Matrotrophic viviparity in the Mabuya genus was first described in Mabuya heathi, a Brazilian representative of the genus. Vitt & Blackburn (1983); Blackburn et al. (1984) describe that this species ovulates very small eggs (1 mm, 0.4 mg) that show a large increase in dry mass over a long period of gestation (neonate: 31 mm, 154 mg). Fertilization occurs in September/October and parturition in October of the following year, this one year gestation period is much longer than the three month period observed in most viviparous reptiles (Flemming & Blackburn, 2003). About 90% of the total dry mass increase of the embryo occurs during the last three months of gestation, once the definitive Mabuya placenta is formed (Vitt & Blackburn, 1983; Blackburn et al., 1984). As Flemming & Blackburn (2003) put it: "In fact, in terms of the total amount of nutrients provided by the placental membranes, these lizards lie within the range of placental mammals". A similar reproduction pattern was then described for many other species of the Mabuya genus, including: M.agilis, M.brachypoda, M. caissara, M. frenata, M. macrorhyncha, M. nigropunctata and the as of yet unnamed but highly studied Mabuya sp. IV, and assumed for M. carvalhoi, M. dorsivittata and M. guaporicola (reviewed in Flemming & Blackburn, 2003). It therefore seems highly likely that the matrotrophic viviparous mode of reproduction is common to the entire genus and results from a single emergence, about 25-30 Mya during the radiation of this clade (Karin et al., 2016; Pereira & Schrago, 2017). The detailed structure and function of the Mabuya placenta has been studied in several Mabuya species(see for example Blackburn & Vitt, 2002; Leal & Ramírez-Pinilla, 2008), but most thoroughly in a population of Mabuya sp. IV from Curití, Colombia (Jerez & Ramírez- Pinilla, 2001, 2003; Ramírez-Pinilla et al., 2006; Vieira et al., 2007; Wooding et al., 2010; Ramírez-Pinilla et al., 2011; Hernández-Díaz et al., 2017). This species was originally referred to as Mabuya mabouya, however following redefinitions of Mabuya phylogeny, this Andean species is now called Mabuya sp. IV (Pinto-Sánchez et al., 2015). It is also of relevance here to mention that Mabuya phylogeny, as well as Scincidae phylogeny in general, still presents a number of uncertainties and older publications often used now deprecated names as some

32 aefsdt omalresnyimcvrn h nieyo h lcnoa nefc,this fetal interface, of placentomal participation precluding the epithelia, of two cells the entirety between epithelial contact the the before covering occur side to syncytium seems maternal large fusion the a form On extremely to nuclei. are fused large cells have giant two the present since always region and this the hypertrophied of in outside exacerbated epithelium are placental differences cells the constitute their binucleated also placentome, populations or two Ramírez-Pinilla mono- these 2001; While thin Ramírez-Pinilla, 2006). & tall al., (Jerez et and cells) cells) (interstitial distinct former two (giant the contains cells between epithelium wedged binucleated fetal wide the large placentome, types, 8B, the (Fig. cell interdigitated In and 2001). folded highly Ramírez-Pinilla, are & tissues fetal Jerez and maternal the which in diameter of In attached. is cord umbilical the which regions. gas-exchange these in distance interhemal the reduce to there layers placentas, cell mammalian maternal many of in loss observed no is is 2001; what Ramírez-Pinilla, to Contrary & 2006). (Jerez al., transfer et gas Ramírez-Pinilla favor to 8 about increased (to in distance interhemal vascularization reduce subepithelial to maternal thin and very fetal and areas the apposed are closely which are plaques in might epithelia absorptive segments, transfer the respiratory Between called lipid exchanges, regions. that gaseous these for suggesting in specialized plaques, occurring absorptive function ex- the important lipid the an of numerous again be sides describe here (2006) both al. that on et likely present Ramírez-Pinilla therefore nature. droplets different is a it of is regions, the material in corresponding changed but the glands in uterine glands of significant cyto- secretions interwoven absorbing plaques of in absorptive involved system placentas, usually mammalian complex are In a 2006). apposed al., instead closely et are but (Ramírez-Pinilla epithelia junctions projections fetal plasmic microvillous and maternal present the not them do of each surface and the in at and plaques placenta absorptive late-term 60 a approximately of be (Jerez to plaques seem absorptive There so-called 2001). the Ramírez-Pinilla, forming & cells exchange hypertrophic tall of of the region regions of nature small Another pole presents the abembryonic 2006). that the al., on probable et found is (Ramírez-Pinilla be it can different though is 1), substrates chap. secreted 2008, of Burton, & fetal Wooding corresponding and in secretion (reviewed maternal uptake active highly of placen- areas synepitheliochorial designate or similarly Ramírez-Pinilla epitheliochorial areolas 2001; in tas, notably Ramírez-Pinilla, placentas, & mammalian (Jerez In 2006). epithelium al., placental et the by and invagination as up the well form taken to as are degenerated having origin both cells maternal epithelial the of of secretions some debris, the with cell in fetal filled with epithelium is uterine cavity the created of The evagination slight region. a corresponding during by form accompanied that gestation, tissue of placental stages the last of the invaginations small are areolas The 8A). (Fig. exchanges Karin 2002; al., species). et Scincidae Mausfeld example re-distributing for publications for (see 2016, genera al., other et into sorted been since have species h lcnoei h ralctda h mroi oe oslyt h mro n to and embryo, the to dorsally pole, embryonic the at located area the is placentome The definitive The Mabuya lcnapeet ortpso ein pcaie o maternofetal for specialized regions of types four presents placenta Mabuya Mabuya p V h lcnoei ml ico mm 6 of disc small a is placentome the IV, sp. ..Vvprt nnnmmainspecies non-mammalian in Viviparity 1.3. lcnaweetepaetlepithelium placental the where placenta Mabuya hr r no are there µ )adthe and m) 33

1.3 1. Placenta

cells in the establishment of the syncytium (Jerez & Ramírez-Pinilla, 2001, 2003; Ramírez- Pinilla et al., 2006). The interface of the two epithelia is microvillous, with both the maternal syncytium and the fetal epithelial giant cells sending out interdigitating microvilli, increasing the exchange surface (Ramírez-Pinilla et al., 2006). Periodically this microvillous border get interrupted by the presence of small cells at the apex of a placental giant cell (Ramírez-Pinilla et al., 2006; Vieira et al., 2007). These small cells occur in groups of variable number and shape and send out thin and very long projections that cross the maternal syncytium and establish an intimate contact with maternal capillaries (Vieira et al., 2007). The projections surround the capillaries and form an extremely complex system of double membranes in direct contact with the maternal syncytial cytoplasm (Fig. 8C, Vieira et al., 2007). The presence of these invasive cells suggests that the relationship between maternal and fetal tissues is much closer than expected 1.3 at first glance. In these regions the maternofetal interface is closer to the endotheliochorial than the epitheliochorial type since there is direct contact between placental tissues and maternal endothelium. The paraplacentome is a small region of non-folded tissues (less than 1 mm) that surrounds the placentome (Fig. 8A, B) and is delimited on the outer side by an invagination of the placental epithelium (Jerez & Ramírez-Pinilla, 2001). At the edge of the placentome and paraplacentome the uterine syncytium stops abruptly and the maternal epithelium is unicellular in the paraplacentome. The paraplacentomal epithelium is virtually identical to that in the placentome, being composed of giant and interstitial cells (Jerez & Ramírez-Pinilla, 2001). There is no microvillous border in between maternal and fetal tissues in the paraplacentome, though at the border to the placentome there is some interdigitation of the two (Ramírez-Pinilla et al., 2006). The maternal side of the paraplacentome presents secreting glands and the presence of lipid droplets in both maternal and fetal cells suggests that this region might also be involved in lipid transfer (Ramírez-Pinilla et al., 2006). All things considered, the placenta of Mabuya sp. IV is remarkably similar to the ones found in mammalians. It presents specialized structures of homologous design (areolas, absorptive plaques, placentome) that seem to accomplish similar functions, even though the nature of the nutrients that are exchanged in these regions seems to differ (perhaps not surprisingly since in reptiles the main source of fetal energy is lipids, as opposed to glucose for mammals, see Thompson & Speake, 2006). While folding and interdigitation is a phenomenon seen in many tissues where an increase of the exchange surface can be advantageous, the Mabuya placenta also presents a syncytial layer, as is the case in many mammalian placentas, as well as mechanisms of fetal invasiveness to establish a more initimate contact with the maternal bloodstream. All these characteristics are remarkable examples of convergence, seeing as these two placentas emerged 150 My apart and since the last common ancestor of mammals and Mabuya would have lived 300 Mya (Kumazawa, 2007; dos Reis et al., 2012; Pereira & Schrago, 2017).

A second origin of extreme matrotrophy in African Scincidae. In addition to the Mabuya genus, the Scincidae family contains the two other lizard species presenting a similarly complex

34 oo ceea n()ad() dpe rmVer ta.(07;Woige l (2010). al. et Wooding (2007); al. same et the Vieira using from Adapted cells, invasive (B). fetal and double (A) the the in by of as vessels Representation scheme blood color (right) double maternal complex projections. around abruptly the formed cell is of invasive projections micrograph syncytium fetal membrane interdigitated. Electron longer from (left) the no placenta. originating paraplacentome are mabuya system the tissues the membrane in the invasion In syncytial but cellular large microvilli, cells. of present a Detail uterine still (C) by Cells the replaced epithelium. of is uterine by fusion epithelium highly formed replaced in the uterine are is The interface by epithelium tissues cells. fetal formed maternofetal fetal The chorionic structure the and cells. binucleated uterine of maternal giant and by placentome, representation fetal primarily between the Detailed microvilli In numerous (B) placentome region. with sides. the interdigitated, paraplacentome the around both exchanges: and band on maternofetal placentome a for present the forms specialized therefore paraplacentome regions The is as areola. and well and as plaques tissues, absorptive fetal placentome, and maternal the apposed of the Structure – 8 Figure myometrium endometrium absorptive plaque areola C A folded fetalepithelium folded uterineepithelium paraplacentome placentome Mabuya p Vplacenta. IV sp. B invasive cellprojection maternal syncytium membrane system maternal capillary ..Vvprt nnnmmainspecies non-mammalian in Viviparity 1.3. A vriwo rvdueu displaying uterus gravid a of Overview (A) placentome fetal invasivecells fetal vessel fetal chorionicepithelium maternal epithelium maternal uterinesyncytium maternal capillary microvillous interface paraplacentome 35 placenta mother

1.3 1. Placenta

placenta: Eumecia anchietae and Lubuya ivensii (formerly known as Mabuya ivensii or Tra- chylepis ivensii, Metallinou et al., 2016). Both of these rare species belong to a single clade of African Scincidae and are the sole members of their respective genus (Metallinou et al., 2016; Pereira & Schrago, 2017). It has now been established that the emergence of viviparity and matrotrophy in this African clade is independent of that observed in the Mabuya genus, though it remains unclear on whether it represents a single common, or two separate origins (Metallinou et al., 2016). The only detailed description of the Eumecia anchietae placenta available presents it as a simple folded epitheliochorial placenta, without the presence of a placentome (Flemming & Branch, 2001). While this study did have access to one late-term embryo (stage 40 of the Dufaure & Hubert (1961) development table), its conservation state was too poor to allow study 1.3 of the placenta (AF Flemming, personal communication) and since in Mabuya sp. IV the highly interdigitated placentome forms only during stages 35-40 (Jerez & Ramírez-Pinilla, 2003), it is possible that the structure of the Eumecia anchietae placenta becomes more complex and invasive later on. The placenta of Lubuya ivensii on the other hand is much more complex and highly invasive, even more so than that of Mabuya, though its study is complicated by the rarity of the species. During implantation the Lubuya placenta sends out knobs of cells that traverse the uterine epithe- lium and reach through to the basement membrane (Blackburn & Flemming, 2010, 2012). The nature of these knobs seems to be syncytial even though this has not been clearly demonstrated so far. Once these placental knobs reach the basement membrane, they begins to send out pro- jections under the uterine epithelium and start to grow between it and the basement membrane, forming an outer placental epithelium that also seems to be syncytial in nature (Blackburn & Flemming, 2012). The growth of placental epithelium beneath the uterine cell layer leads to the shedding of the latter and its replacement by a placental epithelial, possibly syncytial, layer. Near term, most of the uterine epithelium has been destroyed and the outer placental epithelium is in direct contact with the basement membrane and the endothelium of maternal blood vessels, giving rise to an endotheliochorial placenta (Blackburn & Flemming, 2012). The inner placental epithelium from which the knobs originate is in close but loose apposition with the outer layer (Blackburn & Flemming, 2012). The Lubuya invensii placenta does not seem to present distinct regions of histotrophic and respiratory exchanges. In most squamates (as is the case in Mabuya), there is a distinction between regions of nutrient exchange which present large secretory and absorptive cells (areolas, absorptive plaques, placentome) and regions of gas exchange where the cell layers are thinned to permit more efficient diffusion (respiratory segments), the absence of this distinction indicates that the Lubuya ivensii placenta absorbs both gases and nutrients directly from the maternal blood circulation (Blackburn, 1993a; Jerez & Ramírez-Pinilla, 2001; Blackburn & Flemming, 2012). In the end, while they both present invasive fetal cells, the Mabuya and Lubuya placentas are quite different, having taken different paths of ensuring ap- propriate provision of nutrients and gases to the fetus. This would be in accordance with the

36 eaaeoiiso iiaiyadmtorpypooe o hs w clades. two these for proposed matrotrophy and viviparity of origins separate ..Vvprt nnnmmainspecies non-mammalian in Viviparity 1.3. 37

1.3 2. Retroviruses

Retroviruses are the members of the Retroviridae family of viruses, officially created in 1975 (ICTV, 1975) and named after the process of retrotranscription, synthesis of a DNA molecule using an RNA matrix. This concept had long been believed impossible, going against a deeply held belief that the DNA to RNA to protein pathway was strictly one-way, but in 1970 Temin & Mizutani (1970) and Baltimore (1970) demonstrated the existence of enzymes able to synthesize complementary DNA (cDNA) from an RNA molecule. These enzymes have been called reverse transcriptases (RT) and are crucial to the retroviral life cycle, allowing them to convert their RNA genome into a DNA genome that can then insert itself into the genome of the host cell. The particularities of the retroviral replication cycle give rise to two great categories: exogenous (or infectious) retroviruses that exist as viral particles (virions) and infect cells, and endogenous retroviruses (ERV) that exist as integrated DNA genomes. Retroviral virions are spherical particles of about 80-100 nm diameter composed of a lipid envelope that contains a protein core (Fig. 9A). The envelope is of cellular origin but incorporates the viral envelope glycoproteins. The core is of variable shape depending on the retroviral family and contains the RNA genome as well as the proteins necessary to start the viral cycle following entry in a new host cell (reviewed in Coffin et al., 1997, chap. 1 and 2).

As for viruses in general, the origin of retroviruses is unknown. Investigation into endogenous retroviruses, which can act as a sort of fossil trace, show that there is a vast amount of retroviral lineages that infect or have infected hosts all across vertebrate phylogeny, with numerous host- switching events (Hayward et al., 2015). Retroviruses are a very ancient family of viruses, Aiewsakun & Katzourakis (2017) describe a group of retroviruses that are over 450 My old, by comparison the oldest non-retro viruses described to date are 310 My old insect viruses (Theze et al., 2011). An emergence about 450 Mya would put the apparition of retroviruses at about the same time as the apparition of jawed vertebrates during the Palaeozoic (dos Reis et al., 2015). Since the ICTV meeting in Paris in 2002, the Retroviridae family is split into two subfamilies, Orthoretrovirinae and Spumaretrovirinae (Mayo, 2002). The former subfamily contains most known retroviruses and unless noted otherwise will be the subject of the following section. For some details about Spumaretrovirinae see 2.1.1.b.

38 rncitiiito oec fteeeeet Topo ta. 06 e lo222a) The 2.2.2.a.). also see the 2016, illustrating al., nearby, et LTR (Thompson integrated elements types an these tissue of of diverse presence potency of the genes initiation by cellular transcript driven of examples be of can 2017). of number expression Magiorkinis, binding a whose & the been allow Hurst have they in there (rewieved and years factors recent 1985) transcription In al., vertebrate et cellular Graves of 1984; variety al., a et (Laimins described genome elements first MLV was promoter elements the and these in enhancer of presence contains The form, cells. DNA eukaryotic in the transcription in RNA genome driving whole the the to to due 5’ LTR, situated 5’ is the to U5 and LTR 3’ 2.1.3.b.). three the 13 all to Fig. present unique (see LTRs works is U5 both retrotranscription U3 form, and way form, genome nts) RNA DNA (10-100 the the R nts in In nts), while 1000 9B. (100-1000 regions, given Fig. to U3 a in parts: 300 to illustrated three specific from as into is nts), size split LTR (100-200 be the in all of ranging can sequence they and and but length genome, retrovirus, The DNA 2015). DNA Stoye, the & integrated Mager of the in in ends (reviewed CA both in ending at and present TG in form, starting sequences, identical are (LTR) repeats conserved repeats. present terminal Long genomes retroviral proteins, cycle. viral replication the for to code critical are that that domains ORFs non-coding the to addition In features Non-coding a. the between differences minor with DNA, stranded 9 forms. both double Fig. single of either (see or extremities as overlap exist RNA can the can sense genomes thus often retroviral positive and virus, all step, stranded other the retrotranscription each on the to to dependent relating Due is 2.1.3.d.). frames ORFs and different major in the as are of to ORFs arrangement different referred important relative is are the it Additionally they that 2). and note case 1 chap. to this 1997, al., in et Coffin retroviral proteins, in (reviewed retroviruses in accessory complex factor so-called variable encoding A ORFs three 2). additional the present and retroviruses 1 all antigen, while chap. (group-specific 12 encoded, ORFs 1997, major proteins to of al., 7 amount about et the from is Coffin organisation ranging genome in sizes reviewed genome 9B, with organisation retroviruses, (Fig. genomic most general kb in This reading sequence. open described major repeated been three long since presenting bp has kb 600 8 a about by (MLV, of flanked virus genome (ORF) leukemia a frames revealed murine It the 1981). of al., that et was Shinnick genome retroviral sequenced published first The orgarnisation Genomic 2.1.1 Retroviruses Exogenous 2.1 3 h emn htsmwa ofsnl suiu ote3 T fteRAfr,but form, RNA the of LTR 3’ the to unique is confusingly somewhat that segment the U3, rbbytems elkonnncdn oan,teln terminal long the domains, non-coding well-known most the Probably gag polymerase, ; pol envelope, ; ..EoeosRetroviruses Exogenous 2.1. env ,sm eeapresent genera some ), 39

2.1 2. Retroviruses

A immature mature

RNA Env PR MA

CA Gag GagProPol NC RT

B U3 R U5 gag pro pol env RT IN alpha ALV MA CA PR FS SU TM NC FS FS beta MMTV 2.1

gamma MLV R-T FS FS delta BLV

epsilon WDSV R-T

FS lenti HIV-1

spuma SFV

0 kb 5 kb 10 kb

Figure 9 – Organization of retroviral virions and genomes. (A) Micrographs and schematic repre- sentation of the immature and mature virion of HIV-1, showing the arrangement of canonical retroviral proteins and the RNA genomes in the particle. Scale bar 50 nm, adapted from Konvalinka et al. (2015). (B) Structural organization of retroviral genomes. For each family the organization of the LTRs and the canonical retroviral ORFs is represented to scale (see bottom), using the indicated type species. The order of subunits is indicated on the first genome and conserved in all others, faded boxes represent spacer or non-canonical subunits, accessory and regulatory genes are not displayed. Arrows with FS represents synthesis of Pro or Pol by frameshift, with R-T by read-through (see 2.1.3.d.). Adapted from viralzone.expasy.org

40 akn h paiino h auevro seFg ,rvee nMte ta. 2016). al., et Mattei in reviewed rearrangement cleaved, 9, structural is Fig. (see a polyprotein virion triggering Gag mature and the the of proteins 2.1.3) apparition independent and the as 12 marking NC Fig. and (see CA cycle MA, viral releasing the of step virion) fact maturation a the of are the by that components shown particles structural (VLP, the as particles envelope, of viral-like only of lipid formation composed a induce necessary to by sufficient the surrounded is all expression arrangement capsid contains its viral that specific Gag a 2016). virus of Gronenborn, a formation & for in Perilla domains 2016; proteins al., additional canonical et contain can Mattei three matrix It in the structural (reviewed (9). the between order contains that domains always in and and proteins peptides ORF (NC) nucleocapsid most and 5’ (CA) the capsid by (MA), coded is polyprotein Gag The Gag b. optimal for necessary is and genome the 1992). of al., middle the et the (Charneau In in replication 1991). located viral Clavel, is & PPT second (Charneau the PPT HIV human one of like case than retroviruses more Some 2015). present Hughes, (HIV) genome in DNA virus reviewed viral immunodeficiency 13, the Fig. of and synthesis 2.1.3.b. for (see RNAse necessary strand to primer resistance plus domain’s the this generates that RT shown viral later the was absolutely by it is cleavage and domain this RSV that of showed replication (1982) Hughes the & for Sorge critical genome. the of part 3’ the in U3, of tract. Polypurine 1999). al., et (Blond tRNAs associated (W) presents tryptophan example bind for (HERV-W) to W able retrovirus originally PBS endogenous a used Human was ERVs. this classify different amino-acids, tRNA, and bound different name the to to of corresponding nature the tRNAs determines Coffin need PBS in will (reviewed the ACC retroviruses of in sequence fixed finishes the always are Since which nucleotides 1997). end al., three 3’ et first its the using of binds only tRNA sequence retrovirus, the exact studied since The the (TGG) template. on the dependent stranded of highly double synthesis 1977), is Taylor, for a PBS primer needs see the review, RT a polymerases, early as DNA act an 1979). to all (for Dahlberg, like necessary retrotranscription & is since during site Peters strand this 1977; to minus al., tRNA DNA et a genomic (Shine of end U5 3’ of the end of the Binding after nucleotides few a situated site. binding Primer al., et Coffin in for (reviewed U5 essential and is contains R also genome, 6). between It chap. RNA border 1997, 13. the the Fig. forms of and which extremities 2.1.3.b. signal polyadenylation see both which a retrotranscription, at segment, during R present step The LTR switching R. the template and the of U3 part between only border the the definition is by is site initiation transcript h oyuietati hr tec f61 uie meitl upstream immediately purines 6-19 of stretch short a is tract polypurine The h rmrbnigst PS sagnrly1 t ogsequence long nts 18 generally a is (PBS) site binding primer The nvitro in Kioae l,19) During 1995). al., et (Klikova ..EoeosRetroviruses Exogenous 2.1. 41

2.1 2. Retroviruses

MA tends to assemble in a homotrimer of alpha helical folds that creates a flat surface presenting numerous basic residues, thereby allowing binding to lipid membranes (Hill et al., 1996). In some retroviruses such as HIV, the N-terminal glycine residue is myristoylated and the presense of the myristic acid is necessary for membrane association (Bryant & Ratner, 1990). CA is composed of two independent domains joined by a flexible hinge (Campos-Olivas et al., 2000), the N-terminal domain (NTD) and C-terminal domain (CTD). The NTD contains six to seven alpha helices and seems to play a moderate role in immature, but a critical role in mature capsid assembly (reviewed in Mattei et al., 2016). The CTD contains four alpha helices as well as the major homology region (MHR), a stretch of 20 amino acids that is highly conserved between retroviruses and even present in the distantly related Ty3 yeast retrotransposon (Orlinsky et al., 1996). The high degree of conservation of this sequence is particularly striking as Gag sequences are highly divergent, only the 3D structure seems to be conserved among retroviruses. Deletion and mutation experiments of the MHR region suggest that it is implicated in Gag oligomerization during formation of the immature capsid, though its exact function is poorly understood (Chang et al., 2007; Tanaka et al., 2016). Structure analysis of immature Gag conformation supports 2.1 the theory that the MHR is involved in interaction between different Gag molecules (Bharat et al., 2012, 2014; Schur et al., 2015). Finally NC contains positively charged regions and zinc- finger like motifs that allow it to bind to nucleic acids. This binding is necessary to induce Gag oligomerization as in absence of nucleic acids Gag assumes a folded conformation in which NC interacts with MA (Datta et al., 2007).

c. Pro and Pol

The pol ORF codes for several enzymes needed in retroviral replication. It always codes for the reverse transcriptase (RT) and the integrase (IN), and in most retroviruses for the protease (PR or Pro). The pro ORF is always found between gag and pol, and most of the time it is attached 5’ to the latter, but in some retroviruses this coding sequence is instead located in the gag ORF or forms its own independent ORF (see 9, reviewed in Coffin et al., 1997, chap. 2 and 7). The location of pol relative to gag is also variable, in most viruses pol is situated in an overlapping -1 frame when compared to gag (HIV and RSV for example, Jacks & Varmus, 1985; Jacks et al., 1988), while in others it is simply located downstream of gag in the same frame after a stop codon (murine leukemia virus, MLV, for example, Philipson et al., 1978). Expression of polyproteins is a strategy observed in many different virus groups, including retroviruses. Since most retroviral proteins are expressed as part of precursor polyproteins (Gag, Gag-Pro and Gag-Pro-Pol), PR plays an essential role in separating the different proteins (MA, CA, NC, PR, RT, IN) as was first described by von der Helm (1977) and Yoshinaka & Luftig (1977). PR is a small autocatalytic aspartic protease that is active at lower pH values only (Hansen et al., 1988; Billich et al., 1988). Its active conformation is a homodimer in which each monomer contains a catalytic Asp, Thr/Ser, Gly triad (reviewed in Konvalinka et al., 2015). The enzymatic activity of different PR from different retroviruses is variable and probably correlated

42 ulcto TD e i.1 n .... eiwdi ebt ta. 2016). site al., target et characteristic Lesbats the in reviewed then in 2.1.3.c., is resulting and that 14 and break Fig. mechanism strand see repair single (TSD, a DNA duplication in cell resulting host genome, the host strands by of two strand each repaired the one from of one to nucleotides ligates genome six only viral to step the transfer four strand of of The distance species. a 1989; viral al., at the et on occur (Brown depending ligations genome other, host two the The with 1990). genome viral hydroxyl al., the reactive et of these Craigie ends uses 3’ it and 5’ transfer the strand dinucleotides ligate during to CA then groups invariable group, generating trinucleotide -OH genome, reactive or dsDNA exposed di- non-inserted an a with viral cleaves the it of 3’-processing ends during the catalyzes integration, off DNA CCD for in The involved needed 2016). al., reactions is et two Lesbats and 1994; the identity al., et sequence Dyda 1992; low Craigie, with & (Engelman structure DXbinding conserved a a contains shows LTR which in which three (CCD) implicated CTD of domain is composed and core is motif catalytic IN HHCC the 2.1.3.c.). invariant an and recognition, contains 14 allowed which Fig. is NTD (see process the This integration domains, replication. as separate viral known allow and to protein cell IN host the the by of genome the into insert must it 2005). al., duplexes RNA/DNA et (Nowotny of sequence structure any 3D have particular to the seem recognizing not instead cleavage, does its H created for RNase it specificity Indeed, are when 1971). specifically al., that RNA et cleave duplexes (Mölling to DNA RNA-DNA able to associated is the is that endonuclease of an degradation is it the transcription, reverse for during responsible is RT of domain e iue(hmse l,20) diinlyvrlRsse ob eaieyerror-prone relatively be to about seem average on RTs being 1 rates viral mutation Additionally retroviral and 2007). activity proof-reading lacking al., DNApolymerases, et RT (Thomas HIV of minute Analysis per synthesis. DNA needs initiate RT might polymerases, to DNA polymerisation they duplex most primer-template though As a even 2015). recognize polymerases, Hughes, to in cellular (reviewed most ancestor common than a range share template DNA-synthesis, wider DNA-directed of a and exhibiting DNA-synthesis RNA-directed both of polymerase DNA capable retroviral is The domain. domain H RNase the and polymerase DNA the replication: viral asymetrical the recognize to able 2002). be al., et to (Prabu-Jeyabalan the seems substrates of PR natural structure as its of the role Additionally, structures important and 2015). an +2 al., play in et to Glu) Konvalinka seems (especially in substrate acids (reviewed amino hydrophobic observed large been for has as well -2 and as as +1 site, acts in cleavage that Leu) the sequence (especially defined of residues strictly aliphatic -1 large a or be aromatic to for seem preferences only not site, does cleaving Substrate there 1993). as complex al., is et PR (Sedlacek of particles specificity viral abnormal morphologically of formation to lead level expression its with . 5 netertoia eoehsbe ees-rncie noadul-taddDAform, DNA double-stranded a into reverse-transcribed been has genome retroviral the Once T h cncezm htgvstefml t ae otistomjrdmisnee for needed domains major two contains name, its family the gives that enzyme iconic the RT, × 10 − 5 mutations/bp/cycle xvivo ex niae hti sarte lwezm,adn bu 0nucleotides 70 about adding enzyme, slow rather a is it that indicates nvivo in nvivo in uigifcin leaino nyai ciiylvlcan level activity enzymatic of Alteration infection. during rvee nSaosaae l,20) h Ns H RNase The 2003). al., et Svarovskaia in (reviewed ..EoeosRetroviruses Exogenous 2.1. 35 aayi ie n the and site, catalytic D 43

2.1 2. Retroviruses

d. Env

Retroviral envelope proteins are transmembrane glycoproteins that are embedded in the lipid envelope of the virion. Since they are of special interest in the present study, a detailed description will be provided in 3.1, additionally the following does not apply to the Env of epsilonretroviruses, which seem to display significant differences to all other orthoretroviral Env (see Rovnak et al., 2007, as well as 2.1.2.a.). The env ORF is the most 3’ of the three major retroviral ORFs and differs from gag and pol as it is not expressed as part of the retroviral polyproteins generated from a genomic RNA. Env is expressed from a subgenomic transcript that is initiated in the 5’ LTR and spliced between a donor site shortly after the LTR and an acceptor site shortly before env. The transcript terminates in the 3’ LTR at the polyadenylation site, as is usual for retroviral transcripts. Env is composed of two domains that are cleaved by cellular furin-like proteases but remain associated. The surface subunit (SU) is bound to the transmembrane subunit (TM) which contains a helix domain anchoring Env in the viral membrane. They are type 1 viral fusion proteins with the SU being implicated in receptor recognition and binding, and the TM

2.1 containing a hydrophobic fusion peptide that, alongside a number of conformational changes, allows it to induce fusion of the viral and cellular membranes (see 3.1.2, reviewed in Coffin et al., 1997; Henzy & Johnson, 2013; Podbilewicz, 2014; Harrison, 2015). The association between the SU and TM subunits can be either covalent or non-covalent, which splits Env into gamma-type and beta-type respectively (reviewed in Henzy & Johnson, 2013).

e. Regulatory and accessory proteins

Regulatory and accessory (or auxiliary) proteins are found in complex retroviruses and are ex- pressed in addition to the main proteins described above. They are present in lentiviruses such as HIV and deltaviruses such as human T-lymphotropic virus (HTLV, which is now officially called primate T-lymphotropic virus, PTLV), as well as some in some Spumaretrovirinae (reviewed in Coffin et al., 1997; Faust et al., 2017). HIV-1 is considered the model complex retrovirus and is by far the most studied. Its genome encodes two regulatory proteins: trans-activator (Tat) and regulation of expression of viral proteins (Rev), and four accessory proteins: virion infectivity factor (Vif), viral protein r (Vpr), viral protein u (Vpu) and negative factor (Nef). As all reg- ulatory and accessory proteins, they are small ORFs expressed by alternative splicing and/or initiation (Fig. 9 and 10). Many functions have been linked to these proteins, but we will limit the description to their probable primary functions (for a more complete description see Faust et al., 2017).

Regulatory proteins of HIV-1. Transcription of HIV-1 RNA is performed by the host cell machinery, but processivity is low, leading to only short incomplete viral transcripts in the absence of Tat. Tat binds a transient RNA structure at the 5’ end of viral transcripts, called the transactivation response element, and by doing so increases the processivity of the cellular

44 piigeet n rncit r otdit npie a,prilyslcd(-)o ul pie (e-g) spliced fully or (b-d) spliced partially (a), non-canonical viralzone.expasy.org represent unspliced from lines and into Adapted Dotted sorted (a-b) latter. categories. the are the Canonical with transcripts with (bottom) scale and 9, to blue. events genome, Fig. HIV-1 splicing light as the by in code expressed color ORFs transcripts genome. same retroviral regulatory (c-g) the HIV-1 and using complex accessory genome, HIV-1 the of the addition of in ORFs transcripts non-canonical and and Organization canonical – 10 Figure fully spliced partially spliced unspliced g f e d a c b MA MA MA CA CA NC CA CP TIN RT PR NC NC RR IN RT PR vif vif vpr vpr a p e nef rev vpu tat tat vpu SU SU ..EoeosRetroviruses Exogenous 2.1. tp ragmn of Arrangement (top) TM TM rev

nef

transcripts transcripts

non-canonical 45 canonical

2.1 2. Retroviruses

transcription machinery to allow synthesis of full length viral RNAs (Berkhout et al., 1989; Feinberg et al., 1991). Expression of the full complement of viral genes necessitates export of differentially spliced viral RNA from the nucleus. In HIV-1 for example, Gag and Pol are expressed from an unspliced RNA, Env, Vif, Vpr and Vpu from a spliced trancript missing gag and pol and Rev, Tat and Nef from multiply spliced transcripts missing gag, pol and most of env (Fig. 10, reviewed in Pollard & Malim, 1998). Since the cellular machinery only exports fully spliced transcripts, only Rev, Tat and Nef should be expressed. Rev contains a nuclear localization sequence triggering its re-import into the nucleus, where it is able to bind viral RNAs by interacting with the Rev response element contained in env. After binding of viral RNA and oligomerisation, Rev can mediate export of the unspliced or partially spliced viral RNA, as it also contains a nuclear export sequence (Malim et al., 1989).

Accessory proteins of HIV-1. As their name indicates, accessory proteins are not always needed for viral replication, whereas expression of the regulatory proteins is necessary even in highly permissive cells (Balliet et al., 1994). The primary function of most accessory proteins 2.1 is to counter specific aspects of host resistance to viral infection, thus if the cell is already permissive, no action of the accessory proteins is needed. Vif is found not only in HIV-1 but in all lentiviruses except one (equine infectious anemia virus, EIAV) and is necessary to counteract the host cell APOBEC3 restriction factors (Sheehy et al., 2002). These cellular components are cytidine deaminases participating in the innate immune defense against retroviral infections by inducing genome wide C to U mutations (Sheehy et al., 2002; Harris et al., 2003). They are counteracted by Vif which can trigger ubiquitination and proteasome degradation of APOBEC3 (Sheehy et al., 2003). Vpu counteracts the host restriction factor Tetherin, which binds egressing viral particles and prevents their release and dissemination in the host (Neil et al., 2008). It has been shown that Vpu can both bind to Tetherin and prevent its recycling to the membrane once internalized, as well as trigger its ubiquitination and lysosomal degradation (Douglas et al., 2009; Iwabu et al., 2009). Vpr is a conserved protein among all human and primate lentiviruses and of the many reported functions, Faust et al. (2017) consider its conserved ability to stop the cell cycle in G2 phase (Planelles et al., 1996) the primary function. This cell cycle arrest seems to be due to the interaction with and sequestration of the CRL4DCAF1 ubiquitin ligase complex (Hrecka et al., 2007; Belzile et al., 2007) and seems to allow increased LTR driven expression of viral RNA (Goh et al., 1998). Similarly to Vpr, the function of Nef during HIV-1 infection is not well-known at present, but it has repeatedly been implicated in regulating the amount of several cellular proteins (Tetherin, CD4, MHC-I) at the cell membrane (reviewed in Faust et al., 2017). In absence of Nef the incidence of acquired immunodeficiency syndrome (AIDS) after HIV-1 infection seems delayed or absent (Kestier et al., 1991; Kirchhoff et al., 1995).

Regulatory and Accessory proteins in other orthoretroviruses. Interestingly, the presence or absence as well as nature of retroviral regulatory and accessory proteins is very variable even

46 eal oidc rlfrto frsigTcls h agto TV1ifcin(eiwdin (reviewed infection HTLV-1 of target to the seem T-cells, 2011). HTLV-1 al., resting of et of Edwards proteins proliferation accessory induce the to of able all be Additionally, example factors. for can transcription p30 cell cell, host host T-cell the on and Tax from of MHC-I of effects protection export of the prevent provide dampen expression to to downregulating seem p30 by seem and example p13 p8 activation. for form system, protease-cleaved immune its innate Semmes, and the & p12 Giam 2007; in). Jeang, viral reviewed & of (Matsuoka establishment 2016, infections HTLV-1 the transcription. of in viral persistence implicated the inhibiting probably and Rex, is latency and it al., Tax counterproductive, et seem of (Gaudray may activities LTR this 3’ the While the of of many action counteracts under expressed HBZ and 2002). genome viral the of strand antisense The (reviewed 2005). element response Green, spliced Rex & partially specific Younis a or of in unspliced recognition of after export nucleus allowing the Rev, from to transcripts cellular role viral similar trigger a to plays indicates, sufficient name is its alone expression host on its Tax that of transformation wide-reaching effects the so gene fact are viral In physiology progression. of cell cell-cycle trans-activator NF- host strong promotes the and hyper-activates a instability in also is genome Rev Tax but and similar. Tat), Tat are to to (similarly functions similarly p30. expression their events, and splicing and p13 double 10) p12, alternative (Fig. genes: by HIV-1 accessory expressed three are and Rex (HBZ), and more Tax factor a zipper (for homolog basic discussed HTLV-1 Rev be (Tax), and will regulator (Rex) transcriptional basics HTLV-1 Trans-activating of the genes: genome only The regulatory 2016). and three Semmes, work, & contains Giam this 2011; proteins of al., et these scope Edwards see the to description in ascribed complete be functions to of large spectrum too full is the again, Here HTLV-1. by represented host to adaptation particular a denoting genes, these physiology. of and is defenses origin accessory genes recent of more retroviral variability a the major ancestor, indicates common of proteins old function very a and from organization deriving obviously the the similar, from rather While differs degradation 2014). of also instead Sauter, escape sequestration in Tetherin intracellular (reviewed to of limited mechanism being & Tortorec Vpu, the HIV-1 (Le HIV-2, by employed virus and one this SIV one in Env, both role by In additional carried an 2009). is displays Neil, action that Similarly anti-tetherin proteins the 2009). retroviral Nef, al., canonical express et the does (Jia of it here while counteraction and Tetherin Vpu of no Vpu has actor the HIV-2 major present not the does is (SIV) Nef virus instead immunodeficiency gene, Simian mean identical. necessarily HIV-1 is not are does function most it their genus, but the that lentiviruses, across of present lentiviruses, are subset all proteins primate accessory almost the when to Even in specific. common conserved be are to Vif, seem like Vpr, some, like above, others, stated As level. intragenus the at h te ao rhrtoia eu otiigadtoa rtisaetedeltaretroviruses, the are proteins additional containing genus orthoretroviral major other The xvivo ex tax/rex rvee nMtuk en,20;Ga ems 06.e,as 2016).Rex, Semmes, & Giam 2007; Jeang, & Matsuoka in (reviewed RAfo h ulu hl 1 rvnsitrcino a with Tax of interaction prevents p13 while nucleus the from mRNA HBZ eeo TV1i eypriua,biglctdo the on located being particular, very is HTLV-1 of gene κ inln aha,induces pathway, signaling B ..EoeosRetroviruses Exogenous 2.1. 47

2.1 2. Retroviruses

2.1.2 Retroviral phylogeny

The study of retroviral phylogeny is primarily based on the comparison of conserved domains of the pol gene, as RT contains a number of highly conserved motifs (McClure et al., 1988; Poch et al., 1989). Alignments of these RT domains sort retroviruses into two subfamilies and several genera (Fig. 11A). In addition to RT based classification, the TM subunit of Env contains conserved motifs and can also be used for phylogeny purposes (Fig. 11B). The trees established using RT or TM do not overlap, as recombination events have led to associa- tions of Pol and Env from different retroviral genera (see MMTV and SMRV in Fig. 11 and Bénit et al., 2001, for examples). A potent tool for investigating the evolution of the retroviral lineage are endogenous retroviruses, as they can be likened to fossil remains of ancient retro- viruses which can be compared to the genomes of extant retroviruses, most of which are very young (Jern et al., 2005; Feschotte & Gilbert, 2012). The following section will contain a brief overview of the different retroviral clades as they are currently defined by the ICTV (accessible at https://talk.ictvonline.org/taxonomy/). For each genus the type species is named alongside the

2.1 size of its genome, whether it is a simple or complex retrovirus and the type of Env it presents.

a. Subfamily Orthoretrovirinae

Alpharetrovirus Type species: avian leukosis virus (ALV): 7.4 kb, simple, avian gamma-type Env The genus Alpharetrovirus contains 9 species of oncogenic avian retroviruses, among which ALV and RSV. In the canonical alpharetroviral genome, pro is part of the gag ORF, situated at the 3’ end but in the same reading frame (Fig. 9B). This leads to a unique case among retroviruses in which Pro is expressed at a much higher level, similar to that of Gag (reviewed in Konvalinka et al., 2015). Compensating this higher level of expression is a mutation in the active site of RSV Pro which reduces enzymatic activity. If this mutation is reversed, the higher activity combined with the higher expression provoke aberrant virion morphology (Konvalinka et al., 1992; Sedlacek et al., 1993). The Env protein of alpharetroviruses is a variant of the more common gamma-type. In avian gamma-type Env, rather than being located directly at the border of SU and TM, the fusion peptide is situated about nine amino acids after the cleavage site and is flanked by cysteines that form a disulfide bond (Gallaher, 1996).

Betaretrovirus Typespecies: mouse mammary tumor virus (MMTV): 8.8 kb, simple, beta-type Env There are 5 decribed species of betaretroviruses, among which MMTV, Mason-Pfizer mon- key virus (MPMV) and the much studied Jaagskiete sheep retrovirus (JSRV). In betaretroviruses, pro is contained in its own ORF between gag and pol but separated from both by a frameshift (Fig. 9B). While most of the viruses in this genus present a beta-type Env, some like MPMV seem to have acquired a gamma-type Env by recombination events during their evolutionary

48 nmavrs L:ainluoi iu,RV ossroavrs M:ainmeolsoi virus, myeloblastosis retrovirus monkey monkey avian Mason-Pfizer squirrel AMV: MPMV: SMRV: retrovirus, infectious virus, virus, sheep Jaagsiekte equine JSRV: sarcoma EIAV: virus, Rous tumor RSV: virus, mammary mouse virus, MMTV: encephalitis leukosis avian arthritis ALV: WDSV: HIV: virus, caprine virus, CAEV: anemia virus, T-lymphotropic virus, primate hyperplasia PTLV: virus, immunodeficiency epidermal leukemia walleye human bovine BLV: WEHV: virus, virus, sarcoma dermal leukemia walleye feline FLV: murine retrovirus, MLV: koala KoRV: virus, virus, leukemia leukemia virus, ape EFV: reticuloendotheliosis gibbon GaLV: REV: virus, retrovirus, virus, foamy feline foamy below bovine RD-114 simian BFV: scale SFV: RD114: red. virus, (see foamy in site feline FFV: highlighted per virus, substitutions are foamy retroviruses of equine number recombinant average RT/TM the and to tree), proportional each is length program PhyML branch the trees, with obtained both was tree The WAG+ 22). CX Fig. associated the also the (see and alongside ISD gray CKS-17 the in of indicated absence is or RT-based type presence with TM species tree. retroviral the selected was on using tree tree indicated TM-based genera The Unrooted (B) RtREV+ tree. 2017). the the al., et and on (Lefort analysis. program program indicated TM genera PhyML RT-based or the with RT with species, on obtained retroviral based selected genera using retroviral tree of PhyML Phylogeny – 11 Figure B A spuma α

AMV Γ

FFV

IFmdl eetdb h mr oe eeto rga Lfr ta. 07.For 2017). al., et (Lefort program Selection Model Smart the by selected model, +I+F EFV

RSV γ β ALV BF SFV V γ

MPMV RD1

SMRV R EV KoRV

G

GaLV FLV

aL KoRV

1 RD114 BLV 4

MLV V MLV

REV

FLV PTLV-1 δ WEHV-2 CX γ− +ISD

PTLV-2 ε type

6 WDS

0.9 CC V RT WEHV-1 TM MMTV 0.5 β

Γ JSRV Imdlslce yteSatMdlSelection Model Smart the by selected model +I

BLV

F FFV SFV

F EFV PTLV-1

B P FV HIV-2 T HIV LV

CAEV -2 δ -1 spuma

..EoeosRetroviruses Exogenous 2.1. SMRV

MPMV

JSRV A

WEH M

WEHV-1 MMTV ALV

RSV V

A notdRT-based Unrooted (A) V-2

β HIV-2

α HIV-1 β−

CX EIAV DVWDSV -ISD CAEV type type n

EIAV C n C 1-2 ε λ λ oi n the and motif 49

2.1 2. Retroviruses

history. These viruses cluster with other betaretroviruses on an RT tree, but very clearly with gammaretroviruses on a TM tree (Sonigo et al., 1986). Betaretroviral genomes seem to display a bias towards A and T nucleotides (Berkhout et al., 1989; Jern et al., 2005).

Gammaretrovirus Type species: murine leukemia virus (MLV): 8.3 kb, simple, gamma-type Env Gammaretroviruses are the largest genus with 18 described species, among which feline leukemia virus (FLV), gibbon ape leukemia virus (GaLV) and Koala retrovirus (KoRV). In this prolific genus, pro is attached 5’ to the pol ORF and in the same frame as gag but separated by a stop codon (Fig. 9B). By definition, gammaretroviruses present gamma-type envelope genes which seem to confer a broader range of potential hosts than beta-type Env, as gammaretroviruses and their envelopes are found not only in mammals but also in birds, reptiles, amphibians and fish (Henzy & Johnson, 2013, reviewed in). It is also noteworthy in the context of this work that most syncytins are of probable gammaretroviral origin (see 3.3.2). 2.1 Deltaretrovirus Type species: bovine leukemia virus (BLV): 8.4kb, complex, gamma-type Env In addition to BLV, this genus contains the primate retroviruses HTLV-1 to 3. All deltaretro- viruses have gag and pol genes that are more similar to those of betaroviruses and present pro as its own ORF (Fig. 9B), however their Env is clearly gamma-type, indicating that this genus arose from a recombination event (Fig. 11, reviewed in Henzy & Johnson, 2013). It has long been thought that no endogenous deltaretroviral elements existed, but Farkašová et al. (2017) describe a provirus in long-fingered bats that seems to belong to this family. It is important to note that this provirus is highly incomplete, having lost NC and the entirety of pro, pol and env, and that the classification into deltaretroviruses is mostly based on the high similarity of Gag to HTLV-1 Gag, the presence of two putative regulatory protein ORFs, and the nucleotide bias that is characteristic of known deltaretroviral sequences (an over-representation of C, Kypr et al., 1989; Berkhout et al., 2002). While it is frustrating to not have access to the complete genome of this deltaretroviral fossil, it suggest that features such as regulatory genes where already present about 20 Mya in this clade (Farkašová et al., 2017). In any case, deltaretroviruses are conspicuously under-represented in endogenous elements (Hayward et al., 2015).

Epsilonretrovirus Type species: walleye dermal sarcoma virus (WDSV): 12.7 kb, complex, epsilon-type Env This smallest and probably least well know retroviral genus only contains three species: WDSV and walleye epidermal hyperplasia virus 1 and 2. The relative position of pro in relation to gag and env is similar to that in the more well known gammaretroviruses: gag is in the same reading frame as pro-pol but the two ORFs are separated by a stop codon. The Env of epsilonretroviruses however, displays major differences with both beta- and gamma-type Env. With about 1200 amino acids, the WDSV Env is approximately twice as large as classical

50 SV:1. b ope.Acrigt hne l 21) F hudb eae oeastern to of renamed species be type should the SFV become (2018), and al. virus foamy et simian Khan chimpanzee to According complex. kb, 13.3 (SFV): present genera. the new the to between relevance differences limited the of into delve are not spumaretroviruses will since section and this study, yet ratified been not has it co-speciate uispumavirus spumaretroviruses as hosts, phylogeny to hosts: (according the their genera to with five changes into major it introduces splitting (2018) family, al. this et of Khan by proposal accepted recently A other from retro- other The all apart from 2002). separate (Mayo, them subfamily a viruses setting into sorted characteristics officially were they display 2002 in to and retroviruses, know been long have Spumaviruses Subfamily b. genera, retroviral other all viruses. to occurence oncogenic contrary rare 1989; and al., a 2002) et (Kypr be al., nucleotides to A et towards seem Berkhout bias a and show genomes 2007) Lentiviral al., 2015). al., et et (Hayward (Katzourakis recently lentiviruses endogenous only deltaretroviruses, described with been As have 2015). al., et Woodward 16A, (Fig. proteins is HIV-1 society, gag human lentiviruses, on In structures. impact accessory 3D and and important regulatory proteins, to regards its with especially to retrovirus, studied Due actively most AIDS. the probably of agent causative the HIV, Env type Lentivirus 2010). Quackenbush, virus, of & hyperplasia members Rovnak perch in probable virus, reviewed considered sarcoma are swimbladder amphibians salmon retroviruses retrovirus, and fish (snakehead reptiles more species Three mammals, this infect 2014). al., to et used (Brown viruses well of as family to this similarities extant that bearing suggest known elements genus endogenous three this of only analysis the the While viruses, 2010). fish are Quackenbush, epsilonretroviruses & Rovnak in regulation (reviewed and proliferation apoptosis cell expression, of gene viral reading in implicated additional proteins three for code displaying that retroviruses, frames complex the are with viruses coherent these Additionally is 2007). This membranes intracellular domains. with associated transmembrane virions two assembled (as of the observation signal domains between addressing loop transmembrane membrane intra-cellular internal two an the presents least in and genera) at retroviral other contain for one to to opposed seems Env WDSV Env. orthoretroviral h yeseisfrteso ob defunct be to soon the for species type The retrovirus: human well-known most doubt a without the is species lentiviral 10 the Among Fg B.Adtoal h auevro ipasacaatrsi oia ragmn fCA of arrangement conical characteristic a displays virion mature the Additionally 9B). (Fig. yeseis ua muoecec iu HV1:92k,cmlx beta- complex, kb, 9.2 (HIV-1): 1 virus immunodeficiency human species: Type n paigtenmn ovnin.WieteIT cetdti proposal, this accepted ICTV the While conventions. naming the updating and ) Spumaretrovirinae iisuaiu,Poiisuaiu,Bvsuaiu,Flsuaiu,Eq- Felispumavirus, Bovispumavirus, Prosimiispumavirus, Simiispumavirus, Spumaretrovirinae pro-pol ufml otisasnl genus, single a contains subfamily Spumavirus si hfe 1raigfaecmae to compared frame reading -1 shifted a in is eu stesma om virus foamy simian the is genus Simiispumavirus Lentivirus ..EoeosRetroviruses Exogenous 2.1. xvivo ex ontcnanany contain not do Rva tal., et (Rovnak . Spumavirus 51 .

2.1 2. Retroviruses

The differences between spumaretroviruses and orthoretroviruses start with the genomic organisation, the LTRs are much larger (about 1700 nts), there is an additional transcription initiation site (an internal promoter in the 3’ region of env) and pro-pol is expressed as a spliced transcript (reviewed in Rethwilm, 2010). Transcription starts at the internal promoter, which is basally active, and results in the production of the anti-APOBEC accessory protein "bel1 and bel two" (Bet, Löchelt et al., 2005) and the regulatory protein trans-activator of spumaretroviruses (Tas). The latter then both increases transcription from the internal promoter and initiates 5’ LTR driven transcription to allow expression of the other retroviral genes (Keller et al., 1991; Löchelt et al., 1993). Even the canonical retroviral genes gag, pol and env present characteristics that set them apart from those of orthoretroviruses. Spumaretroviral Gag does not present the highly conserved MHR (Orlinsky et al., 1996) and is not cleaved into three proteins, instead remaining mostly intact and conferring an immature appearance to virions, even though they are infectious (reviewed in Rethwilm, 2010). Additionally Gag is unable to form VLPs on its own, the presence of Env is necessary to obtain particles (Fischer et al., 1998). The pro ORF is part of the pol ORF in spumaretroviruses, but there is only a single cleavage event, resulting 2.1 in a PR-RT and an IN molecule. The latter also acts differently than in orthoretroviruses during 3’ processing, only cleaving the 5’ part of the sequence so that once inserted the 5’ LTR does start with TG, but the 3’ LTR does not necessarily end in CA (Enssle et al., 1999). Finally, Env is expressed in an inside-out fashion when compared to orthoretroviruses, with the N-terminal end facing the cytoplasm rather than the lumen of the endoplasmic reticulum (ER). Additionally the exceptionally long signal peptide that adresses the protein to the ER during synthesis is not degraded but instead forms an integral part of the functional, tri-partite spumaretroviral Env (Lindemann et al., 2001).

Beyond the structural differences, one of the major factors distinguishing Spumaretrovirinae and Orthoretrovirinae is the timing of reverse transcription. In the latter, this step occurs just after viral entry into a new cell, the infectious genome is RNA (see Fig. 12 and 2.1.3), while in spumaretroviruses the reverse transcription step occurs between assembly and entry into new cells, meaning that the infectious form of the genome is DNA (Yu et al., 1999). This timing of reverse transcription is remarkably similar to that observed in Hepadnaviridae, one of the other rare families of reverse transcribing viruses. This resemblance to a non-retrovirus is even more intriguing when taking into account the proposed basal location of spumaretroviruses in the retroviral lineage (Jern et al., 2005; Hayward et al., 2015), a hypothesis that was recently strengthened by the discovery of 450 My old spumaretrovirus-like endogenous elements, by far the most ancient known retroviral sequences (Aiewsakun & Katzourakis, 2017). Considering Spumaretrovirinae as the most ancestral retroviruses would suggest that reduction of genome size, use of a single promoter and non-complex genomes are all derived characteristics in retroviral evolution.

52 Nklici ta. 03,ee huhol igeoei eddfrrvretranscription. reverse for needed is one genome single RNA a well the only less of though is copies even 2 transcription 2013), reverse contains RT al., always in proteins, et virion role NC (Nikolaitchik The potential to 2015). whose bound Hughes, IN, are in shell, (reviewed like which known CA proteins genome the RNA viral viral Inside other the 2013). the of al., and viral copies et the two (Rasaiyaah shield contains system to RTC factors immune cell the innate host infection. with the viral interact from can of proteins genome signal CA DNA strong that show a been viral sensors, has cell of it HIV-1 host recognition In by prevent cytoplasm (reviewed might the volume capsid in small the DNA a of double-stranded in preservation confined Additionally, substrate 2015). proteic acid all Hughes, keeping nucleic a as in the of such as virus, presence well the as the for enzymes advantages however theoretical necessary disputed, obvious presents remain RTC the replication around dissociation viral shell of for amount The importance 2001). 1999, its Goff, & and (Fassati during uncoating RTC called the process from disassemble, a dissociate in capsid immediately which step, the this inside not of (RTC) parts does complex though 12), transcription it (Fig. reverse occurs cell, so-called transcription the reverse host form the contents its of and cytoplasm it instead the enters capsid the Once transcription Reverse b. released is enzymes and genome not viral is the entry cytoplasm. contains the viral viral which into the if capsid of infection fusion the Following initiate 2001). membranes, 1999, to cellular al., et and unable Pizzato 1998; are al., they et (Maréchal but Env correct by cell, the mediated presenting the not enter to cells even to specificity Podbilewicz, bind and and can in receptor, Viruses tropism (reviewed infection. for its membrane essential confers absolutely cell that is virus interaction plasmic the SU/receptor the implies The which at 2015). environment, than Harrison, the rather 2014; of endosome acidification an necessitate in retroviral further fusion the on can Depending fusion. process membrane this close and layers membranes species, lipid two the the of brings merging and induce conformation to peptides, changes enough trimer fusion TM After hydrophobic the 2015). step, Harrison, its 2014; attachment Podbilewicz, lipid extends the in (reviewed the which membrane target of the trimer surface into them TM the inserting metastable at SU/receptor present This the attachment cells. trimers target an frees the protein comes on interaction present Env First receptors the specific 3.1.2. bind of in and recognize subunits detail envelope viral SU more the much which in which steps, in and two step, into here decomposed briefly be can a considered and Env, of viral be protein retroviruses, penetration fusion will For the I 2008). type that al., the et shown by (Keele mediated been infection is has acute entry trigger it to HIV-1 enough for be and can 12, virion single (Fig. particle viral cell the of host entry adequate the is the infection retroviral into of step essential first the expected, be might As entry Viral a. cycle Viral 2.1.3 ..EoeosRetroviruses Exogenous 2.1. 53

2.1 2. Retroviruses

maturation

budding and release

receptor binding assembly

2.1 fusion and entry RTC RNA and protein reverse synthesis transcription

integration uncoating and nuclear import PIC

provirus Figure 12 – Representation of the retroviral replication cycle. The infecting virion attaches to the host cell membrane through Env-receptor interaction which also triggers fusion and viral entry. In the cytoplasm, the RNA genome is reverse transcribed into DNA within the reverse transcription complex (RTC). The DNA genome is uncoated (dissociation of the capsid) forming the pre-integration complex (PIC), which is imported into the nucleus. In the nucleus the viral genome is integrated into the cellular genome and viral transcripts are expressed by the host machinery. Viral proteins are synthesized in the cytoplasm or the endoplasmic reticulum/Golgi (Env only) and are addressed towards the plasmic membrane. At the membrane the Gag containing polyproteins assemble and bind two copies of viral genomic RNA. The assembly buds and takes with it an Env containing fragment of the cellular membrane, releasing an immature virion. Finally, PR provokes virion maturation which yields a new infectious virion. Colors are the same used in Fig. 9.

54 hog t Ns ciiy h ieetsesaervee nHge 21)adwl be will and (2015) 13. Hughes Fig. in in reviewed presented as are well steps as here different up The summed activity. degradation H hybrid RNase RNA/DNA and its templates, through DNA and RNA using polymerization DNA both impact negatively would 2015). that Hughes, structures in allow secondary (reviewed of properties transcription formation reverse binding preventing acid by RT nucleic the of its assist presence as to The probably it 1990). transcription, Temin, reverse during & for template (Hu necessary switches differ is RT copies the NC the packaged in otherwise two if genomes the occur or RNA and can nicked two transcription that of reverse of events case recombination presence potential in the the back-up of are a consequences RTC as important serves the probably but copy RNA, damaged second a of presence The Te5-3RU-B-’pu-tadDAfamn a o idt h B hc forms which PBS the to bind now can fragment DNA plus-strand 5’-U3-R-U5-PBS-3’ The 6. synthesized the genome, the of ends both at found definition by is sequence R the Since 3. minus-strand the of synthesis start to primer as duplex tRNA/RNA resulting the uses RT 2. the binds sequence 3’ complementary a with tRNA encapsidated, previously cellular, A 1. TePTsqec o om nioae N/N ulx hc sue sprimers as used is which duplexe RNA/DNA isolated an forms now sequence PPT The 5. paired completely now is which genome RNA remaining the degrades activity H RNase 4. ees rncito tefi ahrcmlctdpoesi hc h Tms perform must RT the which in process complicated rather a is itself transcription Reverse h ’edo h iu-tadDA hsfrsaDADAdpe hc evsas to serves leading which genome. strand, DNA minus-DNA duplex and the DNA/DNA plus- of the a formation both forms of This completion and DNA. elongation for minus-strand primer the of end 5’ the 5’ DNA. the associated of the end frees 5’ This PBS. the the sequence, now RNA is copied and genome hybrid the RNA DNA/RNA of a creates degradation This mediated genome. H viral RNase the provokes of end 5’ the reached has RT short point a of generation the in results This genome. DNA primer. as serving genome, RNA plus-strand viral the on PBS a ece h ’edo h eoe Ns a erddteRAPTsequence, PPT RNA the degraded synthesis has plus-strand H time RNase genome, the DNA fully the by a of Similarly, generating end H. 3’ RNase the by plus- reached the degraded has of thus sequence PBS is DNA the and with strand associated now is minus-strand the of sequence RNA/DNA3’ plus-strand a generates This synthesis. plus-strand for PPT. resistant H RNase relatively the spares it process the In DNA. minus-strand the to PBS. the genome, RNA the of extends end 5’ and current synthesis the minus-strand to DNA up resumes strand RT this The elongated. be can that hybrid -U5-R- rget twihpitte3 n ftemnssrn N srahd h tRNA The reached. is DNA minus-strand the of end 3’ the point which at fragment, 3’ N rgettasest h ’Rsqec,gvn iet e RNA/DNA new a to rise giving sequence, R 3’ the to transfers fragment DNA 5’ -U3-R-U5-PBS- 3’ lssrn fragment. plus-strand 5’ -U5-R- ..EoeosRetroviruses Exogenous 2.1. 3’ N rget twhich at fragment, DNA 5’ -PPT-U3-R-U5-PBS- 55

2.1 2. Retroviruses

R U5 PBS PPT U3 R 1. gag pol env

R U5 PBS PPT U3 R 2. R U5

PBS PPT U3 R 3. R U5

PBS PPT U3 R 4. PBS PPT U3 R U5

2.1 PPT U3 R U5 PBS 5. PBS PPT U3 R U5

U3 R U5 PBS PPT U3 R U5 6. PBS PPT U3 R U5

U3 R U5 PBS PPT U3 R U5

U3 R U5 PBS PPT U3 R U5

Figure 13 – Schematic representation of the reverse transcription steps. RNA is shown in green with the polypurine tract (PPT) in dark green, plus-strand DNA is shown in purple and minus-strand in red. A dotted line represent RNase H mediated RNA degradation and the position of the indicated genomic elements is not to scale. For simplicity gag, pol and env are only displayed for the first step, though they are present throughout. (1.) A complimentary tRNA (green arrow with loops) anneals to the primer binding site (PBS). (2.) RT initiates DNA minus-strand synthesis and copies the viral RNA genome to its R extremity. The resulting RNA/DNA duplex is degraded by RT RNase H activity. (3.) The minus-strand DNA dissociates and re-associated to the 3’ R segment of the viral RNA. (4.) RT elongates the minus-strand DNA until reaching the 5’ extremity of the RNA genome. This forms another DNA/RNA duplex that is degraded by RNase H activity, except for the more resistant PPT. (4.) The PPT forms an RNA/DNA duplex able to serve as primer for plus-strand synthesis, RT elongates this strand until reaching the end of the PBS-complementary sequence of the tRNA sill attached to the minus-strand. This creates an RNA/DNA duplex and the tRNA is degraded. During synthesis, the PPT is finally degraded (5.) The nascent plus-dissociates and re-associates to the minus-strand PBS using its PBS-complementary 5’ end. This allows the RT to finish elongating both DNA strands and obtain a double stranded U3RU5-gag-pol-env-U3RU5 DNA genome. Adapted from Hughes (2015).

56 utmr ihtevrlDAgnm n sesnilfrI ciiyadrtoia genome retroviral The and 2017). Cherepanov, activity several & IN in Engelman for IN 2017; characterized Engelman, essential of & well is interaction Grawenhoff been and in specific (reviewed genome sequence has integration DNA and hand viral structure other the the with the from multimers on arises structure PIC, 2016). al., This the et species. of Lesbats well viral 2015; part not Hughes, a in are a (reviewed intasome, only cell PIC infected is The the per often complex there of this as composition of retroviruses, copy the different single between and differences conversion potential are this neither Fig. of known, (PIC, mechanism complex exact pre-integration into The converted 12). is RTC the transcription, reverse Following integration Genome H RNase c. and polymerase that 2014). seems al., et it (Nowak case subunits this distinct in by although carried RT, are HIV that activities of to subunits similar the conformation of a displays one an subunit of assumes each Ty3 which retrotransposon yeast in related structure distantly (Kohlstaedt homodimeric the role asymetric of structural RT a the Interestingly, plays and polymerase 1992). subunit active al., incomplete an et display smaller to the one that only indicating the is site, subunit RNaseH large of the most and 1986; though even shared, al., different, is et very sequence is (Lightfoote the subunits domain both of H conformation RNase The the 1992). al., all characterized, et Kohlstaedt almost best subunit removed smaller the cleavage a far protease and RT by a complete is which the of RT in consists that HIV subunit The large a 1982). with heterodimer al., a et forming (Hizi especially activity characteristics, enzymatic H distinct RNase display they concerning replicationas viral three in all role preparations, a al., purified play in et form might predominant (Hizi the dimer be to (RT+IN)/(RT+IN) seems an heterodimer the and While heterodimer 1977). RT/(RT+IN) decribed been an a have monomer, or that forms heterodimer, RT three a an the monomer, ALV are In a heterodimer. as a forms assemble RT can HIV it in ALV and homodimer, in 1985), al., et (Roth monomer simple 2008). al., et (Abbondanzieri conformation polymerisation are in RT non-complementary) from the not PPT stabilize the (but to cleave nucleotides also able then complementary to of is but Addition it synthesis DNA. plus-strand how are following for explaining they primer the perhaps as or PPT orientations, PPT the both the use between of both to 3’ transition able quickly nucleotides nucleotides no to the are able if there is configuration if RT H but DNA, RNase RNA, in are using oriented sequence when is PPT Additionally, RT the end. following primer, 5’ 3’ the RNA the to as close to sequence is way. close PPT site other located the polymerisation the is the facing while is domain primer, domain H the H RNase of RNase end the the and complex, primer, primer/template the RNA/DNA of an end With 3’ the polymeri- to DNA to. close the binds is that it site so acid itself sation nucleic orients the RT complex, of primer/template structure DNA/DNA the a on using dependent When highly is RT of orientation binding the an elonga- Using its functions. between degradation discriminate and constantly tion to RT the requires reaction complex this Obviously, eedn ntertoiu,teatv omo Tdffr.Wiei L Tat sa as acts RT MLV in While differs. RT of form active the retrovirus, the on Depending nvitro in ehd bodnir ta.(08 hwthat show (2008) al. et Abbondanzieri method, ..EoeosRetroviruses Exogenous 2.1. 57

2.1 2. Retroviruses

first intasome nucleoprotein complex to be characterized was that of the spumaretrovirus SFV (Hare et al., 2010), which turned out to be similar to the intasomes of orthoretroviruses, but less complex. In SFV, four IN proteins bind the two ends of the viral DNA in a symmetric fashion. Both genomic ends are sandwiched between two IN which use their own CCD alongside the other’s NTD to bind the viral DNA, crosslinking the whole stucture. Both of these inner IN also use their CTD to engage host cell DNA during integration. This structure is stabilized by two outer IN that each bind to one inner IN using a CCD-CCD dimerization interface and seem to only have a structural role (Hare et al., 2010). Characterization of the intasomes of MMTV, RSV, maedi-visna virus (MVV, a lentivirus) and HIV-1 revealed that they contained 8 IN for non-lentiviruses and up to 16 IN for lentiviruses (Ballandras-Colas et al., 2016; Yin et al., 2016; Ballandras-Colas et al., 2017; Passos et al., 2017, respectively). Interestingly, while the dimer of tetramer or tetramer of tetramer stuctures observed for these intasomes sounds very different from that observed for the SFV dimer of dimer, the positions of the four IN at the core of the structure are extremely similar, defining the intasome core as a highly conserved structure among retroviruses (reviewed in Engelman & Cherepanov, 2017). While there are no structures 2.1 currently available for gamma- or epsilonretroviruses, Engelman & Cherepanov (2017) predict their intasomes to contain four IN proteins, based on sequence properties. The intasome core catalyses two subsequent bimolecular nucleophilic substitutions (Fig. 14, Engelman et al., 1991).

1. 3’-processing, during which the inner IN use water as a nucleophil to cleave the ends of the viral DNA. This reaction forms invariant CA-3’ ends with exposed hydroxil groups (the reason why all integrated LTR start in TG and end in CA). Since only the 3’ ends are cleaved, a 5’ overhang is generated at both ends of the DNA genome.

2. Strand transfer, during which the inner IN use the exposed hydroxyl groups of the recessed 3’ ends of the viral genome to attack a phosphdiester bond of the host cell genome and link its 3’ end to the 5’ of host DNA. Being a bimolecular nucleophilic substitution, the breaking of the phosphodiester bond and ligation of the 3’ end occur simultaneously.

There are several interesting aspects of strand transfer. First, the two phosphodiester bonds that are implicated in this reaction are always separated by a defined number of nucleotides, which is determined by the 3D structure of the intasome and thus specific of each retrovirus. The structure of the SFV intasome is such that the two attacks, each perfomed by one of the inner IN, are always four nucleotides apart, while for HIV-1 and MMTV this distance is of 5 and 6 respectively (reviewed in Lesbats et al., 2016). Second, strand transfer links the 3’ end of the viral genome to the host genome by nicking the latter, but does not link the 5’ overhangs and the newly nicked host genome ends, leaving a pair of single stranded gaps. These gaps are probably repaired by the host cell machinery, which degrades the 5’ overhangs using a 5’-Flap endonuclease, fills in the 4-6 missing nucleotides generated by the staggered nucleophilic substitution and finally ligates

58 ligte46b a,gnrtn h S ge rms.Terpeettosaentt scale. to not and are processing), representations The 3’ frames). during (gree lost single-strand TSD those the remaining to of generating the nucleotides gap, ends repair bp complementary 3’ then 4-6 (the the the enzymes flaps binds filling Cellular 5’ This genome. viral here). host the (6 nucleophilic cutting the other bimolecular nicks, each to Two from strands DNA. nucleotides genome viral 4-6 viral processed at the the IN, by alongside catalyzed Strand genome (bottom) are strand). host substitutions minus sequence the the for the LTR binds represent 5’ and nucleotides IN PBS (these and transfer: nucleotides plus-strand (-OH) few the behind for a 3’LTR cleaves group and are hydroxyl and PPT reactive the genome strands between a retroviral leaving the the ends, of of 3’ ends the ends both nucleotide at binds 6 5’ IN and complementary any black, strand, processing: indicate implicated 3’ YYYYYY in the (top) and of XXXXXX DNA sequence. color reaction. the the cellular after in faded red, arrowheads as using shown in show are are minus-strand attack nuclophilic and steps. of Sites purple integration indicated. genome in retroviral shown the is of DNA representation Schematic – 14 Figure IN positionsitselfonthehostgenome strand transfer generates reactiveOHgroups IN hydrolysesviralDNA and 3’ processing the DNA fragments the 5’ flapandligate cellular enzymesexcise the gaps5’->3’ cellular polymerasesfill nucleophilic substitution IN catalysesbimolecular

5’ 5’ 5’ XXXXXX YYYYYY XXXXXX YYYYYY YYYYYY 5’

5’ 5’

5’ 5’ 5’ HO- HO- HO- target siteduplication AC AC AC AC AC XXXXXX YYYYYY ..EoeosRetroviruses Exogenous 2.1. CA CA CA CA CA -OH

-OH YYYYYY XXXXXX YYYYYY XXXXXX XXXXXX

5’ 5’ 5’ 5’ 5’ 5’ lssrn viral Plus-strand 5’ 5’ 5’ 59

2.1 2. Retroviruses

the 5’ ends of viral DNA to the 3’ ends of the host genome (reviewed in Lesbats et al., 2016). This filling in of the stagger is responsible for the TSD, a perfect repetition of a 4-6 nucleotide long sequence on each side of a newly inserted retroviral genome (Fig. 14). The targeting of strand transfer is not a completely random process in retroviruses. While at the sequence level there is only weak indication of specificity (Wu et al., 2005), at the genomic level some retroviral genera present a clear insertion bias (Schröder et al., 2002). Lentiviruses show a preference for actively transcribed and intron-rich genes, gammaretroviruses prefer to insert in strong enhancers and active gene promoter regions, whereas spumaretroviruses prefer to avoid actively transcribed loci. Alpha-, beta- and deltaretroviruses show little to no insertional bias, and the genome-wide distribution of epsilonretroviral insertion has not been well-studied (reviewed in Kvaratskhelia et al., 2014; Lesbats et al., 2016). The insertional bias in both lentiviruses and gammaretroviruses seems to be due to interactions between the PIC and cellular factors, namely LEDGF/p75 and BET respectively. These cellular factors are able to interpret the transcriptional state of host cell chromatin and bind preferentially to a particular state. Exploitation of this property by the viral PIC permits preferential insertion in parts of the 2.1 genome that are more favorable to viral replication. Since the proteins and binding mechanisms involved (as well as the resulting insertional biases) are clearly distinct, usage of cellular factors to guide lenti- and gammaretroviral insertion is a noteworthy example of evolutive convergence (reviwed in Kvaratskhelia et al., 2014; Lesbats et al., 2016).

d. Protein expression

Following integration of the retroviral genome into the cellular genome, forming the so called provirus, gene expression must occur to form new particles (Fig. 12). Viral transcripts are syn- thesized by the cellular RNA Pol II using the transcription initiation site in the 5’ LTR and cell transcription factors or in the case of complex retroviruses regulatory proteins. Since the host cell machinery is responsible for transcription, viral RNAs are capped and polyadenylated. This transcription step yields a complete genomic RNA, which can then either be used directly for translation of gag, pro, and pol, or be spliced (reviewed in Coffin et al., 1997). As stated previously, in simple retroviruses the only spliced viral transcript is the one used for env expres- sion while in complex retroviruses there exist several additional alternatively spliced forms that encode the regulatory and accessory proteins. In addition to serving as template for protein ex- pression, the unspliced genomic RNA can also be packaged into new virions to serve as genome for the next infection (reviewed in Freed, 2015). The importance of unspliced or partially spliced viral RNA is problematic for viral replication, as the host cell machinery inhibits nuclear export of unspliced RNAs. Different strategies have been developed to force export of unspliced or partially spliced RNAs. In complex retroviruses this task is often accomplished by a regulatory protein such as Rev for HIV-1 (Malim et al., 1989) or Rex for HTLV-1 (reviewed in Younis & Green, 2005). In simple retroviruses, the constitutive transport element (CTE) sequence present between env and the 3’ LTR allows export of non-spliced RNA (Bray et al., 1994). This small

60 srlae rmteifce el(i.1) h a oyrti n l he fissubunits its of three assembly all orthoretroviruses, most and In polyprotein process. this Gag a in role The off major 12). a which bud play (Fig. 2.1.1.b.) virion to cell see immature (MA-CA-NC, infected needs the the to forming then from and need which envelope released they lipid is structure, viral synthesized, organized the been highly acquiring have membrane, a genomes cellular into RNA assemble viral and and congregate proteins necessary the Once budding and Assembly e. 1992). al., et Dedera 1992; reaching Roth, before N-glycosylated & additionally (Felkner is Env membrane SU of cell The potential the 2). fusogenic chap. 1997, the the al., enable et free and Coffin to TM in necessary of (reviewed in extremity is network cleavage N-terminal Golgi the This the at TM. through peptide and transits fusogenic Env SU virion. separate the proteases of furin-like surface cellular the which at trimeric transmembrane the present into its be grouping while start will proteins lumen that Env stage form ER this is At the membrane. synthesis ER in the Once is in embedded 2013). Env is al., helix of et domain Nyathi translocation, extracellular co-translational future in for (reviewed the ER protein complete, the transmembrane to protein for nascent common the is directs peptide as ORF signal the hydrophobic of a end of 5’ presence the The at RNA. sub-genomic spliced the using occurs and sion Gag of level the with par on Gag-Pro-Pol. Pro, of of amount amount lower high usual unusually the an of expression Gag-Pro-Pol the or to Gag-Pro leading the of expression RSV, MMTV for in HTLV, Finally, respectively. frameshifts precursor in sequential found two is or disposition one Another necessitating which 2018). in al., example, for et MPMV (Irigoyen and cases occurring readthrough of Gag, 7% to compared leading about when more in expression once Gag-Pro-Pol readthrough, of codon amount stop reduced is MLV, a Pro-Pol to like translate to viruses way other only The In about frame. represents Gag. reading Gag-Pro-Pol of of level expression the of level of is the frameshifting 5% viruses ribosomal these Since in 1995). event, Brierley, rare see review somewhat early a expression an (for and circumvention codon example allowing stop for Gag synthesis, HIV the Gag in of after case frameshifting the ribosomal is by This occurs 5’. Gag-Pro-Pol in of Pro for relative sequence on coding impact the the important contains transcript, an the the has retroviruses This of most 9). In end levels. (Fig. protein Gag-Pro-Pol, organization 5’ just genome or the viral Gag-Pro-Pol, at the and Gag-Pro on starts polyproteins: depending always of form translation the Since in expressed Pol. are latter and Pro Gag, containing of CTE sion the of export nuclear mediate 2003). which Cullen, proteins, in (reviewed of transcript export family (NXF1)/NTF2-related 1 (NXT1) factor 1 export RNA protein nuclear the with directly interacts sequence hl a-r-o sepesdi h yols rmteuslcdtasrp,Evexpres- Env transcript, unspliced the from cytoplasm the in expressed is Gag-Pro-Pol While expres- mediate to needs transcript genomic unspliced the nucleus, the from exported Once gag pol pro , pro R si noelpig- rm eaieto relative frame -1 overlapping an in is ORF satce to attached is and pol pro-pol r ahteronsitdraigframe, reading shifted own their each are gag while , si eaaeOFi h same the in ORF separate a in is ..EoeosRetroviruses Exogenous 2.1. pol ssili hfe frame, shifted a in still is gag and 61

2.1 2. Retroviruses

occurs at the cellular plasmic membrane, the exception to this rule are betaretroviruses in which the capsid assembles in the cytoplasm before being transported to the membrane, this aspect will be discussed below. In non-betaretroviruses, Gag and Gag-Pro-Pol molecules traffic to the cell membrane. In HIV-1 it has been shown that this trafficking is dependent of MA and

a particular membrane phosphoinositide (phosphatidylinositol-4,5-biphosphate, PtdIns(4,5)P2).

Alteration of the MA sequence or depletion of PtdIns(4,5)P2 leads to retargeting of Gag towards intracellular membranes (Ono & Freed, 2004; Ono et al., 2004). Structural studies show that the

binding of HIV-1 Gag to PtdIns(4,5)P2 triggers exposure of the myristric acid linked to Gag, which can insert into the cell membrane and stabilize the interaction (Saad et al., 2006). Once at the membrane Gag induces the formation of lipid raft microdomains by recruiting cholesterol

and sphingolipids, and the over-representation of these two species alongside PtdIns(4,5)P2 in HIV-1 viral envelopes reflects their participation in particle formation (reviewed in Freed, 2015). While Gag multimerization can begin during the transfer from the cytoplasm, the major part of it occurs once Gag is anchored at the membrane (Kutluay & Bieniasz, 2010). This oligomerization process implicates mostly CA-CA domain interactions and is increased when the NC domain of 2.1 Gag binds to a nucleic acid. Gag proteins associate as a continuous network of hexameric rings to form a curved lattice that creates a bud at the membrane. In order to accommodate the amount of curvature necessary to form a 120 nm diameter sphere, the lattice of hexamers contains randomly distributed gaps (reviewed in Freed, 2015; Mattei et al., 2016). While the general hexameric arrangement of immature Gag seems conserved among retroviruses, the details and especially the organization of the CA NTD seem to vary (Schur et al., 2015). Since Gag-Pro-Pol polyproteins contain functional MA and CA domains, they behave the same as Gag with regards to membrane association and lattice assembly, ensuring encapsidation of PR, RT and IN in the virion through their association with Gag domains. RT can also bind to the cellular tRNA needed for initiation of reverse transcription and thus lead to its encapsidation (reviewed in Coffin et al., 1997; Hughes, 2015).

Cytoplasmic unspliced viral RNAs serve as genomes for new virions and thus have to be encapsidated during viral assembly. The trafficking of viral RNA to the cellular membrane probably implicates interaction with the NC domain of Gag, though it has been proposed that it occurs by simple diffusion (Chen et al., 2014). Viral RNAs dimerize in the cytoplasm by using the dimerization initiation signal (DIS) and it is this dimer that is subsequently recognized by Gag NC, which binds to another RNA sequence called packaging signal, or ψ (Nikolaitchik et al., 2013). In HIV-1, DIS sequences are in the 5’ untranslated region (UTR) of the genome and are variable between strains, GCGCGC and GUGCAC being the most common (Hussein et al., 2010). The ψ sequence is also contained in this 5’ UTR and usually contains a region that is absent in spliced transcripts, thus ensuring encapsidation of unspliced genomic RNA (reviewed in Ali et al., 2016). Additionally the PBS of unspliced viral RNAs can bind a matching tRNA primer, leading to its encapsidation. In the cytoplasm and during transfer, Gag is mainly bound to the 5’ UTR and the RRE of HIV-1 viral RNA through its NC domain and this interaction

62 hr hycncodnt sebyo prldm fECTII rvkn progressive a provoking from ESCRT-III, virion, of nascent dome the spiral into a pathway of this assembly of coordinate proteins can membrane they of with where incorporation concurrently starts to recruitment leads ESCRT and which in budding model a propose Lippincott-Schwartz (2017) 2016). al., al. et et Schöneberg in (reviewed to composed monomers leads ESCRT-III dome assembly spiral numerous ESCRT growing of how a involving clear all it proposed, is are models nor Several proteins 2014), scission. stalk ESCRT al., membrane the et recruited around Bleck and 2014; the in al., only not et gather Engelenburg they or (Van whether Votteler whether or in virion unclear nascent (reviewed is the example enter It ESCRT-III) for (notably 2015). domain YPXL Freed, a 2013; present Sundquist, also & auxiliary RSV additional and encode HIV also both retroviruses domains, Most three sorting pathway. are (endosomal transport) ESCRT There for cell 1991). required host al., complexes the and et with RSV HIV, interaction (Gottlinger in allowing own example three for its all (found respectively), on YPXL EIAV and cut PPXY to P(T/S)AP, sequences: unable domain is this late by virus main membrane of the the Deletion to that attached protein. stalk remaining the thin particles proteins a of the Gag release, positions Retroviral virion released. different abolishes HIV-1 be at in to domain located cell domains, the late of that so-called from contain membrane its dissociate 2014). to al., needs et (Pereira buds for virus binds network the and microtubule where the vesicles membrane, cell use small the vesicles of reach the recycling to surface Finally transport assembly. the the anterograde RNA reaching at and before site Gag Golgi assembly immature the and the to reaches ER it the there (Choi through Gag from transfers on endosome, Env present signal MPMV retention aa 1999). 18 al., an encapsulates to et due and cytoplasm forms lattice pericentriolar the Gag in the RNA membrane, viral cellular the at assembling of Instead buds. system. membrane of lattices Gag the to Env a anchor would using which or raft intermediary, directly lipid an either the as interact, involving protein could potentially Env host membrane, and cellular Gag Finally of Gag. area by same established the microdomains to Env addressing and Env specific the Gag all implies both it co-targeting Gag-Env of with moment. takes that virion at bud the the budding on during present proteins that and membrane of the Env generalization across that passively making proposes diffuses incorporation long, Passive several packaging. particularly offers Env (2015) of is Freed mechanisms generally, exclusive species non-mutually More 2015). and this Freed, Env in in (reviewed of problematic tail tail results lattice. the these cytoplasmic Gag the but the between domains, by interaction MA engendered direct bud Gag a membrane suggested the have HIV-1 at in assembles form.Studies Env trimeric Golgi exactly and how glycosylated and cleaved, unclear definitive ER is its the in It membrane through cell transiting the reaches components, it until viral body previous the from pathway synthesis number high a 2015). to Freed, binds in RNA (reviewed the domains membrane NC the Gag At anchored of molecules. Gag of amount small a involves nealtecmoet r sebe tteshrclmmrn u,tensetvirion nascent the bud, membrane spherical the at assembled are components the all Once assembly different a use betaretroviruses MPMV, by illustrated and previously mentioned As different a takes which Env, is virions the of assembly the for needed component final The ..EoeosRetroviruses Exogenous 2.1. 63

2.1 2. Retroviruses

HIV-1 Gag ESCRT-III subunits ESCRT-I, ALIX(?), ESCRT-II, or ESCRT-II-like protein(?) ABCDE

Figure 15 – Representation of the ESCRT mediated budding model proposed by Lippincott- Schwartz et al. (2017). (A) Viral Gag assembles at the membrane, starting the budding process and recruiting cellular proteins of the ESCRT family. ESCRT-III nucleates and encircles the Gag lattice. (inset) Top view of the Gag/ESCRT assembly (B) Continued assembly of the Gag lattice increases the sphericity of the bud, ESCRT-III continues its nucleation. (C) ESCRT proteins remain associated with Gag and are included in the budding virion from where they regulate a continued assembly of ESCRT-III in the stalk. This assembly occurs in a circular pyramid of tighter and tighter circles, leading to an ESCRT-III mediated narrowing of the stalk. At this stage ESCRT-III can still freely diffuse between the viral and cellular compartments (white arrows). (D) The ESCRT-III assembly reaches it’s tightest confor- mation and forms a closed dome. (E) This restricts the stalk to a diameter small enough for spontaneous membrane fission and release of the virion. Some proteins of the ESCRT system are trapped in the virion. 2.1 Adapted from Lippincott-Schwartz et al. (2017).

thinning of the membrane stalk until it is pinched off and the virion is released (Fig. 15). About 5-6 minutes elapse between initial detection of Gag and release of a virion ex vivo (Jouvenet et al., 2008).

f. Maturation

The final crucial step in retroviral replication is the maturation step in which the polyproteins get cleaved so as to release the individual proteins (Fig. 12). Immature virions not only have a very distinct architecture from mature virions (Fig. 16A), they are also not infectious as the viral cycle requires the viral proteins to be separated from one another. The viral polyproteins are cleaved by PR, which itself is part of a polyprotein (most commonly Gag-Pro-Pol), concomitantly or shortly after release, preventing loss of the individualized subunits into the cytoplasm (reviewed in Konvalinka et al., 2015). This mechanism requires a tight regulation of PR activity to prevent premature cleavage. While the exact regulation mechanisms are unknown, the active form of PR is a homodimer, reducing the risk of premature cleavages as the odds of formation of a PR dimer before assembly is generally low. The exception to the rule is found in alpharetroviruses, in which PR is synthesized at the same level as Gag, and thus present at a much higher concentration. This suggests that there might also be other factors, such as interactions with cellular proteins, that could inhibit PR activity until maturation, such a reliance on particular cellular indicators would also explain why in betaretroviruses maturation occurs only once the assembled particle reaches the cell membrane (reviewed in Konvalinka et al., 2015). Additionally, in HIV-1 PR dimers in the context of a polyprotein are less stable than those formed of individualiued PR, and tend to adopt an inactive conformation, reducing the risk of premature activation (Louis et al., 1999).

64 rmsitn.(otm ceai ersnaino h laaesau fGgadascae virion the associated from of Adapted and gray. size dark Gag (2015). in The al. represented of et are 9. status Konvalinka subunits Gag cleavage Fig. non-canonical the in and peptides of as Spacer morphology. representation colors sequence observed cleavage cleavage Schematic same of PR (bottom) the rate the frameshifting. using the of to polyproteins, all representation roportional Gag-Pro-Pol tomography, is (top) and arrow cryo-election sequence. Gag by maturation HIV-1 MLV HIV-1 the (B) and of nm. HIV-1 microscopy, 50 maturation bars electron MPMV scales (gammretrovirus). thin-section MLV by and visualized (lentivirus) HIV-1 is particles. (betaretrovirus), retroviral MPMV of of maturation virions mediated PR – 16 Figure B A ACA MA

mature immature Gag-Pro-Pol NC Gag MPMV ACA MA ACA MA ACA MA NC-TF NC NC mature immature PR ACA MA nvitro in I- MLV HIV-1 TIN RT CT:tasrm etd bandafter obtained peptide transframe NC-TF: . NC A xmlso maueadmature and immature of Examples (A) RNaseH ..EoeosRetroviruses Exogenous 2.1. ACA MA NC 65

2.1 2. Retroviruses

Once maturation is triggered, a highly organized cleavage sequence is initiated (Fig. 16B). A number of studies in HIV have revealed that the first cleavages occur between CA and NC, releasing the RNA associated NC protein as well as Pro-Pol from the still membrane-bound MA-CA. The second cleavage separates MA and CA and MA remains associated to the viral membrane while CA starts to form the mature capsid. Finally, PR, RT and IN are released and small spacer peptides with a regulatory role are removed from CA and NC. Alteration of the kinetics of polyprotein cleavage leads to aberrant mature morphology or non-infectivity (reviewed in Mattei et al., 2016). Once released from the membrane, CA organizes into a fullerene structure of hexamers alongside 12 CA pentamers (which are naturally curved) to allow incurvation of the lattice (Ganser et al., 1999). The exact form of the mature capsid surrounding the viral genome depends on the genus, in HIV for example, it is conical with 5 pentamers at one end and 7 on the other, in MPMV it is conical with 6 pentamers at each end, and in RSV and MLV it is polyhedral (icosahedral) with 12 evenly spaced pentamers forming the vertices (Fig. 16A, reviewed in Mattei et al., 2016). The exact mechanism of transition from the immature to the mature capsid is unknown, but it has been shown that the CA-CA interactions involved are different. While the immature assembly seems to be mainly dependent on CTD-CTD interactions, the formation of the mature assembly is reliant on NTD-NTD intra- hexamer and CTD-CTD inter-hexamer interactions (reviewed in Mattei et al., 2016; Perilla & Gronenborn, 2016). This new mature capsid is less stable than the immature assembly, allowing 2.2 dissociation from the viral membrane after cell entry and disassembly after reverse transcription (reviewed in Konvalinka et al., 2015). In addition to cleaving the polyproteins, PR can also be implicated in activating the fusogenic potential of Env. In MLV and MPMV, PR cleaves the cytoplasmic tail of TM late during maturation, enhancing the fusogenicity of Env (Brody et al., 1994; Rein et al., 1994). In HIV the envelope is not a substrate of PR but Gag interactions with the long cytoplasmic tail of TM inhibit fusion of immature particles (Murakami et al., 2004; Wyma et al., 2004). Maturation seems to also affect the repartition of HIV Env at the surface of the virion, on immature virions Env forms several distinct clusters, but after maturation all Env are gathered in a single cluster and this concentration seems to increase fusogenicity (Chojnacki et al., 2012).

2.2 Endogenous Retroviruses

Endogenization is the process by which a viral genome becomes a permanent part of the genetic material of its host, so that it may be transmitted vertically to the host’s descendants. While there are examples of endogenizations across many viral families (and all seven categories of the Baltimore classification), the very particular replication cycle of retroviruses, with its necessary genomic integration step, leads to their over-representation among endogenous viral elements, especially in animals (reviewed in Feschotte & Gilbert, 2012)

66 ouaino olso agroIln ih lose oelgto h urn tt of state none 2004, current in isolated the and 1920s, on small the light a since mainland some of the shed from study isolated also been The might has insertions. Island population This KoRV. Kangaroo two on or koalas single one of A only population cells. with all endogenization affect recent not very does about presented that koalas infection south-eastern exogenous of than population an less reflecting present and koalas genome, genome south-eastern per koala while copies animals, the these in in copies 2017). endogenous KoRV average al., being of on et number present Legione koalas the north-eastern 2012; measured that al., showed also et (2012) (Simmons al. 25% south-east et about the Simmons in to have while drops studies positive, KoRV progressively Prevalence are number Australia retrovirus. this north-east endogenizing in currently koalas led of a viruses 100% is endogenous that revealed KoRV and that exogenous conclusion in of DNA the characteristics viral was to of harbored there combination cells the units, germline This family into koala genome. within that insertion that their and recent showed insertions (2006) very KoRV al. of a et transmission at Tarlinton vertical hand hinting other and the ORFs On proviruses studies intact genome. inserted subsequent containing the functional, but Additionally, which still 2006). tested, insertions, al., were koalas viral et of all (Tarlinton locations in transmission and vertical genome amounts contradicts different viral present the koalas of different that parts showed amplify to able were (MLV, mice 2006). (JSRV), al., et sheep Tarlinton (KoRV, in koalas found in is notably being it most examples and phenomenon, current 2015) ancient even Kozak, an or is 10% recent endogenization about with While mouse ongoing 2002). in still 2001), al., al., et (Waterston et genome (Lander 8% the about of and ERVs represent of they composed humans is genome In the remnants. of ERV 1-10% nowadays that so accumulation events, an endogenization to & rare led (Aiewsakun of has interaction described animal-retrovirus been of have history long My This 430 2017). as Katzourakis, lost. is old often Endogenization retroviral as the event. endogenizations is with and rare it simultaneously pathway developed rather replication though probably a it would, therefore process, allele is ancient very ERV an a an understandably as of population subsistence the and integration through transmission, vertical The spread and germline then the can in ERV integration Following the 1988). al., et for (Lock mice in time demonstrated long been a to has (ERV), and retroviruses endogenous cells in new germline of resulting to formation spread material, the in to genetic result infections parental retroviral the of potential of) The then part 17). will (Fig. a endogenization and as integration (and during alongside genome vertically cell germline transmitted the be into retrovirus, a inserted by infected be is will cell a genome such viral If moment. the that the at germline by the determined of cells is the reproduction in sexual present genome during genome transmitted vertically the of nature The cells germinal of Infection a. sequences retroviral of Endogenization 2.2.1 oVwsfis lsie sa R sdrn hi nta td,Hne ta.(2000) al. et Hanger study, initial their during as ERV an as classified first was KoRV 1 . 165 5 oisgnm,rpeetn potential a representing copies/genome, oiscl,i codnewt KoRV with accordance in copies/cell, ..Edgnu Retroviruses Endogenous 2.2. 1 . 53 × 10 67 − 3

2.2 2. Retroviruses

horizontal transmission infection germ linesomatic infectious retrovirus

endogenization

endogenous retrovirus

2.2 verticaltransmission

fixation

Figure 17 – Endogenization of exogenous retroviruses. Illustration of the horizontal and vertical transmission pathways of retroviral infection. Infection of somatic cells leads to horizontal transmission through virus production and re-infection of new individuals. This infection is not transmitted hori- zontally. Infection of germline cells leads to endogenization through horizontal transmission. Following endogenization, every cell contains the viral genome. Through continued reproduction and horizontal transmission events, the endogenous retrovirus can become fixed in the population, i.e., present in all individuals.

68 olwdb wtht nitaellrapicto aha copne yacag nGag in change a by accompanied elements, pathway IAPE amplification by intracellular infections an repeated to by switch suggests initiated a This was by followed ER. genome the mouse of the instead of membrane colonization cellular the that extracellular to infectious addressed produce being to Gag able IAPE were of particles, sequences copies IAPE 14 functional about rare contains These also Later functional. genome 2004). mouse al., associated the et an (Dewannieux that with showed elements cell IAP-like (2008) the al. outside et observed ever Ribet by are cistaernae studies particles into no budding 2004; by but al., generated ER, are et the VLPs (Dewannieux Env-less of process, copies this of During number 2008). al., the et augmenting Ribet genome, the in retrotranspose to some genome, mouse functional the contain in which copies of an thousand 700 a containing about element present sequences retroviral IAP IAP-related ERVs. (IAPE) and gene (IAP) particle A-type intra-cisternal the membranes. intracellular to Gag of retargeting the a of necessitates loss and by amplification accompanied gene retrotransposition generally from is switch called retrotransposition by The also amplification 2015). to is re-infection Johnson, by process 2013; Heidmann, this intracellular & genome, form Dewannieux to the in as (reviewed into so expressed re-integrate are then genes viral that instead particles cell, a the in to exits pathway leading never different genome, virus a new the viral uses infect ERVs which the of either of spreading can Intracellular copies copies. present released of already accumulation are progressive that that cells virions germline new them among expression The gene cells, provirus. is integrated point already starting the an except above, from retroviruses exogenous same the for is described pathway re-infection that The as 18). (Fig. either spreading genome, intracellular the through in or copies re-infection proviral by of After number (2012). the al. increase et greatly Simmons often genome each ERVs viral endogenous endogenization, the become of copies has 165 KoRV average which on contains in genome koalas cellular north-eastern the in above, mentioned As ERVs of spreading Intra-genome al., et b. Marchi 2011; al., et (Jha Mya 0.25 or the 0.15 of as 2014). some recently on as estimations insertions and Age suggest 75% 2016). loci over al., unfixed to et 0.05 (Wildschutte from amplifications ranging recent frequency very allelic 36 denoting an found at genomes present human sites, sequenced insertion 2500 & over polymorphic (Medstrand of evolution analyses during comparative radiation insertions Recent their and 1998). after humans Mager, occured and having humans, primates to to specific common are insertions that subgroups: two into split 2012). further al., be et Simmons 2006; al., et (Tarlinton koalas retrovirus tested infectious population 162 active this still the to a spreading of was is 24 virus and 2007, the that in indicating whereas infection, KoRV, KoRV for of evidence positive presented were animals tested 26 the of neapeo h rniinbtenapicto ahascnb on nmc with mice in found be can pathways amplification between transition the of example An can group This group. HERV-K the of members are humans in ERVs active recently most The gag , env pro ee mn hc ny2cpe aermie fully remained have copies 2 only which among gene, and pol ee btno (but genes env ..Edgnu Retroviruses Endogenous 2.2. htalwteesequences these allow that ) env env 69

2.2 2. Retroviruses

re-infection extracellular intracellular

retro- transposition loss of env gag re-addressing

deleteriousneutral impact advantageous 2.2

loss fixation

solo LTR genetic drift purifying selection

Figure 18 – Mechanisms of ERV amplification and fate of ERV insertions. ERVs can amplify their copy number either through an extra-cellular (re-infection, see Fig. 12) or intra-cellular pathway (retrotransposition). The latter is often accompanied by re-addressing of Gag and loss of Env. Each new insertion can be deleterious, neutral or advantageous to the host, influencing its chances of being lost or fixed in the populations. Continually vertically transmitted ERV copies face three main fates: solo LTR formation in which a recombination even excises all the viral genome, leaving a single LTR, genetic drift in which the ERV sequence accumulates mutations and indels, and purifying selection in which one or several ERV sequences are maintained functional following exaptation. Same color scheme as Fig. 9, faded sequences represent loss of function.

70 Fg 9,rvee nJhsn 05.Teetrecassaenteulyrpeetdin represented equally not some are of classes underrepresentation or over- three display clades These vertebrate different 2015). and genomes Johnson, vertebrate in spumaviruses reviewed to III and 19A, to deltaviruses related or (Fig. are beta- ERVs alpha-, I lenti-, Class to viruses. exogenous II epsilonretroviruses, often to and similarity are gamma- RT ERVs their 2.2.3). on (see and based viruses incomplete class exogenous a be of to often assigned those can sequence than their pressure as selective matter, different delicate a a under be species can particular ERV a an of to Assignation genus even genera. or retroviral and exogenous proximity existing genetic the their to reflecting much them groups compare into a to insertions designates sort the to family of useful be part as can are it misleading retroviruses Nevertheless, (all is classification this phylogenetic These however of events. level endogenization "families", higher original related called few numerous often a to sometimes are rise give or groups ERVs one genome, from host originated the all within sequences, amplify to tendency their to Due ERVs of Groups 2015). activity c. al., et integration ERV (Magiorkinis My that 10 fact last the the to over humans linked in probably fourfold is decreased This has ERV humans. replicating in actively no reported viral 2001), intact been al., fully has et and Turner HERV-K113, conserved as highly (such some non-deleterious identified While been of have 2009). fixation loci al., favors et which Brady units, 18, host (Fig. transcription the only of by insertions outsided influenced found is preferentially amplification are that genome revealing human the in current-day the insertions while fixed regions, genomic transcribed virus actively HERV-K towards bias resurrected (Dewannieux insertional mentioned an genome previously discloses the viral two using endogenous infection between functional Interestingly, events a 2006). recombination al., obtain that et to and sufficient family), be (that the would retrovirus of loci infection progenitor unfixed an the to of close production probably allows is sequence enough show consensus copies their HERV-K human-specific that The conservation 2004). al., et (Belshaw re-infection through of including selection, purifying continuous 2013). Heidmann, & complementation Dewannieux prohibits Env or in of (reviewed recombination loss occur subsequent as events unless host, transmission, its of immune horizontal that to the to back ERV of the switch of factors a fate some the ties of irremediably avoidance also it and but factors) system, restriction host insertion of saturating probability the possibly (increasing and particles viral of concentration higher same offer much the a might in permits ERVs Staying cell pathway. of retroviral extracellular Intracellularization traditional 2012). the al., over advantages et clear several (Magiorkinis loci ERV of 80% about for This env amplification. number copy efficient by more confirmed much a later the to was that led show clearly Gag of numbers re-targeting copy and IAPE Env and of IAP loss of Comparisons Env. of loss a and addressing nhmn,aayi fgn osraini h EVKeeet icse bv eel a reveals above discussed elements HERV-K the in conservation gene of analysis humans, In sascae ihhg oynme n htaot2%o R iegsaeresponsible are lineages ERV of 20% about that and number copy high with associated is nsilico in nlssta hwdta nmn amla eoe,ls of loss genomes, mammalian many in that showed that analyses env ugsigta EVKapicto occurred amplification HERV-K that suggesting , ..Edgnu Retroviruses Endogenous 2.2. Retroviridae family). 71

2.2 2. Retroviruses

A α λ spuma γ ε β other Class I Class II Class III B 48.3% 47.0% 4.7% Classified ERVs 4000 2000 0 Spuma

Delta

Lenti

Alpha

Beta

Epsilon

Gamma Panda Cat Ferret Frog Sheep Rhinoceros Tilapia Horse Dog Fugu Pufferfish Stickleback Medaka Turkey Coelocanth lamprey Sea Turtle Lizard Budgerigar Chicken Zebrafinch finch Darwin Cod Zebrafish Megabat hyrax Rock devil Tasmanian Orangutan Marmoset monkey Squirell rat Mole rat Kangaroo rat Nowegian Rabbit Elephant Lemur Galago Pika Dolphin Lama bat Brown Tenrec Wallaby Platypus Human Gorilla macaque Rhesus Tarsier squirrel Ground Sloth Armadillo Opossum Baboon Cattle Pig Hedgehog Shrew Manatee Gibbon Mouse shrew Tree Chimpanzee Olive baboon Olive Guinea pig Guinea Alligator ailMel felCat musFur xenTro equCab oviAri oreNil cerSim canFam tetNig takRub gasAcu oryLat galGal chrPic taeGut latCha anoCar petMar geoFor melUnd melGal gadMor danRer macEug papHam dasNov loxAfr triMan macMul otoGar hetGla ornAna speTri susScr pteVam sorAra proCap echTel homSap ponAbe calJac saiBol micMur musMus ratNor ochPri tupBel bosTau monDom tarSyr panTro nomLeu gorGor dipOrd oryCun turTru vicPac myoLuc eriEur choHof sarHar papAnu cavPor allMis 2.2

Figure 19 – ERVs disproportionately belong to certain retroviral genera. (A) Proportion of ERVs related to each of the major retroviral lineages and corresponding ERV class. Based on the analysis of 35975 ERV genomes in 65 animal species by Hayward et al. (2015). The "other" category contains unclassified retroviral species, often related to spumaviruses. (B) Detailed distribution of ERVs in different host genomes. The trend line indicates the total amount of ERV genomes identified in the species, the bars the proportion each lineage represents among the ERVs of the given genome. A cladogram of retroviral lineages and of the host species is provided as ordinate and abscissa respectively. For simplicity, ERVs not belonging to any of the seven major lineages are not represented. Adapted from Hayward et al. (2015).

of them (Fig. 19B). In general class I and II ERVs are the most numerous elements in vertebrate genomes, especially gamma- and betaretrovirus derived sequences. Delta- and lentiviral ERVs are almost completely absent in all vertebrate genomes and some of the more species specific retroviral genera are present in significant numbers only in their host species: alpharetroviruses in birds and epsilonretroviruses in fish (Hayward et al., 2015).

2.2.2 Impact of retroviral integration on the host

a. LTR integration can drive ectopic gene expression

While LTRs do not code for genes, they are powerful gene expression regulators, being able to initiate transcription and to terminate it by polyadenylation. As such, their presence in the

72 eoecudata otolr o eeepeso osbc ote15sadMcClintock’s and 1950s the to back goes expression gene for controllers as act LTRs. could by genome expression gene of Regulation differentially to LTRs. due active patterns expression ge- differential even present would reproduction, individuals in during identical variations inherited netically imperfectly phenotypic are the regulations explain epigenetic part Since mammals. in might LTRs controlled epigenetically imperfectly mice identical 2006). genetically al., in et genes Maksakova of epialleles in expression (reviewed metastable differential gene generating to driven LTR, leading LTR and the 2002) IAP of al., state humans, et epigenetic (Rakyan in the expression on mecha- oncogene reliant directed mutational highly LTR different is of expression towards LTRs, case bias ETn the a by in show none As could nisms. but groups activity, ERV promoter different LTR that IAP of suggesting by number transcription caused a aberrant initiations reference provoke (2006) transcript to al. aberrant tend et Maksakova ETn Additionally polyadenylation. while and skipping), termination exon to splicing, (aberrant Mak- leading disruption to internal also According in potentially the implicated 20). mostly Interestingly, (Fig. are differ elements 2004). IAP to (2006) al., seem al. et et insertions sakova ETn (Ribet and retrotransposition IAP for of ERV mechanisms dependent mutagenic MusD are and the ORFs coding of their expression lost that (reviewed of elements groups LTR (ETn)/MusD are transposon ETn early 2006). the al., et and Maksakova IAP in the to due them of most mutations, (reviewed cancer in 2016). implicated Trobridge, be may & that Bii genes new in for screen to expression used trigger also of now to expression is integrations expression oncogenes retroviral repress of of the to ability methylated This in 2016). heavily Mager, implicated are & Babaian LTRs they in (reviewed HERV tissues healthy of in reports though of oncogenes, 2006). al., in human number et investigated Maksakova a often in (reviewed are more groups ERV there is active Nevertheless, phenomenon very of this number 2015), a al., present which insertions et mice, ERV (Magiorkinis ectopic recent humans drive Since can in events. intron polyadenylation rare an and in are or splicing gene aberrant the a ORF, near or LTR expression LTR the an gene the of that disrupting insertion implicates gene, possible The insertion host provirus. is following the a of mutagenesis it into sequences insertional provirus While of a mutagenesis. mechanism of insertional common insertion more of the source from simply a results and this process random rather a mutagenesis. Insertional 2.2.3.b.). (see sequences viral accompanying is solo any a aspect without to genome This degenerated have the expression. integrations in gene ERV present older cellular LTR, of of majority immense regulation the the as on important doubly impact major a have can genome htlw&Mri 20)oe h eyitiun yohssta h admisrinof insertion random the that hypothesis intriguing very the offer (2001) Martin & Whitelaw known all of 10% least at for accounts mutagenesis insertional retroviral mice, In hl neto frtoiue a rsn oebae,i sstill is it biases, some present can retroviruses of insertion While h data oierpae lmnsi the in elements repeated mobile that idea The ..Edgnu Retroviruses Endogenous 2.2. enovo de 73

2.2 2. Retroviruses

exon 1 exon 2

sa polyAsd sa sa polyA

aberrant splicing

AAAAA premature termination 2.2 AAAAA

aberrant initiation

Figure 20 – Mutagenic effects of LTR insertions. (top) Pre-insertion locus. Exon 1 and 2 are part of a transcribed gene, the spliced mRNA is represented in gray, spliced sequences as dotted line. (bottom) Post-insertion locus. In this example the inserted ERV contains LTRs presention a cryptic splice acceptor (sa) site, in addition to the canonical polyadenylation (polyA) and transcript initiation functions. The canonical retroviral splice donor (sd) and acceptor sites are also represented. Cellular transcription can be disrupted through several ways, three are presented here: aberrant splicing leads to addition of ERV sequences into the mRNA (green), premature termination leads to interruption of the transcript and polyadenylation (AAAAA) and abberant initiation leads to formation of incomplete transcripts. Adapted form Maksakova et al. (2006).

74 mroi tmcls(atn ta. 02.Adtoal h EVHLRpeet binding presents LTR HERV-H the Additionally in 2012). expressed al., specifically and et RNA) (Santoni polyA cells all stem of (2% embryonic abundantly is which group, are HERV-H the primates and mice in 2012). al., used et ERVs In (Emera the independently primates. exapted but and been expression, have mice can and prolactin in independent this drives prolactin of LTR of illustration an expression stunning specific cases endometrial A species both 2017). the often al., considering and et when host Chuong found the 2016; in be by involved al., acquisition et sequences recent Thompson ERV in a The (reviewed generally 2013). are al., regulations et ERV these mouse (Chuong all the factors that regulatory bind shown to trophoblastic been able cells, key has trophoblast in it activators placentation, transcriptional numerous of provides RLTR13D5 stages group (Macfarlan later previously described Concerning as 2012). elements ERV al., same transient et the this of for control responsible under transcripts are the state of totipotent many the that of Analysis confirmed totipotent). cells (being these mouse stage of of 2-cell subset of the transcriptome involvement of a cells of The like discovery behaved 1999). the that by cells al., strengthened stem embryonic et later was Bénit development of 1995; stage present al., this group ERV-L in et are ERVs the (Cordonnier genes early to mammals belonging of the them placental 6-12% during of most stage, many that in , 2-cell show ERVs mouse the (2004) of to al. control oocyte under et expressed the Peaston from mice, specifically In stages, 2015). pre-implantation al., sequences repeated et the (Lynch of usage promoters through as probably organs, other from recruited them. were among genes of many LTRs these that elements, indicated additionally repeated species by non-viviparous and initiated mammalian viviparous between be in Comparisons involved to genes shown numerous been recently of has expression (2015) placentation endometrial al. et the Lynch that 2016). al., show et example Thompson for in (reviewed or factors cellular conditions specific cellular of particular presence to the responding promoters, tissue-specific as even but promoters, own its over control for cell host the by used replication, be retroviral present also genes. its in can the for transcription role genome regulate In particular to viral very ability use". the a their by new accomplish but provided to a evolved features have for uses LTRs The cell co-opted ends. host is own the [...] when applies function term particular this context, a shaped previously for character, "a selection which by natural process the by describe concept to a (1982) is Vrba domestication) & mostly Gould or by co-option is coined called LTR (also role the physiology, Exaptation cell exapted. mutagenesis this of been insertional part having more the are sequence here to Once discussed contrast events 2017). in regulation the but al., above, LTRs, discussed solo events et and clear Chuong sequences LTR increasingly 2016; the repeated become on al., of dependent has role et transcription the Thompson of then, in regulation Since (reviewed in (1969). ERVs Davidson & especially concept the Britten and in by sequences present forward later put was and batteries 1956), gene 1950, of McClintock, example for (see maize on work h etsuideapeo R T novmn neryhmndvlpetconcerns development human early in involvement LTR ERV of example best-studied The expression gene cell as only not act to able LTRs of reports of number a been now have There ..Edgnu Retroviruses Endogenous 2.2. 75

2.2 2. Retroviruses

sites for the pluripotency transcription factors NANOG and LBP9 (Santoni et al., 2012; Wang et al., 2014). HERV-H driven expression is dependent on the presence of these factors as well as OCT4 (another marker of pluripotency) and 10 % of the transcripts are long non-coding RNAs (lncRNA) (Wang et al., 2014). Knock-down of some of these lncRNAs leads to rapid differentiation (reviewed in Chuong et al., 2017). While the exact function of lncRNAs is not know it has been suggested that they act as scaffolds recruiting OCT4 and the co-activator complexes Mediator and p300 to regulate expression of neighboring genes (Lu et al., 2014).

b. Impact of retroviral gene acquisition on host physiology

Pathologies linked to endogenous retroviral genes. Expression of retroviral genes can have deleterious consequences on both host cell and organism biology. Many retroviral genes have evolved specifically to modify host cell functions so as to create a more favorable environment for viral replication. This is especially true of regulatory and accessory genes (see 2.1.1.e.), though these are rare in ERVs (with the notable exception of HERV-K which presents a complex genome, Dewannieux et al., 2006), but also applies to canonical retroviral proteins. To keep HERV-K as an example, it has been shown that its Env was able to interfere with Tetherin restriction and even to possess oncogenic potential, with a strong impact on cell physiology (Lemaitre et al., 2014; Lemaître et al., 2017). The Env of the related JSRV retrovirus was also shown to be sufficient to

2.2 induce tumorigenesis when expressed at a sufficiently high level (Wootton et al., 2005). Besides these admittedly rare cases in which a structural viral protein can carry oncogenic properties, rodent and avian retroviruses often carry oncogenes such as src, myc or ras (reviewed in Vogt, 2012). While the potential impact of such genes after integration is obvious, endogenizations that present a strong oncogenic or otherwise deleterious impact have a low chance of becoming fixed in the population and as such rarely result in observable ERVs (Fig. 18). This aspect might also explain why there are only a few endogenized complex retroviruses. There are many mostly correlative studies showing that expression of human ERVs is associated with a range of diseases. This is especially true for HERV-K which has for example been associated with melanoma, ovarian, breast, and prostate cancer (Büscher et al., 2005; Wang- Johanning et al., 2007, 2008; Agoni et al., 2013), as well as amyotrophic lateral sclerosis and multiple sclerosis (Li et al., 2015; Bhetariya et al., 2017) and auto-immune diseases (reviewed in Trela et al., 2016). Other studies however have shown that expression of HERV-K derived proteins also occurs in many healthy tissues (Schmitt et al., 2015) and it remains to be seen to what extent (if any) ERV proteins expression plays a direct role in the different pathologies.

Endogenous retroviral genes as restriction factor. A number of the still coding ERV genes have been maintained in a coding state following exaptation. Since exaptation of Env sequences for use as syncytins is critical to the present work, it will be treated in more detail in chapter 3 and other examples will be given here. In most of those cases, ERV genes have been exapted by the cell to protect it from infection by exogenous viruses. Two of the most famous examples are the

76 hsooy ervrlisrin a ebt udnadabesn,bti n aethey case any in but blessing, host a normal and in force. burden role driving evolutionary a a major play both a to domestication represent be able where can material, cases genetic insertions are new Retroviral there with physiology. host, hosts the provides on elements impact viral deleterious of strong a have can events common relatively cross-species are 2017). events eventual Johnson, such an & as Sinha against receptor, in as mechanism (reviewed SLC1A5 defense using a older virus as an another counteracted of serves human Suppressyn transmission it that the possible that in is or activity It retrovirus, Syncytin-1 2000). al., human for & et partner Sinha Blond in necessary and reviewed 3.2.1 the by (see 3.1.2.a., also receptor placenta (see is specific species which a diverse but as in 2017), used retroviruses Johnson, 1 is of family which amount carrier 2013), high al., solute strikingly et the a (Sugimoto with protein Suppressyn proteins for (SLC1A5) Env retrovirus 5 of human member interaction extant blocks known of no Suppressyn binding is counteract. receptor there inhibit to strains, to FLV able exogenous be 2013; of to al., shown subset et been a (Ito has Refrex-1 retroviruses specifically While exogenous can 2013). by They al., infection et 2013). during Sugimoto al., used et receptors surface Sugimoto cell 2013; ERV to factors al., truncated bind et soluble of (Ito two expression discovered humans the recently and from were cats result genes In that respectively) 2003). Suppressyn al., and et (Refrex-1 JSRV (Spencer exogenous block interference cytoplasm to able receptor the Env express by in can JSRV entry particles of its copies ERV viral blocking Other defective 2007). and al., of Gag et to exogenous (Arnaud aggregation able of to is assembly leading (enJS56A1) the addressing, Gag integrating JSRV intra-cellular by of assembly copy endogenous viral an with sheep, interfere In species. other in observed 2001). al., et (Taylor defective infectivity with Env. exogenous diminishing virion the a surface, to of its leads instead at this proteins virus non-fusiogenic, Env it the renders in that integrated mutation a be presents can Fv4 protein Since Fv4 the assembly cell, Env-receptor viral MLV infected during on an Second, only 1991). in al., dependent et (Granowitz is fusion interference require not Receptor does 2005). and binding al., et Nethe infected in already (reviewed an entering cell i.e., super-infection, observed avoid commonly to retroviruses is exogenous interference, allows also receptor and called can the process, Fv4 accessing This from fusion. First, proteins initiating twofold. Env and exogenous is receptor blocking infection receptors, MLV specific restrict Env MLV to with Fv4 interact by employed mechanism The 1983). related MLV an is vivo in sequence strong shows Fv1 ERV-L Fv4. spumaretroviral and to 1 similarity (Fv) susceptibility virus Friend factors restriction mouse ncnlso,i sipratt eebrta hl ervrlifcinadendogenization and infection retroviral while that remember to important is it conclusion, In been since has mechanisms similar using factors restriction as genes viral of Exaptation yitreigwt h ia I Jlcer&Blioe 96.F4o h te hand other the on Fv4 1976). Baltimore, & (Jolicoeur PIC viral the with interfering by env eeta a ensont niisMVifcin(kd Odaka, & (Ikeda infection MLV inhibits to shown been has that gene gag Bs ta. 96 n sal orsrc L infection MLV restrict to able is and 1996) al., et (Best ..Edgnu Retroviruses Endogenous 2.2. env 77

2.2 2. Retroviruses

2.2.3 Fate of ERVs and control by the host

Since production of viral particles or expression of parts of the viral genome is not always in the interest of the host, several ways of dealing with ERV integrations have developed during evolution. Often, but not always, they lead to inactivation of the proviral genome.

a. Epigenetic regulation of ERV expression

The fastest and most direct way for the host cell to regulate ERV expression is through epigenetic control. Cell enzymes are able to chemically modify DNA and the histones it is wound around so as to create zones of enhanced or repressed expression. At the DNA level, this regulation can take the form of CpG dinucleotide methylation to inhibit gene expression, and it has been shown that methylation of CpG in the HERV-K LTR for example can indeed block its transcriptional activity (Lavie et al., 2005). More generally, micro-array based studies reveal that CpG methylation is a silencing mechanism of all HERV in healthy tissues, though there are indications that this mechanism is age-dependent, younger HERV being more methylated than older ones (Szpakowski et al., 2009). At a higher organizational level, epigenetic regulation of gene expression is achieved by chemically modifying histones, especially through acetylation and methylation, though the for-

2.2 mer does not seem to play a major role in HERV expression regulation (reviewed in Hurst & Magiorkinis, 2017). Methylation of histones can either have an activating or inhibiting effect on gene expression and the predominant associated modifications are H3K4me3 and H3K9me3 re- spectively. In silico analyses of IAP and HERV-K integration sites revealed an over-representation of inhibiting and an under-representation of activating histone marks, coherent with repression of ERV expression (Campos-Sánchez et al., 2016). The major cellular actors of histone based epigenetic repression of ERV expression are the Krüppel associated box zinc finger proteins (KRAB-ZFP). The proteins of this family are able to bind to specific DNA sequences with their zinc finger domain and to recruit histone modification complexes through their KRAB domain and interaction with transcription intermediary factor 1-beta (previously called tripar- tite motif-containing protein 28, TRIM28) (reviewed in Hurst & Magiorkinis, 2017). It has been shown that KRAB-ZFP proteins are encoded in genomic regions showing a propensity to recombination events capable of generating new random KRAB-ZFP with altered sequence specificity, and that the apparition and fixation of new copies is correlated with the acquisition of new ERV sequences during evolution (Thomas & Schneider, 2011; Lukic et al., 2014). A more recent comprehensive study of human KRAB-ZFP found that two thirds of them bound sequences of ERVs and retrotransposons and confirmed the correlated emergence of new ERVs and corresponding new KRAB-ZFP proteins (Imbeault et al., 2017). However, the study also showed that a third of all KRAB-ZFP (among which most of the older and most conserved ones) bind to host sequences and that ERV/retrotransposon directed KRAB-ZFP are conserved long after the corresponding element lost its replicative potential, suggesting that KRAB-ZFP play

78 eslkl steEVae n ned tde nHR- aesonta on insertions and young that less shown become have events HERV-K NAHR on that studies reason indeed, to and present ages stands as ERV it well the genome, as as just viral likely LTRs the in less cannot of accumulate itself can parts LTR indels other the and the but mutations in genome, point host Since rest cell. the the ERV, the from an by for disappeared excised form entirely be possible has smallest the genome is viral LTR solo the A of lost. be contain thus not will does such commonly and most centromere NAHR As genome a mediate of 2008). to portion excised capable Coffin, The perfectly 2002). & are Lupski, LTRs Jern & excision identity, (Stankiewicz sequence in the of (reviewed to bp them leads 200-500 between need provirus) processes contained young single material non- a genetic from two bordering the result of LTRs of LTR identity identical Solo sequence two 2001). high the the al., (often which et LTRs in (Lander events (NAHR) 10 recombination about homologous of allelic factor a by elements complete 2.2.1.b.). and an 18 than Fig. rather (see sequences ORF these this of of losing potential of amplification loss probability higher the increased the is to due sequences be ERV might in this of deletion though amount observed this sheer above, commonly the discussed A through silencing mutations. irreversible epigenetic accumulated becomes the quickly to and contrast passively mostly In Parseval occurs 2004). (de process al., expression et autonomous Villesen for 2003; non-functional al., disrupt element et often the which render events and is recombination ORFs genome and viral viral indels the the mutations, neutral integration of a after accumulation to such the leading As host, to 18). the subject (Fig. on sequences effect ERV of deleterious to selection negative neutral or a has insertion ERV cases, most formation LTR In solo and mutations of Accumulation sequence. b. IAP suppressed the of reactivation age-dependent transcription a an rythmic induces of continual gene the vicinity ages, cellular the mouse the the in of As located profile. sequence expression circadian IAP (2002) a mouse with al. mammalian gene a et early of Barbot in demethylation by role progressive investigated a example the play intriguing describes to an seem In often above. sequences discussed ERV as why development reason also the of is This events. part demethylation regulation global be example Epigenetic might for with 2017). reproduction, of Magiorkinis, stages early & cases the by Hurst many impacted in in observed reviewed and is and expression process above ERV reversible (discussed enhanced a accordingly and is reactivated cancers regulation in be epigenetic phenomenon can ii) expression ERV and situations 2012), some sequences al., ERV in of et third a (Pérot about transcribed humans in are that show studies transcriptome expression, ERV all proteins maintain to pressure selective elements. KRAB-ZFP/ERV no ERV the be non-functional would If binding there 2017). race, al., arms an et purely (Imbeault was co-evolution regulation expression gene in roles additional yfrtems omnfr fodEVeeet stesl T,otubrn more outnumbering LTR, solo the is elements ERV old of form common most the far By repress not do mechanisms regulation epigenetic these i) that here note to important seems It 8 × 10 − 5 eobntosprlcsprgnrto,wiei le netosti number this insertions older in while generation, per locus per recombinations nvivo in yoehlto sacommon a is Hypomethylation . ..Edgnu Retroviruses Endogenous 2.2. env 79 ,

2.2 2. Retroviruses

drops to 4 × 10−7recombinations per locus per generation (Belshaw et al., 2007). Using an in silico simulation, the authors further determined that a single nucleotide substitution reduced the odds of NAHR tenfold, two substitutions reduced it a hundredfold (Belshaw et al., 2007). This age-dependent pattern is also observed in other major HERV groups, except for HERV-H which presents an intriguingly low amount of solo LTR elements, suggesting that group-specific factors can influence these excision events (Gemmell et al., 2016). Since the LTR is able to mediate transcription initiation and transcript polyadenylation (Pérot et al., 2012), solo LTR can still have a major impact on host physiology (see 2.2.2.a.) and the only way to completely control the impact of ERV insertion is epigenetic regulation.

c. Exaptation and purifying selection

In the relatively rare case of exaptation of retroviral sequences by the host, these sequences are generally under purifying selection, like would be expected for any other cellular gene. This is for example the case of the exapted syncytin genes (see 3, reviewed in Lavialle et al., 2013) and of members of the HERV-H group (Gemmell et al., 2016). Under purifying selection, the ratio of synonymous to non-synonymous nucleotide mutations is strongly biased towards the former as it conserves amino acid sequence coding and identity. Additionally, deletion of sequences under purifying selection is generally counter-selected on the population level. Exaptation of retroviral

2.2 sequences does not mean that they are free of epigenetic control however, just like many other cellular genes their expression has to be and is often regulated. Syncytins for example are almost exclusively expressed in the placenta (see chapter 3, reviewed in Lavialle et al., 2013; Denner, 2016).

80 3. Syncytins

3.1 Retroviral envelopes

Studies of the human genome revealed that about 50% of still coding ERV ORFs are env genes (Villesen et al., 2004). This might be due to the fact that retroviral Env present, among others, a function that is uncommon among cellular genes: fusogenicity. Among the handful of fusogenic proteins identified in eukaryotic cells, many show strong similarity to viral sequences (see Fédry et al., 2017). The following part describes the structure of orthoretroviral Env, with the exception of epsilonretroviral Env who present a number of differences and are not relevant to this work (see Rovnak et al., 2007, as well as 2.1.2.a.).

3.1.1 Functional domains a. Domains of the SU

Signal peptide. The N-terminus of SU is composed of a hydrophobic sequence of about 20-30 amino-acids forming the signal peptide (Fig. 21). Signal peptides are a commonly used addressing signal to the ER and the secretory pathway, and hydrophobicity plays an important role in their recognition by the signal recognition particle (SRP). Binding of this complex to the nascent Env protein leads to its co-translational insertion into the ER, after which the signal peptide is cleaved (reviewed in Nyathi et al., 2013). While this sequence is therefore not part of the functional mature retroviral Env, it is nevertheless indispensable for its synthesis. Loss of the signal peptide leads to incorrect processing and post-translational modification of the Env polyprotein.

Receptor binding domain. Since the SU is the domain exposed to the exterior, it is responsible for recognition and binding of the cognate receptor domain. This implies the receptor binding domain (RBD) at the surface of SU. The exact nature of the residues implicated in this process is variable depending on the SU and receptor, and often unknown, but seems to often be contained in the N-terminal part of SU (Fig. 21, reviewed in Overbaugh et al., 2001). Studies on Syncytin-1 SU binding to its receptor SLC1A5 for example demonstrated that its 124 N-terminal amino acids constituted the minimal receptor binding domain (Cheynet et al., 2006). A comparison with the SU domain of other retroviruses using SLC1A5 as receptor showed that only about a dozen residues were conserved in this 124 aa, grouped in three cysteine containing motifs (PCXP,

81 3. Syncytins

SU TM SU C C CC CC n CXXC RXKR CX TM N-ter RBDPRR HR1 HR2 Ct C-ter signal fusion extracellular ISD transmembrane peptide peptide membrane intracellular

Figure 21 – Representation of the organization of gamma-type retroviral Env domains. (left) Relative position of the SU (yellow) and TM (orange) domains, from N-ter to C-ter: signal peptide (purple), receptor binding domain (RBD), proline-rich region (PRR), CXXC motif, furin-like cleavage site (RXKR), fusion peptide (green), heptad repeat 1 (HR1), immunosuppressive domain (ISD, blue), CXnCC motif, heptad repeat 2 (HR2), transmembrane region (red) and cytoplasmic tail (Ct). (right) Schematic representation of Env structure using the same color code. The CXnCC motif forms an intra-molecular disulfur bond and induces formation of a loop. In gamma-type Env this motif also establishes a second disulfide bond with a C of the SU CXXC motif, tethering SU and TM. For a beta-type Env, the CXnCC would take the form of CXnC, the SU CXXC and SU/TM disulfide bond would not exist and the ISD would not have the canonical CKS-17 sequence (see also Fig. 22A).

CYX5C and CX8-9CW) and one longer SDGGGX2DX2R motif (Fig. 24, Cheynet et al., 2006). Similarly, the minimal domains needed for receptor binding in MLV and HTLV-1 and -2 are contained in the N-terminal part of SU (Battini et al., 1992, 1995; Kim et al., 2004). Subsequent studies showed that here too very short domains are directly implicated in receptor interaction by MLV (Battini et al., 1998) and HTLV-1 SU (reviewed in Jones et al., 2011). Additionally, replacing the RBD of MLV Env with that of HTLV-1 changes the viral tropism of MLV to that 3.1 of HTLV-1, confirming the importance of this domain (Kim et al., 2000). Importantly, while the RBD is the main determinant of receptor binding, there are other regions in the SU that play an important role in this process (reviewed in Jones et al., 2011)

Proline-rich region. The proline-rich region (PRR) is found in gamma-type envelopes and separates the N-terminal and C-terminal parts of SU (Fig. 21). The PRR seems to form a flexible hinge that is implicated in stability changes of the SU/TM dimer after receptor binding, potentially being a part of the signal inducing TM conformational changes and fusion (Lavillette et al., 1998). The hinge-like structure of the PRR has also been confirmed by structure analyses which show that insertions within this region do not contribute to or majorly impact the structure of Env (Riedel et al., 2017).

CXXC. A conserved CXXC motif (consensus CWLC) is necessary to regulate SU-TM interac- tion in non-alpharetroviral gamma-type Env (Fig. 21, Wallin et al., 2004). The exact localization of the CXXC varies between different retroviruses (see for example de Parseval et al., 2003), but its presence is crucial to fusogenic potential of gamma-type Env. This domain was first discov-

82 flniia M(inil ta. 98 n te eatp Msc sMM n HERV-K and MPMV as such TM beta-type other and 1988) al., regions et al., corresponding (Cianciolo et the TM (Cianciolo lentiviral Env in gamma-type of peptides other immunosuppressive of identified that studies and Following TM FLV 1985). in identified first 21) (Fig. region a forms domain. and Immunosuppressive 2014). smaller al., much et (Aydin is region HR2 tether TM N-terminal gamma-type an to in attached beta-type while domain in helix, variable, helical the more long short of is single HR2 outside a of the structure forms The on it 2014). grooves TM al., et hydrophobic Aydin with 1997; al., helix interacting et 6 (Chan and so-called latter fashion the (forming slanted coiled-coil a HR1 HR2 length inner The in the general 2014). bundle), to al., Aydin its apposed et Aydin are 1997; and 2001; that coiled-coil, al., al., coils inter-TM forms et et domain (Bénit trimeric retroviruses Chan a among conserved 1996; forms is domain al., structure and HR1 et The Fass constitutes 2014). that 21, al., trimer (Fig. Env et the arrangement of fusogenic assembly the functional also the but Env, of formation, conformation structure the coiled-coil for in implicated important are repeats heptad These 2013). Johnson, & CX Henzy the and present) (when domain immunosuppressive the of avian of peptide 1996). repeats. fusion Gallaher, Heptad 21, the (Fig. bond position, disulfide shifted a by its linked cysteines to by extremely flanked 23, addition therefore is (Fig. Env are In membrane gamma-type peptide survival. target fusion the viral the into in to itself Mutations deleterious 2013). insert Johnson, to & able hydrophobic Henzy being aa in 20-30 fusion, reviewed a for is it crucial region which This is in 21). that gamma-type (Fig. stretch downstream avian acids for amino except envelopes, nine retroviral about situated all is in TM of part N-terminal the peptide. Fusion TM the of Domains b. sequences these Env. retroviral and of 1991), cleave domains al., TM proteases et and These SU (Hosaka the motif 2013). distinguish RXRR to Johnson, or serve & RXKR an Henzy after in cellular proteins by (reviewed precursor performed is proteases and network furin-like Golgi or the in furin occurs Cleavage 21). (Fig. SU of end terminal site. cleavage al., Furin-like et Li 2004; al., et Wallin 23, (Fig. fusion and TM and 2008). in this SU carried receptor, group of thiol the separation reactive allowing with a motif, by interaction bond this After disulfide intra-CXXC 2008). an al., into converted et is Li pre-fusion bond SU-TM the 1997; stabilizing CX al., TM and et the subunits (Pinter with both conformation association of TM (in association TM the and strengthening SU below), between see bond disulfide motif, a of part being as ered iutdimdaeyatrtefrnlk laaest,tefso etd is peptide fusion the site, cleavage furin-like the after immediately Situated h Mcnan w etdrpa ein H1adH2,oeo ahside each on one HR2), and (HR1 regions repeat heptad two contains TM The ydfiiin h laaest ewe UadT om h C- the forms TM and SU between site cleavage the definition, By h muoupesv oan(S)i osre 0aa 20 conserved a is (ISD) domain immunosuppressive The n C 1-2 oan(i.2,rvee in reviewed 21, (Fig. domain ..Rtoia envelopes Retroviral 3.1. n CC 83

3.1 3. Syncytins

(Blaise et al., 2001; Morozov et al., 2013). High conservation of the ISD sequence (the so-called CKS17 consensus sequence) is observed only among gamma-type TM (Fig. 22A, Cianciolo et al., 1985; Bénit et al., 2001), and as such the term ISD is sometimes reserved to these envelopes (reviewed in Henzy & Johnson, 2013). In vivo tumor rejection assays in mice (Fig. 22B) seem to confirm the in vitro immunosuppressive properties for a number of envelopes, such as MLV (Mangeney & Heidmann, 1998), MPMV (Blaise et al., 2001), Syncytin-2 and -B (Mangeney et al., 2007) and FLV (Schlecht-Louf et al., 2014) for example. The presence of an ISD is also coherent with the fact that many exogenous retroviral infections lead to immunodeficiencies (reviewed in Denner, 2014). The effects of the ISD on the host immune system are numerous including inhibition of NK cells and cytotoxic T lymphocytes, down-regulation of pro-inflammatory and up-regulation of immuno-suppressive cytokines for example, but the exact mechanisms by which the ISD induces these changes are unknown (reviewed in Denner, 2014).

CXnCC or CXnC motif. These two motifs, collectively referred to herein as CXnC1-2, are present immediately after the ISD in all retroviral envelopes and form one of the major distin- guishing factors between beta- and gamma-type TM (Fig. 22, reviewed in Henzy & Johnson,

2013). For gamma-types the motif is CX6CC, with the two first cysteines forming an intra-TM disulfide bond which creates a loop between the two HR regions so that HR1 and HR2 are apposed (Aydin et al., 2014). The third cysteine forms a disulfide bond with the CXXC motif in the SU, binding both subunits covalently until SU/receptor interaction upon which this bond is isomerized into the intra-CXXC bond (Fig. 23, Pinter et al., 1997). In beta-type envelopes,

this motif takes the form of CX4-7C, presenting the two cysteines necessary to form the disulfide 3.1 bond of the TM loop, but none to bind the SU. As a result in beta-type Env, the SU is not bound covalently to TM (Henzy & Coffin, 2013).

Transmembrane domain and proximal regions. As is to be expected, the TM subunit contains a transmembrane domain, a hydrophobic stretch of about 40 aa that forms a helical domain spanning the membrane (Fig. 21). Deletion or cleavage of this domain turns the remaining N-terminal part of Env into a soluble factor, released from the cell as observed for Refrex-1 (Ito et al., 2013), suppressyn (Sugimoto et al., 2013), ERV3 (Cohen et al., 1985) and human endogenous MER34 ORF (HEMO, Heidmann et al., 2017) for example. The two domains immediately N- and C-terminally of the transmembrane are referred to as the membrane proximal external region (MPER) and the cytoplasmic tail respectively. The length of both regions is variable depending on the virus and can play an important role in regulating virion incorporation, membrane fusion and other Env properties (reviewed in Henzy & Johnson, 2013). The MPER is 20-30 aa longer in beta- than in gamma-type TM (Bénit et al., 2001), and it has been shown that in HIV-1 this stretch plays a role in Env incorporation and during fusion (Vishwanathan & Hunter, 2008). Studies of the cytoplasmic tail of MLV Env have

84 rwbte)adaayi fat-n nioygnrto imnsprsieEvsol lctalower a into elicit should tumors cells Env (immunosuppressive Env-expressing (immunosuppressive cell growth generation the response). tumor antibody tumor antibody anti-Env of Mouse of of injection analysis mice. measurement and before by in better) vector, quantified grow assay expressing are immunosuppression Env responses line Immune an highlighted cell mice. formation with tumor bond transduced the disulfide are in of reflects lines implicated Rationale Green cysteines (B) type. the orange. TM with CX in sequences the (right) retroviral TM. as beta-type the for well residues of consensus as motif the yellow state, Sequences and exogenous/endogenous TM sequence. gamma-type and for CKS-17 identity lineage consensus residue the their to by compared grouped left, the are on brackets by indicated CX retroviruses, and assay. immunosuppression ISD mice the the of of Sequence – 22 Figure beta-type Env gamma-type Env A lenti ERV beta ERV delta gamma alpha HIV-1 FIV CAEV HERV-K Fematrin MMTV JSRV Env-Cav1 pan-Mars-Env2 HEMO EnvV2 Syncytin-Opo1 Syncytin-Ten1 Syncytin-Mar1 Syncytin-Rum1 Syncytin-Car1 Syncytin-Ory1 Syncytin-B Syncytin-A Syncytin-2 Syncytin-1 HTLV-1 BLV REV PERV MPMV MLV GaLV FLV BaEV RSV ALV CKS-17 LQARILAVERYLKDQQLLGIWGCSGKLI--C LQARILAVERYLKDQQLLGIWGCSGKLI--C LVIGLKVEAMEKFLYFAMQELGCNQNQFF-C LEARVARVEAITDRIMLYQELDCWHYHQY-C LRQTVIWMGDRLMSLEHRFQLQCDWNTSDFC LEEAILHIGNELQALKVRLALSCHADYRWIC LEEVVLELGQDVANLKTRMSTRCHANYDFIC LYDVVRVLGEQVQSINFRMKIQCHANYKWIC YQNRLALDYLLAKEGGVCGKFNLTSCC VQSRMALDMKQAQQGGVCA-YN-TMCC SRKDHVLDIPTTQRQTACGTVG-KQCC MDNRLALDYLLAEQGGVCAVIN-KSCC IQNRRSLDLLTLHQRGLCAMHG-MECC LDNRIALDFLLAEQGGVCAIAN-TSCC YQNRLALDYLLAEEGGVCGKFNSSDCC LQNRMALDILTAAQGGTCAIIK-VECC LQNRMALDILTAAQGGTCAIVK-TECC LQNRRGLDLLTAERGGICLALQ-EKCC LQNRRALDLITAEKGGTCLFLQ-EECC LQNRRALDLIVAERGGTCLFLQ-EECC LQNRRGLDMLTAAQGGICLALD-EKCC LQNRRALDLLTAERGGTCLFLG-EECC AQNRRGLDLLFWEQGGLCKALQ-EQCC AQNRRGLDWLYIRLGFQCPTIN-EPCC LQNRRGLDLLTAEQGGICLALQ-EKCC LQNRRGLDLLFLKEGGLCVALK-EECC LQNRRGLDLLTAEQGGICLALQ-EKCC LQNRRGLDLLFLKEGGLCAALK-EECC LQNRRGLDLLFLKEGGLCAALK-EECC LQNRRGLDILFLQEGGLCAALK-EECC LQNRRGLDLLTAEQGGICLALQ-EKCC LQNRAAIDFLLLAHGHGCEDVA-GMCC LQNRAAIDFLLLAQGHGCQDVE-GMCC LQNRRGLDLLFLKEGGL A et S eune fdffrn xgnu n endogenous and exogenous different of sequences ISD left) (A, n C 1-2 oi ndffrn ervrlseisadrationale and species retroviral different in motif CX n C 1-2 B Env transduction of tumor analysis growth ..Rtoia envelopes Retroviral 3.1. of antibody generation injection tumor cell tumor cells Env expressing tumor cells analysis n C 85 1-2

3.1 3. Syncytins

prefusion open extended fold-back hemifusion fusion cell membrane HA1

HA2 virus membrane

prefusion open extended fold-back hemifusion fusion

SU HR1

HR2 TM

Figure 23 – Steps of retroviral Env induced cell fusion and TM rearrangements. (top) Structure of the hemagglutinin trimer during cell fusion. During the pre-fusion state, the fusion peptide (asterisk) is sheltered in a hydrophobic cavity at the center of the trimer. HA1 (the SU equivalent) dissociates from HA2 (the TM equivalent) resulting in the open conformation and initiation of the fusion process. HA2 extends and the fusion peptides insert themselves into the target membrane. HA2 folds back on itself, brining the membranes closer together. Finally once HA2 has entirely folded, a fusion pore is created and fusion is achieved. Structures in color are known, structures in gray are inferred. (bottom) Schematic representation of the same steps for a retroviral Env, using the same color scheme as Fig. 21, except for the heptad repeat domains which are in dark orange here. Adapted from Podbilewicz (2014).

shown that this 32 aa long sequence is cleaved to release the 16 aa R-peptide during maturation and this cleavage is essential to render the protein fusogenic (Januszeski et al., 1997). The tails of JSRV and HERV-K TM have been shown to play a role in their oncogenic activity, probably

3.1 through interaction with cellular signaling factors (Hull & Fan, 2006; Lemaître et al., 2017). Finally the cytoplasmic tail seems to be an important factor in Env incorporation into lentiviral particles (Christodoulopoulos & Cannon, 2001; Freed, 2015).

3.1.2 Env induced membrane fusion

Retroviral Env proteins are class I membrane fusion proteins, the best-studied of which is the influenza hemagglutinin (HA) fusion protein. Interestingly the class I fusion proteins show very little sequence identity, but are structurally highly similar. HA is also synthesized as a precursor before being cleaved into the SU equivalent HA1 and the TM equivalent HA2, which also presents an N-terminal fusion peptide, an intra-molecular disulfide bond and two heptad repeat regions (reviewed in Podbilewicz, 2014; Harrison, 2015).

a. Implication of the cognate receptor in fusion

Necessity of destabilizing the metastable SU/TM. The uncleaved Env folds during synthesis to assume its most stable possible conformation. Separation of SU and TM after cleavage

86 ono,21) hr r vrtodzneoeosadedgnu n hthv ensug- been have that Env endogenous & and Sinha exogenous in dozen (reviewed two interference over receptor are through There re- other 2017). a each Johnson, as block protein to transmembrane able multiple-pass being thus ASCT2) ceptor, (RDR, gamma-type SLC1A5 of number the a use contains that group) envelopes interference RDR called (originally group. group interference terference SLC1A5/RDR the of example specificity: Receptor 2002). al., et conformational (Diaz-Griffero first acidification a endosomal induces after receptor occurs cognate only the fusion to but shift, binding example, for cases ALV trigger most secondary In necessary fusion. a in is for HA, medium While fusogen the of I acidification 2011). class an cases model al., some to in et 2014), opposition Podbilewicz, direct changes Jones (in conformation pH-independant in is (reviewed further fusion envelope provokes process retroviral GLUT-1, triggering called the also completes 2003), and al., with interaction et Subsequent change. (Manel Ghez conformational NRP-1; first SLC2A1 (Neuropilin-1, a receptor through first going a probably with 2006), interacts al., then et cell, the of change surface conformation the proteoglycans final Doranz sulfate at heparan and present 1996; to attaches second al., first the virus et HTLV-1, the inducing In Choe fusion. 1996), for 1996; trigger al., as al., et necessary co-receptor et (Feng CXCR4 the Alkhatib or with 1996; 1996) al., interacts al., et then et Dragic that 1996; domain inducing al., SU 1984), et an al., (Deng freeing et CCR5 and Dalgleish conformation 1984; in al., shift et first (Klatzmann a CD4 are to proteins binds several with first interactions HIV-1 specific necessary. example, for HTLV-1, and HIV-1 both In triggers. most do 2017). as al., molecules, et anchored Greenwood GPI in reviewed or 24A, single-pass (Fig. use retroviruses TM latter beta-type The 2017). (reviewed al., strains highly ALV however most et a of Greenwood are exception 24A), in the that with (Fig. levels, proteins structure proteins and transmembrane sequence (SLC) the pass on carrier divergent multiple presenting solute related Retroviruses cellular functionally fusion. for with of trigger family interact a Commonly usually as retrovirus. TM sufficient the gamma-type be on to depending a assumed partners is cellular interaction more single or one a with interact to triggers. has additional SU and receptor the of 1998). Nature al., et (Chen being trimerization HA change by formed major cavity only sheltered major a the in a peptide interaction, in fusion conformation subunit the of its of burial change effect not stabilizing does precursor the the demonstrating of way, cleavage Studies that 2015). shown Harrison, have the in HA peptide allow (reviewed on fusion to to fusion TM the productive from necessary for of dissociate to is necessary inclusion needs changes prevent TM SU Conversely, conformational to to. and anchored and six is SU state TM the membrane pre-fusion of form the stable into to interaction less interacting The the are in 2015). domains TM Harrison, HR maintain two in TM the (reviewed the which bundle of in conformation helix state, minimum post-fusion energy free its new is the subunit as dimer metastable a yields however, nsm ae,piigo h neoefrfso sabtmr ope n ed several needs and complex more bit a is fusion for envelope the of priming cases, some In orlaeT rmismtsal form, metastable its from TM release To ..Rtoia envelopes Retroviral 3.1. h L15in- SLC1A5 The 87

3.1 3. Syncytins

A receptor virus

multiple pass SLC20A1 KoRV-A, GALV, FeLV-B, FeLV-T, WMV, 10A1-MLV SLC19A2 KoRV-A, FeLV-A SLC20A2 A-MLV, 10A1-MLV, GALV, FeLV-B SLC7A1 E-MLV, BLV SLC5A3 M813, HEMV SLC1A5 RD-114, BaEV, REV, SRV, MPMV, Syncytin-1, Syncytin-Ory1 SLC1A4 BaEV, Syncytin-1 SLC49A1,SLC49A2 FeLV-C SLC53A1/XPR1 X-MLV, P-MLV SLC5A1 PERV-A SL19A1 GLN, MLV SLC9A1 ALV-J SLC2A1 HTLV MFSD2A Syncytin-2 CXCR4 (co) FIV, HIV CCR5 (co) HIV single pass TFRC MMTV TVA ALV-A TVB ALV-B, -D, -E TVC ALV-C CD4 HIV CD134 FIV NRP-1 (co) HTLV GPI-anchored HYAL2 JSRV Ly6E Syncytin-A EFNA1-5 IAPE

Syncytin-1 MALP 13 TLTAPPPC-RCMTS 37 MPRNCYHSATLCMHANTHYWTGKM 1 B BaEV MGFT 26 QKRYGRPC-DCSGG 54 VHSSCYTSYQQCRSGNKTYYTATL 21 SRV-1 MNFN 30 QQKHGKPC-DCAGG 57 VHATCYNHYQQCTIGNKTYLTATM 24 SRV-2 MTVK 30 HQQHGKPC-DCAGG 54 VHSSCYTTYQECFFGNKTYYTAIL 24 MPMV MNFN 30 QQKHGKPC-DCAGG 57 VHASCYNHYQQCNIGNKTYLTATI 24 SRV-4 MNFE 30 QNKHGKPC-ECKGG 57 VHSTCYSSYQQCTIGNKTYFTATI 24 SMRV MLCI 25 TALFGAPC-DCKGG 68 MHSSCYSSFSQCTQGNNTYFTAIL 22 SNV MDCL 35 QQLWGLPC-DCSGG 70 MHSTCYEKTQECTLLGKTYFTAIL 16 Syncytin-Ory1 MALS 31 QQHH--PC-TCSGG 65 VHSSCYTQYQTCNKGNQR LLFAVI 18 Dasy-Env1.1 MHSF 29 KTLYGTPC-ECRGG 65 MHVHCYSEVQSCTRNNKTYYTAIL 25

3.1 TvERV(D) MIST 23 HALFGAPC-DCKGG 64 LHSLCYTSTQQCTGKSGTYLTSRQ 24 REV-A MDCL 35 QQLWGLPC-DCSGG 68 MHSTCYEKAQECTLMGKTYFTAIL 16 DrERV MFSQ 27 DALFGPPC-DCQGG 75 HSACCYYKVYTCSKGGQTYFTARI 28 RD114 MKLP 26 QKQHGKPC-ECSGG 55 MHSSCYTEYRQCRANNKTYYTATL 21 Syncytin-1 NPSCP-GGLGVTVCWTYFTQTGMSDGGGVQDQAREKHVKEVI 180 RNKR BaEV QSPC-NGIKGQSICWSTTAPIHVSDGGGPLDTTRIKSVQRKL 189 RPKR SRV-1 IAGCPENKKGQVVCWNSQPSVHMSDGGGPQDKVREIIVNKKF 197 RPKR SRV-2 SAGC-TGTVGQHICWNPKAPVHISDGGGPQDKAREIAVQKRL 188 RARR MPMV TAGCPNGKKGQVVCWNSRPSVHISDGGGPQDKARDIIVNKKF 196 KAKR SRV-4 SAGCPKDELGKTACWSATPPVHVSDGGGPQDKVREILVQKKF 192 RTKR SMRV QAGC-DGTVGKSVCWNQQAPIHVSDGGGPQDAVRELYVQKQI 185 RQKR SNV QASC-TGTVGKPVCWDPVAPVYVSDGGGPTDMIREESVRERL 191 RHKR Syncytin-Ory1 SPSCP--TKGQVACWYPIAPVGVSDGGGPQDTL TKLKTQKVL 186 RQKR Dasy-Env1.1 QAGC-YGTIGDTVCWNKIPPAHISDGGGPQDQARQIIVQERM 184 RSKR TvERV(D) QASCDKINIGKNVCWSLHAPIHVSDGGGPTDQIREMEVKERV 189 RPKR REV-A QASC-TGIIGKPVCWDPAAPVYVSDGGGPTDMIREESVRERL 191 RHKR DrERV PDKCPG-EIGKPACWKLDSPLG-SDGGGPQDKRRTERVRIQI 179 RHKR RD114 QSPC-RGSINQPVCWSATAPIHISDGGGPLDTKRVWTVQKRL 190 RPKR

Figure 24 – Examples of the nature of known retroviral receptors and example of the SLC1A5 RBD (A) Known receptors of exogenous and endogenous retroviral Env, sorted by receptor type. A small scheme illustrates the types of receptor. Gamma-type Env are in black, avian gamma-type in gray, beta-type env in green and syncytin Env in blue. (B) Alignment of the SU sequences of representatives of the SLC1A5 interference group, highlighting the conserved cysteine containing domains in yellow and the major SDGGGXXDXXR motif in orange. The number indicate distance between the residues and mismatched residues in the conserved domains are highlighted in red. Adapted from Malicorne et al. (2016). 88 rmS-Mt nitaS odbtentetocsenso h XCmtf(alne al., et (Wallin motif CXXC the it of 2008). shifting cysteines al., i.e., two et bond, the Li the between 2004; of al., bond isomerization et intra-SU to (Wallin an leads bond SU-TM to thiol SU-TM the this from in of carried Activation implicated group 2008). not thiol al., is a et that of It Li motif activation 2004; CXXC TM. to SU and due the SU is of this cysteine linking that the HTLV-1 bond by and disulfide MLV both the in of shown case rupture been the implies has In release part 23). (Fig. in this conformation Env, least pre-fusion gamma-type at TM Following of dissociates the 2017). stabilizing it longer al., and of no et changes surface therefore SU (Riedel TM, the from the cavity at of the homo-trimers conformation into hollow the tucked forms interaction, peptide Env SU/receptor fusion form. the unstable with an membrane in the TM the keeps SU the fusion membrane SU/TM. induced Metastable I class protein fusion viral of Model b. rodents. other all in absent presence is the was that to site (2003) which N-glycosylation al. SLC1A4 third et hamster Marin a by just of the linked beyond was by which exemplified BaEV, specificity and further Syncytin-1 receptor for be even of can unusable This notion sequence. the acid extending most amino to group, the resistance interference its for the apparent responsable were of became glycosylations It N-linked members 2000). SLC1A4, al., of et case (Marin the envelopes in these that for as (RDR2, receptor SLC1A4 (such alternative of identification an group the as to interference ASCT1) led SLC1A5 cells, rodent the infect still of could BaEV) members and Syncytin-1 some Env/receptor block that completely stretch observation to short sufficient The is a insertions) interaction. in aa variability 10 sequence includes the also that (which suggest SLC1A5 results of These 2003). al., restore/abrogate replacing completely et to and (Marin able was loop, fusion species extracellular other an the of in that sequence with SLC1A5 sequence aa mouse corresponding 21 the and hypervariable human a of of study presence comparative the A revealed Env. group SLC1A5 most using infection Env using SLC1A5 sequence. gamma-type RT a betaretroviral a which with in apparent associated Syncytin-Ory1, clearly is and are MMTV history envelope’s of examples this the in involvedwith events be Recombination may in 2017). (reviewed it Johnson, recombinations that genome & implying viral Sinha example), after for events to 2015, transmission tropism al., cross-species species numerous et broad Miyaho in very in a confers illustrated receptor (as as virus SLC1A5 the of SDGGGX Use an 2006). especially al., et domains, Cheynet containing 24B, except cysteine similarity conserved sequence As few poor 2017). show a Johnson, group for interference & this Sinha receptor of in the members (reviewed gave the RDR above, that receptor, mentioned D-type virus and RD114 RD114 feline name: the MMTV, original and Syncytin-Ory1, its Syncytin-1, MPMV including (BaEV), fusion, virus for endogenous receptor this baboon use to able be to gested oetclsfr necpint hsbodseistoim en ihyrssatto resistant highly being tropism, species broad this to exception an form cells Rodent h eatbeS/Mcnb iee oasrn-oddsaei which in state spring-loaded a to likened be can SU/TM metastable The ..Rtoia envelopes Retroviral 3.1. 2 DX 2 oi (Fig. motif R 89

3.1 3. Syncytins

Extended intermediate. Release of the now unstable TM subunit leads to its extension, during which the fusion peptides of each unit of the trimer are inserted in the opposite cell membrane, due to their hydrophobic properties (Fig. 23). While there is no direct evidence for the existence of this structure, it is extremely probable (reviewed in Harrison, 2015). In a recent study on MLV, Riedel et al. (2017) show that the length bridged by TM between the viral and cellular membrane is about 120 Å, significantly longer than the 85 Å height of pre-fusion Env.

Six-helix bundle. The intermediate extended structure in which TM is inserted in both mem- branes still does not represent the minimum energy-free state of the system, to reach this

conformation the TM slowly fold in onto themselves, the region around the CXnC1-2 domain serving as hinge (Fig. 23, reviewed in Harrison, 2015). This folding brings the fusion peptide closer to the transmembrane region and draws the viral and cellular membranes closer, until they form a hemifusion stalk and finally a fusion pore (reviewed in Podbilewicz, 2014; Harrison, 2015). Achievement of the fusion pore marks the end of TM refolding, the fusion peptide and transmembrane domains are next to each other and HR1 and HR2 form the stable six helix bun- dle, the free-energy minimum state of TM (Fig. 23, reviewed in Harrison, 2015). Interestingly, this conformational change represents an inversion of TM organization, as in its pre-fusion state the C-terminal part was in the center of the trimer, while in the post-fusion state the N-terminal HR1 form the inner coiled-coil. Such a major rearrangement implies a probable dissociation and re-association of the trimer, but no transitory monomeric form has been observed so far (reviewed in Podbilewicz, 2014).

3.2 Syncytins in mammals

Bridging the gap between the two previous chapters, syncytins are a series of ERV envelope 3.2 genes that have been exapted in different mammalian taxa for a role in placentation (Fig. 25. A syncytin is a captured retroviral envelope that is i) expressed specifically (or at the least preferentially) in the placenta, ii) conserved over the course of evolution, and iii) fusogenic. In this first part I will present a comprehensive list of described syncytins and their particularities and relationship to fused placental cells. As often in biology, the human and mouse counterparts have been studied more extensively and as such will be described in more details.

3.2.1 Human syncytins: Syncytin-1 and Syncytin-2

In humans, two syncytins have been described and named Syncytin-1 and -2. They are exapted env genes from an HERV-W and an HERV-FRD copy, respectively (Blond et al., 2000; Mi et al., 2000; Blaise et al., 2003).

90 rn ftebac.Bac eghi rprinlt ie e cl ttebto dsRi ta. 2012). al., et Reis (dos bottom the at in scale indicated see name time, their to and proportional added exaptation is of been length age have Branch approximate (green) branch. their genes the to of syncytin-like corresponding front selected indicated arrow and is their (purple) tree, placenta syncytins the mammalian known to shared All the arrow. of red emergence ac- a of by positioned age genes supposed syncytin-like The and monotremes. oviparous syncytins exaptation. known of with age mammals to cording of Phylogeny – 25 Figure PLACENTA MAMMALIAN

200 Jurassic

150 Cretaceous 100 50

aml oti h iiaosetein n aspasa ela the as well as marsupials and eutherians viviparous the contain Mammals Tertiary 0 Mya Monotremata Diprotodontia Proboscidae Tenrecidae Bradypodidae Bradypodidae Dasypodidae Ruminantia Perissodactyla Carnivora Insectivora Lagomorpha Rodentia Strepsirrhini Haplorrhini Didelphimorphia Syn-Opo1 Syn-Car1 Fematrin-1Syn-Rum1, Syn-1, -2, Syn-A,-B,-Mar1, Syn-Ory1 Syn-Ten1 HEMO, EnvV2, ERV3 pan-Mars-2 Env-Cav1 ..Snyisi mammals in Syncytins 3.2. marsupials mammals eutherian 91

3.2 3. Syncytins

a. Syncytin-1

Discovery. The screening of a placental cDNA library by Blond et al. (1999) led to the first description of the HERV-W group and the single conserved env ORF with placental expression belonging to it. The next year, both Blond et al. (2000) and Mi et al. (2000) confirm that this Env is indeed expressed in the human placenta and show that expression occurs in the ST. Additionally, they show that transfection of the Env induces formation of large syncytia, demonstrating fusogenic activity. Blond et al. (2000) also identify SLC1A5 as the major cognate receptor of Syncytin-1, with SLC1A4 as alternative receptor (Lavillette et al., 2002). On the basis of these results, Mi et al. (2000) call this protein Syncytin, in reference to its probable role in establishing syncytia in vivo. It will be renamed to Syncytin-1 upon the discovery of Syncytin-2 by Blaise et al. (2003) a few years later. The expression profile of Syncytin-1 was later expanded to include all subpopulations of villous (ST and CT) and extravillous trophoblast cells (Malassiné et al., 2005; Muir et al., 2006). Subsequent studies revealed that the locus and sequence of syncytin-1 as well as the LTR driving its expression are conserved not only among humans, but also other Hominoidea (apes), displaying purifying selection (Mallet et al., 2004; Bonnaud et al., 2004). Interestingly, when compared to a consensus sequence of non-conserved HERV-W Env, Syncytin-1 displays a 4 aa deletion in the cytoplasmic tail which is crucial for its fusogenic activity and also observed in the other investigated Hominoidea species, suggesting it represents an early adaptation to its new role (Bonnaud et al., 2004). The insertion of the Syncytin-1 encoding provirus occurred about 40 Mya (Fig. 25), before separation of Hominoidae from old world monkeys, but the gene was lost in the latter (Bonnaud et al., 2005, see 3.3.1.a. for discussion as to why Syncytin-1 was lost in old world monkeys).

Syncytin-1 and cell-cell fusion. This conservation for over 25 My in Hominoidea strongly suggested that Syncytin-1 plays an important physiological role in these species. Due to its nature

3.2 as an ERV envelope and its expression in a fused cell layer of the placenta, the first proposed role for Syncytin-1 was the establishment of the ST in the placenta (Blond et al., 2000; Mi et al., 2000). Accordingly, suppression of Syncytin-1 expression in primary cytotrophoblasts ex vivo was shown to inhibit cell-cell fusion (Frendo et al., 2003). Investigation of the expression profile of the cognate receptor SLC1A5 in the placenta revealed that it was specifically expressed in the CT layer, but somewhat surprisingly not in the ST (Hayward et al., 2007). The absence of expression in the fused layer might be due to a down-regulation of receptor expression post- fusion, as is observed in trophoblast derived cell lines (Hayward et al., 2007). Hayward et al. (2007) also show that the expression of the receptor is more or less constant during pregnancy, raising the question as to why and how the CT layer persists as individualized cells up to term instead of fusing with the ST. A series of later studies suggested that, in vivo, the CT cells needs to become fusion competent, implicating a number of intracellular changes (reviewed in Gerbaud & Pidoux, 2015). It seems that the fusion competency of CT can be regulated both

92 uto fGM ciiy n hsSnyi- xrsin a enpooe sa explanation an as 2016). proposed al., been et has expression, (Xu Syncytin-1 thus (Strick and activity, expression TGF GCMa A of Syncytin-1 2007). duction al., activate et TGF (Strick syncytialization to of reduce expression and hormones the proliferation to cell steroid augment leads also bind however pathway HERV-W to This 5’ 2007). able the al., In site et 2009). a al., is et (Baczyk itself proliferation LTR cell increase to al., and et differentiation Bolze EVT in reviewed expression a, GCMa homolog of missing Inhibition cells factor glial 2017). name transcription original "chorion-specific its for from sites (GCMa, binding contributes GCMa" presents which and sequence regulation regulatory ERV-derived transcriptory the im- another to of an is upstream suggesting LTR Immediately and 5’ 2009). gestation, al., associated of et Syncytin-1 end (Gimenez expression the of at regulation apoptosis temporal trophoblast portant of decrease observed increase an and with fusion accordance of in the term, during al., towards unmethylated methylated et is but (Matoušková promoter pregnancy, tissues of Syncytin-1 beginning placental the Interestingly, non state 2009). all methylation al., nearly et CpG Gimenez in the 2006; methylated on strongly promoter, dependent other 5’ highly many the for is of As Syncytin-1 2017). al., of et expression Bolze 2014; sequences, al., ERV et Huang in (reviewed factors modifications epigenetic transcription both and through regulated tightly be to needs expression Syncytin-1 tation, expression. Syncytin-1 of Regulation 3.3.3.c.). also placental see of 2017, al., number et a Bolze in in observed cell-cell (reviewed expression just pre-eclampsia abnormal as than such the al., processes pathologies, in et other implicated (Holder in be SU Syncytin-1 also of of region might implication a fusion The by and 2012). part al., responses, in et least immune Tolosa at modulate 2012; mediated to is phenomenon able this was surprisingly exosomes Syncytin- that that derived shown been trophoblast has it on Still, expression 22A, 2007). al., 1 (Fig. et immunosuppression (Mangeney display ISD mice not in conserved does assays it rejection a that tumor showed in study presents later a Syncytin-1 though 2003), 2017). al., et al., Parseval anti-apoptotic de et and proliferation Bolze example as in well for as have (reviewed differentiation studies processes cell Several in fusion. Syncytin-1 cell-cell for al., just role et a than (Malassiné confirmed roles EVT potential the especially other 2006), at al., hints et 2005), (Muir expression tissues its and trophoblastic 2003), unfused al., et other (Frendo in fusion to addition in latter, the of differentiation inhibited placentas. human in Syncytin-1 of roles Other Pidoux, & Gerbaud in (reviewed ST the to closer 2015). in CT increase the an bringing implies fusion probably to adhesion, commitment cell-cell that and secretions ST by negatively and positively xvivo ex u oisdvriyo ucin uighmnplacen- human during functions of diversity its to Due a ensont eraesnyilzto and syncytialization decrease to shown been has yctn1kokdw npiayC cells CT primary in knock-down Syncytin-1 ..Snyisi mammals in Syncytins 3.2. β hs ffcsse to seem effects whose β eitdre- mediated 93

3.2 3. Syncytins

b. Syncytin-2

Discovery. Following identification of Syncytin-1, a genome-wide screening by Blaise et al. (2003) for new fusogenic human ERV Env revealed the existence of Syncytin-2. The same study shows that Syncytin-2 transfection is able to induce syncytialization ex vivo, is expressed specifically in the placenta and that it is highly conserved and fusogenic in all primates at the same locus, suggesting it inserted over 40 Mya (Fig. 25) and has been conserved since under purifying selection (Blaise et al., 2003). Using the GeneBridge4 human/Chinese hamster radiation hybrid panel (Gyapay et al., 1996), Esnault et al. (2008) were able to identify the multiple-pass major facilitator superfamily domain containing 2 A (MFSD2A) carbohydrate transporter as the cognate receptor necessary for Syncytin-2 mediated fusion.

Syncytin-2 and in-fusion. In accordance with a role of Syncytin-2 in CT fusion and ST establishment in the human placenta, siRNA mediated knockdown of both Syncytin-2 or its receptor significantly reduced the fusion of primary trophoblastic cells and trophoblast-derived cells lines (Vargas et al., 2009; Toufaily et al., 2013). Interestingly, the effect of Syncytin- 2 knockdown on cell-cell fusion is much more important than that observed for Syncytin-1 knockdown, suggesting that the former might be the principal mediator of cell-cell fusion in vivo (Vargas et al., 2009). The expression profile of both Syncytin-2 and MFSD2A strongly differ from those of Syncytin-1 and SLC1A5. While Syncytin-1 expression occurs in a more or less constant fashion in all trophoblast cells, Syncytin-2 is expressed transiently at the ST facing membrane of some CT only, suggesting a much tighter control of its spatio-temporal expression (Malassiné et al., 2007). Conversely, while SLC1A5 was expressed in CT but not ST cells, MFSD2A is expressed exclusively in the ST (Esnault et al., 2008). With regards to these expression profiles, Esnault et al. (2008) propose that Syncytin-2 mediated fusion occurs through "in-fusion" in which CT can only fuse into the ST since it is the only layer expressing the cognate receptor. This model also explains why the CT cannot form a second fused layer, as

3.2 the absence of MFSD2A at their surface prevents Syncytin-2 mediated CT-CT fusion.

Other roles of Syncytin-2 in human placentas. Syncytin-2 presents a conserved gamma-type ISD (Fig. 22, de Parseval et al., 2003) and is indeed immunosuppressive in in vivo mouse tumor rejection assays (Mangeney et al., 2007). Its very localized expression profile might indicate that Syncytin-2 has only limited opportunities to be presented to the maternal immune system, but is has been shown that it is also present in trophoblast derived exosomes, alongside Syncytin-1 (Vargas et al., 2014). Similarly to Syncytin-1, Syncytin-2 expression is abnormal in a number of placental pathologies such as pre-eclampsia (reviewed in Bolze et al., 2017).

Regulation of Syncytin-2 expression. While epigenetic controls seem to play a role in restrict- ing expression of Syncytin-2 to the placenta, there is no temporal variation in CpG methylation of the placental LTR as was observed for Syncytin-1, it remains unmethylated from beginning

94 hl h entv ua lcnai eooohra ihafsdcl ae overlying placenta layer hemotrichorial a cell present fused all Muroidae a layer, unfused with an hemomonochorial of remains is discontinuous placenta the human definitive the While layers segregated 2 in Expression b. far. so described classical been otherwise not an has for Syncytin-B receptor of unusual receptor an cognate protein, The envelope. pass the gamma-type multiple showing a by not confirmed and GPI-anchored further a was this formation placenta 2017), for lymphocyte Ly6E al., the of et be importance (Bacquin to assay shown protein this was (Ly6E) Syncytin-A in 6E of properties antigen receptor immunosuppressive the Recently displays 2007). -1 al., not et (Mangeney but only 2005), Syncytin-2 al., where both et situation humans While the Dupressoir to in age parallel Muroidea. 22, remarkable the a (Fig. of assays, rejection sequence back radiation tumor in ISD immunosuppressive the pushing is conserved Syncytin-B during rats), a 25), present mole (Fig. -B (blind Mya and (2014) Syncytin-A Spalacidae al. 40 et about in Vernochet to conserved by vole) endogenization study also (hamster, of subsequent was a Cricetidae is Mya, and 20 that rat) of revealed (mouse, time integration Muridae estimated in an -B with reported only and study Syncytin-A original the of While endogenization 2005). conservation al., distinct et from (Dupressoir deriving retroviruses different syncytins, by human events described previously the to to unrelated able and placenta the fusion in cell-cell expressed mediate specifically are that envelopes an ERV gamma-type through are (2005) genes al. the et in Dupressoir discovered were by genes study syncytin same rodent both syncytins, human of identification Following Discovery Syncytin-B a. and Syncytin-A syncytins: Mouse 3.2.2 factors additional the of one be to 2015). proposed al., cognate been et (Toufaily the has expression and JunD Syncytin-3 genes Recently regulating syncytin them. two of role induce one major to of able a receptor being plays establishment, factor ST transcription placenta this human that in distinct emerges very it present spatio- Nevertheless, their MFSD2A regulating expression. in and temporal role -2 important an and play factors that Syncytin-1 other consider that GCMa, suggesting to patterns, by expression interesting inducible is all It 2010). are al., they et while (Liang promoter Syncytin-2 the of demethylation syncytialization provoking receptor its vivo and ex Syncytin-2 of expression of ectopic expression by induce Ectopic regulated can Syncytin-1. is GCMa of expression al. MFSD2A regulating et factor and Liang transcription Syncytin-2 level, same factor both the transcription GCMa, of the expression On intriguingly 2010). that al., revealed et Liang (2010) 2009; al., et (Gimenez term to Lage l,21) diinly tsesta Capasa motn oei the in role important an plays GCMa that seems it Additionally, 2010). al., et (Liang xvivo ex Dpesi ta. 05.Itrsigy ohevlp ee are genes envelope both Interestingly, 2005). al., et (Dupressoir nvivo in Lnfr ta. 08.Itrsigy yEis Ly6E Interestingly, 2018). al., et (Langford nsilico in ceno h os eoe Both genome. mouse the of screen ..Snyisi mammals in Syncytins 3.2. 95

3.2 3. Syncytins

presenting, in most cases, two internal fused cell layers and an external continuous but loose and fenestrated non-fused layer (see 1.2.2.d. and Vernochet et al., 2014). The presence of two fused layers and two distinct syncytins with different cognate receptors raised the question of whether each syncytin was responsible for the establishment of one ST layer. This hypothe- sis was confirmed by in situ hybridization (Simmons et al., 2008) and through two separate electron microscopy studies showing that knockout of syncytin-A disrupted establishment of ST-I and knockout of syncytin-B did the same for ST-II (Dupressoir et al., 2009, 2011). Sim- ilarly, the knockout of Ly6E prevents syncytialization of ST-I but not ST-II (Langford et al., 2018). It is therefore likely that the independent capture and exaptation of two distinct ERV Env with a distinct receptor preference was instrumental in the establishment of the very particular hemotrichorial placental phenotype observed in mouse and other Muroidea. Interestingly, GCMa has been shown to be necessary in mice for correct establishment of placental structures (Schreiber et al., 2000), and the loci expressing Syncytin-A and -B present several putative GCMa binding sites (Dupressoir et al., 2005). A link between GCMa and expression of Syncytin-B is also suggested by the observation of their co-expression in the same placental cell layer, with GCMa expression preceding that of Syncytin-B (Simmons et al., 2008). Another study reports that Syncytin-A expression might also be under control of GCMa (Schubert et al., 2008) and while this would be in accordance with the fact that GCMa deprived mice display embryonic lethality at the same date as mice knocked out for both syncytin-A and -B (but earlier than in single knockouts, Fig. 26, Schreiber et al., 2000; Dupressoir et al., 2011), no GCMa expression in the ST-I progenitor cells has been reported by Simmons et al. (2008). In any case, the importance of GCMa for syncytin expression and placental morphology in mice is a striking parallel to the case of human syncytins and a noteworthy example of evolutive convergence.

c. Impact of Syncytin knockout on placental physiology

3.2 The knockout experiments discussed above are of immense interest to the study of syncytins, as they are the most conclusive proof of their placental role available today. In other species, such experiments are either inconceivable (humans) or entirely less feasible for technical reasons (other mammals), but in mice it is possible to specifically knockout the syncytin genes and observe the direct effect on placental morphology. Interestingly, knockout of syncytin-A is lethal, all embryos dying at about mid-gestation, while in the case of syncytin-B viable embryos are born, albeit in smaller numbers and with moderate growth retardation (Dupressoir et al., 2009, 2011). The lessened impact of syncytin-B inactivation might be at least partly explained through the observed induction of gap junction beta-6 protein expression, a protein that is not present in the mouse placenta under normal conditions (Dupressoir et al., 2011). Expression of this protein could lead to the establishment of communicating gap junctions, both between cells of the unfused ST-II or with the fused ST-I layer, thereby restoring a syncytium-like state in which there is a certain cytoplasmic continuity across the layer (Dupressoir et al., 2011).

96 ncotpaets aenlbodsae r nagdwieftlvsesaeuaetd dpe from Adapted unaffected. are are 2011). diameter vessels (2009, fetal vessel al. while fetal et enlarged Dupressoir maternal are and spaces indicate spaces blood maternal blood Arrows placentas, maternal placenta. knockout Both mouse vessels. of in fetal sections reduced arrowheads stained and (right) spaces blood thionine or show (left) Arrows cell. eosin-saffron endothelial of an case the ec in and 1 cells 1.2.2.d.) II see ST layer, unfused trophoblast between desmosomes non-fused outer trophoblastic sinusoidal (the identifies stgc layer, cells ST of the giant identify II KO and I ST. placenta. in disrupted the results mouse showing micrographs of This the case individualized. in the remains layers in which ST B layer, embryos ST all the of separate of death a establishment the of on formation knockout prevents syncytin syncytin each of impact the of of Impact – 26 Figure B A o ipiiy h xenlfnsrtdufsdtohbatlyri o ersne B o)Electron top) (B, represented not is layer trophoblast unfused fenestrated external the simplicity, For . µ Mother Syncytin-A ntergt bto)Ipc fsnyi ncoto h aenlbodsae.Hematoxylin- spaces. blood maternal the on knockout syncytin of Impact (bottom) right. the on m ST-I syncytin-A WT Syncytin-A +/+ Syncytin-B ST-II ncotpaets ssonb h qezdrdbodcls In cells. blood red squeezed the by shown as placentas, knockout stgc

Syncytin-A -/- syncytin-A

I Fetus I Syncytin-A -/- or II

-B Mother syncytin-A ncoto os placenta. mouse on knockout Syncytin-A -/- I embryo death growth retardation growth death embryo ri aegot eadto ntecs of case the in retardation growth late in or Syncytin-B ST-II Syncytin-B +/+ syncytin-B

Syncytin-B -/- ncot cl a:2 bar: Scale knockout. Fetus

A ceai representation Schematic (A) Mother yctnAII Syncytin-A ..Snyisi mammals in Syncytins 3.2. Syncytin-B -/- ST-I Syncytin-B -/- µ nteleft, the on m syncytin-B

syncytin- 97 Fetus

3.2 3. Syncytins

A different impact depending on the inactivated syncytin is also observed at the morpholog- ical level. While knockout of each syncytin disrupts the syncytialization of the corresponding layer, they have more far-reaching impacts on other placental tissues (Dupressoir et al., 2009, 2011). Syncytin-A knockout leads to an over-proliferation of the individualized ST-I cells, signif- icantly reducing the volume of maternal blood spaces and probably seriously impacting blood flow and blood supply to the placenta (Dupressoir et al., 2009). Syncytin-B knockout on the other hand leads to an over-expansion of maternal blood spaces, once again impacting the volume and flow speed of placental blood supply, but in this case there is no trophoblast over-proliferation (Dupressoir et al., 2011). A double knockout for both syncytin genes displays complete embry- onic lethality a few days earlier than the syncytin-A knockout, suggesting a cumulative effect of both inactivations (Dupressoir et al., 2011).

3.2.3 Other mammalian syncytins

In the years since the discovery of human and mice syncytins, the investigation of the implication of ERV Env in other mammalian placentas led to rapid identification of new syncytins in many mammalian clades. All presently described syncytins will be briefly presented in this section, classified by the species in which it was first described. Of note, they all correspond to distinct endogenization events that are specific to the clade in question and implicate a variety of viral species (reviewed in Lavialle et al., 2013; Denner, 2016). Unless mentioned otherwise, all these syncytins are specifically expressed in cells of the maternofetal interface and are fusogenic ex vivo.

a. Rabbit: Syncytin-Ory1 3.2 Syncytin-Ory1 was identified in the rabbit (Oryctolagus cuniculus) by Heidmann et al. (2009). It seems to be conserved among Leporidae (rabbits and hares) but not Ochotonidae (pikas), which put the time of insertion around 12 Mya at the time of publication (Heidmann et al., 2009), a date that has been pushed back to about 20 Mya (Fig. 25) according to the TimeTree database (Kumar et al., 2017). Syncytin-Ory1 is a gamma-type Env of the SLC1A5 interference group and therefore uses the same receptor as Syncytin-1. Interestingly, the envelope is close to that of MPMV (reviewed in Sinha & Johnson, 2017), and the syncytin-Ory1 associated RT is also close to that of MPMV as well as other betaretroviral RTs (Malicorne et al., 2016). This suggests that the ancestral retrovirus was a recombinant betaretrovirus that acquired a gamma-type SLC1A5 interference group env, a somewhat remarkable happening as the immense majority of syncytins seem to originate from gammaretroviruses. In this case the use of SLC1A5 as receptor is again linked to hemochorial placentation with a single fused cell layer (see 1.2.2.d.).

98 u1a H70 h atrbigmr uoei talwrp f50(onlse l,2013; al., et (Cornelis 5.0 of pH lower a at fusogenic more being latter the 7.0, pH an at to Rum1 leading 2013) to al., according et 25) Nakaya (Fig. activity fusogenic Mya 2013; 20 al., of et time (Cornelis insertion (sheep) estimated Caprinae in not the and 2013). (cattle) al., linking et potentially (Cornelis placenta and the 1.2.2.b.) of structure (see the placenta with syncytins of sheep of levels level the expression expression in the amount higher observed Interestingly, the with syncytialization 2013). correlating cow, al., of in than et sheep in Cornelis higher se- significantly 25, are purifying Syncytin-Rum1 (Fig. by Mya followed 45 integration about suggesting lection species, mouse-deer) al., and et chevrotain (Cornelis except gammaretrovirus distinct but related a 2013). of infection the from resulting tentially cell trinucleate fetomaternal confusingly somewhat it (Fematrin-1). renamed 1 who inducer later the (2013), months and al. few Syncytin-Rum1 et a cows: investigated Nakaya in also by Env was Cornelis ERV latter two 1.2.2.b.). The BERV-K1/Bos-Env4. of (see identified expression previously interface placental materno-fetal describe (2013) the al. which of et placentation syncytialization synepitheliochorial of limited style only rare displays and particular very the display Fematrin-1Ruminants and Syncytin-Rum1 sheep: decaying and Cattle a as present d. only 2013). is al., et and (Esnault evolution selection purifying of some course displays EnvV2 still the clades, that other over syncytin in Syncytin-1 that seems by It 2013). replaced al., was et (Esnault monkeys World Old in only syncytin 25, simians, more of (Fig. in ancestor Mya discussed common the 43 be in will endogenized about was but gene completeness, envelope ERV for This here 3.3.1.a. in mentioned details is EnvV2 of case curious The EnvV2 monkey: world Old 2012). c. al., et Cornelis 25 to (Fig. according syncytin known Mya oldest 75 carnivoran the Syncyin-Car1 tested about making 26 of and the 2017) date in insertion selection an purifying it under suggesting named conserved species, and and gene orthologous a is such and identify present origin to all able were to (2012) postulated & al. therefore et syncytin-Car1 Wynn were Cornelis 1957; and gene. Morton, 1982) syncytin Neaves, same placenta the & hemochorial Okelo a Oduor present 1964; which Amoroso, hyenas organization except placental 1.2.2.c.; general endotheliochorial (see common a present all dogs) (cats, Carnivorans Syncytin-Car1 dog: and Cat b. earn1i eatp n hti,cnrrl oSnyi-u1 osre nyi Bovinae in only conserved Syncytin-Rum1, to contrarily is, that Env beta-type a is Fematrin-1 po- Syncytin-Car1, with similarity limited displaying env gamma-type a is Syncytin-Rum1 Syncytin-Rum1 yctnCr sagmatp n xrse ya R fgammaretroviral of ERV an by expressed Env gamma-type a is Syncytin-Car1 . xvivo ex speeta notooospsto n1 etdPcr alruminants (all Pecora tested 16 in position orthologous an at present is TimeTree fFmti- saot25fl ihrta htosre ihSyncytin- with observed that than higher fold 2.5 about is Fematrin-1 of ua ta. 07,btcnevdalcaatrsiso a of characteristics all conserved but 2017), al., et Kumar , TimeTree ..Snyisi mammals in Syncytins 3.2. Kmre l,21) The 2017). al., et (Kumar TimeTree Kmre al., et (Kumar 99

3.2 3. Syncytins

Nakaya et al., 2013), the physiological relevance of which has not been established. Interestingly, a subsequent study showed that the fusogenic potential of Fematrin-1 in bovine trophoblast cell lines is also increased at low pH (Koshi et al., 2016). Taking into account the relatively recent and Bovinae-specific acquisition of Fematrin-1, Nakaya et al. (2013) suggest that Syncytin-Rum1 might represent an ancestral syncytin that is currently being replaced by Fematrin-1 in Bovinae and endogenous JSRV Env in Caprinae (see also 3.3.1).

e. Squirrel and marmot: Syncytin-Mar1

While it had been previously shown that some rodents clades presented Syncytin-A and -B, the Sciuridae do not present these two genes. Taking advantage of the recent sequencing of the thirteen-lined ground squirrel (Ictidomys tridecemlineatus) genome, Redelsperger et al. (2014) were able to identify Syncytin-Mar1. This syncytin has been conserved in a coding state in all members of the Marmotini tribe (ground squirrels), but is present as a non-coding sequence in other Sciuridae (Redelsperger et al., 2014), suggesting an endogenization about 35 Mya and an exaptation only in the Marmotini which radiated approximately 27 Mya (Fig. 25, TimeTree, Kumar et al., 2017). This integration date would also mean that Syncytin-A and -B and Syncytin- Mar1 never co-existed in a common ancestor of Marmotini and Muroidea (Redelsperger et al., 2014).

f. Tenrec: Syncytin-Ten1

While at this point a number of syncytins had been described in Euarchontoglires (Syncytin-1, -2, -A, -B, -Ory1, -Mar1) and Laurasiatheria (Syncytin-Car1, -Rum1), no members of the Xenartha or Afrotheria had been shown to present these kind of genes. To investigate the latter group, Cornelis et al. (2014) screened the genome of the lesser hedgehog tenrec (Echinops telfairi) and identified Syncytin-Ten1. It is conserved at an orthologous position and under purifying selection in all Malagasy tenrec and probably present in the non-Malagasy subfamily (Cornelis 3.2 et al., 2014), suggesting an insertion date of 40 to 50 Mya (Fig. 25, TimeTree, Kumar et al., 2017).

g. Short-tailed opossum: Syncytin-Opo1

Marsupials present a very particular and short-lived placenta when compared to that of eutherian mammals. To try to investigate the potential role of syncytins in a marsupial placenta, Cornelis et al. (2015) searched the gray short-tailed opossum (Monodelphis domestica) genome in silico for the presence of coding ERV env and identified syncytin-Opo1. The expression profile of this syncytin is less placenta-specific that for the ones described previously, with non-negligible expression occurring in other tissues such as the spleen (Cornelis et al., 2015). Conservation of the envelope is limited to a subset of the Monodelphis genus, as it is present in a non-coding form in Monodelphis theresa and emiliae and absent in the sister group of Monodelphis (Cornelis

100 okdfraln ie(edane l,21) ti pcfial xrse ntepaet and placenta the in expressed specifically is It 2017). al., et (Heidmann time long a for looked example. HEMO. for has Syncytin-A Env unlike ERV3 essential, that not mean is not it does that proves this it While physiology, population. in function caucasian no the truncated of study early homozygous early 1% of about an presence in from the revealed alleles comes which placentation (1998) Heidmann in main & Env The Parseval ERV3 2007). de al., by of et role (Mangeney assays fundamental rejection a (Fig. mice of against ISD in expression argument an functional the containing be Env to to gamma-type shown contributes was a mentioned that is 22) Env as ERV3 which MFSD2A. an 2005), and Intriguingly, syncytins al., 2000). human et both al., activity Chang transcriptional et 2005; GCMa stimulate (Lin al., to et shown (PKA) later (Knerr A was activity intracellular kinase PKA in protein and cAMP increase of in increase an activity involves the ERV3 and of same levels effect the study, cAMP pro-differentiation follow-up a the In that mRNA. show (hCG) authors gonadotropin chorionic human of synthesis ERV3 and of overexpression that show al., (1999) et al. et Rivera Lin env Bustamante line, in cell (reviewed derived role trophoblast 25), physiological a (Fig. Using important Mya all 2018). an 40 in play about but at to humans endogenization proposed in of was only time it not its adrenal a conservation puts and which apparent in placenta gorillas), its in expression the to (except Due in its Catarrhini 2003). expressed of al., preferentially investigation et is Parseval and it (de 1995), that gland revealed al., al., tissues et et human (Kato (Venables RNA of level the range both protein on the shown was and placenta the 1987) in envelope fusogenic. truncated be cannot this it attachment, of membrane Expression its lacking is protein Env ERV3 Env-like Since secreted 1985). al., the a et yielding that (Cohen theoretically apparent domain, quickly anchoring became transmembrane the it before though 1984), al., et (O’Connell human in ERV3. expression placental with Env ERV Conserved very role, physiological a. a at hint syncytins. of that role selection fusogenic classical purifying the of from distinct signs probably with or conserved one envelopes lacking often these Nevertheless, envelopes discovered. are conserved been have of properties number syncytin canonical a the above, of several described syncytins the to addition In Env Syncytin-like 3.2.4 The 2015). al., et Cornelis 2015). 25, in al., (Fig. interface et only (Cornelis materno-fetal Mya unknown ( the 10 Mya about of 27 for status about conservation syncytialization acquisition, a recent with a 2017) syncytin al., this et makes This 2015). al., et nue eldffrnito,a eeldb eraei rlfrto,mrhlgclchanges morphological proliferation, in decrease a by revealed as differentiation, cell induces R3(locle EVR,wstetidEVdsoee ntehmngenome human the in discovered ERV third the was HERV-R), called (also ERV3 EOi el ecie osre iinERV simian conserved described newly a is HEMO ooepi theresa Monodelphis env ..Snyisi mammals in Syncytins 3.2. htwssmhwover- somehow was that , env R a truncated was ORF TimeTree and emiliae Kumar , 101 is

3.2 3. Syncytins

conserved in all simians (Fig. 25), though it entered the mammalian genome before separation of Laurasiatheria and Euarchontoglires (Heidmann et al., 2017), about 96 Mya (TimeTree, Kumar et al., 2017). HEMO is expressed in CT and EVT and does not disclose any fusogenic activity ex vivo, which is perhaps not surprising considering that the human envelope has lost its furin cleavage site, preventing correct maturation (Heidmann et al., 2017). Intriguingly HEMO has however acquired a new cleavage site shortly upstream of the transmembrane sequence (Heid- mann et al., 2017). Cleavage seems to occur under the action of metalloproteinases and leads to the release of a shed, still linked, SU/most of TM protein from the cell membrane, accordingly HEMO can be detected in the blood of pregnant women (Heidmann et al., 2017). This separation from the membrane anchoring domain is another reason HEMO might not be fusogenic. HEMO Env is also strongly expressed in stem cells, in a way reminiscent of HERV-H, and is related to the marsupial conserved Env pan-Mars-Env2 (Heidmann et al., 2017). It remains to be seen what role, if any, HEMO plays in primate physiology, but its high amount of sequence conservation and peculiar properties suggest a new, non-syncytin role.

b. Non-human syncytin-like genes

enJSRV. The case of endogenous copies of the still active JSRV virus (enJSRV) is complicated by their relatively young age. There have been several descriptions of enJSRV sequences that were exapted and positively selected during sheep domestication, such as the previously mentioned enJS56A1 locus gag which acts as a late restriction factor (Arnaud et al., 2007). While it is possible that enJSRV plays an important placental role by simply protecting the embryo from viral infections, other studies hint at a more complex role. Indeed, enJSRV env sequences from several loci are expressed in specific tissues in the placenta, in particular the BNC and the syncytial plaques (Dunlap et al., 2005). Additionally, inhibition of enJSRV Env in utero leads to reduction in trophoblast cell proliferation, preventing implantation and leading to embryo death (Dunlap et al., 2006). The implication of enJSRV Env in cell proliferation and differentiation is

3.2 intriguing as it has been shown that this protein possesses strong oncogenic properties (Wootton et al., 2005). Implication of enJSRV Env in cell-cell fusion events is less clear, as no copy able to induce cell-cell fusion ex vivo has been identified so far (reviewed in Nakaya & Miyazawa, 2015).

Env-Cav1. The search for non-Muroidea rodent syncytins led Vernochet et al. (2011) to identify Env-Cav1 in the genome of guinea pigs (Cavia procellus). This Env seems conserved in all Caviomorpha under purifying selection, suggesting an endogenization about 30 Mya (Fig. 25), and is expressed in the syncytial streamers, a fused and invasive placental tissue (Vernochet et al., 2011). Ex vivo fusogenicity assays disclosed no fusogenic potential of Env-Cav1, suggesting that its placental role is not that of a classical syncytin (Vernochet et al., 2011). The exact role of this envelope remains to be determined, but Vernochet et al. (2011) suggest that it might be implicated in placental invasion phenomena, drawing a parallel with the presence of Syncytin-1

102 el curdEVEvoe h oreo vlto Fg 5 eiwdi ailee l,2013). al., et Lavialle in reviewed 25, with ancestral (Fig. replacement evolution an its of by course of followed the over placenta, acquisition Env mammalian ERV the the acquired newly of is emergence discrepancy of time apparent the at this independently. syncytin even explain lineages and mentioned to above syncytins hypothesis the of working all endogenization in A after occurred emergence My that this 10-80 unlikely that last is unlikely the It more differentiation. in cell only as arose play such to phenomena placentation, of seem these aspects also case other genes in the syncytin-like role is and important (as syncytin an fusion 1.2.2.a.), cell-cell see without placentas, functional epitheliochorial entirely in be can placentas mammalian While clades, animal gene than all syncytin-like across less conserved is oldest The Syncytin-Car1 2012). syncytin, al., oldest et Env2 the (Cornelis My while 75 2012), at al., old et as about half Reis eutherians, dos and 25 marsupials (Fig. of Mya ancestor 175 common last The clade. the respective in their emerged in genes placenta these mammalian of some of acquisition recent relatively chief the problems, them conceptual among few a poses placentation mammalian in syncytins of involvement The replacement syncytin and syncytin Ancestral 3.3.1 evolution and Syncytins 3.3 membrane 2017). a al., et of Heidmann absence 2015; al., induced et codon (Cornelis latter stop the ways: for and HEMO domain different former both anchoring two that the using but for intriguing medium, is shedding extra-cellular It metalloproteinase the if 2017). into retrovirus, is al., released same are et it the pan-Mars-Env2 (Heidmann of analyses, and endogenization endogenization synteny from same reliable result the While least perform from 2017). at not to al., sequences related two et with the distantly (Heidmann related that too domain probable are are extracellular sequences the simians Env in two and the regions marsupials that conserved 2015). revealed al., of et later number (Cornelis years placentation a few marsupial a in conservation HEMO role strict major of its a fusogenic, Discovery plays be it to that unable implies is heavily pan-Mars-Env2 still al., that et means (Cornelis ERV3 this for While observed is 2015). what all to codon similarly in stop sequence, a transmembrane by site the truncated before least is orthologous just pan-Mars-Env2 at Interestingly, 2015). the endogenization al., its et at Cornelis since 25, present selection (Fig. Mya purifying is 80 under Env coding remained This has 2015). and al., marsupials et (Cornelis pan-Mars-Env2 of pan-Mars-Env2. cells. EVT on saot8 yod(onlse l,21) n h letsilcnevdEVenvelope ERV conserved still oldest the and 2015), al., et (Cornelis old My 80 about is h aesuyta ecie yctnOo lorvae h existence the revealed also Syncytin-Opo1 described that study same The percomORF ecie odt sol 1 yod(ez ta. 2016). al., et (Henzy old My 110 only is date to described , ..Snyisadevolution and Syncytins 3.3. pan-Mars- 103

3.3 3. Syncytins

V to W baton pass Syncytins FRD, V, W apes Syn-1, Syn-2 FRD, V, W

FRD, V, W FRD, V Old World monkeys Syn-2, EnvV2 V remains syncytin

FRD, V New World monkeys Syn-2

prosimians ?

60 40 20 0 Mya

Figure 27 – Compared evolutionary fates of HERV-W Env and EnvV2 in simians. Timed phylogenetic tree of major simian clades, with prosimians as an outgroup (TimeTree, Kumar et al., 2017).The presence of HERV-FRD Env (FRD), EnvV2 (V) and HERV-W Env (W) as fully functional syncytins is indicated on the tree by colored lines and letters. A fading line and gray letter means loss of at least one syncytin property (fusogenicity for EnvV2, all three for HERV-W Env) in this clade. In the last common ancestor of apes and Old World monkey, all three envelopes were present and probably displayed all syncytin characteristics at the same time. Following split of these clades, either EnvV2 was conserved and HERV- W Env lost (lower arrow), or EnvV2 lost its fusogenic properties and HERV-W Env replaced it as syncytin (upper arrow, baton pass event). Present day syncytins are indicated on the right, using the same colors as in the tree. Adapted from Esnault et al. (2013).

a. Example of EnvV2, a decaying syncytin in humans

An example of syncytin replacement and decay can be found when examining the fate of EnvV2 following the primate/Old World monkey split about 45 Mya (Fig. 27, Esnault et al., 2013). EnvV2 is the full-length ORF of a very old ERV-V that integrated the genome of the last common ancestor of simians and is present orthologously and under apparent purifying selection in New World monkeys, Old World monkeys and apes (Fig. 25 and 27, Esnault et al., 2013). Specific placental expression and expression in the ST layer is conserved between humans and the Old World monkey macaque, but fusogenic potential was conserved only in the Old World monkey clade (Esnault et al., 2013). While there are some species with fusogenic EnvV2 in apes and New World monkeys, they are the exception and not the rule (Esnault et al., 2013). These results suggest that EnvV2 had all the properties of a syncytin before the separation of apes

3.3 and Old World monkeys, being fusogenic, expressed in the placenta and conserved in the clade, however it maintained these three properties only in Old World monkeys and not in its ape sister clade (Fig. 27, Esnault et al., 2013). Another study using isolated Rhesus monkey trophoblasts confirmed a probable implication of EnvV2 in Old World monkey ST formation (Kumar et al., 2015). Interestingly, an inverse pattern can be observed for Syncytin-1 which was endogenized by the last common ancestor of these two clades. Syncytin-1 has remained fusogenic in apes, but has lost this property in Old World monkeys in which the ORF has not been conserved at all (Bonnaud et al., 2005; Esnault et al., 2013). From these results, Esnault et al. (2013) propose a model in which the two 45 My syncytin-2 and envV2 genes were the functional syncytins in

104 elcmn,mksi rbbeta hsfniggn a enls oteae,adide no indeed and ages, the to date. lost to been described its has following been gene has selection gene funding purifying such this of that absence probable probable it a makes Sadly with replacement, Mya. associated 175 over origin exapted ancient been have orthol- an must shared such mammals, be marsupial should and species. which eutherian by syncytin, mammalian all is ancestral by placenta of ogously The mammalian organ. number the structured why large diversely to most a as the explanation of far the to placentation contribute the might model in seemingly this the Finally, Env to emergence ERV explanation the parsimonious of after more use My a provides 100 independent also as over it chance such placenta, pure placentation mammalian by for the Lavialle exapted necessary of been in strictly have (reviewed genes could evolution why -B explain of and would Syncytin-A course model the This over essential 2013). longer lost al., no et is is it reproduction, since passes and, and syncytin gene older placentation acquire the newly fashion, for the a to In functionality latter. the placental of of loss baton leading the the but syncytins, former older the the of replace conservation functionally the could to genes new evolutionary an These present host. might the that for properties advantage different slightly to with led genes then new has of events exaptation the appearance evolution, and the endogenization of to course due the genes Env Over new 2013). of the al., acquisition by random et tissue Lavialle fetal in of (reviewed is tolerance system but immune placentation, maternal mammalian since for ISD, necessary the not of is presence fusion the via cell-cell immunosuppression the been functionalities, have new might followed with which organism of placentation the important provided mammalian most that of Env ERV to emergence ancestral hypothesis" that an pass of proposes acquisition "baton formally the term model the The coined it. who (2011) to Imakawa 2013) refer al., & et Nakamura Lavialle of 2012; after- al., team established et the Dupressoir formally and 2011; more by Heidmann, was & earlier (Dupressoir which team years and his several by 2009) wards initially Benkirane, proposed & model (Saib Heidmann a Thierry illustrates EnvV2/Syncytin-1 of case The hypothesis pass Baton b. World non-Old in role another al., plays et (Mangeney still syncytin ISD EnvV2 functional to a 2007). that due contains it fact be as the immunosuppression, either as to might such placentas, or This monkey time, 2013). long al., a et taking (Esnault decay simians all in selection purifying 2013). al., et Esnault 27, (Fig. lost was Syncytin-1 and activity fusogenic envV2 of endogenization Following ancestor. simian common the neetnl,ee huhi otisfsgncpoete,Ev2sildsly in of signs displays still EnvV2 properties, fusogenic its lost it though even Interestingly, spaetlfsoe nae u o l ol oky,i hc nV anandits maintained EnvV2 which in monkeys, World Old not but apes in fusiogen placental as syncytin-1 ..Snyisadevolution and Syncytins 3.3. bu 5Ma treplaced it Mya, 25 about 105

3.3 3. Syncytins

3.3.2 Syncytin diversity and placental diversity

It is important to realise that, as mentioned above, every syncytin described so far has resulted from an independent endogenization event (as illustrated by their non-orthologous status), and this is also the case for most syncytin-like genes except maybe HEMO and pan-Mars-Env2 (Heidmann et al., 2017). Phylogenetic analyses on syncytin TM subunits show that while, except for Fematrin-1, they all cluster (expectedly) with exogenous gamma-type Env, even more specifically the gamma-type Env of gammaretroviruses, they are nevertheless spread across this family of envelopes, clearly originating from distinct exogenous retroviruses (Fig. 28B). The same can be said for an analysis based on the RT sequences associated with syncytins (when they are available), which almost exclusively represent gammaretroviral type RTs, except for the Syncytin-Ory1 and Fematrin-1 associated RT which is clearly betaretroviral (Fig. 28). This suggests that different syncytins possess related but slightly different innate characteristics with regards to fusion, immunosuppression and other potential roles of retroviral envelopes, originating from different retroviral species or strains. Accordingly, shared receptors seem to be the exception rather than the norm amongst syncytins. The diversity of syncytins is further enhanced by their time of acquisition (varying by almost an order of magnitude between 10-80 Mya Cornelis et al., 2012, 2015) and the manner of their exaptation (Fig. 29). Some syncytins such as syncytin-Car1 are associated with a recognizable provirus containing two LTRs, an identifiable TSD and PBS and are expressed following the canonical retroviral pathway of LTR transcript initiation, single splicing and LTR transcript termination (Cornelis et al., 2012). Other syncytins are associated with highly degraded proviruses and are expressed using alternative cellular promoters, or as in the case of syncytin- Rum1 for example a transcript initation site located in the degenerate gag-pol region (Cornelis et al., 2013). The recently described HEMO Env presents an incredibly disrupted provirus that contains deletions and inversions of the the ancestral retroviral sequence and its expression is driven by a cellular CpG island and involves several subsequent and alternative splicing events (Heidmann et al., 2017). The vast diversity of transcription mechanisms and contexts that surround different syncytins and syncytin-like genes doubtlessly can have an impact on the ease and specificity of their transcriptional regulation as well as on their expression level. The different expression levels of Syncytin-Rum1 in sheep and cow placentas has for example been 3.3 proposed to be a causative factor of the varying extent of syncytialization observed in these placentas (Cornelis et al., 2013). Further variation between species is introduced by the potential exaptation of different syncytin genes present and functional concurrently, as was probably the case with EnvV2 and Syncytin-1. Keeping this diversity in mind, it is intriguing to remember the vast diversity of placental structures observed in mammals (Fig. 3 and 4, reviewed in Enders & Carter, 2004; Wooding & Burton, 2008, chap. 1). Placentas vary from the rather simple epitheliochorial pig placenta to the much more complex hemochorial placentas of humans and mice, even though there does

106 eo ahte) n TT eobnn ervrssaehglgtdi e.FrabeitosseFig. see abbreviations ORF. For mer34 red. endogenous in are human scale highlighted HEMO: syncytins (see are and site trees, retroviruses per 11 recombinant both substitutions RT/TM of For and number average green. tree), the each in to below proportional highlighted is are length branch Env blue, Syncytin-like in highlighted 2017). al., et LG+ (Lefort the program and program PhyML alongside the gray with in obtained indicated was is type are TM genera repartitions. CX TM Non-specified syncytin- associated and selected tree. RT the of and the view syncytins in on gammaretroviral all indicated be as to are assumed well gammaretrovirus, as than TM other beta-type genera, and likes.RT-based gamma- selected using and analysis. tree program TM-based PhyML TM the with or obtained is was RT RT tree RtREf+ it The which the on tree. for the syncytins based on all indicated as genera genera well RT-based as retroviral with available, species retroviral among selected using syncytins tree PhyML of RT-based Unrooted Phylogeny – 28 Figure B A

Sync Syn-Opo1

Γ Syn-Rum1 γ FFV oe eetdb h mr oe eeto rga Lfr ta. 07.()Unrooted (B) 2017). al., et (Lefort program Selection Model Smart the by selected model EFV

ytin- SFV

GaLV Syn-Car1 RD114

BLV BFV n

δ PTLV-1 REV

Op C spuma KoR

Syn-1

1-2

o1 Syn-2 MLV V PTLV-2

oi n h rsneo bec fteCS1 S seas i.2) h tree The 22). Fig. also (see ISD CKS-17 the of absence or presence the and motif GaLV CX γ− +ISD

type

KoRV 6 FLV

CC

FLV S yncytin-1 MLV WEH

RD1 ε

Syncytin-Ory1REV WDSV

Syncyt

14 V-2

β in-

B SMRV MPMV WEHV-1

Syncytin-A

S

yncytin-Rum1 Syncytin-2

Syncytin-Ten1 0.5 RT EnvV2 Syncytin-Car1 Γ

IFmdl eetdb h mr oe Selection Model Smart the by selected model, +I+F pan-Mars-Env2HEMO

AMV Syncytin-Mar1 RSV

α BLV ERV-3

PTLV-1PTLV-2 ALV Env-Cav1

TM JSRV

0.8 MPMV

Syn-Ory

M SMRV

MTV evolution and Syncytins 3.3. RSV β− AMVALV CX

-ISD EIAV

type type 1 n

C EIAV

JSRV F emat

H

ri MMTV HIV CAEV n-1 HIV-2 λ HIV-2 CA IV-1

β - λ 1 EV 107 (A)

3.3 3. Syncytins

LTR degenerated gag-pro-pol syncytin-Car1 canonical LTR intiation canonical splicing

degenerated LTR syncytin-Rum1 internal initiation non-canonical splicing

CpG island degenerated pol HEMO cellular promoter initiation alternative splicing 1 kb

Figure 29 – Three examples of different methods of syncytin transcript initiation. Depending on the insertion locus and the evolutive history of this locus, syncytin expression can be initiated in different ways. In some cases like Syncytin-Car1, expression is simply driven by its original LTR promoter and the canonical retroviral splice is present. In other cases like that Syncytin-Rum1, the original promoter LTR has been lost during evolution and transcription is initiated by a cryptic promoter present in the otherwise degenerated gag-pol region. Only the splice acceptor site is canonical. Finally, in the case of the syncytin-like HEMO, transcription is initiated by a cellular CpG island promoter and presents multiple splicing events that can yield alternative transcripts containing different parts of the cellular genome 5’ of the HEMO Env ORF. Faded out boxes represent non-functional ERV sequences, color coding is the same as Fig. 9. The gradient covering the colors of gag, pro and pol indicates that the three cannot be told apart. Green lines represent transcripts and dotted line splicing events.

not seem to be a particular advantage to one form or the other. The involvement of stochastic gene acquisitions through endogenizations might explain the breadth of structures and strategies observed across mammals by introducing the necessary randomness to placental evolution. It has been suggested that following the parsimony principle establishes the hemochorial placenta as an ancestral state while epitheliochorial placentation would be a derived condition (reviewed in Roberts et al., 2016). This suggests a counter-intuitive path for evolution in which the originally separated mother and egg immediately transitioned to a highly invasive phenotype in which the maternal immune system is in direct contact with fetal tissues, before returning to a more distant partnership in which fetal and maternal tissues are merely apposed, closer to an egg/mother relationship. These analyses often do not take into account the potential impact of random gene

3.3 acquisitions on placental structures and the gradual replacements occurring in the baton pass model.

3.3.3 Other roles of syncytins

While the placental functions of syncytins are the best described role and their placenta-specific or preferential expression pattern strongly suggests a major role in placental physiology, some of them have been described as being involved in other processes.

108 bec ffso fotols rcrosit utncetdotols.See l (2011) al. et Søe osteoclast. multinucleated the a in into inefficient precursors is and osteoclast bone resorbed of constantly of fusion is resorption of that The respectively. absence structure osteoblasts dynamic and rather osteoclasts a by is rebuilt bone belief, popular to Contrary fusion Osteoclast b. al., et (Dupressoir fusion abolish far, completely by involved can actors knockout only 2011). the their 2009, not where are placenta they the that but in species, unlike mammalian of number a in fusion points percentage 20 about cases, via both in inhibited similar 2016). were is al., -2 index et fusion (Redelsperger fusion and the the Syncytin-1 of of decrease both the myoblasts, reduction and phenomenon human siRNA point widespread For percentage a 2016). is 20-40 al., fusion et a muscle (Redelsperger in is corresponding syncytins there of the involvement inhibiting case that suggesting and each index, sheep In and time. dog each mouse, syncytin human, from originating cells using 2009). al., et (Dupressoir embryos of of knockout effect the The injury. mice, 2016). after al., female regeneration et not (Redelsperger but species male this in in and Similarly, sex-differences effect specific to sex contribute a is might it Syncytin-B suggesting mice, that female in observed not was effect this Intriguingly, in that showed and cell-cell samples. pre-season increased muscle to strongly that compared show show when expression season also Syncytin-1 competitive who as well (2015), their as al. after fusion et athletes in Frese from by Syncytin-1 taken of confirmed biopsies implication later An was events 2014). contact fusion al., of established that et muscle that filopoda (Bjerregaard with in myotubes colocalized observed and often was fusion myoblasts and myoblast-myotube between Syncytin-1 fusion of of muscle of expression decrease effectors The significant known 2014). other a 2011, to al., triggered leads et fusion expression (Bjerregaard myoblast its that membrane of of show cell the induction inhibition who at (2011) that and expressed specifically al. and is et Syncytin-1 muscles Additionally, Bjerregaard expression. growing by Syncytin-1 in obtained fusiogens. expressed were virus-derived is direction these of this SLC1A5 role in potential results a non-viral cell of first investigation of mediated The the syncytin number to of a led demonstration placentas exist the in process, there fusion this While in myotubes. implicated fusion as into cell-cell described fuse fusiogens controlled myoblasts on reliant which heavily in is muscles processes animal of regeneration and Formation muscles in fusion Cell-cell a. ae oehrteerslssgetta yctn lyasgicn oei uclrcell-cell muscular in role significant a play syncytins that suggest results these together Taken myoblasts primary of fusion assays also (2016) al. et Redelsperger study, same the In fusion muscle mouse to syncytins of contribution the investigated (2016) al. et Redelsperger syncytin-B ncotml ietemsl asi eue yaot20%. about by reduced is mass muscle the mice male knockout syncytin-A ncotwsntsuida ti ehlfrall for lethal is it as studied not was knockout syncytin-B ..Snyisadevolution and Syncytins 3.3. ed osoe muscular slower to leads xvivo ex 109 ,

3.3 3. Syncytins

showed that both Syncytin-1 and SLC1A5 were expressed in differentiating and fusing osteo- clasts, and that inhibition of Syncytin-1 fusion resulted in a significant decrease of the number of nuclei per osteoclast, but did not completely abrogate fusion. This suggests that Syncytin is involved in osteoclast fusion but is not the only actor, and that it is mainly involved in later fusion events that augment the number of nuclei present in an already formed syncytium (Søe et al., 2011). This observation was confirmed by a later time-lapse study which revealed that CD47 was responsible for promoting fusion of mononuclear pre-osteoclasts, while Syncytin-1 mediated fusion of two multinucleated osteoclasts (Møller et al., 2017). Interestingly, Syncytin-1 expression at the cell surface colocalized with f-actin expression, another component necessary for osteoclast fusion, and was especially visible in filopodia involved in cell-cell contact sensing and osteoclast movement (Søe et al., 2011). Another recent ex vivo study by Ethiraj et al. (2018) extends the potential involvement of syncytins to murine osteoclasts in which stimulation of Syncytin-A expression leads to increased osteoclast formation and bone resorption.

c. Syncytins and pathologies

As is the case for non-syncytin ERV env genes (see 2.2.2), abnormal syncytin expression has been linked to a number of pathologies in humans (reviewed in Bastida-Ruiz et al., 2016; Bolze et al., 2017), while their implication in non-human pathologies has not been extensively studied. In most cases the abnormal expression observed in a pathological context seems to be a consequence rather than a cause (reviewed in Bolze et al., 2017). In the placental pathology pre- eclampsia (in which maternal artery remodelling is insufficient, reviewed in Brosens et al., 2011) for example, both Syncytin-1 and -2 expression is significantly down-regulated (correlating with severity of the disease) and cell-cell fusion is impaired (Langbein et al., 2008; Vargas et al., 2009). The reason for this down-regulation might be the hypoxic condition of the placenta (due to insufficient blood supply) which has been shown to induce degradation of GCMa, thereby being able to affect syncytin expression (reviewed in Bolze et al., 2017). The only example of a direct contribution to a disease can be found in some cancers. As with other ERV sequences, ectopic upregulations of syncytin expression is observed in a number of cancers (reviewed in Bastida-Ruiz et al., 2016). It has been shown that in some cancers, tumor cells are implicated in cell fusion events with both tumor and non-tumor cells, thereby creating

3.3 hybrid cancerous cells with novel properties and increasing the overall heterogeneity of the cancer (reviewed in Bastida-Ruiz et al., 2016). As a consequence the cancer can become more difficult to treat and at the same time more invasive and dangerous (reviewed in Bastida-Ruiz et al., 2016). A study by Bjerregaard et al. (2006) shows that about a third of breast cancer derived cell lines express Syncytin-1, and these cells are able to fuse with endothelial cell lines, presenting a potential pathway for a contribution of Syncytin-1 to tumor virulence, though its relevance in vivo has not been confirmed so far.

110 4. Objectives

In view of the results described in the previous section, it has been proposed that syncytin impli- cation in placental evolution is two-fold: i) an ancestral syncytin capture even was responsible for the emergence of the mammalian placenta and ii) subsequent stochastic acquisition of new syncytins and syncytin-like genes led to structural changes resulting in the high diversity of modern day placentas. In the present study, we tried to strengthen the link between syncytin exaptation and both the establishment and subsequent structural transitions of placentas. To do so we approached the two aspects separately and investigated two different placental species, each illustrating one of them. First, we studied placentation in the matrotrophic lizard genus Mabuya. Complex placentas have been described for several members of these genus and they display a highly organized structural arrangement with regions specialized for maternofetal exchanges that are, at least structurally, homologous with those observed in some mammalian placentas (Fig. 8 and see 1.3.2.c.). Of note, in the Mabuya placentome the maternofetal interface presents a syncytial tissue layer, similar to those observed in mammals. In the latter it has been shown that syncytins are indispensible for the establishment of the syncytial layers. The emergence of this placenta is a recent event since the Mabuya genus separated from its closest relatives (the Heremites genus) about 25 Mya (Karin et al., 2016), while the mammalian placenta is thought to have emerged about 175 Mya (dos Reis et al., 2012). This recent emergence led us to investigate if appearance of the Mabuya placenta is linked to the acquisition of a founding syncytin, as is hypothesized for mammals in which the ancient emergence of matrotrophy prohibits identification of a founding exaptation. Identification of a Mabuya syncytin would not only be the first identification of a non-mammalian syncytin gene, but also the first identification of a potential founding syncytin, which contributed directly to a shift to matrotrophic placentation. Second, we investigated the curious case of Hyaenidae placentation. Contrary to every other carnivoran, hyenas present a hemochorial and not endotheliochorial placenta. The hemochorial hyena placenta is much more invasive that that of its carnivoran relatives and it has been repeatedly suggested that syncytins and syncytin-like genes might play a role in placental invasion processes. The major structural endo/hemochorial transition occurred specifically in the Hyaenidae family and has been studied in detail in one of its four extant species: Crocuta crocuta (the spotted hyena). As with the emergence of the Mabuya placenta, this structural transition took place recently in evolution, as Hyaenidae split from their sister family Eupleridae about 30 Mya (Nyakatura & Bininda-Emonds, 2012). Since a direct link between variable syncytin expression

111 4. Objectives

and variable placental organization has been suggested previously in Ruminants (Cornelis et al., 2013), we investigated if the hyena placenta presents a distinct env expression profile when compared to the reference cat and dog placentas, be it through a difference in expression levels of shared genes, or through the acquisition of new hyena-specific genes. 4.0

112 Part II

Results

113 1. Implication of ERV env genes in non-mammalian Mabuya placentation

1.1 Summary

Following the identification of syncytins in most major mammalian clades, we investigated their involvement in non-mammalian placentation. As discussed in 1.3, placentation is not restricted to mammals and is present in a number of lizards, including the Mabuya genus which presents one of the most complex non-mammalian placentas described to date. Due to its structural similarity to mammalian placentas, and in particular the presence of a fused cell layer at the maternofetal interface, we searched for a fusogenic exapted env gene with placental expression, as is observed for mammalian syncytins. Since no Mabuya genome has been published to date and the closest sequenced species is the distantly related green anole (Anolis carolinensis, last common ancestor about 200 Mya), we performed high-throughput RNA sequencing of a Mabuya sp. IV placental transcriptome. Comparison of this transcriptome to an equivalent mammalian placental transcriptome revealed that many implicated pathways are shared, reinforcing the notion that placentation is based on similar processes in these two distantly related clades. An in silico search of the Mabuya transcriptome led to the identification of four coding env ORFs, named Mab-Env1 through 4. These new env can be sorted into two groups: Mab-Env1 and 2 presenting strong similarities, as do Mab-Env3 and 4. All 4 sequences present the major domains of retroviral env and cluster alongside gamma-type envelopes on a TM based phylogenetic tree, however none of them seem to be classical gammaretroviral envelopes. Mab-Env3/4 present characteristics of the avian gamma-type, namely a shifted fusion peptide that is flanked by two cysteines, and cluster with RSV on the TM tree. Mab-Env1/2 show a much more uncommon organization, associating a gamma-type ISD with a beta-type CX7C motif. Their position on the phylogenetic tree puts them among other gamma-type TMs, suggesting that their main phenotype is gamma-like. Further characterization of Mab-Env1 led to the identification of the associated pol gene and allowed construction of an RT based phylogeny, in which it takes an intermediary place, between gammaretroviruses and all other retroviral genera. RT-qPCR based analysis of Mab-Env expression in a panel of tissues show that none displays placenta-specific or preferential expression, but that Mab-Env1 is nevertheless significantly expressed in the placenta while the others show a lower level of expression. We therefore

114 e hshrlto fMZ1 nipratse nissga rndcinadrflcigits reflecting and transduction trig- signal to its sufficient in step is important Mab-Env1 an of MPZL1, a expression of with that phosphorylation cells showed ger MCF10A Mab-Env1 and untransformed with vector transduced interaction expressing we if Mab-Env1 transduction, study signal To MPZL1 contexts. induce physiological can and oncogenic both in cesses (paraplacentome). layers adjacent two in- in supports or Mab-Env1 profile (placentome) expression and paraplacentome. This MPZL1 the expression. in lower of layer indicating teraction unfused stained the less and are placentome tissues the Fetal in layer maternofetal fused the of the layers both maternal placenta. the interface, the in expressed in strongly expression is significant MPZL1 display that does shows hybridization but highly tissues, being of expression, number preferential a or in placenta-specific expressed show not does MPZL1 Mab-Env1, to particles. pseudotyped Mab-Env1 receptor. by infectable cognate cells the rodent restrictive as PZR) related, zero of transmem- protein single Cloning a called (MPZL1, previously 1 transducer; zero-like protein signal myelin brane of identification the the using to led pseudotyped-particles panel Mab-Env1 hybrid of profile the infection identify to The used pair. previously Syncytin-2/MFSD2A was which panel, hybrid irradiation human/hamster GenBridge4 Syncytin-Rum1. assays, for observed fusion was cell-cell as In conditions, titers. pH lower acid in display fusogenic which strongly rodents is Mab-Env1 for except lines, cell to tested (up infectious most highly were particles MLV Mab-Env-pseudotyped that showed genes. syncytin of characteristic third the properties, fusogenic its examined a of insertion that suggesting genera, related closely four Mab-Env1 of DNA genomic from amplified also all of that in Amplification suggesting observed genera. was Scincidae related ORF other coding 10 length in full fragment a internal bp 400 in a as is well as profile ORF expression This tissues maternal with of border apical role physiological the the a at forming with observed accordance layer is fetal expression strong the paraplacentome, of the end In layer. maternal fused the in placentome, the In mice). immunizing Env1 by obtained we antibodies anti-Mab-Env1 (using performed we paraplacentome, adjacent the of layers different on focused aigetbihdthat established Having of conservation study To nmmas rvossuishv mlctdMZ1i elmgainadivso pro- invasion and migration cell in MPZL1 implicated have studies previous mammals, In and RT-qPCR by both assayed was MPZL1 of Expression the using receptor cognate its for search to us allowed Mab-Env1 by conferred tropism The sepesdo ohteftladmtra ie fteitrae ihasrne expression stronger a with interface, the of sides maternal and fetal the both on expressed is otiigpoiu curd u httesqec a o conserved. not was sequence the that but occurred, provirus containing Mabuya Mab-Env1 Mab-Env1 PL rmpaetlcN ofimdta ti nedal orender to able indeed is it that confirmed cDNA placental from MPZL1 Mabuya nti td.T xmn h xrsinpol of profile expression the examine To study. this in a osre nyin only conserved was Mab-Env1 Mab-Env1 lcnoe(h einpeetn ue ellyr n the and layer) cell fused a presenting region (the placentome sepesdi h lcnaadcnevdi h eu,we genus, the in conserved and placenta the in expressed is nvivo in nthe in Mab-Env1 nsitu in Mabuya ste r xrse ntesm iselayer tissue same the in expressed are they as , Mabuya yrdzto swl simmunohistochemistry as well as hybridization Mabuya ttemtroea interface. maternofetal the at pce etdadol ihnta genus, that within only and tested species eu eatmtdt mlf h entire the amplify to attempted we genus h nenlfamn oee was however fragment internal The . nsitu in yrdzto.Similarly hybridization. xvivo Ex Mab-Env1 10 ..Summary 1.1. uinassays fusion 7 um)in ffu/mL) nsitu In nthe in Mab- 115

1.1 1. Implication of ERV env genes in non-mammalian Mabuya placentation

activation. In conclusion, Mab-Env1 is expressed in the placenta, conserved during evolution and fusogenic, it therefore presents the three characteristic of mammalian syncytin genes and was

1.1 renamed syncytin-Mab1. Syncytin-Mab1 is clearly distinct from all other previously identified

syncytins (its unusual ISD/CX7C association sets it apart from all other gamma-type Env), and was exapted in a very distantly related clade. This exaptation of an ERV Env for a potential function in placental physiology is a remarkable example of convergent evolution and might be indicative of the involvement of Env in many more non-mammalian placentas. Additionally, the identification of the Syncytin-Mab1 cognate receptor hints at an involvement of Syncytin-Mab1 in placental invasiveness, a role long proposed for mammalian syncytins and syncytin-likes.

116 ohlohra yn Carnivora / hyena / dotheliochorial § 2 1 France F-70005, Paris, Universités, Sorbonne EPHE, UPMC e France F-94805, d Colombia Bucaramanga, Santander, de dustrial c France b F-94805, Villejuif, Roussy, Gustave 9196, UMR a * Miralles Tarazona jandro Cornelis Guillaume lizard placental a in syncytin A 1.2 ntttd ytmtqe vlto,Boiest,Msu ainldHsor auel,CNRS Naturelle, d’Histoire National Muséum Biodiversité, Evolution, Systématique, de Institut In- Universidad Biologia, de Escuela Vertebrados, de Reproductiva Biologica de Laboratorio CNRS Infectieux, et Endogènes Rétrovirus des Moléculaires Pathologie et Physiologie Unité orsodn uhr [email protected] author: corresponding USA Florida, Gainesville, Florida, of University Institute, Genetics USA UF California, address: Stanford, Present University, Stanford Genetics, of Department address: Present Villejuif, Roussy, Gustave UMS3655, US23/CNRS INSERM, Bioinformatique, de Plateforme France F-91405, Orsay, Paris-Sud, Université 9196, UMR eetrietfidi h iiaosplacental viviparous the in identified receptor qa otiuost hswork this to contributors equal nedgnu ervrlevlp yctnadiscognate its and syncytin envelope retroviral endogenous An ewrs noeosrtoiu neoepoen/paet amcoil/en- / haemochorial / placenta / protein envelope / retrovirus endogenous Keywords: e ataPtii Ramirez-Pinilla Patricia Martha , c,3 ulam Meurice Guillaume , a,b,1,* ahsFunk Mathis , a,b,* d dl Heidmann Odile , éieVernochet Cécile , c her Heidmann Thierry , ..Asnyi napaetllizard placental a in syncytin A 1.2. a,b a,b neDupressoir Anne , a,b,§ rnic Leal Francisca , Mabuya c,2 sa Ale- Oscar , a,b Aurélien , lizard 117

1.2 An endogenous retroviral envelope syncytin and its PNAS PLUS cognate receptor identified in the viviparous placental Mabuya lizard

Guillaume Cornelisa,b,1,2, Mathis Funka,b,1, Cécile Vernocheta,b, Francisca Lealc,3, Oscar Alejandro Tarazonac,4, Guillaume Meuriced, Odile Heidmanna,b, Anne Dupressoira,b, Aurélien Mirallese, Martha Patricia Ramirez-Pinillac, and Thierry Heidmanna,b,5 aUnité Physiologie et Pathologie Moléculaires des Rétrovirus Endogènes et Infectieux, CNRS UMR 9196, Gustave Roussy, Villejuif, F-94805, France; bUMR 9196, Université Paris-Sud, Orsay, F-91405, France; cLaboratorio de Biologia Reproductiva de Vertebrados, Escuela de Biologia, Universidad Industrial de Santander, 680002 Bucaramanga, Colombia; dPlateforme de Bioinformatique, INSERM US23/CNRS UMS3655, Gustave Roussy, Villejuif, F-94805, France; and eInstitut de Systématique, Evolution, Biodiversité, Muséum National d’Histoire Naturelle, CNRS UPMC EPHE, Sorbonne Universités, Paris, F-75005, France

Edited by R. Michael Roberts, University of Missouri-Columbia, Columbia, MO, and approved October 26, 2017 (received for review August 23, 2017)

Syncytins are envelope genes from endogenous retroviruses that Remarkably, placental structures are not restricted to mamma- have been captured during evolution for a function in placentation. lian species. Placentation emerged independently and in a sto- They have been found in all placental mammals in which they have chastic manner in several groups of vertebrates, with the noticeable been searched, including marsupials. Placental structures are not exception of birds (reviewed in refs. 16 and 17). In particular, restricted to mammals but also emerged in some other vertebrates, complex placentas have been described in some South American most frequently in lizards, such as the viviparous Mabuya Scincidae. or African species of Scincidae lizards (18–21; reviewed in refs. 16, Here, we performed high-throughput RNA sequencing of a Mabuya 22, and 23). In one Scincidae genus, Mabuya, placental structures placenta transcriptome and screened for the presence of retroviral env form specialized regions very similar to those found in mammals, genes with a full-length ORF. We identified one such gene, which we e.g., the placentome where maternal and invasive fetal tissues are named “syncytin-Mab1,” that has all the characteristics expected for highly folded and interdigitated. Ultrastructural analysis of this a syncytin gene. It encodes a membrane-bound envelope protein with region further revealed the presence of a syncytial structure at the MICROBIOLOGY fusogenic activity ex vivo, is expressed at the placental level as materno–fetal interface, as observed in numerous mammalian revealed by in situ hybridization and immunohistochemistry, and is species. However, in contrast to most mammals, the syncytial layer conserved in all Mabuya species tested, spanning over 25 My of evo- of the Mabuya placenta is formed by maternal epithelial cells lution. Its cognate receptor, required for its fusogenic activity, was rather than by fetal trophoblast cells, which remain individualized searched for by a screening assay using the GeneBridge4 human/Chi- nese hamster radiation hybrid panel and found to be the MPZL1 gene, Significance previously identified in mammals as a signal-transducing transmem- brane protein involved in cell migration. Together, these results show Retroviral envelope gene capture and exaptation for a placen- that syncytin capture is not restricted to placental mammals, but can tal function has been demonstrated in mammals. Remarkably, also take place in the rare nonmammalian vertebrates in which a vi- placental structures have also emerged on rare occasions in viparous placentotrophic mode of reproduction emerged. It suggests nonmammalian vertebrates, resulting in related modes of re- that similar molecular tools have been used for the convergent evo- production. The Mabuya lizard, which emerged 25 Mya, pos- lution of placentation in independently evolved and highly distant sesses a placenta closely related to that of mammals. Here, vertebrates. we identified a specific retroviral envelope gene capture that shows all the characteristic features of a bona fide mammalian endogenous retrovirus | envelope protein | placenta | syncytin | receptor syncytin, being conserved in Mabuya evolution, expressed in the placenta, and fusogenic. Together with the present identi- yncytins are “captured” genes of retroviral origin that corre- fication of its cognate receptor, these results show that syncytin Sspond to the envelope gene of ancestrally endogenized retro- capture is not restricted to mammals and is likely to be a major viruses. These genes encode fusogenic proteins that are involved in driving force for placenta emergence. the formation, by cell–cell fusion, of the syncytiotrophoblast at the placental materno–fetal interface in eutherian mammals (reviewed Author contributions: G.C., M.F., and T.H. designed research; G.C., M.F., and C.V. per- in refs. 1 and 2). Furthermore, genetically modified mice in which formed research; F.L., O.A.T., A.M., and M.P.R.-P. collected and processed live biological materials and samples; G.M. contributed analytic tools; G.C., M.F., C.V., G.M., O.H., A.D., the two syncytin genes, syncytin-A and syncytin-B, were knocked out A.M., M.P.R.-P., and T.H. analyzed data; and G.C., M.F., and T.H. wrote the paper. showed deficiencies in placental development, with altered struc- The authors declare no conflict of interest. ture of the materno–fetal interface resulting in inhibition of growth This article is a PNAS Direct Submission. or death of the embryo at midgestation (3, 4). Syncytins have been found in all placental mammals in which they have been searched, Published under the PNAS license. with independently captured syncytins occurring across all major Data deposition: RNA sequencing data are deposited in the European Nucleotide Archive, www.ebi.ac.uk/ena (accession no. ERA1116158). Mabuya sequences described in this pa- clades of placental mammals, including Euarchontoglires (primates, per (Mabuya MPZL1, syncytin-Mab1, and Mab-Env2–Mab-Env4) have been deposited in rodents, and lagomorphs), Laurasiatherians (ruminants and carni- the GenBank database (accession nos. MG254887–MG254891, respectively). vores), Afrotherians (tenrec), and even Marsupials (opossum) (Fig. 1G.C. and M.F. contributed equally to this work. 1) (5–15). This has led to the proposal that these genes, which are 2Present address: Department of Genetics, Stanford University, Stanford, CA 94305. absolutely required for placentation as shown by the knockout mice 3Present address: Department of Biology, University of Florida, Gainesville, FL 32611. experiments, are most likely involved in the emergence and evo- 4Present address: Department of Molecular Genetics and Microbiology, University of lution of placental mammals from egg-laying animals (1). Analysis Florida, Gainesville, FL 32611. of the conservation of these genes further indicates that they have 5To whom correspondence should be addressed. Email: [email protected]. been subjected to purifying selection in the course of evolution, as This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. expected for any bona fide cellular gene. 1073/pnas.1714590114/-/DCSupplemental. www.pnas.org/cgi/doi/10.1073/pnas.1714590114 PNAS Early Edition | 1of10 118 Tenrecidae Syn-Ten1 implicated in cell-migration processes (29–32). This identification Proboscidae of a captured syncytin and its receptor in a distant viviparous lizard Dasypodidae with a placentotrophic mode of reproduction shows that syncytin Bradypodidae capture is not restricted to mammals and is likely to be a major Lagomorpha Syn-Ory1 driving force for placenta emergence and evolution. Rodentia Syn-A,-B,-Mar1 eutherian Haplorrhini Syn-1, -2 mammals Results MAMMALIAN Strepsirrhini High-Throughput Sequencing and in Silico Search for Retroviral Env PLACENTA Ruminantia Syn-Rum1 Genes Within the Mabuya Placental Transcriptome. Placental RNA Perissodactyla was extracted from pregnant Andean Mabuya sp. IV females Carnivora Syn-Car1 captured in the wild (Methods). The transcriptome of a placenta at Insectivora embryonic development stage 35 (as defined in ref. 33), presenting Didelphimorphia Syn-Opo1 marsupials a well-established syncytium, was determined by the French Na- Diprotodontia tional Sequencing Center (Genoscope, Evry, France). Placental Monotremata transcripts were assembled using de novo assembly methods in absence of a reference genome (Methods). Transcriptome recon- Gallus gallus (chicken) birds struction identified 72,763 transcripts ranging from 123 bp to Mabuya lizards 24.4 kb [N50 (the length N for which 50% of all bases in the Anolis carolinensis sequences are in a sequence of length L < N) = 2,869 bp; median Danio rerio (zebrafish) size = 1.1 kb], 26,690 of which (36.7%) could be positively Latimeria chalumnae bony fish (coelacanth) matched to a known protein gene from the Refseq Vertebrate 50100150200250300350400 0 Mya database. A set of 21,253 transcripts shows significant expression (more than two tags per million) (34), 16,398 of which were Devonian CarboniferousPermianTriassic Jurassic Cretaceous Tertiary positively annotated (77.2%) and correspond to 13,189 unique genes, among which 10,079 were nonhousekeeping genes. This subset of transcripts contains a range of genes annotated as “placental” genes with key biological functions during placenta Fig. 1. Phylogeny of vertebrates positioning mammals, the Mabuya lizard, development, such as core placental transcription factors, placental and known syncytins. Mammals comprise the monotremes (e.g., platypus) still hormones, iron transporters, gap junction proteins, interleukins and laying eggs, and the marsupials and eutherian mammals, which all possess interleukin receptors, metalloproteinases, and angiogenesis factors a placenta (red font). The lizard Mabuya is shown in red font because it also (SI Transcriptome Analysis and Datasets S2–S4). possesses a placenta. A red arrow indicates the probable time of emergence of To identify putative env-derived syncytin genes in this tran- the mammalian placenta, which has been proposed to correspond to the scriptome, we screened these assemblies using the method we primitive capture of an ancestral syncytin, thereafter replaced in evolution by previously devised to screen mammalian genomes for such genes. the indicated present-day syncytins (reviewed in ref. 1). All currently described syncytin capture events are indicated by arrowheads in purple, together with In short, a BLAST search for env ORFs longer than 400 aa (from the syncytin’s name. Branch length is proportional to time [expressed in My the Met start codon to the stop codon) was performed using a (15, 64, 65)], as indicated in the scale below the tree. set of env sequences including all previously identified syncytins (Methods). It yielded four Mabuya transcripts that we named “Mab-Env1” to “Mab-Env4.” Analysis of the overall structure of cells throughout gestation (19, 20, 24–27). Yet these differences the four identified env genes (Fig. 2 A and B) strongly suggests can be considered minor in view of the tremendous qualitative that they indeed correspond to retroviral Env proteins, with all or transition that must have taken place between egg-laying and most of their characteristic features conserved, including the placental lizards that have a mode of reproduction so closely re- presence of a predicted signal peptide sequence at the N termi- lated to that of placental mammals. Following our previous pro- nus, a putative furin/PACE cleavage site delineating a surface (SU) and a transmembrane (TM) subunit, and a hydrophobic posal that such transitions are likely due to the stochastic > acquisition of new genes via the endogenization of retroviruses and domain 20 aa long located downstream of the highly conserved the exaptation of their envelope gene, a process that has taken C-X5–7-C motif of retroviral envelopes. This hydrophobic domain place on several occasions in mammalian evolution, we investi- is involved in anchoring the Env protein within the plasma mem- brane, a feature required for its fusogenic activity. Finally, all genes gated whether, as in mammals, endogenous retroviral env genes present a canonical immunosuppressive domain (ISD). The pres- could have been specifically captured and exapted for a role in ence of the ISD classifies the four envelopes as gamma-type ret- Mabuya placenta formation in the genus and might even be in- roviral envelopes (reviewed in ref. 35); however only Mab-Env3 and volved in the refined structure of their materno–fetal interface. Mab-Env4 present the gamma-type–specific C-X6-CC motif that Such events would indeed represent a remarkable example of contains the third cysteine necessary for the covalent link between convergent evolution in two major classes of vertebrates. SU and TM observed in gamma-type envelopes, with the other two Currently, no genome within the Mabuya genus has been se- cysteines forming an intra-TM loop. Mab-Env1 and Mab-Env2 quenced, and the most closely related species whose genome is present a C-X7-C motif, a characteristic of beta-type retroviral available is the green anole (Anolis carolinensis), a distant small envelopes in which the SU and TM are not linked by a disulfide oviparous lizard. As a result, we could not screen Mabuya genomes bond and only the intra-TM bond is present (reviewed in ref. 35). to identify candidate syncytin genes as was done previously for Incorporation of Mab-Env1 to Mab-Env4 into a phylogenetic mammalian species. After having captured pregnant females from tree based on the alignment of TM subunits is shown in Fig. 2C aColombianMabuya population [referred to as “Mabuya sp. IV” and demonstrates that they are distinct from previously identi- (28)], we therefore performed high-throughput RNA sequencing fied env genes and that all four of them cluster with gamma-type (RNA-seq) of their placenta transcriptome and searched for the envelopes. Interestingly, Mab-Env1 and Mab-Env2 are grouped presence of expressed, coding retroviral env genes. Remarkably, together and display 82% amino acid identity with strong con- we identified such a gene, syncytin-Mab1,whichprovedtobe servation of most of the functional domains (Fig. 2D), suggesting a syncytin and was characterized for its fusogenic activity, its that they correspond to two genomic copies of the same retroviral conservation during Mabuya evolution, and its expression at the family. The main differences between Mab-Env1 and Mab-Env2 are materno–fetal interface. Using a genetic approach, we identified a 10-aa deletion in the N-terminal domain of the TM subunit of its cognate receptor as the MPZL1 transmembrane protein, pre- Mab-Env2 (the domain presumably harboring the fusion pep- viously characterized as a membrane-bound signal transducer tide critical for Env fusion activity) and a short truncation of the

2of10 | www.pnas.org/cgi/doi/10.1073/pnas.1714590114 Cornelis et al. 119 PNAS PLUS SU SU TM A CC C fusion SU/TM ISD cleavage site peptide JSRV TM MMTV extracellular CXXC RXKR CC membrane N-ter C-ter FIV intracellular signal fusion ISD BLV CAEV HERV-K peptide peptide HTLV1 HIV1

RNAR transmembrane MoMLV Syn-Opo1 B PyERV 3

TgERVF FLV PERVGaLV Mab-Env1 0 Syn-A -3 100 200 300 400 Syn-B RNAR Syn-1 Syn-2 3 REV Mab-Env2 0

Syn-Ory1 -3 BaEV 100 200 300 400 MPMV Mab-Env4RSV RALR Mab-Env3 3 Mab-Env3 0 Syn-Ten1

-3 Syn-Mar1Env-Cav1 100 200 300 400 Fig. 2. Structure of a canonical retroviral Env pro- Mab-Env1 supports values KRRR <25% tein and characterization of the identified Mabuya 3 Syn-Car1 ≥25% candidates. (A) Schematic representation of a retro- Syn-Rum1 CX C Mab-Env4 0 7 ≥50% viral Env protein delineating the SU and TM subunits. Mab-Env2+ISD ≥75% -3 0.7 The furin cleavage site (consensus: R/K-X-R/K-R) be- 100 200 300 400 ≥95% D tween the two subunits, the C-X-X-C motif involved in SU–TM interaction, the hydrophobic signal peptide (purple), the fusion peptide (green), the transmem- brane domain (red), the putative immunosuppressive

domain (ISD) (blue), and the conserved C-X5–7-C motif (CC) are indicated. (B) Characterization of the candi- date Mabuya Env proteins. The hydrophobicity profile for each candidate is shown with the canonical struc- MICROBIOLOGY tural features highlighted in A.(C) Retroviral envelope protein-based phylogenetic tree with the identified Mab-Env protein candidates in red. The unrooted PhyML maximum likelihood tree was obtained using TM subunit amino acid sequences (without the cyto- plasmic tail) from syncytins (in blue) and a series of endogenous and infectious retroviruses (from the dataset in ref. 66). The horizontal scale bar below the tree represents the average number of substitutions per site, and the percent bootstrap values obtained from 1,000 replicates are indicated by circles on each branch (see key in figure). Retroviral families and ma-

jor envelope types, the associated C-Xn-(C)C motif, and the presence or absence of an ISD are indicated on the tree. (D) Amino acid sequences and characteristic structural features of Mab-Env1/Mab-2andMab-Env3/ Mab-4 (GenBank accession nos. MG254888–MG254891). Asterisks indicate amino acid identity; colons indicate amino acid similarity. cytoplasmic tail of Mab-Env1 due to the presence of a premature The expression of Mab-Env2, Mab-Env3, and Mab-Env4 in the stop codon. Although Mab-Env3 and Mab-Env4 are also grouped placenta was 6- to 85-fold lower than that of Mab-Env1. Of note, together, they only display 56% amino acid identity, suggesting that none of these genes was specifically expressed in the placenta. they belong to distinct, but related, retroviral families. Surprisingly, Mab-Env1, in particular, shows high levels of expression in sev- Mab-Env3 and Mab-Env4 cluster with the envelope gene of the eral other organs, including organs of the maternal genital tract Rous sarcoma virus (RSV), an alpharetrovirus presenting an avian (ovary and oviduct), a feature not commonly observed for gamma-type envelope. Closer inspection further reveals that their mammalian syncytins. Due to the high expression levels of Mab- fusion peptide is located about nine amino acids after the putative Env1 in the placenta and the low expression levels of the three furin site, a characteristic of avian gamma-type envelopes (reviewed other env genes, we focused on Mab-Env1 in the present study. in ref. 35). As their name indicates, this subset of gamma-type envelopes was thought to be specific to bird retroviruses. Characterization of the Mab-Env1–Associated Pol Gene. To obtain the sequence of the provirus harboring the Mab-Env1 gene, we Mab-Env1 Is Highly Expressed in Placental Tissues. We performed an rescreened the placental transcriptome in an iterative fashion RT-qPCR analysis of transcript levels for each candidate Mab- using BLAST and found additional overlapping transcripts, each Env gene in different tissues (Fig. 3) using primers that were sharing a 3′ region homologous with the 5′ region of the previous designed to be specific for each env sequence (Table S1). Mabuya transcript, allowing us to reconstruct an internal proviral sequence placental RNA was analyzed at two developmental stages: stage by concatenating the transcripts (Fig. 4A). We confirmed the ex- 30, when the uterine cells start to fuse, forming the syncytial istence of this sequence through PCR on genomic DNA with layer, and stage 37, when the syncytium is well formed. High primers designed at both extremities of the assembled RNA-seq expression levels of Mab-Env1 (similar to that of the ribosomal sequence (Table S1). We were able to amplify and sequence 5.8 kb protein RPL19) could be observed in the placenta at both stages. of internal proviral sequence containing the end of gag and the

Cornelis et al. PNAS Early Edition | 3of10 120 4 Mab-Env1 occurred before the separation of these genera, about 30 Mya (Fig. 5) (36), but that the full-length envelope was conserved only in the 3 Mabuya genus. Of note, we have been unable to amplify even the 400-bp fragment from Heremites vitattus genomic DNA, which is 2 not compatible with the entry date proposed above (Fig. 5). 1 Heremites is a newly created genus (37), and its position on the Trancript level Trancript

(Mab-Env/ RPL19) Scincidae phylogenetic tree is only weakly supported so far, which 0 might explain this apparent anomaly, although sequence divergence

liver could also account for this negative result. ovary ovary heart lungbrain oviduct kidney placenta oviductplacenta intestine Stage 30 Stage 37 In Situ Analyses of Mab-Env1 Transcription and Protein Expression on Mab-Env2 Placental Sections. As described previously in refs. 19, 20, and 24– 1 26, the Mabuya placenta is a complex structure composed of 0 specialized regions (Fig. 6A) including areolae and absorptive pl ovaovi pl ovaovi li in ki he lu br plaques involved in histiotrophic nutrition and, at the embryonic Stage 30 Stage 37 pole, the placentome and paraplacentome. In the placentome Mab-Env3 0,5 region, maternal and fetal epithelia fold and interdigitate dur- 0 ing gestation, and ramifications are highly vascularized. At the pl ovaovi pl ovaovi li in ki he lu br – Stage 30 Stage 37 materno fetal interface, the fetal side is mainly composed of giant Mab-Env4 columnar binucleated chorionic cells. On the maternal side, the 0,5 0 uterine epithelium is replaced by large syncytial structures, most pl ovaovi pl ovaovi li in ki he lu br probably resulting from the cellular fusion of uterine cells. At the Stage 30 Stage 37 cellular interface, maternal and fetal cells form microvilli, further – Fig. 3. Real-time RT-qPCR analysis of candidate env gene expression in Mabuya increasing the exchange surface of the placentome (19, 20, 24 26). sp. IV. Transcript levels are shown as the ratio between the expression levels of Periodically the microvilli are interrupted by the presence of each env gene and those of the RPL19 control gene (SI Methods). Placental tis- sues and maternal ovary and oviduct tissues were recovered at two gestational stages (defined using ref. 33): stage 30, when the uterine syncytium starts to form in the placentome region, and stage 37, when the syncytium is well developed. A pol env The results for the four env candidate genes were obtained using the same series RT TM of tissues. Values are the means of duplicates from three samples ± SEM. RNAseq Genomic DNA entirety of pol and Mab-Env1.Thepol ORF is interrupted at the 3′ 1> end by several stop codons and frameshifts, while the Mab-Env1 2> ORF is intact, as expected (Fig. 4A). Of note, an acceptor splice site 3> can be predicted with a high score in the proviral sequence, about 400 bp upstream of the env start codon (using www.cbs.dtu.dk/ 1000 2000 3000 4000 5000 services/NetGene2), but the corresponding retroviral spliced env B transcript could not be retrieved from the RNA-seq assemblies. support values Syn-Ory1

Identification of the conserved reverse transcriptase (RT) do- MPMV RT <25% JSRV main of the pol gene allowed us to compare it with previously MoMLV ≥25% MMTV – FLV PyERV described RT domains to determine the family of the Mab-Env1 PERV ≥50% ≥75% HERV-K associated virus. A PhyML maximum likelihood tree using RTs GALV 0.4 ≥95% from both exo- and endogenous retroviruses (Fig. 4B)placesthe BaEV RSV Mab-Env1–associated RT closest to the gamma family of RTs. REV TgERVF However, it remains external to all gammaretroviral RTs of the Syn-Opo1 dataset, indicating that it might be a distant member of this family or a member of another, closely related family. In association with the noncanonical combination of an ISD and a C-X -C motif in Syn-Rum1 7 Syn-1 Mab-Env1 (see above and Fig. 2C), this finding suggests that this HTLV1 CAEV Syn-2 lizard retrovirus differs from previously described retroviral families. FIV HIV1 BLV Mabuya

Conservation Within the Genus. To study Mab-Env1 con- Mab-Env1 servation within the Mabuya genus, we tentatively PCR-amplified Syn-Car1 the entire Mab-Env1 ORF from the genomic DNA of several species of Scincidae. Fragments of the expected length were ob- Fig. 4. Characterization of the Mab-Env1–associated RT-containing pol gene tained for all Mabuya species shown in Fig. 5. Sequencing of the and classification among known retroviral families. (A) ORF map of the identified purified PCR products revealed intact reading frames belonging to genomic proviral fragment. Short bars indicate a start codon; tall bars indicate the Mab-Env1 family (Fig. 5). No amplification was obtained in any a stop codon. Green horizontal lines represent overlapping RNA-seq transcripts. non-Mabuya Scincidae, even though amplification of BDNF,acon- Similarities to canonical retroviral genes by BLASTp search are colored in yellow served control gene, indicated that the extracted genomic DNA was for gag and purple for pol. Mab-Env1 is colored red. The conserved RT and TM of good quality. We then used a less stringent approach by designing domains used for the phylogenetic analyses in Figs. 2C and 4B are delineated. degenerate primers in conserved internal Mab-Env1 regions based (B) Unrooted PhyML tree showing the position of retroviral endogenous and exogenous RT domains using the same dataset as in Fig. 2C when available (some on alignments of all available Mabuya sequences. This approach led syncytin-associated RTs were too degenerate to be identified). The Mab-Env1– to the amplification of a 400-bp fragment, ranging from the end of associated RT is highlighted in red, and those associated with previously described the SU to the start of the ISD, in all Mabuya tested, as expected, but syncytins are shown in blue. Major retroviral families are indicated on the tree. also in the Trachylepis, Chioninia, Eumecia,andLubuya genera Bootstraps obtained after 1,000 replicates are indicated by circles on each branch (sequences are shown in Fig. S2). This unexpected result indicates (see key in figure). Branch length is proportional to the average number of that endogenization of the Mab-Env1–associated virus probably substitutions per site (see scale). RT sequences are provided as Dataset S1.

4of10 | www.pnas.org/cgi/doi/10.1073/pnas.1714590114 Cornelis et al. 121 Mab-Env1 MLV particles generated in human 293T cells could infect a panel PNAS PLUS int ORF of target cells of different origins, including human (293T, TE671, Mabuya sp + + SHSY-5Y, HeLa), bovine (MDBK), carnivoran (G355.5, CrFK, Mabuya unimarginata + + MDCK), and rodent (WOP, 208F, A23) cells, in some cases at Mabuya zuliae + + levels similar to those observed using the amphotropic MLV Mabuya dominicana + + (A-MLV) Env protein (although all rodent cells tested exhibited

Mabuya macrorhyncha + + Mabuya a significantly lower titer). Since Mab-Env1 is a protein of reptilian Mabuya nigropalmata + + 18.7 origin, this suggests that its cognate receptor is a highly conserved 24.5 Mabuya croizati + + protein among vertebrates. Rodent cells might be less permissive 26.8 Heremites vittatus - - to infection due to subtle divergences in the evolution of the re- Trachylepis quinquetaeniata + - ceptor in this rapidly evolving clade. Mab-Env2 was found to be 25.6 Chioninia delalandii + - 28.5 Scincidae negative for all target cells tested under the same experimental 29.5 Eumecia anchietae + - conditions. Although we cannot exclude the possibility that Mab- Lubuya ivensii + - Env2 has lost its ability to interact with the mammalian ortholog of 30.1 Eutropis multifasciata - - its cognate receptor or that Mab-Env1 and Mab-Env2 have dif- 24.7 Toenayar novemcarinata - - ferent receptors, these are unlikely. Indeed, as described above, Dasia olivacea - - Mab-Env1 and Mab-Env2 are very similar, but Mab-Env2 presents Tribolonotus gracilis - - a deletion within the putative fusion peptide (Fig. 2D), suggesting Chalcides mionecton - - that Mab-Env2 has lost its fusogenic properties. 200 – Anolis carolinensis* - - Cell cell fusion experiments were performed by transfecting cells with the env expressing plasmids and assaying the formation Fig. 5. Mab-Env1 is conserved only within the Mabuya genus. (Left)Species of syncytia 48 h post transfection in the presence or absence of tree of Scincidae with A. carolinensis as an outgroup. Nodes with black circles an acidic shock (Fig. 7D). Mab-Env1 was able to induce fusion are dated in Mya and placed proportionally to their age (28, 36, 64). Entry of only when the cells were submitted to a low-pH shock, and under Mab-Env1 is indicated by the white-tipped arrow; clades in which the envelope these conditions it induced the formation of large, multinuclear has been conserved are indicated by the black arrow. An asterisk indicates that syncytia (Fig. 7E) in all cell lines tested except rodent cells, as a published genome is available (A. carolinensis). (Right) A plus sign indicates expected from the assay described in Fig. 7B. Mab-Env2 tested amplification of a Mab-Env1 internal 400-bp fragment (int) or the complete negative under all conditions, further confirming the results of env ORF by genomic DNA PCR. Amplification of the complete ORF is restricted

the pseudotyping assay. MICROBIOLOGY to the Mabuya genus, with no amplification in other Scincidae. Of note, both Mab-Env3 and Mab-Env4 were negative in the two assays above. smaller invasive fetal cells located between the fetal and maternal epithelia, allowing a more intimate contact between fetal tissues Search for the Mab-Env1 Receptor. Since Mab-Env1 is able to ef- and the maternal circulatory system (38). The paraplacentome ficiently mediate the infection of human cells but not hamster A23 cells, we took advantage of the human–hamster hybrid ir- surrounds the placentome and is limited on one side by the abrupt radiation panel (Genbridge4; ref. 39) in which fragments of the transition from a uterine syncytium to a uterine epithelium. The human genome had been introduced into a panel of 93 hamster organization of fetal tissues in the paraplacentome does not differ A23 clones (Fig. S3A) and which we had used successfully to significantly from that observed in the placentome, although tis- – identify the receptor of human Syncytin-2 (40). We screened the sues are not folded. The materno fetal interface is still marked by Genebridge4 clones for infection using the pseudotyping assay interdigitated microvilli, although they are smaller than in the – described above. The infection results for the 93 ordered clones placentome (19, 20, 24 26). were [00000100000011100000000010000000100000010100000- To assess the potential physiological relevance of Mab-Env1 000100000000001000000001011000000200000000001], where expression in the placenta, both in situ hybridization and immu- 1 denotes infection (>100 focus-forming units/mL), 0 denotes the nohistochemistry experiments were performed on paraffin sections absence of infection, and 2 indicates that the clone was unavailable of stage 37 Mabuya placentas (Fig. 6B). Specific digoxigenin- for analysis (Fig. S3B). This pattern of infection was compared with labeled antisense riboprobes were synthesized for the detection of a matrix of specific genomic markers by computing a matching Mab-Env1 transcripts, and the corresponding sense riboprobes – score using the RH program (41). The eight best-scoring markers were used as negative controls. Anti Mab-Env1 sera were ob- were located within the same 2-Mb region of the human genome, tained by immunizing mice with Mab-Env1 SU, and preimmune on chromosome 1 (Fig. S3C). Among the genes located within this sera were used as controls (SI Methods). Both methods show that genomic region, 11 encode putative transmembrane proteins. The the expression of Mab-Env1 seems to be widespread, but the capacity of each of these genes to mediate the infection of A23 staining (at both the RNA and protein level) is more intense in the cells by Mab-Env1–pseudotyped particles, and thus to function as cell layers forming the interface between mother and fetus. In a receptor for Mab-Env1, was assayed using expression vectors for the placentome, the staining indicates stronger expression in the the corresponding cloned human cDNAs (Thermo Fisher) and maternal syncytium (ms) than in the fetal chorionic cells (fc). In revealed that only the myelin protein zero-like 1 (MPZL1) protein, the paraplacentome, strong staining is observed at the apical side a previously characterized single-pass transmembrane receptor (29, of the fetal chorionic cell layer, including the microvillous in- 42), was able to increase infection of A23 cells (see below). To terface between mother and fetus. These expression profiles verify that Mab-Env1 can also interact with Mabuya MPZL1, we would support a physiological role for Mab-Env1 at the materno– searched by BLAST for the corresponding cDNA sequence in the fetal interface (as further examined below). Mabuya placental transcriptome and designed primers to amplify MPZL1 from placental cDNA. The Mabuya MPZL1 ORF that Mab-Env1 Is Fusogenic in ex Vivo Assays. The functionality of Mab- could be cloned shares 65% amino acid identity with human Env1 as an ancestrally derived retroviral envelope protein was MPZL1 (Fig. 8A), with the glycosylation sites, the disulfide bond assayed as previously for other syncytins. First, we tested whether predicted for human MPZL1, and the intracellular C-terminal this envelope protein added in trans could render a recombinant domain that contains two immunoreceptor tyrosine-based in- retrovirus deprived of its native env gene able to infect target cells hibitory motifs (ITIMs) (Fig. 8 A and B) (42) being conserved, (Fig. 7A). To do so, the env gene was cloned and introduced into whereas the extracellular N-terminal domains of the two proteins a CMV promoter-containing expression vector that was used to are much more variable. As illustrated in Fig. 8C, when infected generate pseudotyped murine leukemia virus (MLV) particles (SI with Mab-Env1–pseudotyped MLV particles, A23 cells transfected Methods). As illustrated in Fig. 7 B and C,Mab-Env1–pseudotyped with either human or Mabuya MPZL1 exhibit titers 100 times

Cornelis et al. PNAS Early Edition | 5of10 122 paraplacentome A placentome placentome paraplacentome folded uterine epithelium folded fetal epithelium areola fetal maternal

microvillous interface maternal capillary maternal epithelium (me) maternal uterine syncytium (ms) Fig. 6. Structure of the Mabuya sp. IV placenta and absorptive plaque fetal invasive cells expression pattern of Mab-Env1.(A) Schematic rep- endometrium fetal chorionic epithelium (fc) myometrium resentation of the late-stage Mabuya sp. IV placenta fetal vessel (see ref. 27). (Left) Overview of a gravid uterus dis- antisense probe/ sense probe/ playing the apposed maternal and fetal tissues. HES B serum pre-immune serum Maternal and fetal tissues are highly interdigitated ms fc in the placentome region. In addition to the pla- ms fc centome, the Mabuya placenta develops specialized regions for materno–fetal exchanges such as areo- fc ms lae and absorptive plaques. The brown/orange and gray/yellow areas represent the fetal and maternal fc tissues, respectively. (Right) Detailed scheme of the materno–fetal interface in the placentome and para- ms placentome region. In the placentome, maternal and

placentome fetal tissues are highly interdigitated, with numerous ms fc microvilli between fetal and uterine cells. The fetal ep- ithelium is formed by giant binucleated chorionic cells. ms fc The uterine epithelium is replaced by a large syncytial fc structure formed by the fusion of the uterine cells. In the paraplacentome the syncytium is abruptly replaced by uterine epithelium. Cells still present microvilli, but me the tissues are no longer interdigitated. (B, Left)He- fc matoxylin eosin saffron (HES)-stained sections of pla- centa with the maternal uterine syncytium (ms), the me maternal epithelium (me), and the fetal chorionic epi- me thelium (fc) delineated; white arrowheads indicate fc fc maternal and black arrowheads indicate fetal blood vessels. (B, Right) In situ hybridization (odd rows) or me immunohistochemistry (even rows) on serial sections fc for Mab-Env1 expression using digoxigenin-labeled

paraplacentome antisense and sense riboprobes or an anti–Mab-Env1 mouse serum and preimmune serum. (Scale bars: me 50 μm.) Areas marked by rectangles are enlarged on fc the right, and the maternal and fetal domains are me fc delineated.

higher than control cells, indicating that Mabuya MPZL1 can in- directly to the establishment of the maternal syncytium, which deed act as a receptor for Mab-Env1. shows strong staining for both genes. Previous studies have shown that MPZL1 is a signal transducer, Expression and Potential Activation of MPZL1 in Mabuya Placenta. To the activation of which requires phosphorylation of the two tyro- determine if MPZL1 can interact with Mab-Env1 in the placenta, sine residues (Y241 and Y263) contained in its cytoplasmic ITIM we first compared its expression profile with that of Mab-Env1. motifs (Fig. 8 A and B) (29). To further demonstrate the in- Using the same cDNA panel as in Fig. 3, we performed RT-qPCR teraction of Mab-Env1 with MPZL1, we investigated whether and found that MPZL1 is indeed expressed in the placenta but also Mab-Env1 can trigger MPZL1 phosphorylation. To do so, we in other tissues (Fig. 8D). We then performed in situ hybridization transduced untransformed human breast MCF10A cells with on Mabuya placenta sections, as in Fig. 6B, using MPZL1-specific a lentiviral vector expressing Mab-Env1 or a negative control antisense probes and sense probes as negative controls. As illus- (empty vector or a vector expressing the mammalian syncytin- trated in Fig. 8E, a specific antisense-only signal is observed at Car1). As a positive control, we treated the cells with concanavalin A (Con A), which is known to induce homodimerization and the materno–fetal interface. The maternal syncytium in the pla- phosphorylation of MPZL1 (29). Cell lysates were then analyzed centome and the maternal epithelium of the paraplacentome in by Western blot using an antibody directed specifically against the particular show strong staining, while fetal chorionic cells are only phosphorylated form as well as an antibody detecting both forms moderately stained in both regions. This pattern is close to that of MPZL1. In three independent transduction experiments we observed for Mab-Env1 (Fig. 6B), especially for the placentome observed that Mab-Env1 very significantly reduces the total where both genes are more expressed at the level of the maternal amount of MPZL1 without decreasing (rather slightly increasing) syncytium, whereas in the paraplacentome the strongest expression its overall phosphorylation level, compared with the control syn- of each gene takes place in distinct, although interdigitated, tissues cytin-Car1 or empty vector (Fig. 9). This indicates that Mab- of the materno–fetal interface. Therefore Mab-Env1 and MPZL1 Env1 interacts with MPZL1, triggering both its degradation could interact physically in vivo in both the placentome and par- (reminiscent of receptor interference) and, as expected, a strong aplacentome, but in the former this interaction could contribute correlative increase of its phosphorylation.

6of10 | www.pnas.org/cgi/doi/10.1073/pnas.1714590114 Cornelis et al. 123 PNAS PLUS empty Ampho Mab- Mab- these canonical characteristics—i.e., fusion activity, expression in A B — pseudotyping assay vector MLV Env1 Env2 the placenta, and conservation during evolution allow Mab-Env1 nlsLacZ viral RNA “ ” MLV Mab-Env to be now named syncytin-Mab1 and to be added to the series viral core or control 293T of previously characterized syncytins that we and others have transfection infection with identified so far in both Eutherian and Marsupial placental SHSY pseudotyped MLV X-gal human – staining mammals (5 15). 293T target cells The captured Mab-Env1 and associated pol genes have some producer cells TE671 unusual properties that could be linked to the remoteness of the C species in which they were discovered. Although the Env gene falls

8 G355 10

cat within the family of gamma-type envelopes according to both the 6 phylogenetic tree and the presence of an ISD, as shown in Fig. 2, it 10 CrFK possesses a C-X7-C motif in the TM subunit that is not classically 4

10 WOP found in such envelopes. Gamma-type envelopes normally possess three cysteines in this motif (C-X -CC), allowing the formation of 2 6 10 – 208F acovalentSU TM link, whereas a two-cysteine motif is observed viral titer (ffu/mL) rodent 0 in betaretroviruses and lentiviruses, in which the SU and TM are 10 293T HeLa WOP 208F A23 G355 MDCK MDBK A23 not covalently associated (reviewed in ref. 35). The associated Human Rodent Carnivoran Bovine noncoding pol gene displays a characteristic RT domain that DEempty vector MMTV Mab-Env1 clusters more closely to gammaretroviral RTs but forms a sister cell-cell fusion assay clade to all remaining gammaretroviral sequences in our dataset Mab-Env nlsLacZ or control (Fig. 4). This suggests that it is either a remotely related gam- transfection pH 4 maretrovirus or that it belongs to another, potentially unknown, family. The latter might also explain the unusual characteristics of the syncytin-Mab1 gene discussed above. It is also noteworthy that acidic Mab-Env3 and Mab-Env4 present characteristics of avian gamma- shock type envelopes, further reinforcing the idea that the remoteness

X-gal pH 7 of the investigated species should reveal novel categories of staining endogenous retroviruses.

293T Other interesting results of the present investigation concern MICROBIOLOGY the nature of the identified receptor for this retroviral envelope Fig. 7. Mab-Env1 is a fusogenic retroviral envelope protein. (A) Schematic and the properties of this previously characterized gene that representation of the cell infection assay with Mab-Env–pseudotyped virus could be relevant to Mabuya placenta physiology. The MPZL1 particles. Pseudotypes are produced by cotransfecting 293T cells with expression receptor is a single-pass transmembrane glycoprotein belonging vectors for the MLV core, a β-galactosidase encoded by a nlsLacZ-containing to the immunoglobulin superfamily, with an intracellular moiety retroviral transcript, and either a vector expressing Mab-Env proteins or a con- containing two ITIMs. Previous studies have shown that MPZL1 trol vector. The supernatant of the transfected cells is then added to the in- can promote the migration of mesenchymal-derived mouse em- dicated target cells, which are X-gal stained 3 d postinfection to reveal viral – bryonic fibroblasts (30, 31) or hepatocellular carcinoma cells (32) infection. (B) Panel of X-gal stained target cells infected with particles without and can be involved in adhesion-dependent signaling (43, 44). In Env or pseudotyped with Mab-Env1, Mab-Env2, or the Env protein from A-MLV this respect, it will be interesting to determine whether the pres- as positive control. (C) Quantification of viral titers expressed in focus-forming – units per milliliter after infection with Mab-Env1–pseudotyped MLV virions. ently demonstrated syncytin-Mab1 induced phosphorylation of Values shown are the mean of three independent experiments ± SD. MPZL1 could also trigger receptor-mediated signaling in vivo that (D) Schematic representation of the cell–cell fusion assay. 293T cells were could participate in some function essential for Mabuya placen- cotransfected with a plasmid expressing a nuclear β-galactosidase and a plasmid tation in addition to the canonical membrane fusion activity ob- expressing either Mab-Env1 or a control plasmid. After 48 h a 5-min acidic shock served for syncytins. For instance, activation of MPZL1 could was performed (or not) using Dulbecco’s phosphate-buffered saline (DPBS)·HCl enhance placental cell migration/invasion into the maternal tis- at either pH 4 or 7. Cells were X-gal stained 6 h afterward. (E) Mab-Env1 is able sue or even inhibit some specific immune functions via its ITIM to induce cell–cell fusion after an acidic shock. After exposition to acid medium, domains. cells transfected with the Mab-Env1–expressing plasmid show a degree of fusion One remarkable aspect of vertebrate physiology concerns the equivalent to that of cells transfected with a plasmid expressing the Env protein of highly diverse modes of reproduction, with clades such as birds mouse mammary tumor virus (MMTV), which was used as a positive control. having a strictly oviparous mode of reproduction and other clades containing both oviparous and some viviparous species. Even among viviparous species, a range of reproductive strate- Discussion gies exists. In some species, the egg and embryo develop with Analysis of the presently established Mabuya lizard placental very limited maternal supply, being essentially dependent on yolk transcriptome revealed the presence of an endogenous retroviral supply (lecithotrophy), while at the other extreme some species env gene, Mab-Env1. It is highly expressed in placental tissues, show strong dependence on maternal exchanges and almost no especially at the level of the materno–fetal interface, both in the reliance on yolk (matrotrophy) (reviewed in refs. 16 and 45). The maternal syncytium layer in the placentome and in the fetal placental organization of mammals represents the latter ex- epithelium in the paraplacentome. Mab-Env1 is a fusogenic treme, with the mother supplying almost all nutrients necessary protein that is able to functionally replace a present-day retro- for fetal growth. Clearly, the mammalian placenta is likely to be viral env gene within a recombinant infectious retrovirus. It also a monophyletic trait that emerged about 150 Mya and has since can mediate cell–cell fusion, although only under conditions of evolved into several variations on a common theme. In all cases low pH as observed for the ruminant syncytin-Rum1 (11). The in which it has been searched, there is evidence that retroviral cognate receptor of Mab-Env1 was also identified and found to capture has played a role in the diversity of placental structures, be the previously described MPZL1 transmembrane signal and possibly even in the emergence of the mammalian placenta transducer (29). Its expression is not restricted to the pla- via the capture of a primitive retroviral envelope during the ra- centa, but in situ hybridization demonstrates expression at the diation of this family (reviewed in refs. 1 and 2). Of note, there materno–fetal interface, which colocalizes with that of Mab-Env1 is increasing evidence that some nonmammalian vertebrates, in the placentome syncytial layer. Finally, we have shown that among them Mabuya Scincidae and other lizards such as the Mab-Env1 is conserved in the Mabuya species tested, covering African Lubuya ivensii (21, 46), possess a mode of viviparity with ∼25 My of evolution of this specific placental clade (36). Together, placental structures that closely resembles that found in mammals.

Cornelis et al. PNAS Early Edition | 7of10 124 A signal peptide N-glycosylation site 85 Mabuya MPZL1 MAAATAAAAVVVAAALTAAGSGGTGPRAAPLTRWLALLLAVALGLV--QVSAVEVYTPPELSVEN GTEAKLPC TFTSTDVISSAA Human MPZL1 ------MAASAGAGAVIAAPDSRRWLWSVLAAALGLLTAGVSALEVYTPKEIFVAN GTQGKLTC KFKSTSTTGGLT ::*: *: * *** :** ****: ***:***** *: * ***: ** * * ** : transmembrane185 SVTWNFQAEGSTTMVSFFYYSNGKAYPGKDVPFKDRISWAGDLNKKDASISIANMRFQDN GTYTC DVKNPPDIVVVPGEIKLRVVERGNLPVFPIALVAG SVSWSFQPEGADTTVSFFHYSQGQVYLGNYPPFKDRISWAGDLDKKDASINIENMQFIHN GTYIC DVKNPPDIVVQPGHIRLYVVEKENLPVFPVWVVVG **:* ** **: * ****:**:*: * *: ************:****** * **:* **** *********** ** *:* ***: ******: :* * ITIM domains 285 VAVLGVVGLFLVVGLIVCAVYRKKNSKKHYSGCSTSEILVSPVKQAPRKSPSDTEGLVNSVPIRSHQGPVIYAQLDHSGGHYSDKINKSESVVYADI RKN IVTAVVLGLTLLISMILAVLYRRKNSKRDYTGCSTSESL-SPVKQAPRKSPSDTEGLVKSLPSGSHQGPVIYAQLDHSGGHHSDKINKSESVVYADI RKN : *:** *:: :*: :**:****: *:****** * ******************:*:* *****************:****************** Fig. 8. Structure and expression of the Mab-Env1 re- B C D ceptor MPZL1. (A) Alignment of Mabuya MPZL1 (Gen- Ig-like V-type Bank accession no. MG254887) and human MPZL1 106 domain amino acid sequences with major domains and sites 5 8 10 indicated. Asterisks indicate amino acid identities, and 104 6 colons indicate amino acid similarities. (B)To-scale 103 schematic representation of the mature human MPZL1 CCN-Glyc 4 N-Glyc 102 with major domains indicated. (C)HumanandMabuya – 1 2 MPZL1 act as receptors for Mab-Env1 mediated in- viral titer (ffu/mL) 10 fection. Target A23 cells were transfected with a plas- 0 trans- 10 0 mid expressing human or Mabuya MPZL1 or an empty membrane level (MPZL1/RPL19) Trancript liver lung ovary ovary heart brain vector. Twenty-four hours posttransfection, cells were oviduct oviduct kidney placenta placenta intestine infected using MLV particles carrying an nlsLacZ re- Stage 30 Stage 37 MPZL1 Hs ITIM Y241 MPZL1 Hs porter gene and pseudotyped with Mab-Env1 or Mab- MPZL1 Mab MPZL1 Mab Empty vector motives Empty vector Env2 or without Env. Cells were X-gal stained 72 h after Y263 Mab-Env1 Mab-Env2 infection, and infection foci were quantified. Values E HES antisense probe sense probe shown are the mean of three independent experi- ± ms ments SD. (D) MPZL1 is expressed in a wide range of tissues. RT-qPCR analysis of MPZL1 expression levels fc was performed in the same way and on the same series of tissues as in Fig. 3. (E) MPZL1 is expressed at ms the materno–fetal interface. (Left) HES-stained sec-

placentome ms fc tions of placenta with the maternal syncytium (ms), fc the maternal epithelium (me), and the fetal chorionic epithelium (fc) delineated; white arrowheads indi- me cate maternal and black arrowheads indicate fetal blood vessels. (Right) In situ hybridization on serial fc sections for MPZL1 expression using digoxigenin- labeled antisense and sense riboprobes. (Scale bars: me 50 μm.) Areas marked by rectangles are enlarged on me the right, and the maternal and fetal domains are

paraplacentome fc fc delineated.

These species have developed placentas that possess several roviral env capture in the form of the functional fusogenic en- of the characteristic features of the mammalian placenta and velope protein syncytin-Mab1 but also could correlate its capture enable very similar reproductive strategies, including high ma- in time with the emergence in the Mabuya genus of a placental ternal dependence for embryo growth, very tiny eggs almost structure ∼25 Mya (36). Although we acknowledge that the devoid of yolk supply, prolonged pregnancy, and a large increase temporal correlation between the acquisition of a retroviral env in conceptus dry mass during gestation (18, 22, 47). Accordingly, gene possessing all the characteristic features of canonical and taking the present results into account, we propose as mammalian syncytins and the emergence of the Mabuya placenta a working hypothesis an extended version of a model previously is not proof of a causal link, it constitutes a significant hint for proposed for mammals, in which emergence of this reproductive a direct involvement (see also the properties of the syncytin-Mab1 strategy is associated with the stochastic acquisition of a retro- viral envelope gene that provides the corresponding species with syncytin functions. This model can be substantiated as follows. First, it is now Empty Empty Empty established that placenta-like structures have emerged among vector vector vector egg-laying vertebrates throughout vertebrate evolution, ranging -ConA +ConASyncytin-Car1Mab-Env1-ConA +ConASyncytin-Car1Mab-Env1-ConA +ConASyncytin-Car1Mab-Env1 from 400 Mya in fish to 150 Mya in mammals and 25 Mya in the 28 kDa Anti-MPZL1 Mabuya lizards (48; reviewed in refs. 17 and 49). These emer- gences appear to be random and rare. Consistently, retroviral endogenization events have been identified in a large series of Anti-Phospho 28 kDa MPZL1 (Y241) vertebrates, with some of them having occurred over 400 Mya in marine species (50). Envelope gene capture in which this gene γ has been conserved after retrovirus endogenization has recently 50 kDa Anti-tubulin been characterized in spiny-rayed fishes dating back 100 My (51). This indicates that retroviral env capture has most probably Fig. 9. Mab-Env1 induces degradation and phosphorylation of MPZL1. taken place across all of vertebrate evolution, and endogeniza- MCF10A cells were transduced either with an empty lentiviral vector or with tion probably even predates the emergence of vertebrates. the same vector coding for syncytin-Car1 or Mab-Env1. Con A treatment was performed (or not) 1 h before cell lysis as a positive control. Lysates were Therefore, the stochastic emergence of specific modes of re- deglycosylated using PNGase F, and Western blot analysis was performed production during vertebrate evolution could, in principle, be using an antibody recognizing both phosphorylated and unphosphorylated accounted for by retroviral capture events, taking into account MPZL1 (Top) or only phosphorylated MPZL1 (Middle). (Bottom) An anti–tu- the ancestry of these elements. One other hint provided by the bulin-γ antibody was used to compare total protein levels in each well. Each present investigation is that we not only could demonstrate ret- set of blots represents an independent set of transduced cells.

8of10 | www.pnas.org/cgi/doi/10.1073/pnas.1714590114 Cornelis et al. 125 receptor likely to be relevant to such a role, as discussed above). Technologies) and qPCR quantification (MxPro; Agilent Technologies), the PNAS PLUS To be complete, we must also mention that part of the syncytin- library was sequenced using 100-base-length read chemistry in a paired-end Mab1 sequence, although not the complete env, could be identi- flow cell on the Illumina HiSeq2000 instrument (Illumina). Nucleotide se- fied in two clades closely related to the Mabuya genus, namely the quence information has been deposited at the European Nucleotide Archive clade containing the Trachylepis and Chioninia genera and the (ENA; https://www.ebi.ac.uk/ena) under the accession number ERA1116158. clade containing the Eumecia and Lubuya genera, but not in any To maximize the number of reconstructed transcripts, several independent other, more distantly related clades (notably Dasia, Toenayar,or assemblies were performed using two different de novo assembly software Eutropis). This suggests that Mab-Env1 probably entered the programs (Oases and Trinity) as well as a range of k-mer lengths (55, 59, 63, Scincidae family slightly before the divergence of the Mabuya 67, 71, 75 and 79), and only transcripts of >100 bp were conserved. We then genus dated ∼25 Mya, and that the related env gene was not used the EvidentialGene pipeline to remove duplicates and concatenate maintained as a protein-coding sequence in Trachylepis and fragmented transcripts from each individual assembly (54). Reads were Chioninia or in Eumecia and Lubuya. For the former two, this mapped on the nonredundant assembly using Bowtie2 (55), and transcript scenario would be consistent with Trachylepis being oviparous and abundance was quantified using Express (56). To identify the transcripts, we Chioninia being lecithotrophic (52). For the latter two, which used TBLASTX (word_size 3, gapopen 10, gapextend 3, BLOSUM 45) with an display matrotrophic placentation (21, 47, 52), it must be hy- e-value of 10−3 to compare them with the nonredundant protein database pothesized either that the full-length gene could not be amplified of vertebrate genomes (RefSeq database Vertebrates January 5, 2015), and used for technical reasons or that another capture has taken place later the best hit as the transcript identifier for downstream analysis. Gene identifiers in evolution. Along this line, it is theorized that the emergence were converted to gene symbols using the bioDBnet (https://biodbnet-abcc. of viviparity in American (Mabuya)andAfrican(Eumecia and ncifcrf.gov) website for comparison with transcriptomes of other species [human Lubuya) Scincidae results from two independent emergence and mouse placenta (57), Chalcides ocellatus pregnant uterus (58), and Eutherian events (52). uterine decidua (59)]. For functional annotations and enrichment, housekeeping In any event, the identification in the present paper of both genes were first removed (60), and functional annotations [GO Term, Kyoto a syncytin and its cognate receptor in a remote nonmammalian Encyclopedia of Genes and Genomes (KEGG) pathway, and Mouse Genome In- species opens the door to further searches for other syncytin formatics Mammalian Phenotype] were performed using the Enrichr website captures in distant placental vertebrates and to in-depth inves- tigations of their roles. These could include noncanonical func- (amp.pharm.mssm.edu/Enrichr/). tions, as suggested here, that would add to their fusogenic activity. This could lead to a deeper understanding of the mechanisms Database Screening and Sequence Analyses. Retroviral endogenous env gene underlying the rare and stochastic emergence of placentation, sequences were searched in each transcriptome assembly. Sequences con- a remarkable example of evolutionary convergence, and provide taining an ORF longer than 400 amino acids (from start to stop codons) were MICROBIOLOGY hints on species evolution in vertebrates. extracted from the transcriptomes using the getorf program of the EMBOSS package (emboss.sourceforge.net/apps/cvs/emboss/apps/getorf.html)and Methods were translated into amino acid sequences. These sequences were compared with the TM subunit amino acid sequences of 35 retroviral envelope glyco- Animals and Tissues. Mabuya females were captured in the wild by M.P.R.-P. in the municipality of Curiti, Department of Santander, Colombia. This proteins from representative endogenous retroviruses, among which are Colombian population is an as-yet unnamed species described in ref. 28 as known syncytins, and infectious retroviruses using the BLASTp program Mabuya sp. IV. This species corresponds to the one studied in refs. 19, 24, 25, (https://blast.ncbi.nlm.nih.gov/Blast.cgi). Putative envelope protein sequences 27, and 53. Placental and uterine tissues were collected from killed pregnant were then selected based on the presence of a hydrophobic domain (trans- females and were either stored in RNAlater (Qiagen) or processed for in situ membrane domain) located 3′ to a highly conserved C-X5–7-C motif. Multiple hybridization. Other tissues (ovary, oviduct, liver, intestine, kidney, heart, lung, alignments of amino acid sequences were carried out using the MUSCLE al- andbrain)werecollectedfromthesamekilled females. Total RNA was extracted gorithm in AliView (61). Maximum likelihood phylogenetic amino acid trees after tissue lysis in TRIzol reagent (Invitrogen), and genomic DNA was extracted were constructed with PhyML 3.0 (62), with bootstrap percentages computed by phenol-chloroform extraction. Genomic DNA from Mabuya croizati, Mabuya after 1,000 replicates. Substitution models were selected using the SMS pro- dominicana, Mabuya nigropalmata, Mabuya macrorhyncha, Mabuya unima- gram (63): LG+Γ+I for TM sequences and RtREV+ Γ for RT sequences. rginata (sensu lato), and Mabuya zuliae and from Chalcides mionecton, Chioninia delalandii, Eutropis multifasciata, Heremites vittatus, Trachylepis quinquetaeniata, Ethics Statement. This study was carried out in strict accordance with the and Tribolonotus gracilis was obtained by phenol-chloroform extraction from tissue samples provided by A.M. Similarly, genomic DNA was extracted from Dasia French and European laws and regulations regarding animal experimentation olivacea tissues provided by Aaron Bauer (Villanova University, Villanova, PA ) and (Directive 86/609/EEC regarding the protection of animals used for experimental from Lubuya ivensii tissues (Museum of Comparative Zoology numbers R-193864, and other scientific purposes). Mabuya were collected under the permit issued by R-193868, R-193869, and R-193870) provided by Breda Zimkus (The Louis Agassiz the Ministerio de Ambiente y Desarrollo Sostenible (Colombia) and handled in Museum of Comparative Zoology, Cryogenic Collection, Harvard University, accordance with government guidelines on the ethical treatment of animals and Cambridge, MA). Genomic DNA from Eumecia anchietae and Toenayar all applicable regulations (Estatuto Nacional de Proteccion de los Animales, Ley novemcarinata was also provided by Aaron Bauer. 84, December 27, 1989) and following the considerations of the Herpetological Animal Care and Use Committee (HACC, 2004). Mabuya sp. IV Placenta Transcriptome High-Throughput Sequencing and de Other methods are detailed in Supporting Informaton. Novo Assembly. Total RNA was purified from Mabuya sp. IV placenta be- fore sequencing using the Illumina HiSeq. 2000 sequencing system. cDNA ACKNOWLEDGMENTS. We thank Adriana Alberti, Corinne Da Silva, and library construction, sequencing, and transcripts assembly were performed Jean Weissenbach (Genoscope) for Mabuya transcriptome sequencing and by the French National Sequencing Center (Genoscope, Evry, France). Briefly, + assembly; Jean Weissenbach for the gift of the GeneBridge4 RH clones; poly(A) RNA was selected with oligo(dT) beads, chemically fragmented, Aaron Bauer and Breda Zimkus for providing Scincidae samples; Olivia Bawa and converted into single-stranded cDNA using random hexamer priming (Gustave Roussy Preclinical Evaluation Platform) for contributions to the according to the Illumina TruSeq protocol. The second strand was then histological analyses; Mélanie Polrot (Gustave Roussy Animal Care Facilities) generated to create double-stranded cDNA. Next, the paired-end library was for assistance in mouse handling; David Ribet for help in GeneBridge4 panel prepared following Illumina’s protocol: Briefly, fragments were end- data analysis; Cécile Lemaitre, Jhen Tsang, and Marie Dewannieux for help repaired and then 3′-adenylated, and Illumina adapters were added by us- with the MPZL1 phosphorylation assay; and Christian Lavialle for discussion ing NEBNext Sample Reagent Set (New England Biolabs); ligation products and critical reading of the manuscript. This work was supported by the CNRS, were purified, and DNA fragments (>200 bp) were PCR-amplified using by a grant from the Ligue Nationale contre Le Cancer (Equipe Labellisée) (to Illumina adapter-specific primers. After library profile analysis (showing T.H.), and by Agence Nationale de la Recherche Grant ANR RETRO-PLACENTA a typical library size of 200–600 bp) by Agilent 2100 Bioanalyzer (Agilent (to T.H.).

Cornelis et al. PNAS Early Edition | 9of10 126 1. Lavialle C, et al. (2013) Paleovirology of ‘syncytins’, retroviral env genes exapted for 33. Dufaure J, Hubert J (1961) Table de développement du lézard vivipare: Lacerta a role in placentation. Philos Trans R Soc Lond B Biol Sci 368:20120507. (Zootoca) vivipara Jacquin. Arch Anat Microsc Morphol Exp 50:309–328. 2. Denner J (2016) Expression and function of endogenous retroviruses in the placenta. 34. Wagner GP, Kin K, Lynch VJ (2013) A model based criterion for gene expression calls APMIS 124:31–43. using RNA-seq data. Theory Biosci 132:159–164. 3. Dupressoir A, et al. (2009) Syncytin-A knockout mice demonstrate the critical role in 35. Henzy JE, Johnson WE (2013) Pushing the endogenous envelope. Philos Trans R Soc placentation of a fusogenic, endogenous retrovirus-derived, envelope gene. Proc Natl Lond B Biol Sci 368:20120506. Acad Sci USA 106:12127–12132. 36. Pereira AG, Schrago CG (2017) Arrival and diversification of mabuyine skinks (Squa- 4. Dupressoir A, et al. (2011) A pair of co-opted retroviral envelope syncytin genes is mata: Scincidae) in the Neotropics based on a fossil-calibrated timetree. PeerJ 5: required for formation of the two-layered murine placental syncytiotrophoblast. Proc e3194. Natl Acad Sci USA 108:E1164–E1173. 37. Karin BR, Metallinou M, Weinell JL, Jackman TR, Bauer AM (2016) Resolving the 5. Mi S, et al. (2000) Syncytin is a captive retroviral envelope protein involved in human higher-order phylogenetic relationships of the circumtropical Mabuya group (Squa- placental morphogenesis. Nature 403:785–789. mata: Scincidae): An out-of-Asia diversification. Mol Phylogenet Evol 102:220–232. 6. Blond JL, et al. (2000) An envelope glycoprotein of the human endogenous retrovirus 38. Vieira S, de Perez G, Ramírez-Pinilla MP (2007) Invasive cells in the placentome of HERV-W is expressed in the human placenta and fuses cells expressing the type D Andean populations of Mabuya: An endotheliochorial contribution to the placenta? mammalian retrovirus receptor. J Virol 74:3321–3329. Anat Rec (Hoboken) 290:1508–1518. 7. Blaise S, de Parseval N, Bénit L, Heidmann T (2003) Genomewide screening for 39. Gyapay G, et al. (1996) A radiation hybrid map of the human genome. Hum Mol fusogenic human endogenous retrovirus envelopes identifies syncytin 2, a gene Genet 5:339–346. conserved on primate evolution. Proc Natl Acad Sci USA 100:13013–13018. 40. Esnault C, et al. (2008) A placenta-specific receptor for the fusogenic, endogenous 8. Dupressoir A, et al. (2005) Syncytin-A and syncytin-B, two fusogenic placenta-specific retrovirus-derived, human syncytin-2. Proc Natl Acad Sci USA 105:17532–17537. murine envelope genes of retroviral origin conserved in Muridae. Proc Natl Acad Sci 41. Boehnke M, Lange K, Cox DR (1991) Statistical methods for multipoint radiation – USA 102:725 730. hybrid mapping. Am J Hum Genet 49:1174–1188. 9. Heidmann O, Vernochet C, Dupressoir A, Heidmann T (2009) Identification of an en- 42. Zhao ZJ, Zhao R (1998) Purification and cloning of PZR, a binding protein and putative dogenous retroviral envelope gene with fusogenic activity and placenta-specific expres- physiological substrate of tyrosine phosphatase SHP-2. J Biol Chem 273:29367–29372. “ ” sion in the rabbit: A new syncytin in a third order of mammals. Retrovirology 6:107. 43. Eminaga S, Bennett AM (2008) Noonan syndrome-associated SHP-2/Ptpn11 mutants 10. Cornelis G, et al. (2012) Ancestral capture of syncytin-Car1, a fusogenic endogenous enhance SIRPalpha and PZR tyrosyl phosphorylation and promote adhesion-mediated retroviral envelope gene involved in placentation and conserved in Carnivora. Proc ERK activation. J Biol Chem 283:15328–15338. – Natl Acad Sci USA 109:E432 E441. 44. Kusano K, Thomas TN, Fujiwara K (2008) Phosphorylation and localization of protein- 11. Cornelis G, et al. (2013) Captured retroviral envelope syncytin gene associated with zero related (PZR) in cultured endothelial cells. Endothelium 15:127–136. the unique placental structure of higher ruminants. Proc Natl Acad Sci USA 110: 45. Ostrovsky AN, et al. (2016) Matrotrophy and placentation in invertebrates: A new – E828 E837. paradigm. Biol Rev Camb Philos Soc 91:673–711. 12. Cornelis G, et al. (2014) Retroviral envelope syncytin capture in an ancestrally di- 46. Blackburn DG, Flemming AF (2012) Invasive implantation and intimate placental as- verged mammalian clade for placentation in the primitive Afrotherian tenrecs. Proc sociations in a placentotrophic African lizard, Trachylepis ivensi (Scincidae). J Morphol Natl Acad Sci USA 111:E4332–E4341. 273:137–159. 13. Redelsperger F, et al. (2014) Capture of syncytin-Mar1, a fusogenic endogenous ret- 47. Flemming AF, Branch WR (2001) Extraordinary case of matrotrophy in the African roviral envelope gene involved in placentation in the Rodentia squirrel-related clade. skink Eumecia anchietae. J Morphol 247:264–287. J Virol 88:7915–7928. 48. Long JA, Trinajstic K, Young GC, Senden T (2008) Live birth in the Devonian period. 14. Vernochet C, et al. (2014) The captured retroviral envelope syncytin-A and syncytin-B Nature 453:650–652. genes are conserved in the Spalacidae together with hemotrichorial placentation. 49. Wooding P, Burton G (2008) Comparative Placentation - Structures, Functions and Biol Reprod 91:148. Evolution (Springer, Berlin). 15. Cornelis G, et al. (2015) Retroviral envelope gene captures and syncytin exaptation for 50. Aiewsakun P, Katzourakis A (2017) Marine origin of retroviruses in the early Palae- placentation in marsupials. Proc Natl Acad Sci USA 112:E487–E496. ozoic Era. Nat Commun 8:13954. 16. Blackburn DG (2015) Evolution of vertebrate viviparity and specializations for fetal 51. Henzy JE, Gifford RJ, Kenaley CP, Johnson WE (2017) An intact retroviral gene con- nutrition: A quantitative and qualitative analysis. J Morphol 276:961–990. served in spiny-rayed fishes for over 100 My. Mol Biol Evol 34:634–639. 17. Roberts RM, Green JA, Schulz LC (2016) The evolution of the placenta. Reproduction 52. Metallinou M, et al. (2016) A single origin of extreme matrotrophy in African ma- 152:R179–R189. buyine skinks. Biol Lett 12:20160430. 18. Blackburn DG, Vitt LJ, Beuchat CA (1984) Eutherian-like reproductive specializations 53. Ramírez-Pinilla MP, Rueda ED, Stashenko E (2011) Transplacental nutrient transfer in a viviparous reptile. Proc Natl Acad Sci USA 81:4860–4863. during gestation in the Andean lizard Mabuya sp. (Squamata, Scincidae). J Comp 19. Jerez A, Ramírez-Pinilla MP (2001) The allantoplacenta of Mabuya mabouya (Sauria, Physiol B 181:249–268. Scincidae). J Morphol 249:132–146. 54. Gilbert DG (2013) Gene-omes built from mRNA-seq not genome DNA. Seventh Annu 20. Blackburn DG, Vitt LJ (2002) Specializations of the chorioallantoic placenta in the Arthropod Genomics Symposium, 10.7490/f1000research.1112594.1. Brazilian scincid lizard, Mabuya heathi: A new placental morphotype for reptiles. 55. Langmead B, Salzberg SL (2012) Fast gapped-read alignment with Bowtie 2. Nat J Morphol 254:121–131. Methods 9:357–359. 21. Blackburn DG, Flemming AF (2010) Reproductive specializations in a viviparous Af- 56. Roberts A, Pachter L (2013) Streaming fragment assignment for real-time analysis of rican skink and its implications for evolution and conservation. Herpetol Conserv Biol sequencing experiments. Nat Methods 10:71–73. 5:263–270. 57. Hou Z-C, et al. (2012) Elephant transcriptome provides insights into the evolution of 22. Flemming AF, Blackburn DG (2003) Evolution of placental specializations in viviparous – African and South American lizards. J Exp Zool A Comp Exp Biol 299:33–47. eutherian placentation. Genome Biol Evol 4:713 725. 23. Murphy BF, Thompson MB (2011) A review of the evolution of viviparity in squamate 58. Brandley MC, Young RL, Warren DL, Thompson MB, Wagner GP (2012) Uterine gene reptiles: The past, present and future role of molecular biology and genomics. J Comp expression in the live-bearing lizard, Chalcides ocellatus, reveals convergence of Physiol B 181:575–594. squamate reptile and mammalian pregnancy mechanisms. Genome Biol Evol 4: – 24. Jerez A, Ramírez-Pinilla MP (2003) Morphogenesis of extraembryonic membranes and 394 411. placentation in Mabuya mabouya (Squamata, Scincidae). J Morphol 258:158–178. 59. Lynch VJ, et al. (2015) Ancient transposable elements transformed the uterine reg- 25. Ramírez-Pinilla MP, De Pérez G, Carreño-Escobar JF (2006) Allantoplacental ultra- ulatory landscape and transcriptome during the evolution of mammalian pregnancy. – structure of an Andean population of Mabuya (Squamata, Scincidae). J Morphol 267: Cell Rep 10:551 561. 1227–1247. 60. Eisenberg E, Levanon EY (2013) Human housekeeping genes, revisited. Trends Genet – 26. Leal F, Ramírez-Pinilla MP (2008) Morphological variation in the allantoplacenta 29:569 574. within the genus Mabuya (Squamata: Scincidae). Anat Rec (Hoboken) 291:1124–1139. 61. Larsson A (2014) AliView: A fast and lightweight alignment viewer and editor for 27. Wooding FBP, Ramirez-Pinilla MP, Forhead AS (2010) Functional studies of the pla- large datasets. Bioinformatics 30:3276–3278. centa of the lizard Mabuya sp. (Scincidae) using immunocytochemistry. Placenta 31: 62. Guindon S, et al. (2010) New algorithms and methods to estimate maximum- 675–685. likelihood phylogenies: Assessing the performance of PhyML 3.0. Syst Biol 59: 28. Pinto-Sánchez NR, Calderón-Espinosa ML, Miralles A, Crawford AJ, Ramírez-Pinilla MP 307–321. (2015) Molecular phylogenetics and biogeography of the Neotropical skink genus 63. Lefort V, Longueville J-E, Gascuel O (2017) SMS: Smart model selection in PhyML. Mol Mabuya Fitzinger (Squamata: Scincidae) with emphasis on Colombian populations. Biol Evol 34:2422–2424. Mol Phylogenet Evol 93:188–211. 64. Vidal N, Hedges SB (2005) The phylogeny of squamate reptiles (lizards, snakes, and 29. Zhao R, Guerrah A, Tang H, Zhao ZJ (2002) Cell surface glycoprotein PZR is a major amphisbaenians) inferred from nine nuclear protein-coding genes. C R Biol 328: mediator of concanavalin A-induced cell signaling. J Biol Chem 277:7882–7888. 1000–1008. 30. Zannettino AC, et al. (2003) Novel mesenchymal and haematopoietic cell isoforms of 65. Kumazawa Y (2007) Mitochondrial genomes from major lizard families suggest their the SHP-2 docking receptor, PZR: Identification, molecular cloning and effects on cell phylogenetic relationships and ancient radiations. Gene 388:19–26. migration. Biochem J 370:537–549. 66. Henzy JE, Coffin JM (2013) Betaretroviral envelope subunits are noncovalently asso- 31. Roubelakis MG, Martin-Rendon E, Tsaknakis G, Stavropoulos A, Watt SM (2007) The ciated and restricted to the mammalian class. J Virol 87:1937–1946. murine ortholog of the SHP-2 binding molecule, PZR accelerates cell migration on 67. de Parseval N, Lazar V, Casella J-F, Benit L, Heidmann T (2003) Survey of human genes fibronectin and is expressed in early embryo formation. J Cell Biochem 102:955–969. of retroviral origin: Identification and transcriptome of the genes with coding ca- 32. Jia D, et al. (2014) Amplification of MPZL1/PZR promotes tumor cell migration pacity for complete envelope proteins. J Virol 77:10414–10422. through Src-mediated phosphorylation of cortactin in hepatocellular carcinoma. Cell 68. Perro M, et al. (2010) Generation of multi-functional antigen-specific human T-cells by Res 24:204–217. lentiviral TCR gene transfer. Gene Ther 17:721–732.

10 of 10 | www.pnas.org/cgi/doi/10.1073/pnas.1714590114 Cornelis et al. 127 Supporting Information

Cornelis et al. 10.1073/pnas.1714590114 SI Methods with anti–Mab-Env1 serum or the corresponding preimmune serum. Conservation of Mab-Env1 in Scincidae. Amplification of the Mab- Staining was visualized using the peroxidase/diaminobenzidine Env1 ORF or an internal fragment was performed using touch- Mouse PowerVision kit (ImmunoVision Technologies). All slides down PCR with AccuPrime polymerase (Thermo Fisher) on were immunostained in cover plates on the same day to obtain a 200 ng of genomic DNA. The following program was used: standardized intensity of staining. 10 cycles of 10 s at 94 °C, 30 s at 60 °C, −1°Cpercycle,and2min at 68 °C, followed by 40 cycles of 10 s at 94 °C, 30 s at 50 °C, and Mab-Env and MPZL1 Expression Vectors. The Mab-Env1 to Mab- 2 min at 68 °C with a final elongation step of 7 min at 68 °C. Env4 fragments that were PCR-amplified from placental cDNA Amplified fragments were purified and sequenced directly when of Mabuya sp. IV using the cloning primers listed in Table possible or were cloned in pGEM-T Easy (Promega) before se- S1 were digested with EcoRI and MluI, and the PCR products quencing. Primers are described in Table S1. were cloned into the phCMV vector (GenBank accession no. AJ318514, a gift from F.-L. Cosset, Ecole Normale Supérieure Real-Time RT-qPCR. Mab-Env mRNA expression was determined by de Lyon, France). Similarly, Mabuya MPZL1 was amplified from RT-qPCR. Reverse transcription was performed with 500 ng of placental cDNA using the cloning primers listed in Table S1 and DNase-treated RNA as in ref. 67. PCR was carried out with 5 μL was cloned into the phCMV vector. of diluted (1:20) cDNA in a final volume of 25 μL using the Fast For the lentiviral expression vector, Mab-Env1 and syncytin- SYBR Green PCR Master Mix (Qiagen) in an ABI PRISM Car1 ORFs were PCR amplified from phCMV vectors contain- 7000 sequence detection system. Primers are listed in Table S1. ing the respective ORF (from the present study and ref. 10) Primer couple efficiency was tested using a standard curve obtained using the lentiviral vector primers listed in Table S1 and were with serially diluted RT samples (and proved to be >95%), and cloned into a modified CSGW lentiviral vector (68). primer specificity was verified with single-peak dissociation curves. Transcript levels were normalized relative to the amount of the Pseudotyping and Fusion Assays. Mab-Env–pseudotyped MLV par- 6 housekeeping RPL19 gene (ribosomal protein L19), using the ticles were produced by cotransfecting 10 293T cells with 1 μg −ð − Þ μ ΔCt method ð2 Ctenv CtRPL19 Þ. Samples (three individuals for CMVi (a vector expressing MLV gag-pol), 1.5 gMFG-nlsLacZ(a μ each organ) were each assayed in duplicate for each env and for retroviral vector expressing nlsLacZ), and 0.5 gMab-Env1,Mab- the housekeeping gene. Env2, A-MLV, or human Syncytin-1 (HERV-W Env) expression vectors using the jetPRIME transfection kit (Polyplus). Super- In Situ Hybridization. Freshly collected Mabuya sp. IV placentas natants were harvested 48 h after transfection, filtered (PVDF were fixed in 4% paraformaldehyde at 4 °C and were embedded membranes, pore size 0.45 μm; Millipore), supplemented with 4 in paraffin. Serial sections (7 μm) were either stained with H&E Polybrene (8 μg/mL), transferred to target cells (5–8 × 10 cells per or were used for in situ hybridization. Three PCR-amplified well), and spinoculated at 1,200 × g for 2.5 h. X-gal staining was fragments of Mab-Env1 (391 bp, 396 bp, and 482 bp) and two performed 3 d postinfection. Cell–cell fusion assays were performed 6 fragments of Mabuya MPZL1 (268 bp and 262 bp) (primers are by cotransfecting 2 × 10 cells with 2.75 μg of nlsLacZ expression listed in Table S1) were cloned into pGEM-T Easy (Promega) vector and 2.75 μgofMab-Env1 or MMTV Env expression vector for in vitro synthesis of the antisense and sense riboprobes (or empty vector as control), using Lipofectamine LTX reagent generated with SP6 RNA polymerase and digoxigenin 11-UTP (Invitrogen). Forty-eight hours after transfection, cells were treated (Roche Applied Science) after cDNA template amplification. with preheated (37 °C) DPBS·HCl at pH 4 or 7 for 5 min at 37 °C. Sections were processed, hybridized at 42 °C overnight with the DPBS was then replaced by preheated (37 °C) normal supple- pooled riboprobes, and incubated further at room temperature mented growth medium, and cells were incubated at 37 °C for 6 h for 2 h with alkaline phosphatase-conjugated anti-digoxigenin before X-gal staining. All cell lines are described in ref. 11 and antibody fragments (Roche Applied Science). Staining was were grown in DMEM supplemented with 10% FCS (Invitrogen), revealed with nitroblue tetrazolium and 5-bromo-4-chloro-3- 100 mg/mL streptomycin, and 100 U/mL penicillin, at 37 °C and indoyl phosphate phosphatase alkaline substrates, as indicated 6% CO2. by the manufacturer (Roche Applied Science). MPZL1 Phosphorylation Assay. MCF10A cells were obtained from Antibody Production and Immunohistochemistry. ADNAfragment ATCC (CRL-10317) and grown in DMEM/Ham’sF12medium containing 201 amino acids of the Mab-Env1 envelope SU subunit (1:1), supplemented with 5% horse serum (Invitrogen), 100 mg/ (amino acids 30–231; for primers see Table S1) was inserted into mL streptomycin, 100 U/mL penicillin, 10 μg/mL insulin, 20 ng/mL the pET28b (Novagen) prokaryotic expression vector and ex- EGF, 0.5 μg/mL hydrocortisone, and 100 ng/mL cholera toxin, at pressed in BL21(DE3) bacteria. The recombinant C-terminally His- 37 °C and 6% CO2. Populations were obtained by transduction tagged protein was purified from bacteria lysates by nickel affinity using VSV-G–pseudotyped HIV particles and a modified CSGW chromatography. Mouse immunization was performed in accor- lentiviral vector (68) expressing the gene of interest and a dance with standard procedures. Sera containing polyclonal anti- hygromycin resistance factor. Cells were selected for a minimum of bodies were recovered independently from 10 mice and were tested 1 wk in complete culture medium with 50 μg/mL Hygromycin-B by Western blot analyses using lysates of 293T cells transiently (AppliChem) and were serum-starved for 2 h before being ex- transfectedwithaMab-Env1 expression vector. The most specific posed (or not) to 100 μg/mL of Con A (Sigma-Aldrich) for 1 h. sera of the Western blot experiment were then tested for signal Cells were lysed using radioimmunoprecipitation assay (RIPA) strength and absence of background staining by performing im- buffer (Thermo Fisher) supplemented with Halt Phosphatase munohistochemistry on 293T cells transfected with an empty Inhibitor Mixture (Thermo Fisher). Proteins were deglycosylated vector or a Mab-Env1 expression vector. For immunohistochem- using PNGase F (New England Biolabs), separated by SDS/ istry, paraffin sections were processed for heat-induced antigen PAGE, and transferred to nitrocellulose membranes. Tris-buffered retrieval (Tris EDTA, pH9; Abcam) and incubated overnight saline/0.1% Tween/5% BSA buffer was used for blocking and

Cornelis et al. www.pnas.org/cgi/content/short/1714590114 1of7 128 antibody binding. Specific antibodies against MPZL1 (no. 9893) A significant proportion (4,064, 40.3%) of the genes expressed in and phospho-MPZL1 Y263 (no. 8088) were purchased from Cell the Mabuya placenta are also expressed in eutherian placentas. Signaling Technologies. The anti–tubulin-γ antibody was pur- Functional Gene Ontology (GO) annotations highlight biological chased from Sigma-Aldrich (T5326). functions associated with placenta biology, such as embryonic development [embryonic morphogenesis, GO:0048598, adjusted − GenBridge Screening. A23-derived GeneBridge4 RH clones (a gift P value (adjp) = 4.5 × 10 4; organ morphogenesis, GO:0009887, − from J. Weissenbach, Genoscope, Evry, France) (39) were grown in adjp = 7.5 × 10 4; in utero embryonic development, GO:0001701, − the DMEM medium described above, supplemented with HAT adjp = 1.7 × 10 3; embryo development ending in birth or egg μ μ μ −3 (100 M hypoxanthine, 0.4 M aminopterin, and 16 M thymidine). hatching, GO:0009792, adjp = 2.2 × 10 ; chordate embryonic – − Mab-Env1 pseudotyped MLV particles were produced by cotrans- development, GO:0043009, adjp = 2.2 × 10 3]; or regulation of 6 μ μ fecting 10 293T cells with 2.25 g CMVi, 2.25 g MFG-nlsLacZ, cellular migration (negative regulation of cell motility, GO:2000146, μ − and 0.25 g of Mab-Env1 expression vector using the Lipofectamine adjp = 1.5 × 10 4; negative regulation of cell migration, − LTX reagent (Invitrogen). Supernatants from the transfected cells GO:0030336, adjp = 1.5 × 10 4; positive regulation of cell motility, − were harvested 48 h after transfection, filtered through PVDF GO:2000147, adjp = 3.4 × 10 4; positive regulation of cell migra- μ − membranes (pore size 0.45 m) (Millipore), supplemented with tion, GO:0030335, adjp = 3.9 × 10 4; leukocyte migration, μ − Polybrene (8 g/mL), and transferred to the RH clones seeded in 24- GO:0050900, adjp = 4.5 × 10 4). Accordingly, a significant fraction well plates (2.5 × 104 cells per well) the day before infection, followed × of enriched mouse phenotypes linked to this set of genes is asso- by spinoculation at 1,200 g for 2.5 h at 25 °C. X-gal staining was ciated with abnormal development and organ morphology (in- − performed 3 d postinfection. A threshold value for positive infection cluding abnormal extraembryonic tissue, adjp = 1.4 × 10 17; of 100 focus-forming units/mL gave the best logarithm of the odds −14 abnormal embryonic tissue, adjp = 1.6 × 10 ; abnormal em- scores and gene localization results after analysis of the data using the − bryogenesis, adjp = 2.8 × 10 10; and abnormal embryo size, adjp = RH map program (41) (+4 if the presence of the marker correlates − 5.3 × 10 13) and abnormal immune response (abnormal immune with infection or the absence of the marker correlates with absence − system, adjp = 1.2 × 10 14; abnormal adaptive immunity, adjp = of infection, and −2ifitdoesnotcorrelate). − − 4.8 × 10 15; and abnormal antigen presenting, adjp = 3.0 × 10 11). Finally, mouse KEGG pathway analysis reveals the specific en- SI Transcriptome Analysis − richment of the Rap1 signaling pathway (adjp = 2.7 × 10 4) asso- Placental transcripts were assembled using de novo assembly – methods in the absence of a reference genome (Methods). Con- ciated with trophoblast invasion and cell cell fusion. catenated assemblies led to the identification of 4,020,383 highly In addition, since the Mabuya placenta is composed of both redundant transcripts (Fig. S1A). Reduction of the redundancy of maternal and fetal tissues, in Dataset S4 we compared the tran- the transcriptome led to the identification of 72,763 transcripts scriptome of the Mabuya placenta with those of the pregnant ranging from 123 bp to 24.4 kb (N50 = 2,869 bp; median size = uterus of another live-bearing lizard (C. ocellatus)(58)andmam- 1.1 kb) (Fig. S1A). Of these, 26,690 transcripts (36.7%) could be malian endometrial tissues (59). Five hundred twelve genes were positively matched to a known protein gene from the Refseq expressed in all three tissues [i.e., mammalian endometrium, Vertebrate database. A set of 21,253 transcripts shows significant pregnant (but not nonpregnant) C. ocellatus uterus, and Mabuya expression [more than two tags per million (34)], 16,398 of which placenta]. Again, functional annotations reveal specific enrich- were positively annotated (77.2%) and corresponded to 13,189 ments of genes involved in the regulation of cell migration (positive = × −3 unique genes (Fig. S1B), of which 10,079 were nonhousekeeping regulation of cell migration, GO:0030335, adjp 5.8 10 ;reg- genes. Only this subset of transcripts was conserved for further ulation of endothelial cell migration, GO:0010594, adjp = 3.8 × −3 −4 analysis. This subset of transcripts contains a range of genes an- 10 ; leukocyte migration, GO:0050900, adjp = 7.7 × 10 ), blood −4 notated as “placental” genes with key biological functions during coagulation (blood coagulation, GO:0007596, adjp = 2.9 × 10 ; −4 placenta development, such as core placental transcription factors coagulation, GO:0050817, adjp = 2.9 × 10 ; hemostasis, −4 (e.g., Cdx2, Hand1, Gata3, Stat3, Essrb, Tfa2pc,andGcm1), pla- GO:0007599, adjp = 2.9 × 10 ), and response to hormone stim- ulus (cellular response to hormone stimulus, GO:0032870, adjp = cental hormones (e.g., Prl, Crebrf, Ghr,andGhrh), iron transporters − Tfr2 Slc11a2 Fth1 Gja1 2.9 × 10 4; response to steroid hormone, GO:0048545, adjp = (e.g., , ,and ), gap junction proteins (e.g., , − Gja4, Gja5, Gja8,andGjc2), interleukins and interleukin recep- 2.9 × 10 4; cellular response to steroid hormone stimulus, − tors (e.g., Il-15 and Il-11ra), metalloproteinases (e.g., Mmp-2 and GO:0071383, adjp = 6.2 × 10 4). Mmp-9), and angiogenesis factors (e.g., Vegf, Vegfb, Il-11ra, Dll4, Taken together, these strongly argue for the convergent and Pgf)(DatasetS2). evolution of specific transcriptomes within both maternal and We further compared the transcriptome of the Mabuya pla- extraembryonic fetal tissues associated with live-bearing and centa with those of eutherian mammalian placentas [genes embryonic development in the distantly related mammalian expressed in both human and mouse placenta (57)] (Dataset S3). and Mabuya species.

Cornelis et al. www.pnas.org/cgi/content/short/1714590114 2of7 129 A transcript size Counts Counts (all assemblies) (Evigene assembly)

1.5x103

all assemblies assembly 6x104 all transcripts Evigene annotated transcipts

1x103

4x104

0.5x103 2x104

0 0 0.1 1 size (kb) 10 100 B Counts transcript expression

2.0x103 all transcripts

annotated transcripts

1.5x103

1.0x103

0.5x103

0 −5 02 5 10 15 log2(tpm)

Fig. S1. Statistics of Mabuya placenta transcriptomes. (A) Density plot of transcript sizes expressed in kilobases of all merged de novo assemblies (orange, left y axis) or nonredundant Evigene reconstructed transcripts (purple and blue, right y axis). Both the entire Evigene transcriptome (purple) and the subset of BLAST-annotated transcripts (blue) are represented. Note that most unannotated transcripts are among the smallest transcripts (<1 kb), while most transcripts larger than 1 kb could be positively annotated by BLAST. (B) Density plot of nonredundant Evigene transcript (purple) and BLAST-annotated transcript (blue) expression in tags per million (tpm). Again, most unannotated transcripts correspond to poorly expressed (<2 tpm) transcripts.

Cornelis et al. www.pnas.org/cgi/content/short/1714590114 3of7 130 Fig. S2. Recovered Mab-Env internal fragments from non-Mabuya species, as listed in Fig. 5. Functional domains are indicated using the color scheme in Fig. 2, and stop codons or frameshift-inducing deletions are highlighted in red. The corresponding fragment from Mabuya sp. IV is provided as reference; asterisks indicate nucleotide identities; identity percentages range from 79 to 89%.

Cornelis et al. www.pnas.org/cgi/content/short/1714590114 4of7 131 A A23 (tk-) HFL121 hamster cells human fibroblasts lethal nlsLacZ viral RNA irradiation MLV fusion (3,000 rads) Mab-Env1 viral core selection seeding infection with transfection into plate MLV-Mab-Env pseudotype GeneBridge4 X-gal staining and 293T panel of 93 hybrids determination of viral titer producer cells (LacZ+ ffu) B 5 10

4 10

3 10 Viral titer (ffu/ml) titer (ffu/ml) Viral

2 10 1 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 hybrid clone number A23 C Chr.1 p31.1 1q12 q41 4344

167 Mb 168 Mb 169 Mb Marker chr1 (# score) #5 #6 #7#1,2,4 #3 #8 FMO9P POGK GPA33 CREG1 MPZL1 GPR161 XCL2 DPT ATP1B1 SLC19A2 SELL Transmembrane TADA1 MAEL POU2F1 ADCY10 TIPRL XCL1 NME7 F5 genes ILDR2 RBSG4 CD247 DCAF6 SFT2D2 BLZF1 SELP DUSP27 RCSD1 MPC2 TBX19 CCDC181 SELE

Fig. S3. Rationale of the retroviral pseudotyping assay used for screening the GeneBridge4 human/Chinese hamster radiation hybrid panel and mapping of the candidate receptors for Mab-Env1. (A) The 93 hybrid clones of the GeneBridge4 panel generated in ref. 39 as schematized on the Left, were seeded and infected with viral pseudotypes produced by cotransfecting 293T cells with expression vectors for the MLV core, the Mab-Env1 protein, and a lacZ-containing retroviral transcript (Right). Viral titers were determined 3 d after infection by X-gal staining of the cells. (B) Viral titers obtained in a representative experiment simultaneously assaying the 93 clones of the panel (and control A23 cells, at the right end of the x axis). Each cell clone was infected in duplicate and examined for the number of lacZ+ foci. The resulting matrix was derived, and data were compared with a known matrix of specific markers situated across the human genome (SI Methods). (C) Localization of the identified high-scoring genetic markers (nos. 1–8) on chromosome 1, showing the positions of the genes present within 3 Mb, centered on the best-scoring marker, as determined using the University of California, Santa Cruz (UCSC) Genome Browser. The transmembrane candidate receptor genes are highlighted in red.

Cornelis et al. www.pnas.org/cgi/content/short/1714590114 5of7 132 Table S1. Primers Purpose Primer

RT-qPCR Mab-Env1-F 5′-CCCAAGAAATTCATAGGTGT Mab-Env1-R 5′-GCCGCCTCCTTGAACATA Mab-Env2-F 5′-TTTTGACCACACCAGCTATCT Mab-Env2-R 5′-TTTTGGAGACTGCTTGTAAGA Mab-Env3-F 5′-AGCACCAGTCCACATTCTAA Mab-Env3-R 5′-GCAGCTGATAGGTAGGTTTT Mab-Env4-F 5′-CCCTGTTGATCCACCATTA Mab-Env4-R 5′-AGCAATTGAGCATCTGTGT Mabuya-MPZL1-F 5′-TTACCTGGAACTTTCAAGCA Mabuya-MPZL1-R 5′-CCCAACTGATTCTGTCCTT Mabuya-RPL19-F 5′-CCCAATGAAACAAATGAG Mabuya-RPL19-R 5′-TCGTGAGTGAACAGTCACTG Conservation Mab-Env1-Int-F 5′-GCCAATTACCACAAAGGRTGT Mab-Env1-Int-R 5′-GTTTGTTCTGCATAAATYTGTAARGC Mab-Env1-ORF-F 5′-GACAACGATCTCAAGCTACTG Mab-Env1-ORF-R 5′-TTCTGCATAAATCTGTAAGGC Mabuya-BDNF-F 5′-ACCATCCTTTTCCTKACTATGG Mabuya-BDNF-R 5′-CTATCTTCCCCTTTTAATGGTC Cloning Mab-Env1-F-EcoRI 5′-ATACATGAATTCGACAACGATCTCAAGCTACTG Mab-Env1-R-MluI 5′-ATACATACGCGTTTCTGCATAAATCTGTAAGGC Mab-Env2-F-EcoRI 5′-ATACATGAATTCAACGATCTCAAGCTACTGTTC Mab-Env2-R-MluI 5′-ATACATACGCGTAACCCAAGCAAAACTGTAATA Mab-Env3-F-XhoI 5′-GTACACTCGAGTCCTAAACCAATTCATTGTA Mab-Env3-R-MluI 5′-GTACTACGCGTATATTGATTACATACTATTGGA Mab-Env4-F-EcoRI 5′-ATACATGAATTCGAGTGTAAACAGTAGGCCAAA Mab-Env4-R-MluI 5′-ATACATACGCGTCCGTTCCATTCTTATCAGTT Mabuya-MPZL1-F-XhoI 5′-NNNNNNCTCGAGATGGCAGCGGCAACG Mabuya-MPZL1-R-MluI 5′-NNNNNNACGCGTTCAGTTCTTCCGAATGTCAGCA Lentiviral vector Mab-Env1-F-BamHI 5′-NNNNNNGGATCCGCCACCATGAGTTTTTCTGTAGCTAAAAAGTTGTTG Mab-Env1-R-NotI 5′-NNNNNNGCGGCCGCCTAATTAAATCTAAAAGAAGCAGGAAGTGG SynCar1-F-BamHI 5′-NNNNNNGGATCCGCCACCATGAAACTTCTAAGATCTTTCGCTTTTC SynCar1-R-NotI 5′-NNNNNNGCGGCCGCTCAGACATGTGCATGAGGCTT In situ hybridization Mab-Env1-HIS-F1 5′-ATGTTGTAGAAATGTATTCCCC Mab-Env1-HIS-R1 5′-TAACACCTATGAATTTCTTGGG Mab-Env1-HIS-F2 5′-TAACACCTCTAACTGGAACTGG Mab-Env1-HIS-R2 5′-TGGGAGGGTACATAAATTTCTT Mab-Env1-HIS-F3 5′-CTTCATTTGGAATTTCTACCTC Mab-Env1-HIS-R3 5′-TTCAGCTTGATACTCTCTGGC Mabuya-MPZL1-HIS-F1 5′-GAACAATGGGTGGCAAAG Mabuya-MPZL1-HIS-R1 5′-ACGTCTGTCGATGTGAATGTA Mabuya-MPZL1-HIS-F2 5′-AGGACAGAATCAGTTGGGCT Mabuya-MPZL1-HIS-R2 5′-TACACAGCGCACACAATAAGA Antibody production Mab-Env1-SU-F-NcoI 5′-ATACATCCATGGAAGCTCACAATTTGACTTATGG Mab-Env1-SU-R-XhoI 5′-ATACATCTCGAGTACTACTCCTAAAGCTACTGATTC Provirus amplification Mab-Env1-provirus-F1 5′-GTCATGCTCACAGAAGCCAGA Mab-Env1-provirus-R1 5′-CATGGTTTTCTGAAGAAGCAAGCC

Underscoring shows the position of the restriction enzyme sites in the primers.

Dataset S1. RT sequences used for the phylogenetic analysis in Fig. 4B

Dataset S1

Cornelis et al. www.pnas.org/cgi/content/short/1714590114 6of7 133 Dataset S2. Characterization and identification of Mabuya placental transcripts

Dataset S2

Dataset S3. Comparison of Mabuya and mammalian placenta transcriptomes

Dataset S3

Dataset S4. Comparison of Mabuya, Chalcides, and mammalian maternal tissue transcriptomes

Dataset S4

Cornelis et al. www.pnas.org/cgi/content/short/1714590114 7of7 134 2. Implication of ERV env genes in placental structural transitions in Hyaenidae

2.1 Summary

Like all other carnivorans, all extant members of the Hyaenidae species present the shared carnivoran Syncytin-Car1, but their placental organization is very different from that of other carnivorans. Hyaenidae present a highly invasive hemochorial placenta in which the maternal tissues are completely disrupted at term and maternal blood is in direct contact with placental tissues, whereas all other carnivorans present the less invasive endotheliochorial phenotype in which maternal vessels preserve their integrity. This structural transition from an endotheliocho- rial to a hemochorial placenta must have occurred recently as Hyaenidae split from the closest carnivoran family about 30 Mya. To investigate the involvement of ERV env genes in this structural transition, and since there is no sequenced genome available, we performed high-throughput RNA sequencing of placental Crocuta crocuta (spotted hyena) transcripts. The placental transcriptome was then screened in silico leading to the identification of three expressed full-length env ORFs, which contain all major characteristic Env domains. As could be expected, one of these three expressed env was identified as the previously described hyena syncytin-Car1. The other two, Hyena-Env2 and Hyena-Env3 are distinct from previously described syncytins, although Hyena-Env3 closely resembles the non-syncytin Canis-Env3 described previously in dog. Phylogenetic analyses of these two new envelopes identifies them as gamma-type Env. Since the investigated placental structure transition occurred only in Hyaenidae, we studied the conservation state of the two new Hyena-Env among carnivorans (syncytin-Car1 having already been shown to be conserved under purifying selection). PCR amplification of an internal fragment or the whole Hyena-Env2 ORF was possible only using genomic DNA from any of the four extant species of hyena, no amplification was observer outside of Hyaenidae indicating that Hyena-Env2 is hyena-specific. Additionally, sequence conservation among hyenas was high (>98.4% nucleotide identity) with a low ratio of non-synonymous/synonymous mutations, suggesting that Hyena-Env2 is under purifying selection in this group. Hyena-Env3 on the other hand was shown both by PCR and in silico analysis to be present in many other carnivorans,

135 2. Implication of ERV env genes in placental structural transitions in Hyaenidae

mainly Caniformia but also a non-hyena Feliformia. Closer comparison of Hyena-Env3 and the numerous related coding and non-coding copies present in Caniformia genomes revealed that while Canis-Env3 is related to Hyena-Env3, the former is a non-functional copy, having lost the signal peptide necessary for membrane addressing during Env synthesis. A dog provirus associated to Canis-Env3 displays two LTRs with 100% sequence identity, suggesting that the Hyena-Env3/Canis-Env3 associated retrovirus was actively endogenizing until recently, or might even still be active. Since only Hyena-Env2 was shown to be hyena specific, we focused on this gene for our study. Using an inverse PCR approach we identified the 3’ flanking sequence of the inserted Hyena- Env2 containing provirus and using the sequence of the corresponding region in the cat genome we designed primers leading to the amplification of the complete 9 kb proviral genome. The 2.1 target site duplication is preserved on both sides of the provirus and the LTRs display 89% sequence identity, in accordance with a relatively recent integration. The gag and pol ORFs are disrupted by several frameshift or non-sense mutation events. The conserved RT domain associated to Hyena-Env2 could be identified and indicates a gammaretroviral origin. In silico analysis and PCR amplification of the orthologous locus in non-hyena genomes confirmed that it is empty, while the amplification of the locus in hyena genomes confirmed that it is present at the same position in all four extant hyena, having therefore been acquire before radiation of modern day Hyaenidae. Characterization of the placental Hyena-Env2 transcript shows that it presents the canonical retroviral splicing but is not initiated in the 5’LTR and instead presents an additional splicing event. Initiation occurs in a host genome region that presents a CpG island in the orthologous cat genome region, suggesting that transcription of Hyena-Env2 is driven by a host promoter. Accordingly, in the cat genome this CpG island drives transcription of a host gene, albeit in the opposite orientation. Placental expression of the three hyena genes was studied by in situ hybridization and revealed that syncytin-Car1 is expressed in the ST, with a similar expression profile to that described previously in cat and dog. Hyena-Env3 hybridization only led to very faint staining, potentially due to its low level of expression, that hinted at expression mainly in the ST. For Hyena-Env2, the sense RNA probe used as a negative control displayed very strong staining, revealing the presence of a Hyena-Env2 containing anti-sense transcript. Still, the anti-sense RNA probe shows a specific staining of the cell layer underlying the fused ST and containing the unfused CT cells. This was also confirmed by immunohistochemistry using anti-Hyena-Env2 antibodies (obtained by inoculating mice with Hyena-Env2 SU peptides). An expression of Hyena-Env2 in the CT specifically is in accordance with a role in placental physiology. Finally, we investigated the fusogenic potential of all three hyena genes. Syncytin-Car1 behaved as was shown previously, being able to mediate cell infection by pseudotyped virions with a strong tropism towards carnivoran cells. Hyena-Env2 displayed no fusogenic properties in the pseudotyping or cell-cell fusion assay. Hyena-Env3 on the other hand was able to mediate infection of pseudotyped MLV particles, at levels comparable to those achieved with Syncytin-

136 euirifcinpol hti uhmr pcfi hnta fohrEvi hsgroup. this the in on receptor Env/SLC1A5 Env light of understanding shed other binding. the might to of contribute SU and Hyena-Env3 that envelope the of than of domains specificity specific receptor conserved the more in much observed is variations that The displays profile it infection although peculiar ERV group, interference a fusogenic SLC1A5 well-known a the to identified belongs also and hyena-specific have We Envs. non-fusogenic conserved with described case syncytin-like previously the several was the as joins structural Hyena-Env2 mechanism, now potential placental For a Syncytin-Mab1/MPZL1. provide observed might receptor the or cognate expression of its of in reason identification it the be of be carnivorans, implication to An other unlikely transition. of therefore that is from and differ fusogenicity, to Hyena seem role. this not fulfill to does seems Syncytin-Car1 Syncytin-Car1 fusogen placental an the as from placentation hyena transition in for the in as role timeframe same potential the in Hyena-Env2 a occurred with that placenta accordance hemochorial a of in to profile endotheliochorial is expression interface The family. maternofetal this the of at species all among conserved and hyenas binding. domains receptor conserved in highly implicated in be to residues described shifted previously and shows members mutations group several interference presents SLC1A5 Hyena- Hyena-Env3 other with case of that the that not with is SU this its why of unclear alignment is though it Env3, and the envelope usually the only receptor to cognate that tropism as broad confirmed SLC1A5 very of a SCL1A5 confers Use dog lines. clusters cell and restrictive Hyena-Env3 cat of infection hyena, tree, of allowed three phylogenetic human, part latter of based as Cloning SU Env. described an previously other these Env on with of and SUs group sequence the interference SU with SLC1A5 Hyena-Env3 domains the the conserved of shares of infection analysis is mediate Interestingly, that to shows tested. unable lines being cell tropism, non-carnivoran carnivoran any strict a presents Hyena-Env3 Car1. ncnlso,w aeietfidanwycpue yctnlk eeta sseicto specific is that gene syncytin-like captured newly a identified have we conclusion, In sntfsgnci u sa,btfsgnct ih o eterl twsselected was it role the be not might fusogenicity but assay, our in fusogenic not is Hyena-Env2 ncl irto/nainpoessi osbeand possible is processes migration/invasion cell in env ..Summary 2.1. Hyena-Env2 env hti not is that capture. 137

2.1 2. Implication of ERV env genes in placental structural transitions in Hyaenidae 2.1

138 icoil/hea/Carnivora / hyena / liochorial § 1 USA fornia, d USA California, Davis, California, of c France b F-94805, Villejuif, Roussy, Gustave 9196, UMR a Conley Alan Funk Mathis pla- and capture gene envelope retroviral Hyena-specific 2.2 mrec mn anvrn fhmcoilpaetto in placentation hemochorial of carnivorans among emergence eateto ouainHat n erdcin colo eeiayMdcn,University Medicine, Veterinary of School Reproduction, and Health Population of Department CNRS Infectieux, et Endogènes Rétrovirus des Moléculaires Pathologie et Physiologie Unité orsodn uhr [email protected] author: corresponding USA California, Stanford, University, Stanford Genetics, of Department address: Present Cali- Berkeley, California, of University Biology, Integrative and Psychology of Department France F-91405, Orsay, Paris-Sud, Université 9196, UMR lcna xrsinascae neouinwt h unique the with evolution in associated expression placental ewrs noeosrtoiu neoepoen/paet eohra endothe- / hemochorial / placenta / protein envelope / retrovirus endogenous Keywords: atr faheaseicrtoia neoegn with gene envelope retroviral hyena-specific a of Capture et evolution centa a,b c tpe Glickman Stephen , ulam Cornelis Guillaume , ..Heaseicrtoia neoegn atr n lcnaevolution placenta and capture gene envelope retroviral Hyena-specific 2.2. d her Heidmann Thierry , a,b,1 éieVernochet Cécile , Hyaenidae (submitted) a,b,§ a,b dl Heidmann Odile , a,b neDupressoir Anne , 139 a,b ,

2.2 2. Implication of ERV env genes in placental structural transitions in Hyaenidae

Abstract Capture of retroviral envelope genes from endogenous retroviruses has played a role in the evolution of mammals, with evidence for the involvement of these genes in the formation of the materno-fetal interface of the placenta. It has been shown that the diversity of captured genes is likely to be responsible for the diversity of placental structures, ranging from poorly invasive (epitheliochorial) to highly invasive (hemochorial) with an intermediate state (endotheliochorial) as found in carnivorans. The latter recapitulate part of this evolution, with the hyena being the sole carnivoran with a hemochorial placenta. Here we performed RNAsequencing on hyena placental transcripts, and searched for endogenous retroviral envelope genes that have been captured specifically in the Hyaenidae clade and are not found in any other carnivoran. We identified an envelope gene that is expressed in the placenta at the level of the materno-fetal interface, as evidenced by in situ hybridization/immunohistochemistry. The gene entry is coincidental with the emergence of the Hyaenidae clade 30 Mya, being found at the same genomic locus in all 4 extant hyena species: Crocuta crocuta, Hyaena hyaena, Parahyaena brunnea and Proteles cristatus. Its coding sequence has further been maintained —under purifying selection- during 2.2 all of Hyaenidae evolution. It is not found in any of the 30 other carnivorans —both Felidae and Canidae- that we screened. This envelope protein does not disclose any fusogenic activity in a series of cell-cell fusion or pseudotyping assays, at variance with the Syncytin-Car1 gene that is found in all carnivorans —including hyena where it is still present, transcriptionally active in the placenta, and fusogenic. Altogether, the present results illustrate the permanent renewal of placenta-specific genes by retroviral captures, and de facto provide a candidate gene for the endotheliochorial to hemochorial transition of Hyaenidae among carnivorans.

Author Summary. The placenta is the most diverse organ among mammals in terms of struc- tural organization, especially at the level of the maternofetal interface where the exchange of gas, waste and nutrients between maternal and fetal blood takes place. Carnivorans represent a rather unique clade, where capture and exaptation of a retroviral envelope, syncytin-Car1, took place about 80 Mya at the time of their radiation. It is conserved among all species, under purifying selection, expressed at the maternofetal interface and fusogenic, consistent with the syncytial structure of the interface and the "endotheliochorial" architecture of all carnivoran placenta -except of hyenas. Remarkably, the latter disclose a highly invasive hemochorial pla- centa, as found in humans, where a breakdown of maternal blood vessel walls results in maternal blood bathing the syncytial maternofetal interface. Here we identified a retroviral envelope gene capture and exaptation that took place about 30 Mya and is coincident with the emergence of the Hyaenidae clade, being conserved in all extant hyena species. It is expressed at the maternofetal interface in addition to the shared syncytin-Car1 gene. This new env gene, not present in any other carnivoran, is a likely candidate to be responsible for the specific structure of the hyena placenta. Introduction

140 ednl nertdadaatdfrsmlresnilpaetlfntosdrn mammalian during functions placental essential similar for adapted ( and envelope integrated retroviral exogenous pendently distinct As 2016). Roberts in al., (reviewed lagomorphs) et in and bathed rodents directly simians, of or placenta endothelial carnivorans) (hemochorial the of blood maternal placenta with (endotheliochorial contact vessels in fuse blood is cells maternal which of trophoblast layer wall syncytiotrophoblast the invasive types, an complex form more to together the (non-invasive in epithelium uterine whereas intact placentation), the of to epitheliochorial simplest apposed nature the simply In are and 1). cells (Fig. number trophoblast types fetal the placental the major type, is four to placentas leading present-day interface, maternofetal between the at traits layers divergent 2011). major 2009, al., the et (Dupressoir of survival One maternal embryonic separate impaired that in layers resulting syncytiotrophoblast spaces, the blood of fetal level and the at fusion cell-cell of lack a with with placentation, for Identification essential sequence. are their in differ and mouse locations the genomic of separate at integrated are origin, clade, mammalian ( the syncytins Mabuya of lizard outside placental capture viviparous syncytin the a in discovered also ( have Marsupials we Recently related distantly 1). the (Fig. in even and 2014), al., et ( Ruminantia 2012), ( and ( Rodentia (2014) Lagomorpha within al. et namely Vernochet for, (2005); al., al. searched et et been Dupressoir Mangeney have placen- 2003; of they orders al., where other et mammals all Blaise in tal identified 2000; then al., were et counterparts Mi homologous Remarkably, 2000; 2007). al., expressed et specifically (Blond are They activity 2003). suppressive al., in et fusogenic Blaise are 2000; placenta, al., the et in Mi 2000; al., et (Blond ciation and 25 primates 2016). of genome Denner, 2013; formation al., placenta et in Lavialle role key in a (reviewed for mammals retroviruses-, eutherian exogenous in of co-opted entry independently been intracellular have for that required -initially ERVs of the of protein (Env) case as lope the considered is be This physiology. can host genes to and retroviral contributing integration, genes copy their single since of preserved exist been have examples reiterated few which highly a of sequences, pool (ERV) large this retrovirus among However, endogenous 2013). al., et (Lavialle non- are remnants sequences functional retroviral to integrated response of in majority sequences or the Consequently, these passively se- mechanisms. either time, defense retroviral recombinations, host Over of and 2015). indels composed Stoye, mutations, being & genetic Mager via mice) altered 2013; in were al., 10% et and Lavialle human in in (reviewed 8% quences (e.g. genome the of portion neto yeoeosrtoiue hogotvrert vlto a e oasignificant a to led has evolution vertebrate throughout retroviruses exogenous by Infection hr lsl eae ucinlpoete,atog hyhv opeeydistinct completely a have they although properties, functional related closely share syncytin-A syncytin-Ory1 ..Heaseicrtoia neoegn atr n lcnaevolution placenta and capture gene envelope retroviral Hyena-specific 2.2. syncytin-Rum1 and > edane l,20) anvr ( Carnivora 2009), al., et Heidmann , -B 0Marsetvl n eandtercdn aaiydrn spe- during capacity coding their retained and respectively Mya 40 nlyalwdteuabgosdmntainta uhgenes such that demonstration unambiguous the allowed finally xvivo ex onlse l,21) fohras( Afrotherians 2013), al., et Cornelis syncytin sas n n fthem, of one and assays, syncytin-Mab1 Omc ipaigipie lcna functions placental impaired displaying mice KO Syncytin-1 syncytin-Mar1 env syncytin-Opo1 onlse l,21) l identified All 2017). al., et Cornelis syncytin ee pert aebe inde- been have to appear genes ) and syncytin-2 syncytin-Car1 syncytin-2 eesegre l (2014)), al. et Redelsperger ee,ecdn h enve- the encoding genes, syncytin-Ten1 onlse l,2015) al., et Cornelis ipasimmuno- displays , nertdit the into integrated syncytin-A oafide bona onlse al., et Cornelis Cornelis cellular and 141 -B

2.2 2. Implication of ERV env genes in placental structural transitions in Hyaenidae

evolution, we proposed the challenging hypothesis that this stochastic acquisition of genes of exogenous origin has been instrumental in establishing the remarkable structural and functional diversity of the mammalian placenta (reviewed in Lavialle et al., 2013). In the past years, we have substantiated this hypothesis by demonstrating the concomitant acquisition of a syncytin and the emergence of a distinct placental organization during mammalian radiation. The emergence of the very specific ruminant synepitheliochorial placenta is concomitant with the acquisition of syncytin-Rum1 in this clade. The implication of syncytin-Rum1 in establishing this structure was further substantiated by the observation that the more important syncytium formation in ovine species, when compared to bovine species, is accompanied by a higher level of expression of the captured syncytin (Cornelis et al., 2013). The Carnivora clade presents a remarkable and even more challenging situation. Indeed, the carnivoran structural organization of the pla- centa is of the endotheliochorial type (Wooding & Burton, 2008), consistent with the capture of syncytin-Car1 being concomitant with the radiation of this clade about 80 Mya (Cornelis et al., 2012), but there is a remarkable exception to this rule: the spotted hyena (Crocuta crocuta, hereafter referred to as hyena) which belongs to the Feliformia subclade within Carnivora clearly 2.2 possesses a hemochorial type placenta (Morton, 1957; Wynn & Amoroso, 1964; Oduor Okelo & Neaves, 1982; Enders et al., 2006). Since the hyena presents the same syncytin as all other Carnivora, i.e., syncytin-Car1 (Cornelis et al., 2012), we hypothesized that the unique placental structural organization of this species might be associated with a recent de novo capture of a hyena-specific syncytin, which would be responsible for the observed transition. Accordingly, we searched for such a gene in the hyena and since the genome of this species has not yet been sequenced, we first performed a high throughput RNAsequencing of placental transcripts to identify putative env-related syncytins by an in silico search, followed by a structural and func- tional characterization of the identified candidates. These analyses resulted in the identification of a newly captured gene, specific to hyenas and expressed in the placenta, consistent with a role of this gene in the stochastic emergence of the hyena-specific placental structure. An alternative hypothesis, involving for instance a severely modified expression of the shared syncytin-Car1 in the hyena seems to be ruled out with no significant difference observed between the hyena gene and the other carnivoran orthologs, including that of dog and cat. The results are discussed in the wider context of retroviral gene capture and exaptation as a central driving force generating the observed structural placenta diversity. Results

High throughput sequencing and in silico search for retroviral env genes within the hyena Crocuta crocuta placenta transcriptome. Since the hyena genome —at variance with that of some of the other carnivorans —has not been sequenced, we first established the transcriptome of the Crocuta crocuta hyena placenta, in collaboration with the French National Sequencing Center (Genoscope, Evry, France), where assembly of the placental transcripts was also per- formed using de novo assembly methods in the absence of a reference genome (see Methods).

142 ragdi re fivsvns fftltsus iedlmt h aenladftlsides. fetal and maternal the delimits line A scale mammals, tissues. in the fetal found of in interfaces invasiveness maternofetal indicated of of as order types 2012), in four time Bininda-Emonds, arranged the illustrating to & Schemes described proportional Nyakatura (Lower) is tree. 2015; currently length the al., type Branch All below et name. below. placental its Cornelis schemes the with My, together the latter in pink, (expressed as the in key For arrowheads by color placenta. indicated a same are events possess the all using which indicated known mammals, is and eutherian type and placental marsupials with carnivorans, the on emphasis with indicated. mammals syncytins of Phylogeny – 1 Figure cytotrophoblast placenta blood vesselblood blood vesselblood mammals 150 epithelium uterine ..Heaseicrtoia neoegn atr n lcnaevolution placenta and capture gene envelope retroviral Hyena-specific 2.2. Upr aml opietemntee eg,payu)sillyn gs and eggs, laying still platypus) (e.g., monotremes the comprise Mammals (Upper)

100 eutherian mammals epithelio- chorial heterologous syncytiumheterologous 50 Invasion of maternal tissuesInvasion ofmaternal synepithelio- chorial 0 0 Mya Monotrema Marsupiala Afrotheria Haplorrhini Rodentia Lagomorpha Odobenidae Mustelidae Canidae Hyaenidae Pantherinae Felinae Ruminantia Ursidae endothelio-

chorial eiomaCaniformia Feliformia

syncytiotrophoblast

Carnivora

syncytin-Ten1 syncytin-Rum1 syncytin-1, -2syncytin-1, syncytin-Opo1 syncytin-A, -B, -Mar1 syncytin-Ory1 syncytin-Car1 haemo- chorial syncytin

mother fetus capture 143

2.2 2. Implication of ERV env genes in placental structural transitions in Hyaenidae

Transcriptome data was deposited at the European Nucleotide Archive (ENA) under the acces- sion number ERA1116685 and is available for further analyses of placental gene expression. The transcriptome was then screened for putative env-derived syncytin genes as described previously for other mammalian genomes (Bénit et al., 2001). All ORFs of over 400 aa (met codon to stop codon) were compared to a set of selected Env sequences, including previously described syncytins (see Methods), using BLAST. As illustrated in Fig. 2, the search yielded 3 sequences, the previously described syncytin-Car1, which is conserved among all carnivorans, and two novel env sequences that were named Hyena-Env2 and Hyena-Env3. All three contain domains characteristic of canonical gamma-type retroviral Env proteins: a predicted signal peptide at the N-terminus, a putative furin cleavage site (R-X-K-R) delimitating the surface (SU) and trans- membrane (TM) subunits, a hydrophobic domain of over 20 aa located downstream of a highly

conserved C-X5-7-CC motif and a canonical retroviral Env immunosuppressive domain (ISD) (reviewed in Henzy & Johnson, 2013). When placed on a phylogenetic Env tree, does not clus- ter with previously described syncytins. Hyena-Env3 shares conserved regions with the human syncytin-1 (39% aa identities) as well as with the previously described non-syncytin Canis-Env3 2.2 (66% aa identities) (Cornelis et al., 2012). An RT-qPCR analysis of the placental expression of all three envelopes, using primers specific of each gene, discloses that syncytin-Car1 is very strongly expressed in the placenta, its expression being about 5 times higher than that of the ribosomal protein RPL19 housekeeping gene, at levels similar to those observed in cat and dog placentas (Cornelis et al., 2012). The levels of expression of the two newly identified envelopes are much lower than that of syncytin-Car1, namely about 50 fold lower for Hyena-Env2 and 150 fold lower for Hyena-Env3.

Characterization of the identified hyena candidate genes: are they hyena-specific and conserved? Syncytin-Car1 has already been demonstrated to be common to all carnivorans and to be functionally conserved, with evidence for purifying selection. A refined analysis of amino acid identity and non-synonymous vs synonymous mutation rates could not identify any characteristic feature specific to the hyena ortholog (Cornelis et al., 2012) and a thorough BLAST search of the transcriptome did not reveal expression of alternative copies in the hyena placenta. The Hyena-Env2 gene cannot be found by in silico BLAST search within the cat, cheetah, tiger, dog, ferret, walrus, seal, polar bear or panda genome, and as such is a good candidate for being a hyena-specific ERV env gene. To verify this hypothesis on a larger panel of species, a 416 bp internal fragment was tentatively PCR amplified from a panel of genomic DNAs alongside the control amplification of the whole syncytin-Car1 ORF using the primers described in Cornelis et al. (2012) to assay DNA quality (Fig. 3). Amplification of the internal Hyena- Env2 fragment was observed only in the four extant hyena species, confirming the results of our in silico analysis. Finally, we could amplify and sequence the full-length Hyena-Env2 ORF (using primers situated on both sides of the reading frame) from all extant hyenas: Crocuta crocuta, Hyaena hyaena, Parahyaena brunnea and Proteles cristatus (see Fig. 4), revealing that

144 hw stertobtenepeso eeso ahevgn n hs fthe of those and gene env each of (A). levels Methods). expression as between colors ratio the same the in the as expression using shown gene highlighted env features of structural analysis canonical RT-qPCR the (Right) indicated. with are putative shown motif is the (CC) candidate and each C-X5-7-C candidate (red), conserved the domain the of transmembrane with Characterization (B)(Left) the along hydrophobic blue) (green), the (ISD, peptide interaction, domain SU-TM fusion immunosuppressive in the involved (purple), motif C-X-X-C peptide R/K-X-R/K-R, the (consensus: signal subunits, site cleavage two furin the The between subunits. orange) (TM) Transmembrane and (SU) Surface the identified the of characterization envelope and protein crocuta Env retroviral cuta canonical a of Structure – 2 Figure Hyena-Env3 Hyena-Env2 TSKVRELRDRIQARRQDSHIWGLDPH TWRLPNITTTNQTVHTGALGKPIPIYNLSSTQDILTFNASSASSSLPYCTPPGIFLSCPEGIYRCLTASDSLNCIFILLSPSTNIYSHTQLLSLLFPTH 200 WTYLAQIGLSDGGGVQDQAKHHQTKQMLQNLHTQRQAQTTPTYRGLNLDTLQNFLPTDQLTNTLINSSYNLWQNLTGDRDCWVCFPFSTITNIIGIPIPT DSLAEMA QVLENDTLCVLVTLFPQVYYQPASSFFEIQPEQKHS 400 TDCWLCLDSRPPFYVGWAISGQVSRDIEGHCSWGQPPVLTIQEVTGSGLCVLGNGGTLTTFPHLSHLCNQTMTATGSSYLRPPSGAWFACTSGLTPCIHP 300 VWGIRLWLTGHDIGFLFSIQKQLVLPPPVAMGPMAASAANHKPRSTPSVPAPTQAAPSLSATDSPLGGVPIQLRPPRSRPVIYSILNLTYSFLNSTNLTN 200 QCVDCNTFGCRSGADCQHQNLRQQTFYVCPGTGNFDTCGGIEHFFCGSWGCETIAPWVKQPSNDLITLVRASNQTSPSNRNPISIQLTPRGKTENWSVVK AKR MIGPPRWWAQPNFGWYVLLILVLLHPLVWA TL MEGHKEPQKPAQAQAMLLPIFTALLTVCKS membrane Hyena-Env3 Syncytin-Car1 Hyena-Env2 D B A intracellular extracellular NRLLAFMRERLGAIRLMVLRSQYAQPPADQSEDQYVQLGPLKFQEDP fusion peptide AVFVPLLIGAGLATGIATGVAGLG LQNRRRLDLILLHQGGL ISD C C SU TM 0 0 0 0 500 400 300 200 100 ..Heaseicrtoia neoegn atr n lcnaevolution placenta and capture gene envelope retroviral Hyena-specific 2.2. candidates. N-ter CQALGEQCCFY CWIC peptide signal AWVTWLLPLAGPLCLILLSISVA TSINFYYKLSQALNEDMERIADSLTALQTQITSLAAIA 0 0 0 400 300 200 100 0 0 0 400 300 200 100 CWLC CXXC TECNRCMHGSGQRRYFQSYTYMPKICYEGKTPITCTSTHAPGKTFWASELNAQRTTKSPCNQYPKTGTVC 100 TECNRCMHGSGQRRYFQSYTYMPKICYEGKTPITCTSTHAPGKTFWASELNAQRTTKSPCNQYPKTGTVC 100 NPSSHLPNPHQPTTAKWVLRGPLTTPRDLGRTVQELTLTGPASITFPTFHLDLCSLAGDHWNTNPKICKG CWVC A ceai ersnaino ervrlEvpoen delineating protein, Env retroviral a of representation Schematic (A) ANNSGIVQDSLAVVRQHLQERAKIREQNKNWYENIFNWSPWLTA UTM SU RGKR RGKR RAKR RWTR rct crocuta Crocuta cleavage sitecleavage DFRVSA peptide fusion fusion SU/TM RXKR A LPTLIVGTGIEAGVGTGTAALIRGN ISD 600 CC PCLIRYLWERLQEMSRVSVNQLLLQPYSRLPTSDYPYNDAHPSAGSSRN membrane expression rct crocuta Crocuta trans- placental 0.03 5.10 0.11 n rtis h yrpoiiypol for profile hydrophobicity the proteins: Env C-ter 58 C 53 58 55 LQNRRALDLLTAEKGGT lcna rncitlvl are levels Transcript placenta. 52 0.6 62 60 30 QQFDALAQAIDFDLAQLENSTRHIRGSL 500 QQFDALAQAIDFDLAQLENSTRHIRGSL RPL19 99 38 40 85 45 59565 58 99 98 47 oto ee(see gene control LITALAGPLALLLLLLTLGPY LITALAGPLALLLLLLTLGPY 96 24 100 26 51 47 19 Syn-Ten1 80 98 99 CLYLKEECCYY 63 MPMV Syn-2 BaEV PyERV Syn-B TgERVF 99 Syn-A MoMLV Syn-Ory1 Syn-1 REV Hyena-Env3 77 PERV FLV Hyena-Env2 GaLV HIV1 Syn-Rum1 Syn-Car1 dog Syn-Car1 FIV Syn-Car1 cat Syn-Car1 Syn-Car1 hyena Syn-Car1 Syn-Opo1 Cro- 145 Syn-Mab1 Syn-Mar1 CAEV RSV HERV-K VNQSGIV 400 Env-Cav1 MMTV HTLV1 JSRV BLV R R 300 600 2.2 2. Implication of ERV env genes in placental structural transitions in Hyaenidae

they share a very high sequence similarity (over 98.4% nucleotide identities). The ratio of non- synonymous to synonymous mutations (dN/dS) is well below unity (between 0.20 and 0.67), strongly suggesting that the gene is under purifying selection, even though the small dataset and low number of variable sites may be a statistical limitation. Altogether these results indicate that Hyena-Env2 endogenization occurred after the separation of the Hyaenidae family from the rest of carnivorans about 30 Mya, and that has been conserved since the emergence of this clade. Interestingly, sequences bearing high similarity to Hyena-Env3 can be found in silico within many Caniformia genomes (e.g., Canis familiaris (dog), Ailuropoda melanoleuca (panda), Ursus maritimus (polar bear), Mustela putorius furo (ferret)), at a variable number of copies, but not Feliformia genomes (e.g., Felis catus (cat), Acinonyx jubatus (cheetah), Panthera tigris (tiger)). The dog genome presents a full-length protein-coding copy, previously identified as Canis-Env3 (Cornelis et al., 2012), alongside a large number of non-coding copies. However, in Canis- Env3 the methionine which precedes the signal peptide in the Hyena-Env3 ORF is not present, suggesting that Canis-Env3 is not a functional copy of this gene, lacking a membrane addressing signal. A dog provirus (canFam3 genome assembly, chr3: 82,194,219-82,202,071) associated 2.2 with a non-coding Canis-Env3 copy presents two 100% identical LTRs and almost intact gag, pol and env ORFs suggesting that this family was active until very recently. In addition, most conserved copies in the dog genome are not present in the orthologous regions of the panda or ferret genomes, indicating that numerous insertions occurred after separation of the three lineages, about 60 Mya (Fig. 3). To assay whether this family of env genes could have entered the Caniformia branch after the Feliformia/Caniformia split about 65 Mya, and still be specific to hyenas within Feliformia, we attempted to amplify a 674 bp fragment internal to the Hyena-Env3 ORF from genomic DNA of the species in our panel. As indicated in Fig. 3, PCR fragments of the expected size were obtained in many Caniformia species tested and sequencing confirmed them being analogous to Hyena-Env3. In Feliformia the PCR assay revealed presence of Hyena- Env3 in all four extant hyena species but also, unexpectedly, in Arctictis binturong. This env gene seems therefore to have been captured on multiple occasions and at different time points in evolution, but is clearly not specific to hyenas.

Structure of the Hyena-Env2 associated provirus and placental transcript. To further characterize the hyena-specific Hyena-Env2 gene, we searched for its insertion locus by inverse PCR. Genomic DNA from Crocuta crocuta was digested, using restriction enzymes with only one cutting site in the Hyena-Env2 gene, and self-ligated to generate circular genomic DNA fragments. PCR was performed on these circular fragments using divergent primers located within the Hyena-Env2 gene fragment and the obtained PCR products were sequenced. One of them contained the expected part of the env gene and extended 3’ to it for over 1000 bp, with evidence for the presence of an LTR and flanking genomic sequence. The identified flanking sequence mapped to a locus of the cat genome (the closest sequenced carnivoran genome), most probably corresponding to the ortholog of the integration site of the Hyena-Env2 associated

146 yrdarw.(ih)Peec rasneo h three the of absence are the or available both Presence is (from genome (Right) of sequenced arrows. species dates a red capture Carnivora which by Approximate 34 for asterisk. Species the an families. together of by indicated corresponding are indicated names their genes env of The three names the top). of the the presence with the at for tested bar length suborders) Caniformia branch (scale and Horizontal Feliformia time 2012). to Bininda-Emonds, & proportional (Nyakatura is Carnivora. indicated in outgroup genes Pholidota env of the conservation with and date Entry – 3 Figure lsdntspeec ftefl-eghcdn R n iu eoe bec ftegn.For gene. the of absence denotes minus a by and both ORF gene, the coding full-length the of Hyena-Env2 presence denotes plus a rgeto h niecdn R,wieamnsdntsisasne lsdntsta sequence a that denotes plus A absence. its denotes of minus a while this ORF, to corresponding coding by entire seen the as or gene fragment this of presence denotes plus a Hyena-Env2 Syncytin Car1

lsdntsapicto ftefl-eghcdn R n iu eoe bec of absence denotes minus a and ORF coding full-length the of amplification denotes plus a , PHOLIDOTA CARNIVORA 80 srsrce oHande(red). Hyaenidae to restricted is 70 nsilico in CANIFORMIA FELIFORMIA 60 ..Heaseicrtoia neoegn atr n lcnaevolution placenta and capture gene envelope retroviral Hyena-specific 2.2. env Million yearsMillion 50 eewsapie rmgnmcDAb C ridentified or PCR by DNA genomic from amplified was gene erhwe vial rPRo 1 pitra rget For fragment. internal bp 416 a of PCR or available when search Carnivora phylogenyCarnivora 40 30 Hyena-Env2 20 10 0 Manis pentadactylaManis javanicaManis Tremarctos ornatus melanoleucaAiluropoda grypusHalichoerus rosmarusOdobenus Zalophus californianus fulgensAilurus Aonyx cinerea Procyon lotor mephitisMephitis binotataNandinia Arctictis binturong crocutaCrocuta feroxCryptoprocta suricattaSuricata pardicolorPrionodon Panthera tigris Panthera saxicolor pardus Panthera leo Panthera krugeri leo Panthera Leptonychotes weddelliiLeptonychotes cristatusProteles hyaenaHyaena brunneaParahyaena parvulaHelogale catusFelis Ursus maritimusUrsus furo putorius Mustela megalotisOtocyon vulpesVulpes familiaris lupus Canis Neofelis nebulosa nebulosa Neofelis serval Leptailurus concolorPuma jubatusAcinonyx nsilico in *

erho C mlfiaino 7 pinternal bp 674 a of amplification PCR or search syncytin-Car1 * *

env *

* * * ee nec pce.For species. each in genes

* *

- - - - - +- +- - + - -+ - -+ - -+ - -+ - ++ - ++ - ++ - ++ - ++ - ++ - ++ - ++ - + - -+ ++ - + - + + + + + + + + + -+ + -+ - -+ - -+ - -+ - -+ - -+ - -+ - -+ - -+ - -+ - -+ - -+ - + - + + Syncytin-Car1 Lf)Crioapyoeei tree phylogenetic Carnivora (Left) and Hyen

Hyena-Env2 a -E n Hyena-Env2 Prionodontidae Procyonidae Odobenidae Pantherinae Herpestidae Hyaenidae Nandiniidae Mephitidae Mustelidae Eupleridae Viverridae Phocidae Otariidae Ailuridae Manidae Canidae v3 Ursidae Felinae nsilico in r represented are syncytin-Car1 Hyena-Env3 Presence . 147 , ,

2.2 2. Implication of ERV env genes in placental structural transitions in Hyaenidae

80 Crocuta crocuta MEGHKEPQKPAQAQAMLLPIFTALLTVCKSNPSSHLPNPHQPTTAKWVLRGPLTTPRDLGRTVQELTLTGPASITFPTFH Parahyaena brunnea MEGHKEPQKPAQAQAMLLLIFTALLTVCKSNPSSQLPNPHQPTTAKWVLRGPLTTPRDLGRTVQELTLTGPASITFPTFH Hyaena hyaena MEGHKEPQKPAQAQAMLLLIFTALLTVCKSNPSSHLPNPHQPTTAKWVLRGPLTTPRDLGRTVQELTLTGPASITFPTFH Proteles cristatus MEGHKEPQKPTQAQAMLLLIFDALLTVCKSNPSSQLPNPHQPTTAKWVLRGPLTTPRDLGRTVQELTLTGPASITFPTFH **********:******* ** ************:********************************************* 180 LDLCSLAGDHWNTNPKICKGQCVDCNTFGCRSGADCQHQNLRQQTFYVCPGTGNFDTCGGIEHFFCGSWGCETIAPWVKQPSNDLITLVRASKQTSPSNR LDLCSLAGDHWNTNPRICKGQCVDCNTFGCRSGADCQHQNLRQQTFYVCPGTGNFDTCGGIEHFFCGSWGCETIAPWVKQPSNDLITLVRASNQTSPSNR LDLCSLAGDHWNTNPRICKGQCVDCNTFGCRSGADCQHQNLRQQTFYVCPGTGNFDTCGGIEHFFCGSWGCETIAPWVKQPSNDLITLVRASNQTSPSNR LDLCSLAGDHWNTNPRICKGQCVDCNTFGCRSGADCQHQNLRQQTFYVCPGTGNFDTCGGIEHFFCGSWGCETIAPWVKQPSNDLITLVRASNQTSPSNR *************** ****************************************************************************:******* 280 NPISIQLTPRGKTENWSVVKVWGIRLWLTGHDIGFLFSIQKQLVLPPPVAMGPMAASAANHKPRSTPSVPAPTQAAPSLSATDSPLGGVPIQLRPPRSRP SU NPISIQLTPRGKTENWSVAKVWGIRLWLTGHDIGFLFSIQKQLVLPPPVAMGPMAASAANHKPRSTPSVPAPTQAAPSLSATDSPLGGVPIQLRPPRTRP 2.2 NPISIQLTPRGKTENWSVAKVWGIRLWLTGHDIGFLFSIQKQLVLPPPVAMGPMAASAANHKPRSTSSVPAPTQAAPSLSATDSPLGGVPIQLRPPRTRP NPISIQLTPRGKTENWSVAKVWGIRLWLTGHDIGFLFSIQKQLVLPPPVALGPMAASAANHKPRSTPSVPAPTQAPPSLSSTDSPLGGVPIQLRPPRSRP ****************** *******************************:*************** ******** ****:****************:** 380 VIYSILNLTYSFLNSTNLTNTDCWLCLDSRPPFYVGWAISGQVSRDIEGHCSWGQPPVLTIQEVTGSGLCVLGNGGTLTTFPHLSHLCNQTMTATGSSYL VIYSILNLTYSFLNSTNLTNTDCWLCLDSRPPFYVGWAISGQVSRDIEGHCSWGQPPVLTIQEVTGSGLCVLGNGGTLTTFPHLSHLCNQTMTATGSSYL VIYSILNLTYSFLNSTNLTNTDCWLCLDSRPPFYVGWAISGQVSRDIEGHCSWGQPPVLTIQEVTGSGLCVLGNGGTLTTFPHLSHLCNQTMTATGSSYL VIYSILNLTYSFLNSTNLTNTDCWLCLDSRPPFYVGWAISGQVSRDIEGHCSWGRPPVLTIQEVTGSGLCVLGNGGTLTTFPHLSHLCNQTMTATGSSYL ****************************************************** ********************************************* 480 RPPSGAWFACTSGLTPCIHPQVLENDTLCVLVTLFPQVYYQPASSFFEIQPEQKHSRGKRDFRVSAALPTLIVGTGIEAGVGTGTAALIRGNQQFDALAQ RPPSGAWFACTSGLTPCIHPQVLKNDTLCVLVTLFPQVYYQPASSFFEIQPEQKHSRGKRDFRVSAALPTLIVGTGIEAGVGTGTAALIRGNQQFDALAQ RPPSGAWFACTSGLTPCIHPQVLKNDTLCVLVTLFPQVYYQPASSFFEIQPEQKHSRGKRDFRVSAALPTLIVGTGIEAGVGTGTAALIRGNQQFDALAQ RPPSGAWFACTSGLTSCIHPQVLKNDTLCVLVTLFPQVYYQPASSFFEIQPEQKHSRGKRDLRVSAALPTLIVGTGIEAGVGTGTAALIRGNQQFDALAQ *************** *******:*************************************:************************************** 580 AIDFDLAQLENSTRHIRGSLDSLAEMALQNRRRLDLILLHQGGL CQALGEQCC FYANNSGIVQDSLAVVRQHLQERAKIREQNKNWYENIFNWSPWLTAL AIDFDLAQLENSTHHIRGSLDSLAEMALQNRRRLDLILLHQGGL CQALGEQCCFY ANNSGIVQDSLAVVRQHLQERAKIREQNKNWYENIFNWSPWLTA L AIDFDLAQLENSTRHIRGSLDSLAEMALQNRRRLDLILLHQGGL CQALGEQCCFY ANNSGIVQDSLAVVRQHLQERAKIREQNKNWYENIFNWSPWLTA L TM AIDFDLVQLENSTRHIRGSLDSLAEMALQNRRRLDLILLHQGGL CQALGEQCCFYASNSGIVQDSLAVVRQHLQERAKIREQNKNWYENIFNWSPWLTAL ****** ****** ****************************************** ******************************************* 649 ITALAGPLALLLLLLTLGPYTLNRLLAFMRERLGAIRLMVLRSQYAQPPADQSEDQYVQLGPLKFQEDP ITALAGPLALLLLLLTLGPYTLNRLLAFMRERLGAIRLMVLRSQYAQPPADQSEDQYVQLGPLKFQEDP ITALAGPLALLLLLLTLGPYTLNRLLAFMRERLGAIRLMVLRSQYAQPPADQSEDQYVQLGPLKFQEDP ITALAGPLALLLLLLTLGAYTLNRLLAFMRERLSAIRLMVLRSQYAQPPADQSEDQYVQLGPLKFQEDP ****************** ************** ***********************************

Figure 4 – Alignment of Hyena-Env2 genes from all four extant hyena species. Functional domains are highlighted using the same color code as in Fig. 2 and the SU and TM subunits are delineated. Asterisks indicate amino acid identities and colons indicate amino acid similarities.

148 rncit eeo ervrloii,i rvnb rmtro ellrorigin. cellular of promoter placental a the as by driven of region is expression this origin, that retroviral present suggests of gene to This a seem 5A). transcript, (Fig. others ap1m1) while predicted (e.g., SLC35E1), are intron CHERP, transcripts other an (e.g., Several nearby island. CpG initiated a at be site in donor be to splice to predicted appears a and presents position Hyena- region the expected this of the genome locus cat insertion the the to In sequence orthologous provirus. The site associated assembly). the (felCat8 Env2 of genome upstream cat kb the 1 algorithm in about BLAT transcript region the the a of matches used end we 5’ published, the been position sequence to has the try genome match to hyena not no does transcript Since instead LTR. the proviral transcripts, of end retroviral the 5’ for of the expected since as place LTR takes event 5’ splicing the another transcript in The occur 5C). not transcripts. (Fig. retroviral does for codon observed initiation classically However, start as envelope signal, polyadenylation the a with before LTR just 3’ site site the in acceptor donor ends an a with and LTR sites, 5’ splicing the transcript envelope after retroviral canonical the the identify of the to obtained with sequence able was transcript band complete obtained major the the single aligning A providing By cDNA. mRNA. end, the samples, of each ends RNA for 3’ placenta sequenced or total 5’ and the using amplify (RACE) to Ends PCR by cDNA followed of Amplification Rapid performed deletion kb 2 a presents genome dog the note Of region. cases. this all in in site empty an discloses which 5B), to internal by primers confirmed with obtained results the confirming ( empty, Hyaenidae to closest species suricatta Suricata four the and dog, cat, classifies domains 6). RT (Fig. catalytic gammaretrovirus retroviral exogenous, a as as pol well provirus of as this domain RT endogenous catalytic other, the database, of Comparing tRNA set C). a a 5B, with using (Fig. BLAST tRNA by glycine A a identified matching sequences. be sequence can non-coding the (PBS) altered site severely binding duplication, with primer although site tRNA genes, target putative full-length pol bp At the and contains 4 C). gag provirus 5B, recognizable a The and (Fig. observed integration. above proviral we a identified DNA for locus genomic expected the as and LTRs within between inserted junctions and the identity) (89% LTRs of fragment 3’ large and A controls. the as dog of and characterization cat from in amplified be as could well kb to 9 as about side, species, 5’ related the closely for in sequence from presence genome provirus cat whole the the using amplify designed, tentatively were Primers 5A). (Fig. provirus nodrt hrceiefrhrtesqec fteplacental the of sequence the further characterize to order In from DNA genomic and primers of set same the using sequencing and amplification PCR nsilico in or ..Heaseicrtoia neoegn atr n lcnaevolution placenta and capture gene envelope retroviral Hyena-specific 2.2. eoaeparvula Helogale nlsso h rhlgst nalsqecdcrioa eoe (Fig. genomes carnivoran sequenced all in site ortholog the of analyses Hyena-Env2 rct crocuta Crocuta otiigpoiu,bree yceryrcgial 5’ recognizable clearly by bordered provirus, containing hwta h rhlgstsi hs pce r indeed are species these in sites ortholog the that show ) rct crocuta Crocuta eoi N n,oc eune,allowed sequenced, once and, DNA genomic Hyena-Env2 rtci itrn,Cytpot ferox, Cryptoprocta binturong, Arctictis eoi N n oasyits assay to and DNA genomic Hyena-Env2 soitdpoiu,w were we provirus, associated Hyena-Env2 Hyena-Env2 Hyena-Env2 hswsfurther was This . rncit we transcript, Hyena-Env2 placental gene, 149

2.2 2. Implication of ERV env genes in placental structural transitions in Hyaenidae

A ap1m1 CHERP SLC35E1 chrA2 12,370,000 12,375,000 Cat Dog Panda Ferret transcription CpG rich region B exon 50 0 Mya intron **** ***** ******* ** * ***** ** placental transcript initiation dog ( C. familiaris ) AGAGAT------repeated elements ferret ( M. putorius ) AGAGGTAGCACATTGTTGTACCTATATGTATTTTCATG SINE walrus ( O. rosmarus ) AGAGATAGCACACTGTTGTACCTGTATGTATTTTCGTG LINE seal ( L. weddellii ) AGAGATAGCACACTGTTGTACCTGTATGTATTTTCGTG synteny panda ( A. melanoleuca ) AGAGATAGCACACTGTTGTACCTATAT--GTTTTCGTG highly conserved polar bear ( U. maritimus ) AGAGATAGCACACTGTTGTACCTATATGTGTTTTCGTG conserved cat ( F. catus ) AGAGATAGCACGCTGTTGTACCTATACGTGTTTTCGTG gap cheetah ( A. jubatus ) AGAGATAGCACGCTATTGTACCTATATGTGTTTTCGTG tiger ( P. tigris ) AGAGATAGCACGCTGTTGTACCTATATGTGTTTTCGTG fossa ( C. ferox ) AGAGAGAGCACACCGTTGTACCCCTATGTATTTTCGTG 2.2 target site duplication P. cristatus AGAGAGAGCACATTGTTGTACTGAAAG//ATAACATT TGGTGCGTTGGCCAGGAACCCCA//TGAAAG//ATAACAGTACCTGTAAGGGTTTTCGTG H. hyaena AGAGAGAGCACACTGTTGTACTGAAAG//ATAACATT TGGTGCATTGGCCAGGAACCCCA//TGAAAG//ATAACAGTACCTGTAAGGGTTTTCGTG H. brunnea AGAGAGAGCACACTGTTGTACTGAAAG//ATAACATT TGGTGCATTGGCCAGGAACCCCA//TGAAAG//ATAACAGTACCTGTAAGGGTTTTCGTG C. crocuta AGAGAGAGCACATTGTTGTACTGAAAG//ATAACATT TGGTGCGTTGGCCAGGAACCCCA//TGAAAG//ATAACAGTACCTGTAAGGGTTTTCGTG 5’LTR PBS (Gly) 3’LTR C Hyena-Env2 associated provirus Gag Pol Hyena-Env2 LTR CAACCG |gtca aaag |GTCCAC placental AATAAA//CA AAAAAAAAA... transcript TGCGAG |gtga gcag |GTCTCA

Figure 5 – Characterization of the Hyena-Env2 associated provirus and its genomic location. (A) Hyena-Env2 insertion occurred in a relatively conserved and transcriptionally active region. (Upper) Predicted transcripts at the syntenic cat locus with exons and introns delineated. (Lower) 10 kb region of the cat genome containing the ortholog site to the Hyena-Env2 insertion locus, with repeated sequences (as indicated by RepeatMasker, http://www.repeatmasker.org/) and CpG islands marked, as well as con- servation of this region in other carnivorans. The ortholog region to the hyena transcript initiation site is also indicated. See key on the bottom right. (B) The insertion site of the Hyena-Env2 associated provirus is empty in other Carnivora. Timed phylogenetic tree and nucleotide alignment of the syntenic loci of sequenced Carnivora, the four extant Hyaenidae, as well as Cryptoprocta ferox, the closest relative to Hyaenidae. The target site duplication is highlighted by red boxes showing that insertion occurred only in Hyaenidae. For the latter the nucleotide sequence of the extremities of both LTRs as well as the PBS are shown. Asterisks indicate sequence identities in the locus alignment, excluding the dog sequence. (C) Structure of the Hyena-Env2 associated provirus and its placental transcript in Crocuta crocuta. LTRs are indicated in purple, gag in orange, pol in yellow and Hyena-Env2 in red. Vertical bars indicate stop codons or frameshifts in the three reading frames. The spliced placental transcript is drawn to scale below with the exons aligned to their corresponding proviral regions. The sequence of the splicing sites is provided.

150 acltdatr100rpiae n r niae nec rnh(e e nfiue.Bac eghis length Branch figure). in scale). key (see site (see per branch substitutions each of on number average indicated the are to and proportional replicates 1,000 after when calculated 2C Fig. in as dataset described same the using domains The RT available. exogenous and endogenous retroviral of position The – 6 Figure

Syn-Rum1

Syn-Opo1

γ Hy

syncytins ena-E

REV nv2 Hyena-Env2

Hyena-Env2 Syn1

Syn2

Syn-Car1 dog BaEV GALV PERV

r hw nbu.Mjrrtoia eeaaeidctdo h re ottaswere Bootstraps tree. the on indicated are genera retroviral Major blue. in shown are Syn-Car1 cat ..Heaseicrtoia neoegn atr n lcnaevolution placenta and capture gene envelope retroviral Hyena-specific 2.2. FLV

47

52 66

65

53 MoMLV

84

100 6 ascae Ti ihihe nrd n hs soitdwt previously with associated those and red, in highlighted is RT -associated 5

soitdvrsi gammaretrovirus. a is virus associated 86

63 79

71 Syn-Mab1 0.4

100 797

80

BLV

HTLV 1

88 notdPyLte hwn the showing tree PhyML Unrooted δ

40

0 100 88

91 49 75 68

MPMV 4

6 Sy JSRV n-O

RSV

TgERVF HERV-K MMTV ry1 CAEV

PyERV

I FIV HIV1 β α λ 151

2.2 2. Implication of ERV env genes in placental structural transitions in Hyaenidae

In situ hybridization of the identified Hyena-Env genes on placental sections. The hyena is the only carnivoran with a placenta of the hemochorial type (Morton, 1957; Wynn & Amoroso, 1964; Oduor Okelo & Neaves, 1982). In all other carnivorans, the placenta is of the less invasive endotheliochorial type, where the fetal tissues break down the maternal epithelium but leave the maternal blood vessels intact (Fig. 7 top, Wooding & Burton, 2008). In the hemochorial hyena placenta, maternal blood vessels have been destroyed and the fetal tissues form villi that are in direct contact with maternal blood (Fig. 7 bottom, Morton, 1957; Wynn & Amoroso, 1964; Oduor Okelo & Neaves, 1982). Each primary villus contains fetal blood vessels and is covered by a layer of fused cells: the syncytiotrophoblast (Oduor Okelo & Neaves, 1982; Enders et al., 2006). Under this layer are the cytotrophoblasts, which do not form a continuous layer and have a cytoplasm filled with numerous lipid droplets (Oduor Okelo & Neaves, 1982; Enders et al., 2006). The villi present frequent ramifications (not schematized in Fig. 7) forming a complex labyrinthine organization with numerous linked maternal blood spaces. The surface of these ramifications is also covered in syncytiotrophoblast (Enders et al., 2006). Maternal blood is pumped into the placental blood spaces at the fetal side and flows down through the labyrinth towards the 2.2 maternal side where it drains into maternal veins, thus establishing an efficient counter-current exchange system between maternal and fetal blood (Wynn & Amoroso, 1964). At the point of contact between the fetal villi and the maternal endometrium, the syncytiotrophoblast layer peels off and the tip of the villus expands and is covered by tall columnar cytotrophoblasts that often contain phagocytized erythrocytes (Enders et al., 2006). To determine the expression pattern of the three hyena envelopes in the placenta we performed in situ hybridization (ISH) on placenta tissue sections (gestation day 62, about mid-pregnancy) using three specific digoxigenin-labeled antisense RNA probes for each envelope, as well as the corresponding sense probes as negative controls. As previously described for the cat and dog placenta (Cornelis et al., 2012), hyena syncytin- Car1 is expressed specifically in the syncytiotrophoblast and not in the underlying cytotro- phoblast, indicating that the hyena syncytin-Car1 expression locus does not differ from that observed in other carnivorans. Unexpectedly, for Hyena-Env2 the sense probe used as a negative control resulted in a much stronger signal than the antisense probe detecting the sense envelope transcript. Sense/antisense specific RT-PCR on placental RNA with the primer pairs used to synthesize ISH probes revealed the existence of an antisense transcript that covered at least all 3 regions detected by the probes, since all pairs allowed amplification from sense and antisense cDNA. This antisense transcript most probably initiates downstream from the proviral insertion site, at a CpG island (by anal- ogy with the cat genome, see Fig. 5A) that could drive antisense transcription. Nevertheless, the signal observed using the antisense probes is differential between the different placental cell layers: staining is observed only in the layer that contains the cytotrophoblasts, just below the syncytiotrophoblast. This localization of expression is clearly different from that of hyena syncytin-Car1 and suggests a distinct role for Hyena-Env2. To confirm this ISH result, we gen-

152 hte hsi u oteasneo h rprrcpo nalcl ie etd huhwe though tested, lines cell all in receptor proper the of absence the to for due tropism is strong this a whether displays and 10). infection and 9B viral Fig. mediate 2012, al., to et able (Cornelis is lines cell it carnivoran test: this in genes Car1 9A). (Fig. foci infection stained cell X-Gal target counting various by infect assayed to was used infection then and were lines particles Viral gene. reporter nls-LacZ an containing MLV the expressing plasmid the a for coding plasmid a with co-transfected where recombinant envelope render the could whether tested proteins which assay, pseudotyping a using assayed first was envelopes hyena identified the of Fusogenicity a above), observed. was see RT-qPCR, syncytiotrophoblast by the measured of as labeling expression faint low specific its with (consistent strength signal low the of level the at observed syncytiotrophoblast. is staining the strong of a not ISH, and the - As cytotrophoblasts with Methods). consistent (see and controls 8 negative Fig. as immu- in sera illustrated post pre-immune the corresponding with the immunohistochemistry using performed sera, nization and SU Hyena-Env2 the from polypeptides derived bacteria-synthesized with mice immunizing by antibodies anti-Hyena-Env2 erated primary the Only spaces. blood blood maternal types. large maternal placental of the both presence for placenta, represented the hyena are in the hemochorial villi resulting presents the right disrupted In the been interface. on have drawing syncytial vessels The the row). of (bottom organization placenta hyena cellular hemochorial and row) hemo- (top and placenta carnivoran endotheliochorial of placentas differences hyena structural chorial the of Representation – 7 Figure hemochorial endotheliochorial

yn-n2tse eaiei h suoyigasy(e i.9 n 0.I sunknown is It 10). and 9B Fig. (see assay pseudotyping the in negative tested Hyena-Env2 Syncytin- Feliformia other like behave to previously shown been has Syncytin-Car1 Hyena for ISH by obtained be could result conclusive no although note Of (hyena) (cat, dog) ..Heaseicrtoia neoegn atr n lcnaevolution placenta and capture gene envelope retroviral Hyena-specific 2.2. ceai ersnaino h i-etto anvrnendotheliochorial carnivoran mid-gestation the of representation Schematic gag-pol Hyena-Env env ee n lsi xrsigadfcieMVgenome MLV defective a expressing plasmid a and genes ls L-eie ia atce netos 9Tcells 293T infectious. particles viral MLV-derived -less in xvivo ex env eeudrcnrlo M promoter, CMV a of control under gene assays. syncytiotrophoblast (ST)syncytiotrophoblast fetal blood vessel (fbv) vessel fetal blood maternal blood vessel blood maternal cytotrophoblast (CT)cytotrophoblast maternal blood (mb) blood maternal syncytiotrophoblast fetal blood vessel fetal blood cytotrophoblast fetal villus (fv) fetal villus fetal villus (fv) fetal villus endometrium endometrium uoeiiyo h identified the of Fusogenicity Hyena-Env3 u to due 153

2.2 2. Implication of ERV env genes in placental structural transitions in Hyaenidae

HES sense probe antisense probe CT fv fbv ST

Syncytin-Car1 mb mb fv fbv

2.2 CT mb ST mb HES pre-immune serum anti-Hyena-Env2 serum

fv Hyena-Env2 fbv CT

ST mb mb

Figure 8 – Expression pattern of syncytin-Car1 and Hyena-Env2 in Crocuta crocuta placenta. (top rows) Hematoxylin eosin (H&E)-staining and in situ hybridizations on serial sections of Crocuta crocuta placenta using specific digoxigenin-labeled antisense probes for syncytin-Car1 and Hyena-Env2 (and corresponding sense probes as a control). Specific staining is observed in the syncytiotrophoblast (ST) for syncytin-Car1 and in the cytotrophoblasts (CT) for Hyena-Env2. Strong staining with the Hyena-Env2 sense probes indicates existence of an antisense transcript. Areas marked by a rectangle are enlarged on the right, all scale bars: 50 µm. (bottom row) H&E-staining and immunohistochemistry on serial sections of Crocuta crocuta placenta using an anti-Hyena-Env2 mouse serum (and the pre-immune serum as control). The region shown is equivalent to that of the Hyena-Env2 ISH. Specific staining is again observed in the cytotrophoblasts (CT). Scale bars: 50 µm.

154 a1o yn-n3adtreigcrioa ellns niaigseistoim yn-n2tested Hyena-Env2 tropism. species indicating assay. lines, this cell in carnivoran syncytin- negative hyena targeting using when and detected Hyena-Env3 were the foci without or Infection from (G355.5), cells. Car1 particles cat (A23) protein hamster were for or Env cells (293T) Target cells the human control. (A72), target X-gal positive or dog as are stained Hyena-Env3 (Ampho-MLV) which Hyena-Env2, Virus X-gal Leukemia cells, syncytin-Car1 (B) Murine target hyena entry. amphotropic indicated with viral the pseudotyped reveal to or to added Env post-infection a then control days by a is 3 or encoded cells proteins transfected stained ß-galactosidase Hyena-Env the a expressing of vector core, MLV Supernatant a by vector. either the produced and for are transcript vectors Pseudotypes retroviral expression nlsLacZ-containing particles. with viral cells proteins. Hyena-Env-pseudotyped 293T envelope with hyena co-transfecting assay three pseudotyping the the for assay of Fusogenicity – 9 Figure A producer cells target cellstarget 293T pseudotypingassay viral coreviral MLV nlsLacZ viral RNA viral nlsLacZ staining X-gal yping assay assay yping aining vi ..Heaseicrtoia neoegn atr n lcnaevolution placenta and capture gene envelope retroviral Hyena-specific 2.2. ra transfection transfection l l or control Hyena-Env or controlor Hyena-Env RN MLV-Hyena-Env MLV-Hyena-Env infection withwithinfection pseudotype pseudotype A

hamster human catca t dog B B A23A2 3 293T G355.5 A72 vectorvector empty Ampho Ampho MLV MLV Syncytin- Syncytin Car1 Car1 A ceai representation Schematic (A) - Hyena- Hyena- Env2 Env2 Hyena- Hyena- Env3 En 155 v3

2.2 2. Implication of ERV env genes in placental structural transitions in Hyaenidae

5 Hyena- 10 none Env2 SU+TM 10 4 64 kDa

10 3

2 Syncytin-Car1 hyena 10 Hyena-Env2 Hyena-Env3

viral(ffu/mL) titer 10 1

10 0 G355.5 A72 A23 293T cat dog hamster human

Figure 10 – Viral titers obtained in the pseudotyping assay using all three hyena envelopes. Quan- tification of viral titers expressed in focus-forming units per milliliter after infection with syncytin-Car1

2.2 or Hyena-Env pseudotyped MLV virions. An arrow indicates values below background staining, values shown are the mean of three independent experiments ± SD. (inset) Western blot using the anti-Hyena- Env2 mouse serum used in Fig. 8, and lysates of 293T producer cells transfected with a control empty vector (none) or a Hyena-Env2 expression vector. A specific band is observed at the expected size for the full-length Hyena-Env2.

tested several cell lines from cat and dog which should present proteins closely related to hyena proteins, or if an intrinsic loss of fusogenic activity is the cause of this result. Of note, similar results were obtained with HIV- and SIV-pseudotyped particles. These negative results might also be due to a lack of incorporation of Hyena-Env2 into the viral particles. We therefore performed a more straightforward cell-cell fusion assay: 293T cells were transfected with a Hyena-Env2 expressing and plasmid put into contact with a diverse panel of non-transfected target cells 24 h post-transfection. No syncytium formation was observed for any of the target cells used in this assay (cat, dog, human or hamster cells, see Fig. 9B). Finally, Hyena-Env3 was found to confer infectivity to MLV-derived viral particles for some target cell lines. Among tested cells, only carnivoran cell lines were permissive to viral infection and presented significant numbers of infection foci 72 hours post infection, reaching approx. 104 focus forming units/mL (see Fig. 10). These results suggest that Hyena-Env3 confers a carnivoran-specific tropism to the pseudotyped particles. Sequence comparison further showed that the SU of Hyena-Env3 seems to be related to the SU of human Syncytin-1, rabbit Syncytin- Ory1 and of other Env from retroviruses of the SLC1A5 interference group (Fig. 11A), a group of viruses all using the membrane transporter SLC1A5, previously known as ASCT2 or RDR, as a common receptor (Malicorne et al., 2016; Sinha & Johnson, 2017). Alignments reveal that Hyena-Env3 and SLC1A5 group Envs share conserved domains that are essential for interaction with the receptor (Cheynet et al., 2006). Transfection of non-permissive human HEK293T cells

156 lyarl a es npr frtetmn ftemtra muersos gis the against response immune maternal the of taming to the likely —for is part syncytins (Cornelis in captured gene least by syncytin —at layer carried a role also cell of a activity fused capture play immunosuppressive the a in The with of that 2017). associated shown establishment al., also been et the is has placenta, interface it maternofetal a recently the with More at 2011). species 2009, non-mammalian al., rare knocked et mice some (Dupressoir in role genes this these of for demonstration unambiguous out interface an maternofetal with the species, at layer mammalian many cell syncytiotrophoblast in for the responsible of directly formation is and and fusion genes cell-cell syncytin through Fusion exapted response. virus been immune unambiguously during anti-viral has host fusion activity the retroviral membrane inhibit of to mediate activity properties to immunosuppressive an canonical activity and the fusogenic entry, a to include linked which directly al., genes, is et envelope impact Lavialle in Their (reviewed 2016). formation syncytin Denner, placenta the includes in 2013; involved latter be The to 2017). emphasis proved al., which special et genes, with Chuong envelope in site, (reviewed development integration in ERV role the their near on located genes transcription of expression sequences retroviral coordinated protein-coding the retroviral i) the this either ii) to of or exceptions se sequences), exaptation of (LTR series categories, enhancers a major and are two promoters there into but physiology, fall host endogenization They cases the rule. most to In "neutral" 2015). Stoye, across is & spread retroviruses Mager has 2006; of retrovirus al., exogenous et endogenized (Tarlinton infectious being country the whole presently after the is population which koala ongoing, retrovirus Australian still leukemia the are koala in captures Such the 10% KoRV, 2015). to Stoye, instance up & for in Mager with structures in (reviewed proviral genome remnant a vertebrate of is the presence retroviruses of the of for endogenization accounts that that is illustration process, additional date permanent an time-resolved entry This is sequence. Its capture coding carnivoran. gene intact an env other with retroviral species, at any hyena present all is in in It found site Mya. integration 30 not same about Feliformia, the is within Hyaenidae and the of species, emergence the hyena with family.coincident all Hyaenidae the in to present unique is is that origin It retroviral of gene captured newly a revealed which might which Discussion Hyena-Env3, in positions two or 11C). one (Fig. by tropism RDR carnivoran shifted the peculiar al., or the of et explain missing (Cheynet envelopes be binding different receptor to the or appear conformation among 2006) protein conserved in implicated acids and amino group assay the interference pseudotyping the of in Some observed Hyena-Env3 11B). of the tropism both (Fig. the confirming and MLV infection, receptor allow as pseudotyped not SLC1A5 Hyena-Env3 of did cDNA by role SLC1A5 placental infection human hyena with to or transfection permissive dog while cells particles, cat, the from cloned rendered we Methods) that (see SLC1A5 for vectors expression with ms omnythe commonly (most eew aepromdahg hogptaayi fteheapaet transcriptome, placenta hyena the of analysis throughput high a performed have we Here ..Heaseicrtoia neoegn atr n lcnaevolution placenta and capture gene envelope retroviral Hyena-specific 2.2. env ee.Tefre a endmntae ob epnil o the for responsible be to demonstrated been has former The gene). 157 per

2.2 2. Implication of ERV env genes in placental structural transitions in Hyaenidae

A18 6 6 B 14 10 45 13 1 35 72 74 21 40 4 23 90 10 30 86 99 92 39 94 51 61 95 51 98 99 32 3 10 MPMV Hyena-Env2 Syn-Ory1 GaLV

92 BaEV MoMLV PyERV PERV FLV 99 Syn-1 BLV RSV REV Syn-Opo1 Hyena-Env3 Syn-Mar1 MMTV TgERVF

HLTV1 2 HERV-K Syn-2 Syn-A Syn-Ten1 93 Syn-B 10

HIV1 60 Syncytin-Car1 dog Syncytin-Car1 Env-Cav1 Syncytin-Car1 cat Syncytin-Car1 Syncytin-Car1 hyena Syncytin-Car1 JSRV Syn-Rum1 1

CAEV 10 viral(ffu/mL) titer FIV Hyena-Env3-MLV Syn-Mab1

RDR interference 0.8 10 0 group empty human cat hyena dog vector SLC1A5 C SU subunit Syn-1 MALP 13 TLTAPPPC-RCMTS 37 MPRNCYHSATLCMHANTHYWTGKM 1 NPSCP-GGLGVTVCWTYFTQTGMSDGGGVQDQAREKHVKEVI 180 RNKR BaEV MGFT 26 QKRYGRPC-DCSGG 54 VHSSCYTSYQQCRSGNKTYYTATL 21 QSPC-NGIKGQSICWSTTAPIHVSDGGGPLDTTRIKSVQRKL 189 RPKR SRV-1 MNFN 30 QQKHGKPC-DCAGG 57 VHATCYNHYQQCTIGNKTYLTATM 24 IAGCPENKKGQVVCWNSQPSVHMSDGGGPQDKVREIIVNKKF 197 RPKR SRV-2 MTVK 30 HQQHGKPC-DCAGG 54 VHSSCYTTYQECFFGNKTYYTAIL 24 SAGC-TGTVGQHICWNPKAPVHISDGGGPQDKAREIAVQKRL 188 RARR MPMV MNFN 30 QQKHGKPC-DCAGG 57 VHASCYNHYQQCNIGNKTYLTATI 24 TAGCPNGKKGQVVCWNSRPSVHISDGGGPQDKARDIIVNKKF 196 KAKR 2.2 SRV-4 MNFE 30 QNKHGKPC-ECKGG 57 VHSTCYSSYQQCTIGNKTYFTATI 24 SAGCPKDELGKTACWSATPPVHVSDGGGPQDKVREILVQKKF 192 RTKR SMRV MLCI 25 TALFGAPC-DCKGG 68 MHSSCYSSFSQCTQGNNTYFTAIL 22 QAGC-DGTVGKSVCWNQQAPIHVSDGGGPQDAVRELYVQKQI 185 RQKR SNV MDCL 35 QQLWGLPC-DCSGG 70 MHSTCYEKTQECTLLGKTYFTAIL 16 QASC-TGTVGKPVCWDPVAPVYVSDGGGPTDMIREESVRERL 191 RHKR Syn-Ory1 MALS 31 QQHH--PC-TCSGG 65 VHSSCYTQYQTCNKGNQRLLFAVI 18 SPSCP--TKGQVACWYPIAPVGVSDGGGPQDTLTKLKTQKVL 186 RQKR Dasy-Env1.1 MHSF 29 KTLYGTPC-ECRGG 65 MHVHCYSEVQSCTRNNKTYYTAIL 25 QAGC-YGTIGDTVCWNKIPPAHISDGGGPQDQARQIIVQERM 184 RSKR TvERV(D) MIST 23 HALFGAPC-DCKGG 64 LHSLCYTSTQQCTGKSGTYLTSRQ 24 QASCDKINIGKNVCWSLHAPIHVSDGGGPTDQIREMEVKERV 189 RPKR REV-A MDCL 35 QQLWGLPC-DCSGG 68 MHSTCYEKAQECTLMGKTYFTAIL 16 QASC-TGIIGKPVCWDPAAPVYVSDGGGPTDMIREESVRERL 191 RHKR DrERV MFSQ 27 DALFGPPC-DCQGG 75 HSACCYYKVYTCSKGGQTYFTARI 28 PDKCPG-EIGKPACWKLDSPLG-SDGGGPQDKRRTERVRIQI 179 RHKR RD114 MKLP 26 QKQHGKPC-ECSGG 55 MHSSCYTEYRQCRANNKTYYTATL 21 QSPC-RGSINQPVCWSATAPIHISDGGGPLDTKRVWTVQKRL 190 RPKR Hyena-Env3 MIGP 21 PLVWATECNRCMHG 12 MPKICYEGKTPITCTSTHAPGKTF 12 KSPCNQYPKTGTVCWTYLAQIGLSDGGGVQDQAKHHQTKQML 171 RAKR * * * ** * ** ***** * * *

Figure 11 – Hyena-Env3 belongs to the SLC1A5 interference group of envelopes. (A) The Hyena- Env3 SU is closely related to envelopes previously described as using SLC1A5 (also known as RDR and ASCT2) as receptor. Maximum likelihood tree using SU subunit amino-acid sequences from syncytins and a series of endogenous and infectious retroviruses. Previously described syncytin genes are highlighted in blue and the RDR interference group is highlighted in orange. Branch length represents the average number of substitutions per site (scale bar top left), and the percent bootstrap values obtained from 1,000 replicates are indicated on the branches. (B) Carnivoran but not human SLC1A5 allow Hyena-Env3 mediated infection of non-permissive cells. 293T cells were transfected with either an empty vector or expression vectors for SLC1A5 from human, cat, hyena or dog. Transfected target cells were then infected using Hyena-Env3-pseudotyped MLV particles and viral titer in focus forming units per mL were quantified after LacZ staining. Arrows indicate titers below background, values shown are the mean of three independent experiments ± SD. (C) Sequence alignment of SU protein subunits, from the methionine to the putative furin cleavage site, of endogenous and exogenous envelopes using SLC1A5 as a receptor as well as Hyena-Env3 (in bold). Variable regions are represented by the number of omitted residues; dashes correspond to deletions/insertions. The conserved motif shown to be essential for the Syncytin-1/SLC1A5 interaction (SDGGGX2DX2R) is highlighted in orange, and the three cysteine- containing motifs (PCXC, CYX5C, and CX7-9CW) and amino acids that are highly conserved among retroviral members of the RDR interference group are highlighted in yellow (Cheynet et al., 2006). Amino acid mismatches in conserved regions are highlighted in red, asterisks indicate amino acid identities.

158 a1 en o-uoei,btrte em ob oercnl atrdado ohyena to add-on captured recently more a be to seems rather but non-fusogenic, being Car1, in those to similar could are we expression placental and of 2012), dog. pattern al., or and et cat activity (Cornelis fusogenic selection its purifying that under here gene show imply a to of seem behavior normal not the does but elucidated be shared of to acquisition needs case, specific still this In role fusion. cell-cell biological the which for genes, identified not captured presently likely through is "exapted" activity fusion function that unique suggest membrane strongly the cell data the these to Altogether, exported are 2017). they al., after shed et actively (Cornelis (Heidmann are codon or stop 2013) al., premature et conserved Lavialle 2015; a al., to et (due for attachment since membrane evidence either cell fusogenic with lacking be evolution, are cannot clearly they of they Mya as mentioning, 100 worth over are selection, for purifying conservation strong strong with and captures placenta gene the retroviral of in examples expression additional Two vitro. streamer" in "syncytial activity fusogenic the any in specifically expressed the and of selection purifying under conserved is a it identified named previously thus had —and syncytins we of line, this it Along which below). for (see retroviral property for captured sole selected the is be still not and/or might been this activity has However, fusogenic 2013). its syncytins that al., other suggesting et thus two selection, (Esnault where active humans- —including and primates present higher are in gene Env-V activity primate fusogenic a a its of that lost features discovery characteristic the 2009; the by all Benkirane, sustained in discloses & further today that found recently (Saib genes was group syncytin It the manner. our of simple by diversity a the proposed for first accounts 2011), model, Heidmann, This & Dupressoir 2013). al., previously et (reviewed the process Lavialle selection of in Darwinian place classical in a following selected them their were replacing on envelopes, functions, depending captured placental that, fulfilling genes to env Mya, regards retroviral 150 with new about "fitness" of gene, captures envelope reiterated retroviral by ancestral followed an was of for which capture account primitive mammals -to the placental proposes of by emergence initiated also that was scheme captures- gene The such mammals. of evolution placental convergent observed of the different clades that in major captures resulting the evolution, gene in mammalian retroviral of present independent course the to over correspond place taken simply have for, searched been present-day have al., they the et that Lavialle showing in data (reviewed the of pregnancy lines, during these placenta Along and 2013). fetus the of graft semi-allogenic codnl,the Accordingly, Cavia syncytin-Car1 lcna ugsigi ly oei lcnain u ecudntdemonstrate not could we but placentation, in role a plays it suggesting placenta, env ..Heaseicrtoia neoegn atr n lcnaevolution placenta and capture gene envelope retroviral Hyena-specific 2.2. Hyena-Env2 Hyena-Env2 ee nedteheaotoou osntdsls n eito from deviation any disclose not does orthologue hyena the indeed gene, nthe in Cavia Hyena-Env2 syncytin eecnb de oteaoemnindls f"orphan" of list above-mentioned the to added be can gene eede o emt ufi h aefnto ssyncytin- as function same the fulfil to seem not does gene eu htdslssol ato h etrscharacteristic features the of part only discloses that genus syncytin -like atr al ihntenwwl-salse set well-established now the within falls capture env-Cav1 ee,fudi l lcna aml where mammals placental all in found genes, oafide bona env Hyena-Env2 ee ihpaetlepeso,adthe and expression, placental with genes Vrohte l,21) hsenvelope This 2011). al., et (Vernochet yctni l ol oky,has monkeys, World Old in syncytin env sntascae ihls fthe of loss with associated not is eei tl ne purifying under still is gene syncytin genes 159

2.2 2. Implication of ERV env genes in placental structural transitions in Hyaenidae

placentation. The absence of a demonstrated fusogenic activity of Hyena Env-2 associated with its conservation during Hyaenidae evolution, suggests another role. We therefore propose that Hyena-Env2 is directly involved in the transition from an endotheliochorial carnivoran placenta to the hemochorial hyena placenta, that differs from the former mainly by the disruption of maternal blood vessel walls by invading fetal cells, resulting in maternal blood bathing the syncytial maternofetal interface, as observed in all hemochorial placentas, including human (Wooding & Burton, 2008). In this respect it is worth mentioning that a recent analysis of the properties of the envelope protein of the HERV-K family of human endogenous retroviruses has provided hints that this envelope could promote an epithelio-mesenchymal phenotypic transition as well as cellular invasion (EMT) (Lemaître et al., 2017) and that the recently described syncytin- Mab1 might be able to modulate cell migration phenomena through interaction with its cognate receptor (Cornelis et al., 2017). Although preliminary assays carried out with Hyena-Env2 could not demonstrate a similar effect, a more refined experimental setup, or even better, identification of the Hyena-Env2 receptor, should now be carried out to understand in molecular terms how the captured envelope contributes to the transition in placentation type. 2.2

Acknowledgments. We wish to especially thank Adriana Alberti, Corinne Da Silva and Jean Weis- senbach (Genoscope, Evry, France) for the Crocuta crocuta transcriptome sequencing and assembly. We also thank Géraldine Véron (MNHN), Frédéric Delsuc (Université de Montpellier, France) and Baptiste Mulot (Zooparc de Beauval, France) for providing us with some of the DNA and/or tissue samples, Olivia Bawa (Gustave Roussy Preclinical Evaluation Platform, France) for contribution to the histologi- cal analyses and Mélanie Polrot (Gustave Roussy Animal Care Facilities, France) for assistance in mice handling. We thank Christian Lavialle for discussion and critical reading of the manuscript. This work was supported by the CNRS and by grants to TH from the Ligue Nationale contre Le Cancer (Equipe Labellisée) and the Agence Nationale de la Recherche (ANR "RETRO-PLACENTA"), as well as a grant to MF by the Fondation pour la Recherche Médicale (FDT20170437061). Material and Methods

Animals and tissues, ethical agreement. Genomic DNA of dog (Canis familiaris) and cat (Felis catus) were extracted by phenol/chloroform extraction from nitrogen frozen organs taken from pregnant females and supplied by the ENVA-INRA. Genomic DNA of spotted hyena (Cro- cuta crocuta), fossa (Cryptoprocta ferox), binturong (Arctictis binturong), meerkat (Suricata suricatta) and mongoose (Helogale parvula) were extracted from blood samples obtained from B. Mulot and R. Potier (Zooparc de Beauval, France) using the DNA Blood Kit II (PaxGene). Genomic DNA of striped hyena (Hyaena hyaena) was obtained from G. Véron (Muséum Na- tional d’Histoire Naturelle, France, sample id MNHN-AM60). Genomic DNA of brown hyena (Parahyena brunnea) was extracted from blood samples provided by the Aspinall Foundation (Lympne, England). Genomic DNA of aardwolf (Proteles cristata) was provided by F. Delsuc (Université Montpellier 2, France, sample id TS307). Total RNA extracts and paraffin included

160 rm(aso,21)adteMSL rtcl aiu ieiodpyoeei amino phylogenetic likelihood Maximum protocol. MUSCLE the and 2014) (Larsson, gram C-X (transmembrane conserved domain highly hydrophobic a a to of 3’ presence located the domain) on for based Center selected then National ex- were the sequences of protein were envelope program Putative BLASTp (https://blast.ncbi.nlm.nih.gov/Blast.cgi). codon) Information the Biotechnology using retroviruses), stop sequences syn- known infectious acid package which —and amino —among to ERVs EMBOSS cytins subunit representative TM (from contain- start the glycoproteins the envelope retroviral with from 35 (from compared Sequences of were program sequences acids, amino These assembly: getorf sequences. amino into the acid transcriptome translated using 400 and transcriptomes each than (http://emboss.sourceforge.net/apps/cvs/emboss/apps/getorf.html) the longer in from ORF tracted for an searched analyses. ing sequence were and quences screening Database 59), (55, lengths k-mer two as of well transcripts as only Trinity) and and (Oases programs software different assembly two novo using de performed were assemblies independent several transcripts, reconstructed number of the maximize To ERA1116685. number (ENA; accession Archive the Nucleotide under European https://www.ebi.ac.uk/ena) the at deposited (Illumina). been instrument has HiSeq2000 information Illumina sequence the Nucleotide on cell flow paired-end a and in Technologies) 100-base- chemistry using (Agilent read sequenced length Bioanalyzer was 2100 library the Agilent Technologies), Agilent by (MxPro; bp) quantification 200-600 qPCR of (showing size analysis library profile typical library a After primers. adapter-specific Illumina ( using Reagent fragments DNA PCR-amplified Sample and NEBNext purified, were using products ligation by Biolabs); added England (New were Set adapters Illumina and repaired 3’-adenylated, end- then were fragments and Next, protocol: Illumina’s cDNA. following double-stranded prepared create was library to paired-end generated the then was Illumina strand the second to The according protocol. priming TruSeq hexamer random using and cDNA fragmented, single-stranded chemically into beads, converted oligo(dT) with Evry, tran- selected (Genoscope, was RNA and Center poly(A)+ Sequencing Briefly, sequencing, National France). French construction, the by library performed were cDNA assembly system. scripts sequencing 2000 HiSeq Illumina sembly. crocuta Crocuta scientific other and experimental for used (Direc- animals purposes). experimentation of animal protection regarding the accordance regulations regarding strict and 86/609/EEC in tive laws out European carried and was French study the This with USA). School, Veterinary California of versity of samples tissue utpeainet faioai eune eecridotb sn h lVe pro- AliView the using by out carried were sequences amino-acid of alignments Multiple oa N a uie from purified was RNA Total rct crocuta Crocuta lcnoetasrpoehg hogptsqecn and sequencing throughput high transcriptome placentome ..Heaseicrtoia neoegn atr n lcnaevolution placenta and capture gene envelope retroviral Hyena-specific 2.2. > 0 pwr conserved. were bp 100 lcnawr rvddb .Cne n .Gika (Uni- Glickman S. and Conley A. by provided were placenta rct crocuta Crocuta 5-7 Cmotif. -C ervrlendogenous Retroviral lcnabfr euniguigthe using sequencing before placenta env > 0 p were bp) 200 enovo de eese- gene 161 as-

2.2 2. Implication of ERV env genes in placental structural transitions in Hyaenidae

acid trees were constructed with PhyML 3.0 (Guindon et al., 2010), with bootstrap percentages computed after 1,000 replicates. Substitution models were selected using the SMS program (Lefort et al., 2017): LG+Γ+I for TM sequences, WAG+Γ+I+F for SU sequences and RtREV+Γ for RT sequences. For all genomes available at http://genome-euro.ucsc.edu/index.html, or- thology and other genome wide in silico studies were performed on the same site using the BLAT algorithm. Assemblies used were: Felis catus felCat8 (2014), Canis familiaris canFam3 (2011), Mustela putorius furo musFur1 (2011) and Ailuropoda melanoleuca ailMel1 (2009). The BLASTn algorithm (https://blast.ncbi.nlm.nih.gov/Blast.cgi) was used to screen the genomes of Panthera tigris PanTig1.0, Ursus maritimus UrsMar_1.0, Acinonyx jubatus aciJub1, Odobenus rosmarus divergens Oros_1.0 and Leptonychotes weddellii LepWed1.0, all of them available at http://www.ncbi.nlm.nih.gov/genome/?term=carnivora.

Entry date in the hyena species. For both Hyena-Env2 and 3 presence or absence was assayed by PCR of an internal fragment in all species indicated in the carnivoran family tree. PCR was performed on 100 ng genomic DNA, using conserved primers (see Table ) 1 and Accuprime 2.2 polymerase (Invitrogen) for 50 cycles (30 s at 94 ◦C, 30 s at 55 ◦C, 1 min at 68 ◦C). For species where a sequenced genome is available BLASTn was used to assay the presence of either of the envelopes. Finally, for Hyena-Env2 the full-length reading frame of all four species of Hyaenidae and the empty locus in cat, dog and the 4 species closest to Hyaenidae according to our tree (Arctictis binturong, Cryptoprocta ferox, Suricata suricatta and Helogale parvula) were amplified by PCR using the above protocol. In all sequenced carnivora the sequence of the empty locus was determined by BLASTn search.

Identification and sequencing of the Hyena-Env2 associated provirus and integration site. Inverse PCR was performed by digesting 2 µg of Crocuta crocuta genomic DNA with NcoI (New England Biolabs) and inactivating the enzyme at 65 ◦C for 30 min before DNA purification using the Gel and PCR clean up kit (Macherey-Nagel) and self-ligation using T4 DNA Ligase (New England Biolabs). PCR amplification was performed with divergent primers (see Table 1) and Accuprime DNA Polymerase (Invitrogen) for 50 cycles (30 s at 94 ◦C, 30 s at 55 ◦C, 7 min at 68 ◦C) followed by a nested PCR using the same conditions with nested divergent primers (see Table 1). Fragments of >1 kb were gel purified and cloned into a pGEM-T Easy vector (Promega) before sequencing. Sequences were checked for presence of the Hyena-env gene and BLAT analysis was performed to identify the syntenic region of the Felis catus genome. Locus primers were designed based on the syntenic region (see Table 1) and used to amplify the locus in hyenas and other carnivora. The full Hyena-Env2 associated provirus was amplified from 100 ng of Crocuta crocuta genomic DNA using Long Expand DNA polymerase (Roche) with Buffer 2, 0.32 µM of each primer and the following program: 10 cycles of 10 s at 94 ◦C, 30 s at 55 ◦C and 8 min at 68 ◦C, followed by 30 cycles of 15 s at 94 ◦C, 30 s at 55 ◦C and 8 min +20 s/cycle at 68 ◦C. The PCR

162 ieidctdi h rmrname. primer the study. in this indicated in site used primers of List – 1 Table HeaEv-URXo 5’-ATACATCTCGAGAGAGTGTTTTTGTTCAGGCTG Hyena-Env2-SU-R-XhoI 5’-ATACATTCATGAATCATAAACCTAGATCCACCCC Hyena-Env2-SU-F-BspHI Antibody production 5’-TGGGGTTTAGGCAAGAGGTGAG Hyena-Env2-5’RACE-R2 5’-CCCAGGAGCCACAAAAGAAATG Hyena-Env2-5’RACE-R1 5’-CTTACTCCTTGCCTCCACCCTCA Hyena-Env2-3’RACE-F2 5’-CAACCCCGACAGCCCTGAAA Hyena-Env2-3’RACE-F1 RACE 5’-CCCCTCAGGTGTCATTCC Hyena-Env2-InvPCR-R2 5’-GCACCAAATGTTATGTCTGAA Hyena-Env2-InvPCR-R1 5’-CTTCACTCTGCTGCCGC Hyena-Env2-InvPCR-F2 5’-GTAAAAACCCCGCTCACTAT Hyena-Env2-InvPCR-F1 5’-GATTAAGCATAGGGGTCCCInverse PCR 5’-GTTTTCGTTCCCCTTCTCA 5’-TTCCAGGAGGGGTACAGTAG Hyena-Env3-HIS-R3 5’-GCCCAAATAGGACTCTCTGA Hyena-Env3-HIS-F3 5’-ATTAAGTTCAGAGGCCCAAAA Hyena-Env3-HIS-R2 5’-CAAACCTTACTCCCTGGAGAC Hyena-Env3-HIS-F2 5’-GGGGGACCAGTTAAAGATG Hyena-Env3-HIS-R1 5’-CCTAGTCACCCTCTTTCCA Hyena-Env3-HIS-F1 5’-CTGTAGCCGTCATGGTCTG Hyena-Env2-HIS-R3 5’-GGCTGCAAATCATAAACCT Hyena-Env2-HIS-F3 5’-AAAAGCCTATGTCGTGACC Hyena-Env2-HIS-R2 5’-ATTGACATTAACCGGACCA Hyena-Env2-HIS-F2 Hyena-Env2-HIS-R1 Hyena-Env2-HIS-F1 In situhybridization 5’-NNNNNNACGCGTTTACATAACYGAYTCCTTCTCGCAGGG 5’-ATACATACGCGTTTTTACTGTATGCACCCTT 5’-NNNNNNGAATTCGCCACCATGGTGGCCGATCCSCC Carni-SLC1A5-R-MluI 5’-ATACATCTCGAGATCAATAACCTTAAGCGGG Carni-SLC1A5-F-EcoRI 5’-ATACATACGCGTAGTGAAGCCCCGAACAATA Hyena-Env3-R-MluI Hyena-Env3-F-XhoI 5’-ATACATGAATTCCAAAGAGGACTACCGGGCA Hyena-Env2-R-MluI Hyena-Env2-F-EcoRI Cloning 5’-TGCATTTTACTGGTTTCCTTA 5’-CRGGGYTTTAATTTGTGATC Hyena-Env2-Locus-R1 Hyena-Env2-Locus-F1 5’-GTGGGCAGGCGAGAGTA Locus amplification 5’-CTCCTGGAATATTTCTCTCATG 5’-AGTGAAGCCCCGAACAATA Hyena-Env3-cons-R 5’-AACCCCGACAGCCCTGAAA 5’-CTGTAGCCGTCATGGTCTG Hyena-Env3-cons-F 5’-GGCTGCAAATCATAAACCT Hyena-Env2-ORF-R Hyena-Env2-ORF-F Hyena-Env2-Int-R Hyena-Env2-Int-F 5’-TCCATGAGAATCCGCTTGT 5’-CAGTGGGGAGAAAATTTTGTAConservation 5’-ACCGTGAATCTAAGAAGATTG 5’-AGACCAAGCTAAACACCATCA 5’-GCCCCACACCTTCACTACT Carnivora-RPL19-R 5’-ACGACCTAATTACCCTTGTTC Carnivora-RPL19-F Hyena-Env3-R 5’-AGGAGTAAGCCACATAAATTC Hyena-Env3-F 5’-CAGACTGCCAAAATCATGA Hyena-Env2-R Hyena-Env2-F Syncytin-Car1-R Syncytin-Car1-F RT-qPCR ..Heaseicrtoia neoegn atr n lcnaevolution placenta and capture gene envelope retroviral Hyena-specific 2.2. nelndsqecscrepn otersrcinenzyme restriction the to correspond sequences Underlined 163

2.2 2. Implication of ERV env genes in placental structural transitions in Hyaenidae

product was purified and sequenced using primers designed to anneal about every 600 bp.

Characterization of the placental Hyena-Env2 transcript. Rapid amplification of cDNA ends (RACE) was performed using the SMARTer RACE cDNA Amplification Kit (Clontech). cDNA was generated from 100 ng Crocuta crocuta placental RNA and the placental Hyena-Env2 transcript 5’ and 3’ ends were amplified using the "RACE" primers in Table 1 and the Advantage 2 Polymerase Kit (Clontech) with the following program: 5 cycles of 30 s at 94 ◦C and 5 min at 72 ◦C, followed by 5 cycles of 30 s at 94 ◦C, 30 s at 70 ◦C and 5 min at 72 ◦C and finally 30 cycles of 30 s at 94 ◦C, 30 s at 68 ◦C and 5 min at 72 ◦C.

Real-time RT-qPCR. Hyena-env RNA expression was determined by RT-qPCR. Reverse transcription was performed with 100 ng of DNase-treated RNA as in (de Parseval & Heidmann, 1998). PCR was carried out with 10 µl of diluted (1:10) cDNA in a final volume of 25 µl using the FastSYBR Green PCR Master Mix (Qiagen) in an ABI PRISM 7000 sequence detection system. Transcript levels were normalized relative to the amount of the housekeeping RPL19 gene 2.2 (ribosomal protein L19). Samples were assayed in triplicate, see Table 1 for primer sequences.

In situ hybridization. Freshly collected Crocuta crocuta placentas were fixed in 4% paraformaldehyde at 4 ◦C and were embedded in paraffin. Serial sections (4 µm) were ei- ther stained with H&E or were used for in situ hybridization. Three PCR-amplified fragments of Hyena-Env2 (452 bp, 416 bp, and 496 bp) and three fragments of Hyena-Env3 (415 bp, 448 bp and 414 bp) (primers are listed in Table 1) were cloned into pGEM-T Easy (Promega) for in vitro synthesis of the antisense and sense riboprobes generated with SP6 RNA polymerase and digoxigenin 11-UTP (Roche Applied Science) after cDNA template amplification. In situ hybridization of hyena syncytin-Car1 transcripts was performed using two sets of sense and antisense riboprobes directed against cat syncytin-Car1 (Cornelis et al., 2012) Sections were processed, hybridized at 42 ◦C overnight with the pooled riboprobes, and incubated further at 4 ◦C overnight with alkaline phosphatase-conjugated anti-digoxigenin antibody fragments (Roche Applied Science). Staining was revealed with nitroblue tetrazolium and 5-bromo-4-chloro-3- in- doyl phosphate phosphatase alkaline substrates, as indicated by the manufacturer (Roche Applied Science).

Antibody production and immunohistochemistry. A DNA fragment containing 197 amino acids of the Hyena-Env2 envelope SU subunit (amino acids 240-437; for primers see Table 1) was inserted into the pET28b (Novagen) prokaryotic expression vector and expressed in BL21(DE3) bacteria. The recombinant C-terminally His-tagged protein was purified from bacteria lysates by nickel affinity chromatography. Mouse immunization was performed in accordance with stan- dard procedures, using female 8-week-old BALB/c mice. Pre-immunization sera as well as sera containing polyclonal antibodies were recovered independently from 10 mice and were tested

164 Crei ta. 03 n eegoni MMsplmne ih1%FS(Invitrogen), FCS 10% 37 with at penicillin, supplemented U/mL in DMEM 100 described in and are streptomycin, grown mg/mL lines were 100 cell and All 2013) microscope. al., under et assayed (Cornelis syncytia of absence or presence transfected:target of and 50:50 or 30:70 20:80, of ratio at a and at plates well 24 in seeded lines and cell resuspended target were alongside cells transfection post h 4 Technologies). (Life LTX using plasmid) 2.75 infection. before h 48 technologies) (Life LTX relevant, using When transfected microscope. were under cells counted target foci infection and post-infection h 72 performed was µ ( plates 0.45 well harvested, size 24 were pore supernatants membranes, containing (PVDF kit virus transfection filtered post-transfection, JetPRIME h the 48 using transfection). vectors expression (Polyplus Env (A-MLV) virus leukemia murine 0.5 and nlsLacZ), expressing cells 293T 106 cotransfecting 1 by with produced were above. as particles vector PCR phCMV viral and same MLV the Hyena-Env-pseudotyped 1) into Table cloned and (see MluI and primers crocuta EcoRI conserved using Crocuta digested were using and fragments cDNA dog placental Cat, from France). amplified Lyon, was de SLC1A5 Supérieure Normale Ecole ( Cosset, EcoRI F.-L. with from digested were They 1). ( Table (see primers "Cloning" the assays. and fusogenicity and vectors Expression staining. day of same intensity the standardized on a plates obtain cover to in immunostained were slides kit PowerVision All Technologies). Mouse peroxidase/diaminobenzidine (ImmunoVision the using visualized was pre-immune Staining corresponding serum. the or serum anti-Hyena-Env2 with immunohisto- 9; overnight pH incubated For EDTA, and (Tris vector. Abcam) retrieval expression antigen heat-induced Hyena-Env2 for a processed were or sections paraffin vector chemistry, empty an 293T with on immunohistochemistry transfected performing by cells staining sig- background for of tested absence then were and experiment strength Hyena-Env2 blot nal Western a the with of transfected sera transiently specific most cells The vector. 293T expression of lysates using analyses blot Western by Hyena-Env3 /L.Siouainwspromda 20gfr25ha omtmeaue -a staining X-Gal temperature. room at h 2.5 for g 1200 at performed was Spinoculation g/mL). o ocluecl-elfso assays, fusion cell-cell co-culture For Hyena-Env3 µ 2 yn-n xrsinvco n 2.75 and vector expression Hyena-Env g µ × Mi(L a-o xrsinvco) 1.5 vector), expression gag-pol (MLV CMVi g 10 5 n lIadcoe notepCVvco GnakacsinA381,gift AJ318514, accession (GenBank vector phCMV the into cloned and MluI and ) el/elttl -a tiigwspromd2 r4 fe tr fco-culture of start after h 48 or 24 performed was staining X-Gal total. cells/well 5 − Rswr mlfidfrom amplified were ORFs ..Heaseicrtoia neoegn atr n lcnaevolution placenta and capture gene envelope retroviral Hyena-specific 2.2. 8 × 10 4 µ el/el,splmne ihPlbee(ia ocnrto 8 concentration (Final Polybrene with supplemented cells/well), yn yctnCr,HeaEv,HeaEv3o amphotropic or 3 Hyena-Env Hyena-Env2, syncytin-Car1, hyena g 2 µ × ;Mlioe n rnfre otre el in cells target to transferred and Millipore) m; rct crocuta Crocuta 10 µ C rget otiigthe containing fragments PCR fRS a l-aZepesn reporter expressing nls-LacZ (an R9SA of g 6 E23 el eec-rnfce with co-transfected were cells HEK293T µ ◦ n %CO 6% and C F-lLc rtoia genome (retroviral MFG-nlsLacZ g eoi N yPRusing PCR by DNA genomic 2 Hyena-Env2 . Hyena-Env2 rXhoI or ) 165

2.2 2. Implication of ERV env genes in placental structural transitions in Hyaenidae 2.2

166 Part III

Discussion

167 1. Two novel examples of ERV Env exapted for a potential role in placentation and a new member of the SLC1A5 interference group

Our investigation into the implication of endogenous retroviral env in placentation of Hyaenidae and Mabuya have led us to the identification of a non-mammalian syncytin, a new mammalian syncytin-like and an endogenous envelope belonging to a well-known interference group. In this section, I will first discuss a few interesting aspects of each of these envelopes, while the next sections will consider their links to placental evolution and potential non-canonical roles in placentation.

1.1 Hyena-Env2 is a conserved syncytin-like envelope gene expressed during hyena placentation

Hyena-Env2 is the first of two novel envelopes identified following our high-throughput se- quencing of a hyena placental transcriptome. Hyena-Env2 presents all major functional domains of a retroviral gamma-type envelope, including a very well conserved ISD, and we have been able to characterize the entire associated provirus, including the conserved RT sequence which supports a gammaretroviral origin for this gene. Conservation studies show that Hyena-Env2 is a gene present only in hyenas, BLAST analysis in 3 Feliformia and 6 Caniformia genomes yielded no sequence bearing significant similarities and all attempts to amplify the Hyena-Env2 ORF, or parts of it, were negative in these species. The identification of a single insertion locus in Crocuta crocuta and the fact that this locus presents the Hyena-Env2 associated provirus in all extant hyena species (with an identical TSD sequence) confirmed that the single insertion event occurred before radiation of modern day hyenas. Conversely, all investigated syntenic non-hyena loci were empty, with the non-duplicated TSD sequence readily identifiable (except in the case of dog in which this genomic region is deleted). The exact age of the Hyena-Env2 insertion into the Hyaenidae genome is difficult to pinpoint due to the evolutive history of hyenas. This family split of from other carnivorans about 30 Mya

168 ih10 dnia Tsi o losget eetatvt fti aiy Nevertheless, family. this of activity recent suggests also the dog of in many provirus LTRs intact identical almost an 100% of every with detection almost The in endogenizations. of recent loci insertions suggesting orthologous detectable at species, present many different not are the which of by conserved revealed most the as genome, Caniformia active) maypublished still it is active expression, been (or has placental recently retrovirus associated very weak Hyena-Env3 the very that seems shows It investigation. and further hyena-specific warrant not interference is Hyena-Env3 SLC1A5 While the of member a is Hyena-Env3 1.2 in sections, following the in discussed be potential will 3.2. The in syncytin-likes syncytin-likes. particular other non-fusogenic conserved and Hyena-Env2 the as of well roles as syncytins could that by roles placental fulfilled other of be number a are there however placentation, hyena in role syncytin 2005). al., et (Dupressoir day gestation on gestation both depending that distinct expression shown from differential been highly samples has it placental available, where readily are mice, periods In attained. is phenotype hemochorial that hemochorial. possible already is sequence is that to placenta It used mid-gestation sample a RNA to the corresponds ii) transcriptome placental and the expression, in important keep similarly to a important necessitate as is not expression for may it placental role but proposed Its Syncytin-Car1, the physiology. of i) placenta that that in than mind lower role much possible is RT-qPCR a by at assessed layers hint unfused the still in expression placenta and the conservation its of loss), fusion intrinsic an not and cells target both ORFs, ERV of canonical characterization other the the and of hyenas extant fossil ORF all in Hyena-Env2 the in provirus the in state associated case, coding any indication the functional In no hyenas. a extinct is in 70 conserved there the been tissue, of has mode soft reproductive of the composed to common as is 10-30 last record imprecise placenta very a a the share at Hyena-Env2 hyenas since of extant Sadly insertion of 4 Mya. age the The the of puts 1992). one which Mya, now Gauthier, 10 is about & families ancestor Queiroz carnivoran (de species-rich most poor the subsist, species clade of most this one of once representatives was of extant which record 4 family fossil only this the but Nowadays species, 1992). known Gauthier, 70 & comprises Queiroz Hyaenidae (de 20-30 Mya comprising 10 diversity, about its of species, height different the reached and 2012) Bininda-Emonds, & (Nyakatura h o-uignct fHeaEv biul rvnsi rmacmlsigaclassical a accomplishing from it prevents obviously Hyena-Env2 of non-fusiogenicity The While ..HeaEv samme fteSCA nefrnegopwt euirtropism peculiar a with group interference SLC1A5 the of member a is Hyena-Env3 1.2. ru ihapcla tropism peculiar a with group Hyena-Env2 env Rsascae ihteeisrin r o-oigadteoedsrbdas described one the and non-coding are insertions these with associated ORFs Hyena-Env2 snnfsgnci u sa wihmgtb u otentr four of nature the to due be might (which assay our in non-fusogenic is rct crocuta Crocuta xrsini aibedrn etto n erae nea once decreases and gestation during variable is expression Hyena-Env2 gag ae traiyaprn htti a o h case the not was this that apparent readily it makes and sdsic fta ftecrioa yctnand syncytin carnivoran the of that of distinct is pol rsniganme fdisruptions. of number a presenting syncytin-A and -B rncit display transcripts 169

1.2 1. Two novel Env with a potential placental role and a new SLC1A5 Env

Canis-Env3 by Cornelis et al. (2012) has lost its initiation site when compared to the Hyena-Env3 ORF, losing its signal peptide sequence in the process and rendering the envelope non-functional. Hyena-Env3 on the other hand still displays significant fusogenic activity and while it might not serve a role in placentation currently, it is possible that this envelope comes to play a more important role in the future, following the example of the EnvV2/HERV-W substitution and the baton pass model, after all the HERV-W insertion that gave rise to Syncytin-1 also occurred in Old World monkeys, but the envelope was not conserved in that clade. The relationship between Hyena-Env3 and its cognate receptor SLC1A5 might also prove to be interesting in the context of both endogenous and exogenous retrovirology. Most SLC1A5 using envelopes do not show any particular tropism apart from the group-wide shared difficulty with infecting rodent cells. In the case of Hyena-Env3 the tropism is much more restricted as 1.3 the Env seems to be able to interact only with carnivoran SLC1A5. Analysis of the sequence of Hyena-Env3 SU and carnivoran SLC1A5 have only revealed that in the former some of the motives conserved among SLC1A5 using Env SUs are absent or, more often, present insertions of one or two amino acids. It is unclear whether these differences are able to explain the particular tropism of Hyena-Env3, but further investigation might shed more light on the receptor specificity of retroviral envelopes and in particular those of the well-known SLC1A5 group, which also contains two previously described mammalian syncytins.

1.3 Syncytin-Mab1: characterization of a non-mammalian syncytin and its receptor

Syncytin-Mab1 is the first syncytin identified in a non-mammalian species. This identification suggests that the process of ERV Env exaptation for a role in placentation is not limited to the mammalian clade, but that it might have occurred in some (or all) of the non-mammalian placental species. Similarly to what is observed for all presently described mammalian syncytins, the expression of syncytin-Mab1 co-localizes with the presence of a fused cell layer in the Mabuya placenta. This observation, alongside the fusogenic potential observed in ex vivo assays, suggests that Syncytin-Mab1 participates in the establishment of the placental syncytium. Interestingly, while Syncytin-Mab1-pseudotyped MLV particles display very high titers in our infection assay, cell- cell fusion is observed only in presence of an acidic shock. This is similar to the behavior of Syncytin-Rum1 (Cornelis et al., 2013), but here again the physiological relevance of this property is unknown. While it is possible that fusion in vivo is at least partly controlled by a temporary acidification of the placental medium, it is also possible that pH-dependence is an artifact of our assay. Indeed, the ex vivo assays were performed using mammalian cells to express and assay proteins from a distantly related lizard clade. It is possible that the cells we used do not express certain lizard specific factors which might result in differential fusion-impacting

170 u s fsc oeuea eetri aeeog ob oe here. noted be to enough rare is below, receptor detail as more molecule in a discussed such be of will use physiology pathways. but placental signaling for cellular nature important this several of in for implications implicated it rule been The gamma-type), has the that not than transducer is rather signal exception Syncytin-Mab1 a the that is is hypothesis (which the protein supporting transmembrane envelopes, pass gamma-type single a it is only our so. that do also to but able mammals, be in to observed seems able classically method be those to screening necessary from is differ it that that Env shows detect only not to it as species, non-mammalian in envelopes syncytin-likes and of identified profile has peculiar screen The the our envelopes. in gamma-type but found avian birds, lizard in fusion of exclusively examples predicted TM. almost coding gamma-type found shifted two avian usually in a is observed present commonly type is they envelope as This that residue, cysteine and by flanked analysis is phylogenetic that our peptide in TM RSV type activity. assay fusion pseudotyping Env canonical the its in to potential detrimental fusogenic not high is its it but that Syncytin-Mab1, indicates to specific retroviruses, is lizard it among common if is or characteristics asso- envelope particular gamma-type this and whether beta- of unknown of is formation ciation for It envelopes. partner gamma-type SU in the Additionally, bond provides SU/TM associated. covalent that covalently the motif not CXXC are the TM present and not SU does the Syncytin-Mab1 which cysteine, in third TM the beta-type present in CX not situation does of bond form SU/TM the the of taking establishment tree. for phylogenetic essential based is TM and CX TMs a the on that reveals Env however gamma-type inspection spumaretro- other or Closer with epsilon- an clustering be a and to presenting envelope, short envelopes, too viral gamma-type much being standard ISD, to divergent somewhat close retroviral if very known recognizable seems other it all glance to first compared At when envelopes. characteristics unusual very unknown. some is presents role Mab1 this of nature exact the but interaction, pattern maternofetal This in role interface. a maternofetal the support would forming membrane expression the sub-cellular at particular exclusively very localized a reveals Analysispattern, physiology. paraplacentome the placental in in pattern function expression another the accomplish of also might it that suggest would or Syncytin-Mab1 of processing post-translational h lcna xrsino yctnMb naespeetn nyufsdcl layers cell unfused only presenting areas in Syncytin-Mab1 of expression placental The ial,tentr ftecgaercpo fSnyi-a1i oehtsrrsn.Not surprising. somewhat is Syncytin-Mab1 of receptor cognate the of nature the Finally, gamma- avian the with cluster -4 and Mab-Env3 that see to interesting is it note, similar a On Syncytin- represents, it evolution convergent of example remarkable the disregarding Even ..Snyi-a1 hrceiaino o-amla yctnadisreceptor its and syncytin non-mammalian a of characterization Syncytin-Mab1: 1.3. Mabuya p Vtasrpoei fpriua eeac otesre o syncytins for screen the to relevance particular of is transcriptome IV sp. 7 nta.Tepeec fol w ytie srmnseto the of reminiscent is cysteines two only of presence The instead. C n Cmtfta scnevdi l gamma-type all in conserved is that motif CC Mabuya PL ntetocl types. cell two the in MPZL1 nsilico in 171

1.3 2. Implication of syncytins and syncytin-likes in placental evolution

2.1 Retroviral insertions have been a constant driving force of vertebrate evolution

In this study, we describe several recent endogenization events, the syncytin-Mab1 containing provirus was first endogenized by an ancestral lizard about 30 Mya and Hyena-Env2 capture occurred between 10 and 30 Mya in the last common ancestor of all extant hyenas. Additional examples of recent and even still ongoing endogenizations can be observed in mice, sheep and koalas (reviewed in Dewannieux & Heidmann, 2013; Greenwood et al., 2017). On the other hand, the oldest known endogenization took place in early Palaeozoic fishes, over 450 Mya, at about the same time as the apparition of jawed vertebrates and long before the ocean-land transition (Aiewsakun & Katzourakis, 2017). Between these two extremes reside a number of known endogenization events, Hayward et al. (2015) identify almost 100,000 ERVs among 65 published vertebrate genomes and suggest that no jawed vertebrate clade has been entirely spared from retroviral endogenizations, though the total amount observed in each clade shows great deal of variability. Each new retroviral insertion is a source of genetic diversity, both through the retroviral sequences it brings with it and the randomness of the insertion locus. In mice where ERV activity is still high, it has been shown that at least 10% of de novo mutations observed in laboratory mice during the last 30 years are due to new ERV insertions (Maksakova et al., 2006). It is also important to remember that the 8% of the human and 10% of the mouse genome occupied by sequences of retroviral origin (Lander et al., 2001; Waterston et al., 2002) represent only the fraction of retroviral insertions that have been fixed in the population, and the evolutive impact of all endogenizations was without a doubt more important. Even so and as discussed in the introduction, there are numerous examples of retroviral insertions still affecting cellular gene expression in animals to this day and a growing amount of evidence shows that retroviral gene exaptation is an ongoing and widespread phenomenon. Interestingly, early development and placentation seem particularly affected by these pro- cesses. This might be in part due to the fact that compared to many other animal traits, pla- centation evolved relatively late, allowing for easier comparative analysis between placental and

172 fnwhs niEVfcos htcnltras eue o h salsmn flnaespecific lineage 2017). of al., establishment the et for (Imbeault used networks be expression also gene later can that sequences the factors, ERV anti-ERV generation for host through case new indirectly pair of and this KRAB-ZFP/ERV insertion In their through the processes. variability genetic of epigenetic to directly through co-option contribute expression the gene to host these due of of be regulation Conservation 2017). to al., seems et (Imbeault proteins genome corresponding KRAB-ZFP the the integrate after actively to long ceased selection has purifying ERV that under shown has conserved study been recent have a only host, KRAB-ZFP their was some and insertions invaders genomic retroviral between new race arms and an KRAB-ZFP and of reflective of acquired co-evolution newly observed of the been that silencing had it assumed While a 2017). Magiorkinis, allowing & Hurst expression, in (reviewed ERV elements ERV pathogenic of potentially implicated regulation been have epigenetic proteins These the proteins. of in family KRAB-ZFP the of example the in found al., LTR et lineage-specific (Chuong network from regulation 2016). complex derived and extensive number be an A constitute to now innate-immunity. and shown events in insertion been response have interferon promoters of non-placental regulation A interferon-inducible 2012). the of al., in et found (Emera be regulatory primates can and element example mice repeated in LTRs species-specific precisely all by more in driven and that sequence is shows expression network regulatory endometrial underlying hormone cases the this three of of expression studies endometrial and clades, independently three emerged endometrium these has the In in elephants. prolactin and of expression mice the primates, is examples of famous most the of One (Lynch 2015). evolution al., mammalian et by on sequences elements ERV regulatory of new amplification of subsequent acquisition and of endogenization impact major the includ- illustrating sequences, sequences, repeated ERV ex- from ing promoters originally tissue-specific by them regulated is of genes expression these most Endometrial of 2015). many tissue, al., of et (Lynch endometrial systems hemolymphoid in and gut expression brain, for the in recruited pressed be thousands to that had reveal mammals genes placental Studies of of tissues. new that the with in monotremes expression of gene transcriptomes regulate to comparing able be to as so networks, expression gene might expression ERV of effects deleterious consequences. and dire organs less have other placenta therefore the than Finally, short-lived functionality. more their much contribute is to them allowing otherwise elements, of expression ERV allow repressed have which events germline Additionally, demethylation population. the with general in associated the expressed is in ERVs development fixation early that of and reason amplification to that number stands copy fact of It odds the cells. better retrovirus these to a of endogenized part one be in infect to since to and relationship, needs changes, special a necessary share retroviruses the and decrypt cells germline to order in animals non-placental nsmay ervrssadvrertshv oeitdfraot40M n retroviral and My 450 about for co-existed have vertebrates and retroviruses summary, In is variation genetic generate can insertions retroviral which in way indirect more final, A new of establishment the necessitates placenta, the as such organ, new a of apparition The ..Rtoia netoshv enacntn rvn oc fvrert evolution vertebrate of force driving constant a been have insertions Retroviral 2.1. 173

2.1 2. Implication of syncytins and syncytin-likes in placental evolution

insertions have been a constant on-going phenomenon in that timespan, contributing to genetic diversity and potentially enabling major evolutive transitions, notably in the domain of reproduc- tion. The insertion of the lineage specific syncytin-Mab1 and Hyena-Env2 containing proviruses are yet more examples of this phenomenon.

2.2 Syncytin-Mab1 and convergent evolution of animal pla- centas

2.2.1 Convergent use of ERV env genes in placentation

Viviparous matrotrophy has evolved several times independently in animals. The association of these two modes of reproduction has emerged repeatedly in fishes, once in mammals and at least 6 times in squamates (Blackburn, 2015). In all cases structures involved in gas and nutrient transfer have been co-opted to serve a new purpose at the maternofetal interface and often derive from modified oviduct or uterine tissues (Blackburn, 2015). The reasons behind the emergence of this reproductive mode are supposed to be diverse, ranging from thermal advantages in the case of squamata to increased control over fetal development in eutherians (Blackburn, 2015; Griffith & Wagner, 2017), but some of the resulting structures show high functional and 2.2 sometimes structural homology. In placental viviparous matrotrophic species, the membranes of mother and fetus have evolved to be in contact and often display specializations for maternofetal exchange. The apparition of this organ specialized for maternofetal exchanges in clades as disparate as fishes, lizards and mammals is a noteworthy case of evolutive convergence. In this study we have focused on the placenta of a particular group of lizards, the neotropical genus Mabuya. The Mabuya placenta is a recent emergence (about 25 Mya) and presents many specialized structures that are also found in mammals: areola, absorptive plaques and a region of strong maternofetal interdigitation that also presents a fused cell layer. The existence of this fused cell layer was of particular interest to us, as it seemed to originate through cell-cell fusion events in the maternal epithelium (Leal & Ramírez-Pinilla, 2008), reminiscent of the cell-cell fusion events observed at the maternofetal interface in many mammalian placentas. Such fusion events need presence of a fusogen and in many cases animal fusogens are derived from viral proteins. In the context of mammalian placentas in particular, it has now been well-established that ERV derived fusogenic syncytins are necessary for establishment of fused cell layers at the maternofetal interface (Dupressoir et al., 2009, 2011). Syncytins are env genes that have been acquired stochastically, are expressed in the placenta and are fusogens. They are under purifying selection, as can be expected for a gene with an important role in placentation. Here we have shown that the Mabuya placenta expresses such a captured retroviral envelope, notably in the fused cell layer, and that this envelope presents fusogenic properties. The discovery of

174 oeei ee,aogmmaspaetsaehgl iegn.Temmainpaet is placenta mammalian The divergent. phy- highly high a are at placentas seen mammals when convergence among of level, amounts logenetic high shows placentas of evolution the While of evolution divergent and syncytin-likes and Syncytins 2.3 genus. each the of explaining a placentas syncytins the of acquired of identification differentially of particularities the that to to lead potentially only also but not syncytin, could founding situation This present to organization. seem placentas in current differences the but probable marked Mya, 30 is about clade, It this syncytins. in once non-mammalian emerged viviparity of that investigation continued the for interest paramount the of of complexity placenta the the While matrotrophic. highly is that Eumecia the of apparition the with emergence. associated processes placental the of on light placenta. time some mammalian properties the shed the might towards at and clues genes acquired new such yield and model. of might genus syncytin pass founding this baton this of of the characterization species Further with tested accordance all in in in syncytins placentation present unidentifiable, other being of by of during emergence replaced point degenerated recent the been has The to it since placenta, that has mammalian probable the it is of as it apparition and since My date 175 to the identified mammalian been with not associated has syncytin mammalian founding placentation its The the provided viviparity. observed matrotrophic of that is to (as syncytin, emergence oviparity monotremes) matrotrophic founding in the from ancestral transition that the an allowing now functionalities of new time capture with host the long with a associated for was proposed placenta been founding has a potential already a It of is description species syncytin. first remote the also this is of it evolution, placenta convergent the of example in remarkable syncytin syncytin a founding of a involvement potential of the example While first the as Syncytin-Mab1 2.2.2 mammalian to assigned function fusogenic placenta the the of accomplish syncytins. layer nevertheless cell could fused it the in that expression suggests its but placentation, association mammalian the in presenting CX implicated Env, with ERV unusual ISD an conserved is a Syncytin-Mab1 of mammals. in case the be to complex a of emergence the the in that placenta indicates Syncytin-Mab1, syncytin, non-mammalian first this ope lcna r lopeeti cnia eeacoeyrltdwith related closely genera Scincidae in present also are placentas Complex h amla placenta mammalian the and ..Snyisadsnyi-ie n iegn vlto ftemmainplacenta mammalian the of evolution divergent and syncytin-likes and Syncytins 2.3. Lubuya Mabuya Lubuya eeaaepr ftesm rnhadec oti igespecies single a contain each and branch same the of part are genera eu a copne ytecpueo ervrlEv si thought is as Env, retroviral a of capture the by accompanied was genus rsnsahgl naieftlsnyim hscaewudb of be would clade This syncytium. fetal invasive highly a presents 7 oi,stigi pr rmtegma rbtrtoia Env betaretroviral or gamma- the from apart it setting motif, C Mabuya loe st n igesnyi gene syncytin single a find to us allowed Eumecia lcnai oryknown, poorly is placenta Mabuya The . 175

2.3 2. Implication of syncytins and syncytin-likes in placental evolution

assumed to be monophyletic, having emerged in the last common ancestor of all mammalian species about 175 Mya. From this ancestral placenta have spawned a number of different pla- cental organizations and have turned the placenta into the organ showing the highest structural diversity among mammals. It is not entirely clear why placental organization is this divergent and there does not seem to be a clear evolutive sequence to the transition. While the epithe- liochorial placenta of pigs is much less complex than the hemochorial placenta of mice, both allow appropriate provision of nutrients to the embryo over the course of gestation, and some indicators even suggest that the much simpler pig placenta is more efficient than more complex ones. Interestingly, the repartition of placental types among mammals is clearly paraphyletic, major organizational categories being found in several independent branches. This presents a problem when trying to reconstitute the evolutionary history of mammalian placentation, and parsimonious reconstructions result in the counter-intuitive hypothesis that ancestral mammalian placentation was of the highly invasive hemochorial type, while the derived state is the epithe- liochorial type in which placental membranes are simply apposed. There are several major limitations to this method however. First, classification of mammalian placentas is a highly com- plex process as even within the great categories of epithelio-, synepithelio-, endothelio-, and hemochorial placentation there exists a great amount of variability, both in terms of the struc- tural organization of placental tissue layers and in terms of placental invasiveness and actually proximity of maternal and fetal tissues. These finer details are too cumbersome to be used in a large-scale parsimony analysis and thus get lost in the reductive simplification of mammalian placental categories. Second, analyses based only on the structure of current day definitive mam- malian placentas do not take into account the impact of stochastic gene acquisition reflected by syncytins and syncytin-likes. The acquisition of different new genes implicated in placental 2.3 phylogeny in the different branches of mammals is extremely likely to have an important impact on the emergence of different modes of placentation. The random nature of ERV acquisition and exaptation by different branches is in stark opposition to the concept behind parsimony phylogeny. Of note, the newly emerged placentas of squamates are closer to the less invasive ep- itheliochorial type, with only the placenta of Lubuya ivensii presenting a significant degradation of maternal tissue. While it is risky to infer the status of the original mammalian placenta based on these recent non-mammalian placentas, they hint at a smoother transition from two epithelia separated by an egg-shell to two apposed epithelia, instead of jumping from separated epithelia to direct trophoblast-maternal blood contact.

The description in the present study of two new env acquired specifically at the time of placental emergence (Mabuya placenta) or a major structural transition (hemochorial hyena placenta) are two more examples of the potential impact the acquisition of such genes can have on placental physiology. The transition from an endotheliochorial to hemochorial phenotype in hyenas would represent a reversion in the hemo- to epitheliochorial model, the reason of which would be unclear. Evolution of the endotheliochorial placenta of carnivorans seems to be associated with the acquisition of the shared Syncytin-Car1, which has been conserved under

176 oei iewt h lentv oe rpsdoe h er o yctn n syncytin-likes and syncytins for years be the would over placentation. proposed and mammalian roles placentation, in alternative mammalian the in with genes line syncytin in of more role canonical of the from implication different potential characteristic The spaces blood placentas. maternal large hemochorial the the create is of to placenta as probably this so in most vessels difference blood is major maternal fusion the of but breaching Cell-cell gene, layer. Syncytin-Car1 cell shared fused the novel on dependent a still include not does hyenas in observed 2011). 2009, structural al., major et to (Dupressoir leads embryos genes all necessitates these of placenta of death absence hemotrichorial potential the the and that deficiencies of and layer syncytin different fused a each of that in presence shown the placentation, hemochorial been described of even emergence syncytins has the of it with number time mice a in been associated have is There capture whose phenotype. to-date placental new Syncytin-Car1 a by provided in already resulting those to and adding functions its the placentation, hyena with to associated add-on potentially as function, of new gain to a a endotheliochorial of it through an capture such is from has hyenas As transition in it function. the placenta explain that of hemochorial to loss, suggesting a hypothesis than promoter, probable rather cellular most gain, the a with like by associated seems usually driven process mainly a be exapted, to been newly seems the expression hyenas extant presents its all family among and conserved Hyeanidae is a present gene the This to still expression. contrary, reversion placental is with explain the Hyena-Env2 gene acquired could On This that type. My. differences placental 75 functional present remarkable discernible previously a no for presents species and hyenas these in of all in selection purifying h custo of acquisition The ..Snyisadsnyi-ie n iegn vlto ftemmainplacenta mammalian the of evolution divergent and syncytin-likes and Syncytins 2.3. env ihpaetlepeso uhas such expression placental with Hyena-Env2 sacs fpriua neeta h tutrltransition structural the as interest particular of case a is Hyena-Env2 Hyena-Env2 hsnwgn ol serve could gene new This . ol hrfr every be therefore would 177

2.3 3. Syncytin-likes and alternative envelope functions in the placenta

3.1 Growing evidence of conservation of non-fusogenic Env

Even though it is non-fusogenic and thus cannot fulfill a classical syncytin role in placentation, Hyena-Env2 shows signs of conservation and continued expression in hyena placentas. It is therefore part of the growing list of somewhat orphan genes that can be referred to as syncytin- likes. This group of env also comprises previously identified genes such as HEMO, pan-Mars- Env2, Env-Cav1, ERV3 and EnvV2 in humans. All of these genes are ERV env genes expressed in the placenta and conserved as coding ORFs for 30-80 My. None of them are fusogenic, either due to a probable intrinsic loss of fusogenicity (Hyena-Env2, Env-Cav1, EnvV2) or loss of membrane attachement (HEMO, pan-Mars-Env2, ERV3). Evolutive models predict that ERV sequences that do not fulfill a role in host physiology should quickly degenerate to non-coding sequences or solo LTRs. Analyses in the human genome show that in most ERV groups 90% of integrations degenerate to solo LTRs within 5-10 My (Gemmell et al., 2016). Conservation of coding ORFs with low non-synonymous to synonymous mutation rates therefore strongly suggest that these non-fusogenic syncytin-likes fulfill an important role in host physiology. Examples illustrating the loss of non-exapted env genes can be found in Old World monkey in which the HERV-W Env was not exapted to serve as syncytin and was instead lost during evolution. While the actual nature of the roles of non-fusogenic syncytin-likes in placentation (or other processes) are still unknown, there are hints available based on their nature as retroviral Env, their pattern of expression, and the implication of some of them in physiological processes.

3.2 Alternative placental roles for Env

3.2.1 Immunosuppression

Since most syncytins and syncytin-likes are derived from gamma-type Env containing retro- viruses, many present a conserved ISD sequence and could present immunosuppressive proper- ties in vivo. In humans and mice it has been shown that in each case one of the two syncytins displays immunosuppressive properties in the in vivo mouse tumor rejection assay (Syncytin-1

178 ih rvn h s famuebsdasy diinly iadpaetto osntse to seem not Env does placentation retroviral lizard Additionally, lizard assay. based of mouse nature a of its use immunosup- though the the prevent Syncytin-Mab1, assay might founding and the try its to of for interesting properties selected extremely pressive be not would was cell-cell it syncytin substantial context founding this present the In not fusogenicity. that did hypothesis placenta the mammalian strengthening ancestral events, evolution, the fusion placental that for probable pathway even hemochorial to is epitheliochorial it the placentation. to for prerequisite suscribes absolute one an if not is Indeed, layer syncytialized a of presence the that this dicating to contribute could tolerance. in -B maternofetal involved of and actors processes Syncytin-2 only as of the such not number are syncytins they a while phenomenon, are and fetus there the placentas of mammalian tolerance day ancestral the current of expression In the immuno- syncytin. through particular potentially a established, foreign be contact, of to this tissues needed accommodate To context with suppressive barrier. tissues egg-shell maternal inert of an of contact instead oviparous direct origin, previous the The the therefore monotremes. is and in reproduction placenta observed of mammalian mode early still the is between as distinction reproduction, major of of probable logistics mode lead-up the oviparous the in part) matrotrophic such in least the as (at solved to and been already mammals, had fetus placental and mother and between transfer monotremes nutrient of ancestor present probably common very the was Matrotrophy in placenta. mammalian the of emergence during syncytin maternal of dampening placenta. a the of to vicinity contributing they the peptide, factors in released responses diffusible the immune of secreted part as is HEMO) act for theoretically ISD the could of part (or ISD the Since syncytin-likes. EnvV2 functional. the nevertheless that is mention but sequence to consensus important the is from divergence it strong though shows divergent, ISD more ISD slighly the while is sequence, Syncytin-Car1 CKS-17 canonical of the of residues of conserved case the the to In compared viral it mutations fusogenicity, proteins. to to syncytin-like linked of responses not role is immune potential function suppress a this is Since to retroviruses. thought infectious of are case they the in as particles just system, immune imagine maternal to complicated the not is therefore blood is maternal It the tissue. fetal example, that foreign for genetically placenta with hyena contact hemochorial direct the in in true, particularly placentation invasive is more In this backgrounds. genetic different from tissues of contact direct is there the in cells 2007). tumor al., against et responses immune reduce to vivo able in also syncytin- were the that EnvV2 showed and study ERV3 same likes The 2007). al., et Mangeney respectively, Syncytin-B and nadto,cretdyeihlohra lcna ontpeetcl-elfso vns in- events, fusion cell-cell present not do placentas epitheliochorial day current addition, In founding the by provided function major the was immunosuppression that probable is It soluble truncated the of case the in intriguing particularly is role immunosuppressive An as role, key a plays system immune maternal the of regulation placentation, of context the In env sa,dmntaigta hspoet sidpneto uoei oeta (Mangeney potential fusogenic of independent is property this that demonstrating assay, ee ihpaetlepeso lyarl npoetn h lcnafo takby attack from placenta the protecting in role a play expression placental with genes ..AtraiepaetlrlsfrEnv for roles placental Alternative 3.2. Hyena-Env2 h S hw nytwo only shows ISD the , 179

3.2 3. Syncytin-likes and alternative envelope functions in the placenta

present the same focus on creation of an immunosuppressive context as observed in mammalian placentas (Griffith & Wagner, 2017). The exact mechanisms of fetal protection from the maternal immune system in squamates are not known, but they might implicate the presence of retroviral ISD sequences.

3.2.2 Differentiation and proliferation

Another proposed role of ERV derived genes in placental physiology concerns differentiation of trophoblast cells. Creation of new cell differentiation pathways is instrumental in the evolution of a new organ and the placenta presents a number of specific differentiated cell populations, both on the maternal and fetal side. Human trophoblasts for example can persist as isolated unfused CT, presenting certain stem-like properties such as continued proliferation during gestation, can fuse into the ST, a very active secretory tissue, or can differentiate into invasive EVT cells which then split into the endovascular and insterstitial sub-populations. It has been shown that Syncytin-1 is implicated in some of these differentiation pathways (Frendo et al., 2003), but once again a role in differentiation does not seem to be restricted to syncytins sensu stricto. Among syncytin-likes, the human ERV3 Env has been shown to be implicated in trophoblast cell differentiation ex vivo (Lin et al., 1999, 2000) and sheep enJSRV env seem to play an important role in the formation and differentiation of BNC cells (Dunlap et al., 2005). Knock-down of enJSRV transcripts prevents implantation of the sheep conceptus and leads to embryo death (Dunlap et al., 2006). Contrarily to fusogenic and immunosuppressive functions, roles in differentiation are difficult to predict as there is no particular retroviral Env motif or domain associated with them whose absence or presence can be examined. Of note, expression of the HEMO syncytin-like was on the contrary associated with an undifferentiated stem-cell phenotype, in a way reminiscent of that observed for HERV-H expression (Heidmann et al., 2017). It is possible that this Env contributes to a maintenance of an undifferentiated state. The presence of undifferentiated and proliferating trophoblast cells is an important aspect of placentation as the surface area of the organ expands massively in a relatively short time, necessitating a constant supply of new cells. In human placentas both the amount of nuclei in the ST and that in the CT layer show a constant increase during gestation (Simpson et al., 1992; Mayhew et al., 1994). It is unknown if Syncytin-Mab1 or Hyena-Env2 play a role in cell differentiation, as men-

3.2 tionned above such a role is hard to predict and to investigate. The expression of Hyena-Env2 in the CT layer would be in accordance with a potential role in CT differentiation or proliferation.

3.2.3 Invasion

Invasion of maternal tissues by invasive fetal cells is a common phenomenon in placentation. It is observed both in mammals, where it ranges from invasion by a few specialized cells to invasion of the whole conceptus and degradation of all maternal tissue layers, and in non-mammalian placentas. In the Mabuya placenta a subset of fetal cells send out projections that cross the

180 nainosre nterrsetv lcna.Teheaseicepeso of expression hyena-specific The placentas. respective their in observed invasion 2011). al., in et expressed (Vernochet been strongly placenta has but pig guinea This non-fusogenic the is tissues. of which these tissues example, syncytial of invasive for invasiveness Env-Cav1 the in of to role implicated as contribute also suggested therefore be could Env to syncytin-like placenta and shown the syncytin of been in Expression have 2007). al., invasion et pathways (Ferretti and from processes the invasive allograft migration of placental an Several cell than 2017). cancer cancer al., a et in like Macaulay more 2016; implicated is al., whole et a (Beaman as perspective placenta mother’s similarities the the of that number argued a been share has cells it trophoblast and to and able tumor Interestingly, be 2017). al., to et and related (Lemaître properties invasion The oncogenic and ability. possess migration invasive cell also and enhance to migratory shown increased was with Env (Wootton HERV-K associated for cells endogenous process JSRV of exogenous transformation a ongogenic of 2005), induce envelope al., to The sufficient et exist. be properties to shown such been of has examples example few a though far, localized so a of formation the in involved uterine distance. are the interhemal they inside the that reducing establish phenotype, likely they endothelial seems network it membrane unknown, extensive is the exact syncytium and the While cells times. all these at of epithelium placental purpose the of with body contact in the however, remains cells placentation cells invasive synepitheliochorial few these or a epitheliochorial Contrary only vessels. in blood observed and maternal is term with what contact to until direct subsist establish layers that projections tissues the invasive out maternal While of send all mass. that which category, cell in second placental state the the main intermediate into of invasion the squarely one falls increase of hyenas show least independently of that at tissues placenta cells which maternal of in into subsets placentas placentation. present migrate hemochorial also all and and placentas in invasive endotheliochorial case more i.e., blood degraded, these the maternal is Often is between term This at distance layers whole. the cell reduces a maternal invasion as other of the placenta type on the contents other their and The secrete barrier. to this them the allows these which of epithelium in by side maternal placentas concerned the synepitheliochorial with are in BNC endometrium also cells of the but fusion into specialized hormones, migrate few BNC placental in a which secrete system, in they only placentation circulatory where a as horse maternal plays in true observed the it is not to cases This is access few processes. opposite direct a the more In cases objectives. a main these secretions particular two fetal somewhat maternal allowing with last, the in occur this of of role to exception wall seems the the invasion With through chamber. placental jet breaks case, the conceptus in whole itself the salps, implant non-vertebrate to as ovary of role placentation major the a in plays Even invasion epithelium. maternal invasive entire more the detach the layers in and epithelium uterine iial,w rps htHeaEv n yctnMb ol otiuet h fetal the to contribute could Syncytin-Mab1 and Hyena-Env2 that propose we Similarly, retroviral of involvement The env xvivo ex nivsv rcse a o ensuidextensively studied been not has processes invasive in yuigtesm aha sta sdb JSRV by used that as pathway same the using by uuaivensii Lubuya ..AtraiepaetlrlsfrEnv for roles placental Alternative 3.2. lcnaivsv syncytialized invasive placenta Mabuya rsnsacurious a presents Hyena-Env2 181

3.2 3. Syncytin-likes and alternative envelope functions in the placenta

at the maternofetal interface would be an elegant explanation to the hyena-specific invasive phe- notype displayed by this organ, even though our preliminary transwell migration and invasion assays performed with Hyena-Env2 transduced untransformed MCF10A cells proved inconclu- sive. It is possible that this type of assay with this type of cell is not adapted to the probably limited transformative power of a syncytin-like Env, this function seeming to be lessened in endogenous Env when compared to their exogenous counterpart. Neither HERV-K nor enJSRV Env were able to induce oncogenic transformation in 208F cell ex vivo for example, while the exogenous JSRV led to the formation of transformed foci (Lemaître et al., 2017). The invasive potential of syncytin-likes such as Hyena-Env2 will therefore most probably need to be studied using highly sensitive migration and invasion assays. An additional way of studying a potential role of an env gene in invasion is the identification of its cognate receptor (see discussion of Syncytin-Mab1/MPZL1 below). Search for the Hyena-Env2 receptor is currently ongoing, using cDNA library based screening assays similar to those described in Bianchi et al. (2014) and Bacquin et al. (2017).

In the Mabuya placenta, the invasive cells represent a small sub-population that sends out thin processes into the uterine syncytium. These complex but small structures are difficult to observe using optical microscopy, develop only very late during gestation, and seem to be very labile during sample preparation (Leal & Ramírez-Pinilla, 2008). Accordingly we have not been able to conclusively investigate the expression of Syncytin-Mab1 in this cell population, though its expression profile would certainly allow it to play a role in the invasive fetal cells. The identification of the MPZL1 as the Syncytin-Mab1 receptor even provides a possible mechanism for an invasive potential of this syncytin. Indeed, MPZL1 has been previously shown to increase cell invasion and migration in mouse (with a potential role in early embryo formation, Roubelakis et al., 2007) and in human hepatocellular carcinoma (Jia et al., 2014). The activity of MPZL1 as a signal transducer is mediated by its two intracellular immunoreceptor tyrosine-based inhibitory motives (ITIM). Dimerization of MPZL1 induces phosphorylation of Y241 and Y263 (one in each ITIM) which leads to signal transduction through the SHP-2/Src pathway (Zhao & Zhao, 1998; Zhao et al., 2002). Interestingly SHP-2 is also implicated in the ERK1/2 pathway which has been shown to be to be important for HERV-K/JSRV oncogenicity (Lemaître et al., 2017). Transduction experiments have shown that expression of Syncytin-Mab1 is sufficient to trigger

3.2 MPZL1 phosphorylation and activation. This activation can also be achieved using concanavalin A, which binds to several MPZL1 monomers and brings them into close contact, triggering activation (Zhao et al., 2002), or simply overexpression of MPZL1, increasing the odds of dimer formation (Jia et al., 2014). We therefore propose that interaction of a Syncytin-Mab1 Env trimer with MPZL1 might trigger its activation in a similar fashion, presenting three SU subunits with MPZL1 binding sites. This mechanism could occur at the membrane surface (within the same cell membrane or two adjacent membranes) or within the cell, though the latter seems to also lead to a decrease in the total amount of MPZL1 as observed in our transduction experiment, possibly through a mechanism similar to receptor interference. Syncytin-Mab1 expressing cells did not

182 ue aes(si h aeo h rsnl ecie yctnMb,btas Syncytin-1 also but non- Syncytin-Mab1, also described but presently fused the in of expressed case be the can is syncytins (as that layers fact fused previously The to genes. related seem syncytin even which described of some functions, syncytin-like for candidates potential and events, insertion placentation. mammalian of types other all in in role do important probably an the they play as and might placentation, genes roles, epitheliochorial syncytin-like that different propose It to of above. unreasonable discussed number not as therefore fusogenicity, a is for selected play not theoretically probably was could syncytin Env mammalian ancestral ERV placenta, of type this structure. simple its comparatively of and with invasion many type, Additionally, of epitheliochorial amount investigated. the low been resemble relatively closely not most has placentas placentation non-mammalian syncytins mammalian simpler of last the involvement high the the a represents with which and species in horses, of and type number a pigs in a observed such is importance placenta economic of syncy- type no This which observed. in is type tissue only tialized the is placentation epitheliochorial placentas, mammalian Among placentation epitheliochorial in Syncytin-likes 3.3 capabilities. impact transport direct a modifying having by is membrane, physiology cell interaction cell the Env-receptor on at Still, latter processes. the of signaling down-regulation cell by accompanied in often role direct receptors, less as a transporters playing pass multiple probably use of mostly most which that gamma-type, however the mind of in are kept envelopes be might these to syncytin-likes has It soluble interaction. even is receptor fusion, through Syncytin-Mab1 physiology not host by affect and MPZL1 aggregation of receptor activation on previously dependent the of probably involvement Since non-fusogenic syncytin-likes. to the and on led identification syncytins light The already described additional (2017). have shed al. and might et Bacquin receptors underway by more are Ly6E of receptor syncytins Syncytin-A of large-scale the Accordingly, of receptors partner. identification host cognate the their the of those identify on to as efforts properties depend can intrinsic genes their these of on role much the as that suggesting physiology, host on effects their and syncytins reasons. same the potentially for MPZL1, of effect downstream significant a potential reveal failed its same to cells MCF10A lessening the transduced activation, using in assays following invasion MPZL1 and latter migration and transwell Similarly, the Syncytin-Mab1 impact. of of of expression amount degradation that as to total fact little leads the the as cell by to to increase or phosphorylation signaling assays), Src their MPZL1 that in of report 15% (2014) contribution al. small et (Jia relatively effector a phosphoryated to This ERK1/2). due (Src, be MPZL1 from might downstream situated molecules of activation increased show en etbae pteicoilseishv etil o ensae rmretroviral from spared been not certainly have species epitheliochorial vertebrate, Being in role a play not do syncytins that suggest might layers cell fused of absence the While of study general the in interest particular of is MPZL1 and Syncytin-Mab1 of case The nsilico in cen ftepbihdhreadpggnm eelanme of number a reveal genome pig and horse published the of screens ..Snyi-ie neihlohra placentation epitheliochorial in Syncytin-likes 3.3. 183

3.3 3. Syncytin-likes and alternative envelope functions in the placenta

for example) indicates that absence of cell-cell fusion does not necessarily imply absence of syncytins, and some epitheliochorial placentas, like that of horse, present differentiated and invasive cell subpopulations, two processes that implicate syncytins and syncytin-likes in other mammalian placentas. While the investigation of the implication of retroviral envelopes in these placentas will prob- ably be much more complicated, since the non-fusogenic roles proposed for these envelopes are harder to demonstrate and less well-known, these studies nevertheless seem crucially impor- tant in order to really understand the potential contributions of ERVs to all of mammalian and non-mammalian placentation. 3.3

184 4. Conclusion

In the present study we have investigated the implication of syncytins and syncytin-like genes in two exceptional placentas. The placenta of hyenas shows an extremely invasive phenotype when compared to that of other carnivorans, while the neotropical Mabuya lizard genus presents one of the few truly complex non-mammalian placentas. The acquisition of Hyena-Env2 correlates in time with the major structural transition of the placenta of Hyaenidae, while the characteristics and expression of the previously present Syncytin-Car1 do not display any signifcant changes. Hyena-Env2 was shown to be present exclusively in Hyaenidae and conserved in a coding state in all extant members of this family, at a syntenic chromosomal locus. This species-specific acquisition at the same time as a major species-specific structural transition in the placenta suggests that Hyena-Env2 serves as an add- on to Hyaenidae placentation, adding its functionality to that of Syncytin-Car1 and leading to the establishment of the invasive hemochorial placenta that is exclusive to this family among carnivorans. Syncytin-Mab1 is the first identified non-mammalian syncytin, extending this now commonly admitted concept beyond simply mammals, and suggesting that the use of novel genes of retroviral origin might be a wide-spread phenomenon in animals. The use of syncytins for the establishment of placental structures in very distantly related species is a remarkable example of convergent evolution and strengthens the hypothesis that emergence of complex placentas necessitates the capture of a founding syncytins. The presently identified non-mammalian syncytins might be the first example of this group, having been acquired in the same timespan as that of placental emergence. Additionally, Syncytin-Mab1 seems to be a unique envelope associating properties from different retroviral clades and the identification of its receptor suggests an implication in placental invasion as well as cell-cell fusion. The identification in the present study of several new endogenous retroviral envelopes con- tributes to the already known diversity and importance of these genes in placentation, but also extends these concepts to new contexts, reinforcing their involvement in placental structural organization and the emergence of new placentas.

185 4. Conclusion 4.0

186 Part IV

Bibliography

187 Bibliography

Abbondanzieri, E. A., Bokinsky, G., Rausch, J. W., Zhang, J. X., Le Grice, S. F. J., & Zhuang, X. (2008). Dynamic binding orientations direct activity of HIV reverse transcriptase. Nature, 453(7192), 184–189. Adam, S., Elfeky, O., Kinhal, V., Dutta, S., Lai, A., Jayabalan, N., Nuzhat, Z., Palma, C., Rice, G. E., & Salomon, C. (2017). Review: Fetal-maternal communication via extracellular vesicles - Implications for complications of pregnancies. Placenta, 54, 83–88. Agoni, L., Guha, C., & Lenz, J. (2013). Detection of Human Endogenous Retrovirus K (HERV- K) Transcripts in Human Prostate Cancer Cell Lines. Frontiers in Oncology, 3, 180. Aiewsakun, P., & Katzourakis, A. (2017). Marine origin of retroviruses in the early Palaeozoic Era. Nature Communications, 8, 13954. Ali, L. M., Rizvi, T. A., & Mustafa, F. (2016). Cross- and co-packaging of retroviral RNAs and their consequences. Viruses, 8(10), 11–20. Alkhatib, G., Combadiere, C., Broder, C. C., Feng, Y., Kennedy, P. E., Murphy, P. M., & Berger, E. A. (1996). CC CKR5: A RANTES, MIP-1alpha , MIP-1beta Receptor as a Fusion Cofactor for Macrophage-Tropic HIV-1. Science, 272(5270), 1955–1958. Allen, W., & Moor, R. M. (1972). The origin of the equine endometrial cups. Journal of Reproduction and Fertility, 29, 313–316. Allen, W. R., Hamilton, D. W., & Moor, R. M. (1973). The origin of equine endometrial cups. II. Invasion of the endometrium by trophoblast. The Anatomical Record, 177(4), 485–501. Apps, R., Murphy, S. P., Fernando, R., Gardner, L., Ahad, T., & Moffett, A. (2009). Human leucocyte antigen (HLA) expression of primary trophoblast cells and placental cell lines, determined using single antigen beads to characterize allotype specificities of anti-HLA antibodies. Immunology, 127(1), 26–39. Aristotle (ca -350). Tonperi ta zoia historion. Using the D’ArcyWentworth Thompson translation (1910) available at http://classics.mit.edu/Aristotle/history_anim.html. Arnaud, F., Murcia, P. R., & Palmarini, M. (2007). Mechanisms of late restriction induced by an endogenous retrovirus. Journal of virology, 81(20), 11441–11451. Aydin, H., Cook, J. D., & Lee, J. E. (2014). Crystal Structures of Beta- and Gammaretrovirus Fusion Proteins Reveal a Role for Electrostatic Stapling in Viral Entry. Journal of Virology, 88(1), 143–153. Babaian, A., & Mager, D. L. (2016). Endogenous retroviral promoter exaptation in human cancer. Mobile DNA, 7(1), 24–45. Bacquin, A., Bireau, C., Tanguy, M., Romanet, C., Vernochet, C., Dupressoir, A., & Heidmann, T. (2017). A Cell Fusion-Based Screening Method Identifies Glycosylphosphatidylinositol- Anchored Protein Ly6e as the Receptor for Mouse Endogenous Retroviral Envelope Syncytin- A. Journal of Virology, 91(18), e00832–17. Baczyk, D., Drewlo, S., Proctor, L., Dunk, C., Lye, S., & Kingdom, J. (2009). Glial cell missing- 1 transcription factor is required for the differentiation of the human trophoblast. Cell Death and Differentiation, 16(5), 719–727.

188 Ballandras-Colas, A., Brown, M., Cook, N. J., Dewdney, T. G., Demeler, B., Cherepanov, P., Lyumkis, D., & Engelman, A. N. (2016). Cryo-EM reveals a novel octameric integrase structure for betaretroviral intasome function. Nature, 530(7590), 358–361. Ballandras-Colas, A., Maskell, D. P., Serrao, E., Locke, J., Swuec, P., Jónsson, S. R., Kotecha, A., Cook, N. J., Pye, V. E., Taylor, I. A., Andrésdóttir, V., Engelman, A. N., Costa, A., & Cherepanov, P. (2017). A supramolecular assembly mediates lentiviral DNA integration. Science, 355(6320), 93–95. Balliet, J. W., Kolson, D. L., Eiger, G., Kim, F. M., McGann, K. A., Srinivasan, A., & Coll- man, R. (1994). Distinct Effects in Primary Macrophages and Lymphocytes of the Human Immunodeficiency Virus Type 1 Accessory Genes vpr, vpu, and nef: Mutational Analysis of a Primary HIV-1 Isolate. Virology, 200(2), 623–631. Baltimore, D. (1970). Viral RNA-dependent DNA Polymerase: RNA-dependent DNA Poly- merase in Virions of RNA Tumour Viruses. Nature, 226(5252), 1209–1211. Barbot, W., Dupressoir, A., Lazar, V., & Heidmann, T. (2002). Epigenetic regulation of an IAP retrotransposon in the aging mouse: progressive demethylation and de-silencing of the element by its repetitive induction. Nucleic Acids research, 30(11), 2365–73. Barbour, L. A., Shao, J., Qiao, L., Pulawa, L. K., Jensen, D. R., Bartke, A., Garrity, M., Draznin, B., & Friedman, J. E. (2002). Human placental growth hormone causes severe insulin resistance in transgenic mice. American Journal of Obstetrics and Gynecology, 186(3), 512–517. Barrau, M. D., Abel, J. H., Torbit, C. A., & Tietz, W. J. (1975). Development of the implantation chamber in the pregnant bitch. American Journal of Anatomy, 143(1), 115–129. Bastida-Ruiz, D., Van Hoesen, K., & Cohen, M. (2016). The Dark Side of Cell Fusion. International Journal of Molecular Sciences, 17(5), 638. Battaglia, F. C., & Meschia, G. (1978). Principal Substrates of Fetal Metabolism. Physiological Reviews, 58(2), 499–527. Battini, J. L., Danos, O., & Heard, J. M. (1995). Receptor-binding domain of murine leukemia virus envelope glycoproteins. Journal of Virology, 69(2), 713–719. Battini, J. L., Danos, O., & Heard, J. M. (1998). Definition of a 14 amino-acid peptide essential for the interaction bet the murine leukemia virus amphotropic envelope glycoprotein and its receptor. Journal of Virology, 72(1), 428–435. Battini, J. L., Heard, J. M., & Danos, O. (1992). Receptor choice determinants in the envelope glycoproteins of amphotropic, xenotropic, and polytropic murine leukemia viruses. Journal of Virology, 66(3), 1468–1475. Beaman, K. D., Jaiswal, M. K., Katara, G. K., Kulshreshta, A., Pamarthy, S., Ibrahim, S., Kwak- Kim, J., & Gilman-Sachs, A. (2016). Pregnancy is a model for tumors, not transplantation. American Journal of Reproductive Immunology, 76(1), 3–7. Belshaw, R., Pereira, V., Katzourakis, A., Talbot, G., Paces, J., Burt, A., & Tristem, M. (2004). Long-term reinfection of the human genome by endogenous retroviruses. Proceedings of the National Academy of Sciences, 101(14), 4894–4899. Belshaw, R., Watson, J., Katzourakis, A., Howe, A., Woolven-Allen, J., Burt, A., & Tristem, M. (2007). Rate of Recombinational Deletion among Human Endogenous Retroviruses. Journal of Virology, 81(17), 9437–9442. Belzile, J. P., Duisit, G., Rougeau, N., Mercier, J., Finzi, A., & Cohen, É. A. (2007). HIV-1 Vpr- mediated G2 arrest involves the DDB1-CUL4AVPRBP E3 ubiquitin ligase. PLoS Pathogens, 3(7), 0882–0893. Benirschke, K., Burton, G. J., & Baergen, R. N. (2012). Pathology of the Human Placenta. Berlin, Heidelberg: Springer Berlin Heidelberg.

189 Bénit, L., Dessen, P., & Heidmann, T. (2001). Identification, Phylogeny, and Evolution of Retroviral Elements Based on Their Envelope Genes. Journal of Virology, 75(23), 11709– 11719. Bénit, L., Lallemand, J. B., Casella, J. F., Philippe, H., & Heidmann, T. (1999). ERV-L elements: a family of endogenous retrovirus-like elements active throughout the evolution of mammals. Journal of Virology, 73(4), 3301–8. Berkhout, B., Grigoriev, A., Bakker, M., & Lukashov, V. V. (2002). Codon and amino acid usage in retroviral genomes is consistent with virus-specific nucleotide pressure. AIDS research and human retroviruses, 18(2), 133–141. Berkhout, B., Silverman, R. H., & Jeang, K. T. (1989). Tat trans-activates the human immunod- eficiency virus through a nascent RNA target. Cell, 59(2), 273–282. Best, S., Tissier, P. L., Towers, G., & Stoye, J. P. (1996). Positional cloning of the mouse retrovirus restriction gene Fvl. Nature, 382(6594), 826–829. Bharat, T. A. M., Castillo Menendez, L. R., Hagen, W. J. H., Lux, V., Igonet, S., Schorb, M., Schur, F. K. M., Krausslich, H.-G., & Briggs, J. A. G. (2014). Cryo-electron microscopy of tubular arrays of HIV-1 Gag resolves structures essential for immature virus assembly. Proceedings of the National Academy of Sciences, 111(22), 8233–8238. Bharat, T. A. M., Davey, N. E., Ulbrich, P., Riches, J. D., De Marco, A., Rumlova, M., Sachse, C., Ruml, T., & Briggs, J. A. G. (2012). Structure of the immature retroviral capsid at 8Å resolution by cryo-electron microscopy. Nature, 487(7407), 385–389. Bhetariya, P.J., Kriesel, J. D., & Fischer, K. F. (2017). Analysis of Human Endogenous Retrovirus Expression in Multiple Sclerosis Plaques. Journal of emerging diseases and virology, 3(2), 1–17. Bianchi, E., Doe, B., Goulding, D., & Wright, G. J. (2014). Juno is the egg Izumo receptor and is essential for mammalian fertilization. Nature, 508(7497), 483–7. Bii, V.,& Trobridge, G. (2016). Identifying Cancer Driver Genes Using Replication-Incompetent Retroviral Vectors. Cancers, 8(11), 99. Billich, S., Knoop, M.-T., Hansen, J., Strop, P., Sedlacek, J., Mertz, R., & Moelling, K. (1988). Synthetic Peptides as Substrates and Inhibitors of Human Immune Deficiency Virus-1 Pro- tease. The Journal of biological chemistry, 263, 17905–17908. Bjerregaard, B., Holck, S., Christensen, I. J., & Larsson, L. I. (2006). Syncytin is involved in breast cancer-endothelial cell fusions. Cellular and Molecular Life Sciences, 63(16), 1906– 1911. Bjerregaard, B., Talts, J. F., & Larsson, L.-I. (2011). The Endogenous Envelope Protein Syn- cytin Is Involved in Myoblast Fusion. In Cell Fusions, (pp. 267–275). Dordrecht: Springer Netherlands. Bjerregaard, B., Ziomkiewicz, I., Schulz, A., & Larsson, L.-I. (2014). Syncytin-1 in differen- tiating human myoblasts: relationship to caveolin-3 and myogenin. Cell and tissue research, (pp. 355–362). Björkman, N. H. (1969). Light and electron microscopic studies on cellular alterations in the normal bovine placentome. The Anatomical Record, 163(1), 17–29. Blackburn, D. G. (1993a). Chorioallantoic placentation in Squamate reptiles: Strcutre, function, development, and evolution. Blackburn, D. G. (1993b). Histology of the late-stage placentae in the matrotrophic skink Chalcides chalcides (Lacertilia; Scincidae). Journal of Morphology, 216(2), 179–195. Blackburn, D. G. (1993c). Standardized Criteria for the Recognition of Reproductive Modes in Squamate Reptiles. Herpetologica, 49(1), 118–132.

190 Blackburn, D. G. (1994). Standardized criteria for the recognition of embryonic nutritional patterns in squamate reptiles. Copeia, 4(1), 925–935. Blackburn, D. G. (1999). Are viviparity and egg-guarding evolutionarily labile in squamates? Herpetologica, 55(4), 556–573. Blackburn, D. G. (2015). Evolution of vertebrate viviparity and specializations for fetal nutrition: A quantitative and qualitative analysis. Journal of Morphology, 276(8), 961–990. Blackburn, D. G., & Flemming, A. F. (2010). Reproductive specializations in a viviparous African skink and its implications for evolution and conservation. Herpetological Conserva- tion and Biology, 5(2), 263–270. Blackburn, D. G., & Flemming, A. F. (2012). Invasive implantation and intimate placental associations in a placentotrophic African lizard, Trachylepis ivensi (Scincidae). Journal of Morphology, 273(2), 137–159. Blackburn, D. G., & Vitt, L. J. (2002). Specializations of the chorioallantoic placenta in the Brazilian scincid lizard, Mabuya heathi: a new placental morphotype for reptiles. Journal of Morphology, 254(2), 121–31. Blackburn, D. G., Vitt, L. J., & Beuchat, C. A. (1984). Eutherian-like reproductive specializations in a viviparous reptile. Proceedings of the National Academy of Sciences, 81(15), 4860–4863. Blaise, S., de Parseval, N., Bénit, L., & Heidmann, T. (2003). Genomewide screening for fusogenic human endogenous retrovirus envelopes identifies syncytin 2, a gene conserved on primate evolution. Proceedings of the National Academy of Sciences, 100(22), 13013–8. Blaise, S., Mangeney, M., & Heidmann, T. (2001). The envelope of Mason–Pfizer monkey virus has immunosuppressive properties. Journal of General Virology, 82(7), 1597–1600. Blankenship, T. N., & Given, R. L. (1992). Penetration of the uterine epithelial basement membrane during blastocyst implantation in the mouse. The Anatomical Record, 233(2), 196–204. Bleck, M., Itano, M. S., Johnson, D. S., Thomas, V. K., North, A. J., Bieniasz, P. D., & Simon, S. M. (2014). Temporal and spatial organization of ESCRT protein recruitment during HIV-1 budding. Proceedings of the National Academy of Sciences, 111(33), 12211–12216. Blond, J.-L., Besème, F., Duret, L., Bouton, O., Bedin, F., Perron, H., Mandrand, B., & Mallet, F. (1999). Molecular characterization and placental expression of HERV-W, a new human endogenous retrovirus family. Journal of Virology, 73(2), 1175–1185. Blond, J.-L., Lavillette, D., Cheynet, V., Bouton, O., Oriol, G., Chapel-Fernandes, S., Mandrand, B., Mallet, F., & Cosset, F.-L. (2000). An Envelope Glycoprotein of the Human Endogenous Retrovirus HERV-W Is Expressed in the Human Placenta and Fuses Cells Expressing the Type D Mammalian Retrovirus Receptor. Journal of Virology, 74(7), 3321–3329. Bolze, P.-A., Mommert, M., & Mallet, F. (2017). Chapter Five - Contribution of Syncytins and Other Endogenous Retroviral Envelopes to Human Placenta Pathologies. In W. R. Huckle (Ed.) Molecular Biology of Placental Development and Disease, vol. 145, (pp. 111–162). Academic Press. Bone, Q., Pulsford, A. L., & Amoroso, E. C. (1985). The placenta of the salp (Tunicata: Thaliacea). Placenta, 6(1), 53–63. Bonnaud, B., Beliaeff, J., Bouton, O., Oriol, G., Duret, L., & Mallet, F. (2005). Natural history of the ERVWE1 endogenous retroviral locus. Retrovirology, 2, 1–7. Bonnaud, B., Bouton, O., Oriol, G., Cheynet, V., Duret, L., & Mallet, F. (2004). Evidence of selection on the domesticated ERVWE1 env retroviral element involved in placentation. Molecular Biology and Evolution, 21(10), 1895–1901. Brady, T., Lee, Y. N., Ronen, K., Malani, N., Berry, C. C., Bieniasz, P. D., & Bushman, F. D.

191 (2009). Integration target site selection by a resurrected human endogenous retrovirus. Genes & Development, 23(5), 633–642. Bray, M., Prasad, S., Dubay, J. W., Hunter, E., Jeang, K. T., Rekosh, D., & Hammarskjold, M. L. (1994). A small element from the Mason-Pfizer monkey virus genome makes human immunodeficiency virus type 1 expression and replication Rev-independent. Proceedings of the National Academy of Sciences, 91(4), 1256–1260. Brierley, I. (1995). Ribosomal frameshifting on viral RNAs. Journal of General Virology, 76(8), 1885–1892. Britten, R. J., & Davidson, E. H. (1969). Gene Regulation for Higher Cells: A Theory. Science, 165(3891), 349–357. Brody, B. A., Rhee, S. S., & Hunter, E. (1994). Postassembly cleavage of a retroviral glycoprotein cytoplasmic domain removes a necessary incorporation signal and activates fusion activity. Journal of Virology, 68(7), 4620–7. Brosens, I., Pijnenborg, R., Vercruysse, L., & Romero, R. (2011). The “Great Obstetrical Syn- dromes” are associated with disorders of deep placentation. American Journal of Obstetrics and Gynecology, 204(3), 193–201. Brosens, I., Robertson, W. B., & Dixon, H. G. (1967). The physiological response of the vessels of the placental bed to normal pregnancy. The Journal of Pathology and Bacteriology, 93(2), 569–579. Brown, K., Emes, R. D., & Tarlinton, R. E. (2014). Multiple Groups of Endogenous Epsilon-Like Retroviruses Conserved across Primates. Journal of Virology, 88(21), 12464–12471. Brown, P. O., Bowerman, B., Varmus, H. E., & Bishop, J. M. (1989). Retroviral integration: structure of the initial covalent product and its precursor, and a role for the viral IN protein. Proceedings of the National Academy of Sciences, 86(8), 2525–2529. Bryant, M., & Ratner, L. (1990). Myristoylation-dependent replication and assembly of human immunodeficiency virus 1. Proceedings of the National Academy of Sciences, 87(2), 523–7. Burton, G. J., & Fowden, A. L. (2015). The placenta: a multifaceted, transient organ. Philo- sophical Transactions of the Royal Society B: Biological Sciences, 370(1663), 20140066– 20140066. Burton, G. J., & Jauniaux, E. (1995). Sonographic, stereological and Doppler flow velocimet- ric assessments of placental maturity. BJOG: An International Journal of Obstetrics and Gynaecology, 102(10), 818–825. Burton, G. J., & Tham, S. W. (1992). Formation of vasculo-syncytial membranes in the human placenta. Journal of developmental physiology, 18(1), 43–7. Burton, G. J., Woods, A. W., Jauniaux, E., & Kingdom, J. C. (2009). Rheological and Physio- logical Consequences of Conversion of the Maternal Spiral Arteries for Uteroplacental Blood Flow during Human Pregnancy. Placenta, 30(6), 473–482. Büscher, K., Trefzer, U., Hofmann, M., Sterry, W., Kurth, R., & Denner, J. (2005). Expression of Human Endogenous Retrovirus K in Melanomas and Melanoma Cell Lines. Cancer Research, 65(10), 4172–4180. Bustamante Rivera, Y. Y., Brütting, C., Schmidt, C., Volkmer, I., & Staege, M. S. (2018). Endogenous retrovirus 3-History, physiology, and pathology. Frontiers in Microbiology, 8, 1–9. Campos-Olivas, R., Newman, J. L., & Summers, M. F. (2000). Solution structure and dynamics of the Rous sarcoma virus capsid protein and comparison with capsid proteins of other retroviruses 1 1Edited by P. E. Wright. Journal of Molecular Biology, 296(2), 633–649.

192 Campos-Sánchez, R., Cremona, M. A., Pini, A., Chiaromonte, F., & Makova, K. D. (2016). Inte- gration and Fixation Preferences of Human and Mouse Endogenous Retroviruses Uncovered with Functional Data Analysis. PLoS Computational Biology, 12(6), 1–41. Carter, A. M., Gothlin, J., & Olin, T. (1971). An angiographic study of the structure and function of the uterine and maternal placental vasculature in the rabbit. Journal of Reproduction and Fertility, 25(2), 201–210. Chan, D. C., Fass, D., Berger, J. M., & Kim, P. S. (1997). Core structure of gp41 from the HIV envelope glycoprotein. Cell, 89(2), 263–273. Chang, C.-W., Chuang, H.-C., Yu, C., Yao, T.-P., & Chen, H. (2005). Stimulation of GCMa Transcriptional Activity by Cyclic AMP/Protein Kinase A Signaling Is Attributed to CBP- Mediated Acetylation of GCMa. Molecular and Cellular Biology, 25(19), 8401–8414. Chang, Y. F., Wang, S. M., Huang, K. J., & Wang, C. T. (2007). Mutations in Capsid Major Homology Region Affect Assembly and Membrane Affinity of HIV-1 Gag. Journal of Molecular Biology, 370(3), 585–597. Charneau, P., Alizon, M., & Clavel, F. (1992). A second origin of DNA plus-strand synthesis is required for optimal human A Second Origin of DNA Plus-Strand Synthesis Is Required for Optimal Human Immunodeficiency Virus Replication. Journal of Virology, 66(5), 2814– 2820. Charneau, P., & Clavel, F. (1991). A single-stranded gap in human immunodeficiency virus unintegrated linear DNA defined by a central copy of the polypurine tract. Journal of Virology, 65(5), 2415–21. Chavatte-Palmer, P., & Tarrade, A. (2016). Placentation in different mammalian species. Annales d’endocrinologie, 77(2), 67–74. Chen, J., Grunwald, D., Sardo, L., Galli, A., Plisov, S., Nikolaitchik, O. A., Chen, D., Lockett, S., Larson, D. R., Pathak, V. K., & Hu, W.-S. (2014). Cytoplasmic HIV-1 RNA is mainly transported by diffusion in the presence or absence of Gag protein. Proceedings of the National Academy of Sciences, 111(48), E5205–E5213. Chen, J., Lee, K. H., Steinhauer, D. A., Stevens, D. J., Skehel, J. J., & Wiley, D. C. (1998). Struc- ture of the hemagglutinin precursor cleavage site, a determinant of influenza pathogenicity and the origin of the labile conformation. Cell, 95(3), 409–417. Cheynet, V., Oriol, G., & Mallet, F. (2006). Identification of the hASCT2-binding domain of the Env ERVWE1/syncytin-1 fusogenic glycoprotein. Retrovirology, 3, 41. Choe, H., Farzan, M., Sun, Y., Sullivan, N., Rollins, B., Ponath, P. D., Wu, L., Mackay, C. R., LaRosa, G., Newman, W., Gerard, N., Gerard, C., & Sodroski, J. (1996). The β-chemokine receptors CCR3 and CCR5 facilitate infection by primary HIV-1 isolates. Cell, 85(7), 1135– 1148. Choi, G., Park, S., Choi, B., Hong, S., Lee, J., Hunter, E., & Rhee, S. S. (1999). Identification of a cytoplasmic targeting/retention signal in a retroviral Gag polyprotein. Journal of Virology, 73(7), 5431–7. Chojnacki, J., Staudt, T., Glass, B., Bingen, P., Engelhardt, J., Anders, M., Schneider, J., Müller, B., Hell, S. W., & Kräusslich, H. G. (2012). Maturation-dependent HIV-1 surface protein redistribution revealed by fluorescence nanoscopy. Science, 338(6106), 524–528. Christodoulopoulos, I., & Cannon, P. M. (2001). Sequences in the Cytoplasmic Tail of the Gibbon Ape Leukemia Virus Envelope Protein That Prevent Its Incorporation into Lentivirus Vectors Sequences in the Cytoplasmic Tail of the Gibbon Ape Leukemia Virus Envelope Protein That Prevent Its Incorporation i. Journal of Virology, 75(9), 4129–4138. Chuong, E. B., Elde, N. C., & Feschotte, C. (2016). Regulatory evolution of innate immunity through co-option of endogenous retroviruses. Science, 351(6277), 1083–1087.

193 Chuong, E. B., Elde, N. C., & Feschotte, C. (2017). Regulatory activities of transposable elements: from conflicts to benefits. Nature Reviews Genetics, 18(2), 71–86. Chuong, E. B., Rumi, M. A., Soares, M. J., & Baker, J. C. (2013). Endogenous retroviruses function as species-specific enhancer elements in the placenta. Nature Genetics, 45(3), 325– 329. Cianciolo, G., Copeland, T., Oroszlan, S., & Snyderman, R. (1985). Inhibition of lymphocyte proliferation by a synthetic peptide homologous to retroviral envelope proteins. Science, 230(4724), 453–455. Cianciolo, G. J., Bogerd, H., & Snyderman, R. (1988). Human retrovirus-related synthetic peptides inhibit T lymphocyte proliferation. Immunology letters, 19(1), 7–13. Coan, P. M., Ferguson-Smith, A. C., & Burton, G. J. (2004). Developmental Dynamics of the Definitive Mouse Placenta Assessed by Stereology. Biology of Reproduction, 70(6), 1806–1813. Coan, P. M., Ferguson-Smith, A. C., & Burton, G. J. (2005). Ultrastructural changes in the interhaemal membrane and junctional zone of the murine choriollantoic placenta across gestation. Journal of Anatomy, 207(6), 783–796. Coffin, J. M., Hughes, S. H., & Varmus, H. E. (1997). Retroviruses. Cold Spring Harbor: Cold Spring Harbor Laboratory Press. Cohen, M., Powers, M., O’Connell, C., & Kato, N. (1985). The nucleotide sequence of the env gene from the human provirus ERV3 and isolation and characterization of an ERV3-specific cDNA. Virology, 147(2), 449–458. Cordonnier, A., Casella, J. F., & Heidmann, T. (1995). Isolation of novel human endogenous retrovirus-like elements with foamy virus-related pol sequence. Journal of Virology, 69(9), 5890–7. Cornelis, G., Funk, M., Vernochet, C., Leal, F., Tarazona, O. A., Meurice, G., Heidmann, O., Dupressoir, A., Miralles, A., Ramirez-Pinilla, M. P., & Heidmann, T. (2017). An endogenous retroviral envelope syncytin and its cognate receptor identified in the viviparous placental Mabuya lizard. Proceedings of the National Academy of Sciences, 114(51), E10991–E11000. Cornelis, G., Heidmann, O., Bernard-Stoecklin, S., Reynaud, K., Véron, G., Mulot, B., Dupres- soir, A., & Heidmann, T. (2012). Ancestral capture of syncytin-Car1, a fusogenic endogenous retroviral envelope gene involved in placentation and conserved in Carnivora. Proceedings of the National Academy of Sciences, 109(7), E432–41. Cornelis, G., Heidmann, O., Degrelle, S. a., Vernochet, C., Lavialle, C., Letzelter, C., Bernard- Stoecklin, S., Hassanin, A., Mulot, B., Guillomot, M., Hue, I., Heidmann, T., & Dupressoir, A. (2013). Captured retroviral envelope syncytin gene associated with the unique placental structure of higher ruminants. Proceedings of the National Academy of Sciences, 110(9), E828–37. Cornelis, G., Vernochet, C., Carradec, Q., Souquere, S., Mulot, B., Catzeflis, F., Nilsson, M. a., Menzies, B. R., Renfree, M. B., Pierron, G., Zeller, U., Heidmann, O., Dupressoir, A., & Heidmann, T. (2015). Retroviral envelope gene captures and syncytin exaptation for placentation in marsupials. Proceedings of the National Academy of Sciences. Cornelis, G., Vernochet, C., Malicorne, S., Souquere, S., Tzika, A. C., Goodman, S. M., Catze- flis, F., Robinson, T. J., Milinkovitch, M. C., Pierron, G., Heidmann, O., Dupressoir, A., & Heidmann, T. (2014). Retroviral envelope syncytin capture in an ancestrally diverged mam- malian clade for placentation in the primitive Afrotherian tenrecs. Proceedings of the National Academy of Sciences, 111(41), E4332–41. Craigie, R., Fujiwara, T., & Bushman, F. (1990). The IN protein of Moloney murine leukemia

194 virus processes the viral DNA ends and accomplishes their integration in vitro. Cell, 62(4), 829–837. Crawford, L. V., & Crawford, E. M. (1961). The Properties of Rous Sarcoma Virus Purified by Density Gradient Centrifugation. Virology, 13, 227–232. Cullen, B. R. (2003). Nuclear mRNA export: Insights from virology. Trends in Biochemical Sciences, 28(8), 419–424. Dalgleish, A. G., Beverley, P. C., Clapham, P. R., Crawford, D. H., Greaves, M. F., & Weiss, R. A. (1984). The CD4 (T4) antigen is an essential component of the receptor for the AIDS retrovirus. Nature, 312(5996), 763–767. Dantzer, V. (1985). Electron microscopy of the initial stages of placentation in the pig. Anatomy and Embryology, 172(3), 281–293. Dantzer, V., Leiser, R., Kaufmann, P., & Luckhardt, M. (1988). Comparative Morphological Aspects of Placental Vascularization. In Placental Vascularization and Blood Flow, (pp. 235–260). Boston, MA: Springer US. Datta, S. A., Curtis, J. E., Ratcliff, W., Clark, P. K., Crist, R. M., Lebowitz, J., Krueger, S., & Rein, A. (2007). Conformation of the HIV-1 Gag Protein in Solution. Journal of Molecular Biology, 365(3), 812–824. de Parseval, N., & Heidmann, T. (1998). Physiological knockout of the envelope gene of the single-copy ERV-3 human endogenous retrovirus in a fraction of the Caucasian population. Journal of Virology, 72(4), 3442–5. de Parseval, N., Lazar, V., Casella, J.-F., Benit, L., & Heidmann, T. (2003). Survey of human genes of retroviral origin: identification and transcriptome of the genes with coding capacity for complete envelope proteins. Journal of Virology, 77(19), 10414–10422. de Queiroz, K., & Gauthier, J. (1992). Phylogenetic Taxonomy. Annual Review of Ecology and Systematics, 23(1), 449–480. Dedera, D. A., Gu, R., & Ratner, L. (1992). Role of asparagine-linked glycosylation in human immunodeficiency virus type 1 transmembrane envelope function. Virology, 187(1), 383–387. Deng, H., Liu, R., Ellmeier, W., Choe, S., Unutmaz, D., Burkhart, M., Marzio, P. D., Marmon, S., Sutton, R. E., Hill, C. M., Davis, C. B., Peiper, S. C., Schall, T. J., Littman, D. R., & Landau, N. R. (1996). Identification of a major co-receptor for primary isolates of HIV-1. Nature, 381(6584), 661–666. Denner, J. (2014). The transmembrane proteins contribute to immunodeficiencies induced by HIV-1 and other retroviruses. Aids, 28(8), 1081–1090. Denner, J. (2016). Expression and function of endogenous retroviruses in the placenta. APMIS, 124(1-2), 31–43. Dewannieux, M., Dupressoir, A., Harper, F., Pierron, G., & Heidmann, T. (2004). Identification of autonomous IAP LTR retrotransposons mobile in mammalian cells. Nature Genetics, 36(5), 534–9. Dewannieux, M., Harper, F., Richaud, A., Letzelter, C., Ribet, D., Pierron, G., & Heidmann, T. (2006). Identification of an infectious progenitor for the multiple-copy HERV-K human endogenous retroelements. Genome research, 16(12), 1548–56. Dewannieux, M., & Heidmann, T. (2013). Endogenous retroviruses: acquisition, amplification and taming of genome invaders. Current Opinion in Virology, 3(6), 646–656. Diaz-Griffero, F., Hoschander, S. A., & Brojatsch, J. (2002). Endocytosis is a critical step in entry of subgroup B avian leukosis viruses. Journal of Virology, 76(24), 12866–12876. Donaldson, W.L., Zhang, C. H., Oriol, J. G., & Antczak, D. F. (1990). Invasive equine trophoblast expresses conventional class I Major Histocompatibility Complex antigens. Development, 110, 63–71.

195 Doranz, B. J., Rucker, J., Yi, Y., Smyth, R. J., Samson, M., Peiper, S. C., Parmentier, M., Collman, R. G., & Doms, R. W. (1996). A dual-tropic primary HIV-1 isolate that uses fusin and the β-chemokine receptors CKR-5, CKR-3, and CKR-2b as fusion cofactors. Cell, 85(7), 1149–1158. dos Reis, M., Inoue, J., Hasegawa, M., Asher, R. J., Donoghue, P. C. J., & Yang, Z. (2012). Phylogenomic datasets provide both precision and accuracy in estimating the timescale of placental mammal phylogeny. Proceedings of the Royal Society B: Biological Sciences, 279(1742), 3491–3500. dos Reis, M., Thawornwattana, Y., Angelis, K., Telford, M. J., Donoghue, P. C., & Yang, Z. (2015). Uncertainty in the Timing of Origin of Animals and the Limits of Precision in Molecular Timescales. Current Biology, 25(22), 2939–2950. Douglas, J. L., Viswanathan, K., McCarroll, M. N., Gustin, J. K., Fruh, K., & Moses, A. V. (2009). Vpu Directs the Degradation of the Human Immunodeficiency Virus Restriction Factor BST-2/Tetherin via a TrCP-Dependent Mechanism. Journal of Virology, 83(16), 7931–7947. Dragic, T., Litwin, V., Allaway, G. P., Martin, S. R., Huang, Y., Nagashima, K. A., Cayanan, C., Maddon, P. J., Koup, R. A., Moore, J. P., & Paxton, W. A. (1996). HIV-1 entry into CD4+ cells is mediated by the chemokine receptor CC-CKR-5. Nature, 381(6584), 667–673. Dufaure, J., & Hubert, J. (1961). Table de développement du lézard vivipare: Lacerta (Zootoca) vivipara Jacquin. Archives D’Anatomie Microscopique Et De Morphologie Experimentale, 50(3), 309. Dunlap, K. A., Palmarini, M., Adelson, D. L., & Spencer, T. E. (2005). Sheep Endogenous Betaretroviruses (enJSRVs) and the Hyaluronidase 2 (HYAL2) Receptor in the Ovine Uterus and Conceptus. Biology of Reproduction, 73(2), 271–279. Dunlap, K. A., Palmarini, M., Varela, M., Burghardt, R. C., Hayashi, K., Farmer, J. L., & Spencer, T. E. (2006). Endogenous retroviruses regulate periimplantation placental growth and differentiation. Proceedings of the National Academy of Sciences, 103(39), 14390–5. Dupressoir, A., & Heidmann, T. (2011). Les syncytines des protéines d’enveloppe rétrovirales capturées au profit du développement placentaire. Medecine/Sciences, 27(2), 163–169. Dupressoir, A., Marceau, G., Vernochet, C., Benit, L., Kanellopoulos, C., Sapin, V.,& Heidmann, T. (2005). Syncytin-A and syncytin-B, two fusogenic placenta-specific murine envelope genes of retroviral origin conserved in Muridae. Proceedings of the National Academy of Sciences, 102(3), 725–730. Dupressoir, A., Vernochet, C., Bawa, O., Harper, F., Pierron, G., Opolon, P., & Heidmann, T. (2009). Syncytin-A knockout mice demonstrate the critical role in placentation of a fusogenic, endogenous retrovirus-derived, envelope gene. Proceedings of the National Academy of Sciences, 106(29), 12127–32. Dupressoir, A., Vernochet, C., Harper, F., Guégan, J., Dessen, P., Pierron, G., & Heidmann, T. (2011). A pair of co-opted retroviral envelope syncytin genes is required for formation of the two-layered murine placental syncytiotrophoblast. Proceedings of the National Academy of Sciences, 108(46), E1164–73. Dupressoir, S., Lavialle, C., & Heidmann, T. (2012). From ancestral infectious retroviruses to bona fide cellular genes: role of the captured syncytins in placentation. Placenta, 33(9), 663–71. Duval, M. (1892). Le placenta des rongeurs. Paris: Felix Alcan. Dybowski, B. (1873). Ueber Comephorus baicalensis Pall. Verhandlungen der Zoologisch- Botanischen Gesellschaft in Wien, 23, 474–484.

196 Dyda, F., Hickman, A., Jenkins, T., Engelman, A., Craigie, R., & Davies, D. (1994). Crystal structure of the catalytic domain of HIV-1 integrase: similarity to other polynucleotidyl transferases. Science, 266(5193), 1981–1986. Edwards, D., Fenizia, C., Gold, H., de Castro-Amarante, M. F., Buchmann, C., Pise-Masison, C. A., & Franchini, G. (2011). Orf-I and Orf-II-encoded proteins in HTLV-1 infection and persistence. Viruses, 3(6), 861–885. Emera, D., Casola, C., Lynch, V. J., Wildman, D. E., Agnew, D., & Wagner, G. P. (2012). Conver- gent evolution of endometrial prolactin expression in primates, mice, and elephants through the independent recruitment of transposable elements. Molecular Biology and Evolution, 29(1), 239–247. Enders, A., & Carter, A. (2004). What Can Comparative Studies of Placental Structure Tell Us?—A Review. Placenta, 25(SUPPL. A), S3–S9. Enders, A. C. (1965). A comparative study of the fine structure of the trophoblast in several hemochorial placentas. American Journal of Anatomy, 116(1), 29–67. Enders, A. C. (1989). Trophoblast differentiation during the transition from trophoblastic plate to lacunar stage of implantation in the rhesus monkey and human. American Journal of Anatomy, 186(1), 85–98. Enders, A. C., Blankenship, T. N., Conley, A. J., & Jones, C. J. P. (2006). Structure of the midterm placenta of the spotted hyena, Crocuta crocuta, with emphasis on the diverse hemophagous regions. Cells Tissues Organs, 183(3), 141–155. Enders, A. C., & Schlafke, S. (1971). Penetration of the uterine epithelium during implantation in the rabbit. American Journal of Anatomy, 132(2), 219–239. Engelman, A., & Craigie, R. (1992). Identification of conserved amino acid residues critical for human immunodeficiency virus type 1 integrase function in vitro. Journal of Virology, 66(11), 6361–9. Engelman, A., Mizuuchi, K., & Craigie, R. (1991). HIV-1 DNA integration: Mechanism of viral DNA cleavage and DNA strand transfer. Cell, 67(6), 1211–1221. Engelman, A. N., & Cherepanov, P. (2017). Retroviral intasomes arising. Current Opinion in Structural Biology, 47, 23–29. Enssle, J., Moebes, A., Heinkelein, M., Panhuysen, M., Mauer, B., Schweizer, M., Neumann- Haefelin, D., & Rethwilm, A. (1999). An active foamy virus integrase is required for virus replication. Journal of General Virology, 80(6), 1445–1452. Esnault, C., Cornelis, G., Heidmann, O., & Heidmann, T. (2013). Differential Evolutionary Fate of an Ancestral Primate Endogenous Retrovirus Envelope Gene, the EnvV Syncytin, Captured for a Function in Placentation. PLoS Genetics, 9(3), e1003400. Esnault, C., Priet, S., Ribet, D., Vernochet, C., Bruls, T., Lavialle, C., Weissenbach, J., & Heidmann, T. (2008). A placenta-specific receptor for the fusogenic, endogenous retrovirus- derived, human syncytin-2. Proceedings of the National Academy of Sciences, 105(45), 17532–17537. Ethiraj, P., Link, J. R., Sinkway, J. M., Brown, G. D., Parler, W. A., & Reddy, S. V. (2018). Microgravity modulation of syncytin-A expression enhance osteoclast formation. Journal of Cellular Biochemistry. Farkašová, H., Hron, T., Pačes, J., Hulva, P., Benda, P., Gifford, R. J., & Elleder, D. (2017). Discovery of an endogenous Deltaretrovirus in the genome of long-fingered bats (Chiroptera: Miniopteridae). Proceedings of the National Academy of Sciences, 114(12), 3145–3150. Fass, D., Harrison, S. C., & Kim, P. S. (1996). Retrovirus envelope domain at 1.7 angstrom resolution. Nature Structural biology, 3(5), 465–9.

197 Fassati, A., & Goff, S. P. (1999). Characterization of intracellular reverse transcription complexes of Moloney murine leukemia virus. Journal of Virology, 73(11), 8919–8925. Fassati, A., & Goff, S. P. (2001). Characterization of Intracellular Reverse Transcription Com- plexes of Human Immunodeficiency Virus Type 1. Journal of Virology, 75(8), 3626–3635. Faust, T. B., Binning, J. M., Gross, J. D., & Frankel, A. D. (2017). Making Sense of Multifunc- tional Proteins: Human Immunodeficiency Virus Type 1 Accessory and Regulatory Proteins and Connections to Transcription. Annual Review of Virology, 4(1), 241–260. Fédry, J., Liu, Y., Péhau-Arnaudet, G., Pei, J., Li, W., Tortorici, M. A., Traincard, F., Meola, A., Bricogne, G., Grishin, N. V., Snell, W. J., Rey, F. A., & Krey, T. (2017). The Ancient Gamete Fusogen HAP2 Is a Eukaryotic Class II Fusion Protein. Cell, 168(5), 904–915.e10. Feinberg, M. B., Baltimore, D., & Frankel, A. D. (1991). The role of Tat in the human immunodeficiency virus life cycle indicates a primary effect on transcriptional elongation. Proceedings of the National Academy of Sciences, 88(9), 4045–9. Felkner, R. H., & Roth, M. J. (1992). Mutational analysis of the N-linked glycosylation sites of the SU envelope protein of Moloney murine leukemia virus. Journal of Virology, 66(7), 4258–4264. Feng, Y.,Broder, C. C., Kennedy, P. E., & Berger, E. A. (1996). HIV-1 Entry Cofactor: Functional cDNA Cloning of a Seven-Transmembrane, G Protein-Coupled Receptor. Science, 272(5263), 872–877. Ferretti, C., Bruni, L., Dangles-Marie, V., Pecking, A. P., & Bellet, D. (2007). Molecular circuits shared by placental and cancer cells, and their implications in the proliferative, invasive and migratory capacities of trophoblasts. Human Reproduction Update, 13(2), 121–141. Feschotte, C., & Gilbert, C. (2012). Endogenous viruses: Insights into viral evolution and impact on host biology. Nature Reviews Genetics, 13(4), 283–296. Fischer, N., Heinkelein, M., Lindemann, D., Enssle, J., Baum, C., Werder, E., Zentgraf, H., Müller, J. G., & Rethwilm, a. (1998). Foamy virus particle formation. Journal of Virology, 72(2), 1610–1615. Flemming, A. F., & Blackburn, D. G. (2003). Evolution of placental specializations in viviparous African and South American lizards. Journal of experimental zoology. Part A, Comparative experimental biology, 299(1), 33–47. Flemming, A. F., & Branch, W. R. (2001). Extraordinary case of matrotrophy in the african skink Eumecia anchietae. Journal of Morphology, 247(3), 264–287. Fowden, A. L. (2010). Comparative aspects of fetal carbohydrate metabolism. Equine Veterinary Journal, 29(S24), 19–25. Freed, E. O. (2015). HIV-1 assembly, release and maturation. Nature Reviews Microbiology, 13(8), 484–496. Freemark, M. (2006). Regulation of maternal metabolism by pituitary and placental hormones: Roles in fetal development and metabolic programming. Hormone Research, 65(SUPPL. 3), 41–49. Frendo, J.-l., Olivier, D., Cheynet, V., Blond, J.-l., Bouton, O., & Vidaud, M. (2003). Direct Involvement of HERV-W Env Glycoprotein in Human Trophoblast Cell Fusion and Differen- tiation. Molecular and Cellular Biologyology, 23(10), 3566–3574. Frese, S., Ruebner, M., Suhr, F., Konou, T. M., Tappe, K. a., Toigo, M., Jung, H. H., Henke, C., Steigleder, R., Strissel, P. L., Huebner, H., Beckmann, M. W., van der Keylen, P., Schoser, B., Schiffer, T., Frese, L., Bloch, W., & Strick, R. (2015). Long-Term Endurance Exercise in Humans Stimulates Cell Fusion of Myoblasts along with Fusogenic Endogenous Retroviral Genes In Vivo. PloS ONE, 10(7), e0132099.

198 Friess, A. E., Sinowatz, F., Skolek-Winnisch, R., & Trautner, W. (1980). The placenta of the pig. Anatomy and Embryology, 158(2), 179–191. Gallaher, W. R. (1996). Similar Structural Models of the Transmembrane Proteins of Ebola and Avian Sarcoma Viruses. Cell, 85(4), 477–478. Ganser, B. K., Li, S., Klishko, V. Y., Finch, J. T., & Sundquist, W. I. (1999). Assembly and analysis of conical models for the HIV-1 core. Science, 283(5398), 80–83. Gaudray, G., Gachon, F., Basbous, J., Biard-Piechaczyk, M., Devaux, C., & Mesnard, J.-M. (2002). The complementary strand of the human T-cell leukemia virus type 1 RNA genome encodes a bZIP transcription factor that down-regulates viral transcription. Journal of Virol- ogy, 76(24), 12813–22. Gaunt, M., & Ockleford, C. D. (1986). Microinjection of human placenta: II: Biological application. Placenta, 7(4), 325–331. Gemmell, P., Hein, J., & Katzourakis, A. (2016). Phylogenetic Analysis Reveals That ERVs "Die Young" but HERV-H Is Unusually Conserved. PLoS Computational Biology, 12(6), e1004964. Gerbaud, P., & Pidoux, G. (2015). Review: An overview of molecular events occurring in human trophoblast fusion. Placenta, 36, S35–S42. Ghez, D., Lepelletier, Y., Lambert, S., Fourneau, J.-M., Blot, V., Janvier, S., Arnulf, B., van Endert, P. M., Heveker, N., Pique, C., & Hermine, O. (2006). Neuropilin-1 Is Involved in Human T-Cell Lymphotropic Virus Type 1 Entry. Journal of Virology, 80(14), 6844–6854. Giam, C.-Z., & Semmes, O. (2016). HTLV-1 Infection and Adult T-Cell Leukemia/Lymphoma—A Tale of Two Proteins: Tax and HBZ. Viruses, 8(6), 161. Gilmore, R. G. (1993). Reproductive biology of lamnoid sharks. Environmental Biology of Fishes, 38(1-3), 95–114. Gimenez, J., Montgiraud, C., Oriol, G., Pichon, J. P., Ruel, K., Tsatsaris, V., Gerbaud, P., Frendo, J. L., Evain-Brion, D., & Mallet, F. (2009). Comparative methylation of ERVWE1/syncytin-1 and other human endogenous retrovirus LTRs in placenta tissues. DNA Research, 16(4), 195–211. Goh, W. C., Rogel, M. E., Kinsey, C. M., Michael, S. F., Fultz, P. N., Nowak, M. A., Hahn, B. H., & Emerman, M. (1998). HIV-1 Vpr increases viral expression by manipulation of the cell cycle: A mechanism for selection of Vpr in vivo. Nature Medicine, 4(1), 65–71. Gottlinger, H. G., Dorfman, T., Sodroski, J. G., & Haseltine, W. A. (1991). Effect of mutations affecting the p6 gag protein on human immunodeficiency virus particle release. Proceedings of the National Academy of Sciences, 88(8), 3195–3199. Gould, S. J., & Vrba, E. S. (1982). Exaptation—a Missing Term in the Science of Form. Paleobiology, 8(01), 4–15. Granowitz, C., Colicelli, J., & Goff, S. P. (1991). Analysis of Mutations in the Envelope Gene of Moloney Separation of Infectivity from Superinfection Murine Leukemia Virus : Resistance. Virology, 554, 545–554. Graves, B. J., Eisenman, R. N., & McKnight, S. L. (1985). Delineation of transcriptional control signals within the Moloney murine sarcoma virus long terminal repeat. Molecular and cellular biology, 5(8), 1948–58. Grawenhoff, J., & Engelman, A. N. (2017). Retroviral integrase protein and intasome nucleo- protein complex structures. World journal of biological chemistry, 8(1), 32–44. Greenwood, A. D., Ishida, Y., O’Brien, S. P., Roca, A. L., & Eiden, M. V. (2017). Trans- mission, Evolution, and Endogenization: Lessons Learned from Recent Retroviral Invasions. Microbiology and Molecular Biology Reviews, 82(1), e00044–17.

199 Griffith, O. W., & Wagner, G. P. (2017). The placenta as a model for understanding the origin and evolution of vertebrate organs. Nature Ecology & Evolution, 1(4), 72. Grosser, O. (1909). Vergleichende Anatomie und Entwicklungsgeschichte der Eihäute under Placenta: mit besonderer Berücksichtigung des Menschen. Lehrbuch für Studierende und Aerzte. Wien: Braumüller. Grueso, E., Rocha, M., & Puerta, M. (2001). Plasma and cerebrospinal fluid leptin levels are maintained despite enhanced food intake in progesterone-treated rats. European Journal of Endocrinology, 144(6), 659–665. Guindon, S., Dufayard, J. F., Lefort, V., Anisimova, M., Hordijk, W., & Gascuel, O. (2010). New algorithms and methods to estimate maximum-likelihood phylogenies: Assessing the performance of PhyML 3.0. Systematic Biology, 59(3), 307–321. Gyapay, G., Schmitt, K., Fizames, C., Jones, H., Vega-Czarny, N., Spillett, D., Muselet, D., Prud’homme, J. F., Dib, C., Auffray, C., Morissette, J., Weissenbach, J., & Goodfellow, P. N. (1996). A radiation hybrid map of the human genome. Human molecular genetics, 5(3), 339–46. Hafez, S. (2017). Chapter One - Comparative Placental Anatomy. In W. Huckle (Ed.) Molecular Biology of Placental Development and Disease, vol. 145, (pp. 1–28). Academic Press. Haines, A. N., Flajnik, M. F., & Wourms, J. P. (2006). Histology and Immunology of the Placenta in the Atlantic Sharpnose Shark, Rhizoprionodon terraenovae. Placenta, 27(11-12), 1114–1123. Hanger, J. J., Bromham, L. D., McKee, J. J., O’Brien, T. M., & Robinson, W. F. (2000). The Nucleotide Sequence of Koala (Phascolarctos cinereus) Retrovirus: a Novel Type C Endogenous Virus Related to Gibbon Ape Leukemia Virus. Journal of Virology, 74(9), 4264–4272. Hansen, J., Billich, S., Schulze, T., Sukrow, S., & Moelling, K. (1988). Partial purification and substrate analysis of bacterially expressed HIV protease by means of monoclonal antibody. The EMBO Journal, 7(6), 1785–1791. Hare, S., Gupta, S. S., Valkov, E., Engelman, A., & Cherepanov, P. (2010). Retroviral intasome assembly and inhibition of DNA strand transfer. Nature, 464(7286), 232–236. Harris, R. S., Bishop, K. N., Sheehy, A. M., Craig, H. M., Petersen-Mahrt, S. K., Watt, I. N., Neuberger, M. S., & Malim, M. H. (2003). DNA deamination mediates innate immunity to retroviral infection. Cell, 113(6), 803–809. Harrison, S. C. (2015). Viral membrane fusion. Virology, 479-480, 498–507. Hayward, A., Cornwallis, C. K., & Jern, P. (2015). Pan-vertebrate comparative genomics unmasks retrovirus macroevolution. Proceedings of the National Academy of Sciences, 112(2), 464–469. Hayward, M. D., Potgens, A. J. G., Drewlo, S., Kaufmann, P., & Rasko, J. E. J. (2007). Distri- bution of human endogenous retrovirus type W receptor in normal human villous placenta. Pathology, 39(4), 406–412. Heidmann, O., Béguin, A., Paternina, J., Berthier, R., Deloger, M., Bawa, O., & Heidmann, T. (2017). HEMO, an ancestral endogenous retroviral envelope protein shed in the blood of pregnant women and expressed in pluripotent stem cells and tumors. Proceedings of the National Academy of Sciences, 114(32), E6642–E6651. Heidmann, O., Vernochet, C., Dupressoir, A., & Heidmann, T. (2009). Identification of an endogenous retroviral envelope gene with fusogenic activity and placenta-specific expression in the rabbit: a new "syncytin" in a third order of mammals. Retrovirology, 6, 107. Henzy, J. E., & Coffin, J. M. (2013). Betaretroviral Envelope Subunits Are Noncovalently Associated and Restricted to the Mammalian Class. Journal of Virology, 87(4), 1937–1946.

200 Henzy, J. E., Gifford, R. J., Kenaley, C. P., & Johnson, W. E. (2016). An Intact Retroviral Gene Conserved in Spiny-Rayed Fishes for over 100 My. Molecular Biology and Evolution, 34(3), 634–639. Henzy, J. E., & Johnson, W. E. (2013). Pushing the endogenous envelope. Philosophical Transactions of the Royal Society B: Biological Sciences, 368(1626), 20120506–20120506. Hernández-Díaz, N., Torres, R., & Ramírez-Pinilla, M. P. (2017). Proteomic Profile of Mabuya sp. (Squamata: Scincidae) Ovary and Placenta During Gestation. Journal of Experimental Zoology Part B: Molecular and Developmental Evolution, 328(4), 371–389. Hernandez-Verdun, D. (1974). Morphogenesis of the syncytium in the mouse placenta - Ultra- structural study. Cell and Tissue Research, 148(3), 381–396. Hernandez-Verdun, D., & Legrand, C. (1975). In vitro study of chorionic and ectoplacental tro- phoblast differentiation in the mouse. Journal of embryology and experimental morphology, 34, 633–644. Heron, A. C. (1972). Population ecology of a colonizing species: The pelagic tunicate Thalia democratica - I. Individual growth rate and generation time. Oecologia, 10(4), 269–293. Hiby, S. E., Apps, R., Sharkey, A. M., Farrell, L. E., Gardner, L., Mulder, A., Claas, F. H., Walker, J. J., Redman, C. C., Morgan, L., Tower, C., Regan, L., Moore, G. E., Carrington, M., & Moffett, A. (2010). Maternal activating KIRs protect against human reproductive failure mediated by fetal HLA-C2. Journal of Clinical Investigation, 120(11), 4102–4110. Hill, C. P., Worthylake, D., Bancroft, D. P., Christensen, A. M., & Sundquist, W. I. (1996). Crystal structures of the trimeric human immunodeficiency virus type 1 matrix protein: implications for membrane association and assembly. Proceedings of the National Academy of Sciences, 93(7), 3099–3104. Hizi, A., Gazit, A., Guthmann, D., & Yaniv, A. (1982). DNA-processing activities associated with the purified α, β2, and αβ molecular forms of avian sarcoma virus RNA-dependent DNA polymerase. Journal of Virology, 41(3), 974–981. Hizi, A., Leis, J. P., & Joklik, W. K. (1977). RNA-dependent DNA polymerase of avian sarcoma virus B77. II. Comparison of the catalytic properties of the alpha, beta2, and alphabeta enzyme forms. The Journal of biological chemistry, 252(7), 2290–5. Holder, B. S., Tower, C. L., Forbes, K., Mulla, M. J., Aplin, J. D., & Abrahams, V. M. (2012). Immune cell activation by trophoblast-derived microvesicles is mediated by syncytin 1. Im- munology, 136(2), 184–191. Hosaka, M., Nagahama, M., Kim, W. S., Watanabe, T., Hatsuzawa, K., Ikemizu, J., Murakami, K., & Nakayama, K. (1991). Arg-X-Lys/Arg-Arg motif as a signal for precursor cleavage catalyzed by furin within the constitutive secretory pathway. Journal of Biological Chemistry, 266(19), 12127–12130. Hrecka, K., Gierszewska, M., Srivastava, S., Kozaczkiewicz, L., Swanson, S. K., Florens, L., Washburn, M. P., & Skowronski, J. (2007). Lentiviral Vpr usurps Cul4-DDB1[VprBP] E3 ubiquitin ligase to modulate cell cycle. Proceedings of the National Academy of Sciences, 104(28), 11778–11783. Hu, D., & Cross, J. C. (2010). Development and function of trophoblast giant cells in the rodent placenta. International Journal of Developmental Biology, 54(2-3), 341–354. Hu, W. S., & Temin, H. M. (1990). Genetic consequences of packaging two RNA genomes in one retroviral particle: pseudodiploidy and high rate of genetic recombination. Proceedings of the National Academy of Sciences, 87(4), 1556–60. Huang, Q., Chen, H., Li, J., Oliver, M., Ma, X., Byck, D., Gao, Y., & Jiang, S. W. (2014). Epigenetic and non-epigenetic regulation of syncytin-1 expression in human placenta and cancer tissues. Cellular Signalling, 26(3), 648–656.

201 Hughes, R. L. (1993). Monotreme development with particular reference to the extraembryonic membranes. Journal of Experimental Zoology, 266(6), 480–494. Hughes, S. H. (2015). Reverse Transcription of Retroviruses and LTR Retrotransposons. Mi- crobiology Spectrum, 3(2), 1–25. Hull, S., & Fan, H. (2006). Mutational analysis of the cytoplasmic tail of jaagsiekte sheep retrovirus envelope protein. Journal of Virology, 80(16), 8069–80. Hunt, D. M., Fitzgibbon, J., Slobodyanyuk, S. J., Bowmaker, J. K., & Dulai, K. S. (1997). Molecular Evolution of the Cottoid Fish Endemic to Lake Baikal Deduced from Nuclear DNA Evidence. Molecular Phylogenetics and Evolution, 8(3), 415–422. Hurst, T. P., & Magiorkinis, G. (2017). Epigenetic control of human endogenous retrovirus expression: Focus on regulation of long-terminal repeats (LTRs). Viruses, 9(6), 1–13. Hussein, I. T. M., Ni, N., Galli, A., Chen, J., Moore, M. D., & Hu, W.-S. (2010). Delineation of the preferences and requirements of the human immunodeficiency virus type 1 dimerization initiation signal by using an in vivo cell-based selection approach. Journal of Virology, 84(13), 6866–75. Huxley, T. H. (1851). Observations upon the Anatomy and Physiology of Salpa and Pyrosoma. Philosophical Transactions of the Royal Society, 2(January), 567–593. ICTV (1975). Minutes of the third meeting of ICTV. Avaibable at https://talk.ictvonline.org/ICTV/proposals/Ratification_1975.pdf. Ikeda, H., & Odaka, T. (1983). Cellular expression of murine leukemia virus gp70-related antigen on thymocytes of uninfected mice correlates with Fv-4 gene-controlled resistance to friend leukemia virus infection. Virology, 128(1), 127–139. Imbeault, M., Helleboid, P. Y., & Trono, D. (2017). KRAB zinc-finger proteins contribute to the evolution of gene regulatory networks. Nature, 543(7646), 550–554. Irigoyen, N., Dinan, A. M., Brierley, I., & Firth, A. E. (2018). Ribosome profiling of the retrovirus murine leukemia virus. Retrovirology, 15(1), 1–8. Ito, J., Watanabe, S., Hiratsuka, T., Kuse, K., Odahara, Y., Ochi, H., Kawamura, M., & Nishigaki, K. (2013). Refrex-1, a Soluble Restriction Factor against Feline Endogenous and Exogenous Retroviruses. Journal of Virology, 87(22), 12029–12040. Iwabu, Y., Fujita, H., Kinomoto, M., Kaneko, K., Ishizaka, Y., Tanaka, Y., Sata, T., & Tokunaga, K. (2009). HIV-1 accessory protein Vpu internalizes cell-surface BST-2/tetherin through transmembrane interactions leading to lysosomes. Journal of Biological Chemistry, 284(50), 35060–35072. Jacks, T., Power, M. D., Masiarz, F. R., Luciw, P. A., Barr, P. J., & Varmus, H. E. (1988). Characterization of ribosomal frameshifting in HIV-1 gag-pol expression. Nature, 331(6153), 280–283. Jacks, T., & Varmus, H. E. (1985). Expression of the Rous sarcoma virus pol gene by ribosomal frameshifting. Science, 230(4731), 1237–1242. Januszeski, M. M., Cannon, P. M., Chen, D., Rozenberg, Y., & Anderson, W. F. (1997). Func- tional analysis of the cytoplasmic tail of Moloney murine leukemia virus envelope protein. Journal of Virology, 71(5), 3613–3619. Jerez, A., & Ramírez-Pinilla, M. P. (2001). The allantoplacenta of Mabuya mabouya (Sauria, Scincidae). Journal of Morphology, 249(2), 132–46. Jerez, A., & Ramírez-Pinilla, M. P. (2003). Morphogenesis of extraembryonic membranes and placentation in Mabuya mabouya (Squamata, Scincidae). Journal of Morphology, 258(2), 158–78. Jern, P., & Coffin, J. M. (2008). Effects of Retroviruses on Host Genome Function. Annual Review of Genetics, 42(1), 709–732.

202 Jern, P., Sperber, G. O., & Blomberg, J. (2005). Use of endogenous retroviral sequences (ERVs) and structural markers for retroviral phylogenetic inference and taxonomy. Retrovirology, 2, 50. Jha, A. R., Nixon, D. F., Rosenberg, M. G., Martin, J. N., Deeks, S. G., Hudson, R. R., Garrison, K. E., & Pillai, S. K. (2011). Human endogenous retrovirus K106 (HERV-k106) was infectious after the emergence of anatomically modern humans. PLoS ONE, 6(5), 1–8. Jia, B., Serra-Moreno, R., Neidermyer, W., Rahmberg, A., Mackey, J., Fofana, I. B., Johnson, W. E., Westmoreland, S., & Evans, D. T. (2009). Species-Specific Activity of SIV Nef and HIV-1 Vpu in Overcoming Restriction by Tetherin/BST2. PLoS Pathogens, 5(5), e1000429. Jia, D., Jing, Y., Zhang, Z., Liu, L., Ding, J., Zhao, F., Ge, C., Wang, Q., Chen, T., Yao, M., Li, J., Gu, J., & He, X. (2014). Amplification of MPZL1/PZR promotes tumor cell migration through Src-mediated phosphorylation of cortactin in hepatocellular carcinoma. Cell Research, 24(2), 204–217. Johnson, W. E. (2015). Endogenous Retroviruses in the Genomics Era. Annual Review of Virology, 2(1), 135–159. Jolicoeur, P., & Baltimore, D. (1976). Effect of Fv-1 gene product on proviral DNA formation and integration in cells infected with murine leukemia viruses. Proceedings of the National Academy of Sciences, 73(7), 2236–40. Jones, C. J., Harris, L. K., Whittingham, J., Aplin, J. D., & Mayhew, T. M. (2008). A Re-appraisal of the Morphophenotype and Basal Lamina Coverage of Cytotrophoblasts in Human Term Placenta. Placenta, 29(2), 215–219. Jones, K. S., Lambert, S., Bouttier, M., Bénit, L., Ruscetti, F. W., Hermine, O., & Pique, C. (2011). Molecular aspects of HTLV-1 entry: Functional domains of the HTLV-1 surface subunit (SU) and their relationships to the entry receptors. Viruses, 3(6), 794–810. Jouvenet, N., Bieniasz, P. D., & Simon, S. M. (2008). Imaging the biogenesis of individual HIV-1 virions in live cells. Nature, 454(7201), 236–240. Karin, B. R., Metallinou, M., Weinell, J. L., Jackman, T. R., & Bauer, A. M. (2016). Resolving the higher-order phylogenetic relationships of the circumtropical Mabuya group (Squamata: Scincidae): An out-of-Asia diversification. Molecular Phylogenetics and Evolution, 102, 220–232. Kato, N., Pfeifer-Ohlsson, S., Kato, M., Larsson, E., Rydnert, J., Ohlsson, R., & Cohen, M. (1987). Tissue-specific expression of human provirus ERV3 mRNA in human placenta: two of the three ERV3 mRNAs contain human cellular sequences. Journal of Virology, 61(7), 2182–91. Katzourakis, A., Tristem, M., Pybus, O. G., & Gifford, R. J. (2007). Discovery and analysis of the first endogenous lentivirus. Proceedings of the National Academy of Sciences, 104(15), 6261–6265. Keele, B. F., Giorgi, E. E., Salazar-Gonzalez, J. F., Decker, J. M., Pham, K. T., Salazar, M. G., Sun, C., Grayson, T., Wang, S., Li, H., Wei, X., Jiang, C., Kirchherr, J. L., Gao, F., Anderson, J. A., Ping, L.-H., Swanstrom, R., Tomaras, G. D., Blattner, W. A., Goepfert, P. A., Kilby, J. M., Saag, M. S., Delwart, E. L., Busch, M. P., Cohen, M. S., Montefiori, D. C., Haynes, B. F., Gaschen, B., Athreya, G. S., Lee, H. Y., Wood, N., Seoighe, C., Perelson, A. S., Bhattacharya, T., Korber, B. T., Hahn, B. H., & Shaw, G. M. (2008). Identification and characterization of transmitted and early founder virus envelopes in primary HIV-1 infection. Proceedings of the National Academy of Sciences, 105(21), 7552–7557. Keller, A., Partin, K. M., Löchelt, M., Bannert, H., Flügel, R. M., & Cullen, B. R. (1991). Characterization of the transcriptional trans activator of human foamy retrovirus. Journal of Virology, 65(5), 2589–94.

203 Kestier, H. W., Ringler, D. J., Mori, K., Panicali, D. L., Sehgal, P. K., Daniel, M. D., & Desrosiers, R. C. (1991). Importance of the nef gene for maintenance of high virus loads and for development of AIDS. Cell, 65(4), 651–662. Khan, A. S., Bodem, J., Buseyne, F., Gessain, A., Johnson, W., Kuhn, J. H., Kuzmak, J., Linde- mann, D., Linial, M. L., Löchelt, M., Materniak-Kornas, M., Soares, M. A., & Switzer, W. M. (2018). Spumaretroviruses: Updated taxonomy and nomenclature. Virology, 516(December 2017), 158–164. Kim, F. J., Manel, N., Garrido, E. N., Valle, C., Sitbon, M., & Battini, J.-L. (2004). No Title. Retrovirology, 1(1), 41. Kim, F. J., Seiliez, I., Denesvre, C., Lavillette, D., Cosset, F. L., & Sitbon, M. (2000). Definition of an amino-terminal domain of the human T-cell leukemia virus type 1 envelope surface unit that extends the fusogenic range of an ecotropic murine leukemia virus. The Journal of biological chemistry, 275(31), 23417–20. King, A., Birkby, C., & Loke, Y. W. (1989). Early human decidual cells exhibit NK activity against the K562 cell line but not against first trimester trophoblast. Cellular Immunology, 118(2), 337–344. King, G. J., Atkinson, B. A., & Robertson, H. A. (1979). Development of the bovine placentome during the second month of gestation. Reproduction, 55(1), 173–NP. Kirchhoff, F., Greenough, T. C., Brettler, D. B., Sullivan, J. L., & Desrosiers, R. C. (1995). Ab- sence of Intact nef Sequences in a Long-Term Survivor with Nonprogressive HIV-1 Infection. New England Journal of Medicine, 332(4), 228–232. Klatzmann, D., Champagne, E., Chamaret, S., Gruest, J., Guetard, D., Hercend, T., Gluckman, J. C., & Montagnier, L. (1984). T-lymphocyte T4 molecule behaves as the receptor for human retrovirus LAV. Nature, 312(5996), 767–768. Klikova, M., Rhee, S. S., Hunter, E., & Ruml, T. (1995). Efficient in vivo and in vitro assembly of retroviral capsids from Gag precursor proteins expressed in bacteria. Journal of Virology, 69(2), 1093–8. Knerr, I., Schubert, S. W., Wich, C., Amann, K., Aigner, T., Vogler, T., Jung, R., Dötsch, J., Rascher, W., & Hashemolhosseini, S. (2005). Stimulation of GCMa and syncytin via cAMP mediated PKA signaling in human trophoblastic cells under normoxic and hypoxic conditions. FEBS Letters, 579(18), 3991–3998. Kohlstaedt, L. A., Wang, J., Friedman, J. M., Rice, P. A., & Steitz, T. A. (1992). Crystal structure at 3.5 A resolution of HIV-1 reverse transcriptase complexed with an inhibitor. Science, 256(5065), 1783–1790. Konvalinka, J., Horejsi, M., Andreansky, M., Novek, P., Pichova, I., Blaha, I., Fabry, M., Sedlacek, J., Foundling, S., & Strop, P. (1992). An engineered retroviral proteinase from myeloblastosis associated virus acquires pH dependence and substrate specificity of the HIV- 1 proteinase. EMBO Journal, 11(3), 1141–1144. Konvalinka, J., Kräusslich, H. G., & Müller, B. (2015). Retroviral proteases and their roles in virion maturation. Virology, 479-480, 403–417. Koshi, K., Nakaya, Y., Kizaki, K., Ishiguro-Oonuma, T., Miyazawa, T., Spencer, T. E., & Hashizume, K. (2016). Induction of ovine trophoblast cell fusion by fematrin-1 in vitro. Animal Science Journal, 87(3), 419–422. Kozak, C. A. (2015). Origins of the endogenous and infectious laboratory mouse gammaretro- viruses. Viruses, 7(1), 1–26. Krohn, M. (1846). Observations sur la génération et le développement des Biphores (Salpa). Annales des sciences naturelles, 3(6), 110–131.

204 Kumar, P., Thirkill, T. L., Ji, J., Monte, L. H., & Douglas, G. C. (2015). Differential Effects of Sodium Butyrate and Lithium Chloride on Rhesus Monkey Trophoblast Differentiation. PLOS ONE, 10(8), e0135089. Kumar, S., Stecher, G., Suleski, M., & Hedges, S. B. (2017). TimeTree: A Resource for Timelines, Timetrees, and Divergence Times. Molecular biology and Evolution, 34(7), 1812–1819. Kumazawa, Y. (2007). Mitochondrial genomes from major lizard families suggest their phylo- genetic relationships and ancient radiations. Gene, 388(1-2), 19–26. Kutluay, S. B., & Bieniasz, P. D. (2010). Analysis of the Initiating Events in HIV-1 Particle Assembly and Genome Packaging. PLoS Pathogens, 6(11), e1001200. Kvaratskhelia, M., Sharma, A., Larue, R. C., Serrao, E., & Engelman, A. (2014). Molecular mechanisms of retroviral integration site selection. Nucleic Acids Research, 42(16), 10209– 10225. Kypr, J., Mrázek, J., & Reich, J. (1989). Nucleotide composition bias and CpG dinucleotide content in the genomes of HIV and HTLV. Biochimica et Biophysica Acta (BBA) - Gene Structure and Expression, 1009(3), 280–282. Ladyman, S. R., Augustine, R. A., & Grattan, D. R. (2010). Hormone interactions regulating energy balance during pregnancy. Journal of Neuroendocrinology, 22(7), 805–817. Laimins, L. A., Gruss, P., Pozzatti, R., & Khoury, G. (1984). Characterization of enhancer elements in the long terminal repeat of Moloney murine sarcoma virus. Journal of Virology, 49(1), 183–9. Lander, E. S., Linton, L. M., Birren, B., Nusbaum, C., Zody, M. C., Baldwin, J., Devon, K., Dewar, K., Doyle, M., Fitzhugh, W., Funke, R., Gage, D., Harris, K., Heaford, A., Howland, J., Kann, L., Lehoczky, J., Levine, R., McEwan, P., McKernan, K., Meldrim, J., Mesirov, J. P., Miranda, C., Morris, W., Naylor, J., Raymond, C., Rosetti, M., Santos, R., Sheridan, A., Sougnez, C., Stange-Thomann, N., Stojanovic, N., Subramanian, A., Wyman, D., Rogers, J., Sulston, J., Ainscough, R., Beck, S., Bentley, D., Burton, J., Clee, C., Carter, N., Coulson, A., Deadman, R., Deloukas, P., Dunham, A., Dunham, I., Durbin, R., French, L., Grafham, D., Gregory, S., Hubbard, T., Humphray, S., Hunt, A., Jones, M., Lloyd, C., McMurray, A., Matthews, L., Mercer, S., Milne, S., Mullikin, J. C., Mungall, A., Plumb, R., Ross, M., Shownkeen, R., Sims, S., Waterston, R. H., Wilson, R. K., Hillier, L. W., McPherson, J. D., Marra, M. A., Mardis, E. R., Fulton, L. A., Chinwalla, A. T., Pepin, K. H., Gish, W. R., Chissoe, S. L., Wendl, M. C., Delehaunty, K. D., Miner, T. L., Delehaunty, A., Kramer, J. B., Cook, L. L., Fulton, R. S., Johnson, D. L., Minx, P. J., Clifton, S. W., Hawkins, T., Branscomb, E., Predki, P., Richardson, P., Wenning, S., Slezak, T., Doggett, N., Cheng, J. F., Olsen, A., Lucas, S., Elkin, C., Uberbacher, E., Frazier, M., Gibbs, R. A., Muzny, D. M., Scherer, S. E., Bouck, J. B., Sodergren, E. J., Worley, K. C., Rives, C. M., Gorrell, J. H., Metzker, M. L., Naylor, S. L., Kucherlapati, R. S., Nelson, D. L., Weinstock, G. M., Sakaki, Y., Fujiyama, A., Hattori, M., Yada, T., Toyoda, A., Itoh, T., Kawagoe, C., Watanabe, H., Totoki, Y., Taylor, T., Weissenbach, J., Heilig, R., Saurin, W., Artiguenave, F., Brottier, P., Bruls, T., Pelletier, E., Robert, C., Wincker, P., Rosenthal, A., Platzer, M., Nyakatura, G., Taudien, S., Rump, A., Smith, D. R., Doucette-Stamm, L., Rubenfield, M., Weinstock, K., Hong, M. L., Dubois, J., Yang, H., Yu, J., Wang, J., Huang, G., Gu, J., Hood, L., Rowen, L., Madan, A., Qin, S., Davis, R. W., Federspiel, N. A., Abola, A. P., Proctor, M. J., Roe, B. A., Chen, F., Pan, H., Ramser, J., Lehrach, H., Reinhardt, R., McCombie, W. R., De La Bastide, M., Dedhia, N., Blöcker, H., Hornischer, K., Nordsiek, G., Agarwala, R., Aravind, L., Bailey, J. A., Bateman, A., Batzoglou, S., Birney, E., Bork, P., Brown, D. G., Burge, C. B., Cerutti, L., Chen, H. C., Church, D., Clamp, M., Copley, R. R., Doerks, T., Eddy, S. R., Eichler, E. E., Furey, T. S., Galagan, J., Gilbert, J. G., Harmon, C., Hayashizaki, Y., Haussler,

205 D., Hermjakob, H., Hokamp, K., Jang, W., Johnson, L. S., Jones, T. A., Kasif, S., Kaspryzk, A., Kennedy, S., Kent, W. J., Kitts, P., Koonin, E. V., Korf, I., Kulp, D., Lancet, D., Lowe, T. M., McLysaght, A., Mikkelsen, T., Moran, J. V., Mulder, N., Pollara, V. J., Ponting, C. P., Schuler, G., Schultz, J., Slater, G., Smit, A. F., Stupka, E., Szustakowki, J., Thierry-Mieg, D., Thierry-Mieg, J., Wagner, L., Wallis, J., Wheeler, R., Williams, A., Wolf, Y. I., Wolfe, K. H., Yang, S. P., Yeh, R. F., Collins, F., Guyer, M. S., Peterson, J., Felsenfeld, A., Wetterstrand, K. A., Myers, R. M., Schmutz, J., Dickson, M., Grimwood, J., Cox, D. R., Olson, M. V., Kaul, R., Raymond, C., Shimizu, N., Kawasaki, K., Minoshima, S., Evans, G. A., Athanasiou, M., Schultz, R., Patrinos, A., & Morgan, M. J. (2001). Initial sequencing and analysis of the human genome. Nature, 409(6822), 860–921. Langbein, M., Strick, R., Strissel, P. L., Vogt, N., Parsch, H., Beckmann, M. W., & Schild, R. L. (2008). Impaired cytotrophoblast cell–cell fusion is associated with reduced Syncytin and increased apoptosis in patients with placental dysfunction. Molecular Reproduction and Development, 75(1), 175–183. Langford, M., Outhwaite, J., Hughes, M., Natale, D., & Simmons, D. (2018). Deletion of the Syncytin A receptor Ly6e impairs syncytiotrophoblast fusion and placental morphogenesis causing embryonic lethality in mice. Scientific Reports, 8(1), 1–11. Larsen, J. F. (1963). Electron Microscopy of the Implantation Site in the Rabbit. Acta Obstetricia et Gynecologica Scandinavica, 42(6 S), 45–45. Larsson, A. (2014). AliView: A fast and lightweight alignment viewer and editor for large datasets. Bioinformatics, 30(22), 3276–3278. Lavialle, C., Cornelis, G., Dupressoir, A., Esnault, C., Heidmann, O., Vernochet, C., & Heid- mann, T. (2013). Paleovirology of ’syncytins’, retroviral env genes exapted for a role in pla- centation. Philosophical Transactions of the Royal Society B: Biological Sciences, 368(1626), 20120507. Lavie, L., Kitova, M., Maldener, E., Meese, E., & Mayer, J. (2005). CpG methylation directly regulates transcriptional activity of the human endogenous retrovirus family HERV-K (HML- 2). Journal of Virology, 79(2), 876. Lavillette, D., Marin, M., Ruggieri, A., & Kabat, D. (2002). The Envelope Glycoprotein of Human Endogenous Retrovirus Type W Uses a Divergent Family of Amino Acid Transporters / Cell Surface Receptors. Journal of Virology, 76(13), 6442–6452. Lavillette, D., Maurice, M., Roche, C., Russell, S. J., Sitbon, M., & Cosset, F. L. (1998). A proline-rich motif downstream of the receptor binding domain modulates conformation and fusogenicity of murine retroviral envelopes. Journal of Virology, 72(12), 9955–9965. Le Tortorec, A., & Neil, S. J. D. (2009). Antagonism to and intracellular sequestration of human tetherin by the human immunodeficiency virus type 2 envelope glycoprotein. Journal of Virology, 83(22), 11966–78. Leal, F., & Ramírez-Pinilla, M. P. (2008). Morphological variation in the allantoplacenta within the genus Mabuya (squamata: Scincidae). Anatomical record (Hoboken, N.J. : 2007), 291(9), 1124–39. Lefort, V., Longueville, J.-E., & Gascuel, O. (2017). SMS: Smart Model Selection in PhyML. Molecular Biology and Evolution, 34(9), 2422–2424. Legione, A. R., Patterson, J. L., Whiteley, P., Firestone, S. M., Curnick, M., Bodley, K., Lynch, M., Gilkerson, J. R., Sansom, F. M., & Devlin, J. M. (2017). Koala retrovirus genotyping analyses reveal a low prevalence of KoRV-A in Victorian koalas and an association with clinical disease. Journal of Medical Microbiology, 66(2), 236–244. Leiser, R. (1979). Blastocystenimplantation bei der Hauskatze Licht- und elektronen-

206 mikroskopische Untersuchung. Anatomia, Histologia, Embryologia: Journal of Veterinary Medicine Series C, 8(1), 79–96. Leiser, R., & Koob, B. (1993). Development and characteristics of placentation in a carnivore, the domestic cat. Journal of Experimental Zoology, 266(6), 642–656. Lemaitre, C., Harper, F., Pierron, G., Heidmann, T., & Dewannieux, M. (2014). The HERV-K Human Endogenous Retrovirus Envelope Protein Antagonizes Tetherin Antiviral Activity. Journal of Virology, 88(23), 13626–13637. Lemaître, C., Tsang, J., Bireau, C., Heidmann, T., Dewannieux, M., & Cofer, S. (2017). A human endogenous retrovirus-derived gene that can contribute to oncogenesis by activating the ERK pathway and inducing migration and invasion. PLOS Pathogens, 13(6), e1006451. Lesbats, P., Engelman, A. N., & Cherepanov, P. (2016). Retroviral DNA Integration. Chemical Reviews, 116(20), 12730–12757. Li, K., Zhang, S., Kronqvist, M., Wallin, M., Ekstrom, M., Derse, D., & Garoff, H. (2008). Intersubunit Disulfide Isomerization Controls Membrane Fusion of Human T-Cell Leukemia Virus Env. Journal of Virology, 82(14), 7135–7143. Li, W., Lee, M.-H., Henderson, L., Tyagi, R., Bachani, M., Steiner, J., Campanac, E., Hoffman, D. A., von Geldern, G., Johnson, K., Maric, D., Morris, H. D., Lentz, M., Pak, K., Mammen, A., Ostrow, L., Rothstein, J., & Nath, A. (2015). Human endogenous retrovirus-K contributes to motor neuron disease. Science Translational Medicine, 7(307), 307ra153. Liang, C.-Y., Wang, L.-J., Chen, C.-P., Chen, L.-F., Chen, Y.-H., & Chen, H. (2010). GCM1 Regulation of the Expression of Syncytin 2 and Its Cognate Receptor MFSD2A in Human Placenta1. Biology of Reproduction, 83(3), 387–395. Lightfoote, M. M., Coligan, J. E., Folks, T. M., Fauci, A. S., Martin, A., & Venkatesan, S. (1986). Structural characterization of reverse transcriptase and endonuclease polypeptides of the acquired immunodeficiency syndrome retrovirus. Journal of Virology, 60(2), 771–775. Lillegraven, J. A., Thompson, Steven, D., McNab, B. K., & Patton, J. L. (1987). The origin of eutherian mammals. Biological Journal of the Linnean Society, 32(3), 281–336. Lin, L., Xu, B., & Rote, N. S. (1999). Expression of endogenous retrovirus ERV-3 induces differentiation in BeWo, a choriocarcinoma model of human placental trophoblast. Placenta, 20, 109–118. Lin, L., Xu, B., & Rote, N. S. (2000). The cellular mechanism by which the human endogenous retrovirus ERV-3 env gene affects proliferation and differentiation in a human placental trophoblast model, BeWo. Placenta, 21, 73–78. Lindemann, D., Pietschmann, T., Picard-Maureau, M., Berg, A., Heinkelein, M., Thurow, J., Knaus, P., Zentgraf, H., & Rethwilm, A. (2001). A particle-associated glycoprotein signal peptide essential for virus maturation and infectivity. Journal of Virology, 75(13), 5762–71. Lippincott-Schwartz, J., Freed, E., & van Engelenburg, S. (2017). A Consensus View of ESCRT- Mediated Human Immunodeficiency Virus Type 1 Abscission. Annual Review of Virology, 4(1), 309–325. Löchelt, M., Muranyi, W., & Flügel, R. M. (1993). Human foamy virus genome possesses an internal, Bel-1-dependent and functional promoter. Proceedings of the National Academy of Sciences, 90(August), 7317–21. Löchelt, M., Romen, F., Bastone, P., Muckenfuss, H., Kirchner, N., Kim, Y.-B., Truyen, U., Rösler, U., Battenberg, M., Saib, A., Flory, E., Cichutek, K., & Münk, C. (2005). The antiretroviral activity of APOBEC3 is inhibited by the foamy virus accessory Bet protein. Proceedings of the National Academy of Sciences, 102(22), 7982–7. Lock, L. F., Keshet, E., Gilbert, D. J., Jenkins, N. a., & Copeland, N. G. (1988). Studies of

207 the mechanism of spontaneous germline ecotropic provirus acquisition in mice. The EMBO Journal, 7(13), 4169–77. Long, J. A., Trinajstic, K., Young, G. C., & Senden, T. (2008). Live birth in the Devonian period. Nature, 453(7195), 650–652. Louis, J. M., Marius Clore, G., & Gronenborn, A. M. (1999). Autoprocessing of HIV-1 protease is tightly coupled to protein folding. Nature Structural Biology, 6(9), 868–875. Lu, X., Sachs, F., Ramsay, L. A., Jacques, P. É., Göke, J., Bourque, G., & Ng, H. H. (2014). The retrovirus HERVH is a long noncoding RNA required for human embryonic stem cell identity. Nature Structural and Molecular Biology, 21(4), 423–425. Lukic, S., Nicolas, J. C., & Levine, A. J. (2014). The diversity of zinc-finger genes on human chromosome 19 provides an evolutionary mechanism for defense against inherited endogenous retroviruses. Cell Death and Differentiation, 21(3), 381–387. Lynch, V. J., Nnamani, M. C., Kapusta, A., Brayer, K., Plaza, S. L., Mazur, E. C., Emera, D., Sheikh, S. Z., Grüzner, F., Bauersachs, S., Graf, A., Young, S. L., Lieb, J. D., DeMayo, F. J., Feschotte, C. C., Wagner, G. P., Grützner, F., Bauersachs, S., Graf, A., Young, S. L., Lieb, J. D., DeMayo, F. J., Feschotte, C. C., & Wagner, G. P. (2015). Ancient transposable elements transformed the uterine regulatory landscape and transcriptome during the evolution of mammalian pregnancy. Cell Reports, 10(4), 551–562. Macaulay, E. C., Chatterjee, A., Cheng, X., Baguley, B. C., Eccles, M. R., & Morison, I. M. (2017). The Genes of Life and Death: A Potential Role for Placental-Specific Genes in Cancer. BioEssays, 39(11), 1700091. Macfarlan, T. S., Gifford, W. D., Driscoll, S., Lettieri, K., Rowe, H. M., Bonanomi, D., Firth, A., Singer, O., Trono, D., & Pfaff, S. L. (2012). Embryonic stem cell potency fluctuates with endogenous retrovirus activity. Nature, 487(7405), 57–63. Mager, D. L., & Stoye, J. P. (2015). Mammalian Endogenous Retroviruses. Microbiology spectrum, 3(1), 1–20. Magiorkinis, G., Blanco-Melo, D., & Belshaw, R. (2015). The decline of human endogenous retroviruses: extinction and survival. Retrovirology, 12(1), 8. Magiorkinis, G., Gifford, R. J., Katzourakis, A., De Ranter, J., & Belshaw, R. (2012). Env-less endogenous retroviruses are genomic superspreaders. Proceedings of the National Academy of Sciences, 109(19), 7385–7390. Maksakova, I. A., Romanish, M. T., Gagnier, L., Dunn, C. A., Van De Lagemaat, L. N., & Mager, D. L. (2006). Retroviral elements and their hosts: Insertional mutagenesis in the mouse germ line. PLoS Genetics, 2(1), 1–10. Malassiné, A., Blaise, S., Handschuh, K., Lalucque, H., Dupressoir, A., Evain-Brion, D., & Heidmann, T. (2007). Expression of the fusogenic HERV-FRD Env glycoprotein (syncytin 2) in human placenta is restricted to villous cytotrophoblastic cells. Placenta, 28(2-3), 185–91. Malassiné, A., Handschuh, K., Tsatsaris, V., Gerbaud, P., Cheynet, V., Oriol, G., Mallet, F., & Evain-Brion, D. (2005). Expression of HERV-W Env glycoprotein (syncytin) in the extravil- lous trophoblast of first trimester human placenta. Placenta, 26(7), 556–562. Malicorne, S., Vernochet, C., Cornelis, G., Mulot, B., Delsuc, F., Heidmann, O., Heidmann, T., & Dupressoir, A. (2016). Genome-Wide Screening of Retroviral Envelope Genes in the Nine-Banded Armadillo (Dasypus novemcinctus, Xenarthra) Reveals an Unfixed Chimeric Endogenous Betaretrovirus Using the ASCT2 Receptor. Journal of Virology, 90(18), 8132– 49. Malim, M. H., Hauber, J., Le, S.-Y., Maizel, J. V., & Cullen, B. R. (1989). The HIV-1 rev trans- activator acts through a structured target sequence to activate nuclear export of unspliced viral mRNA. Nature, 338(6212), 254–257.

208 Mallet, F., Bouton, O., Prudhomme, S., Cheynet, V., Oriol, G., Bonnaud, B., Lucotte, G., Duret, L., & Mandrand, B. (2004). The endogenous retroviral locus ERVWE1 is a bona fide gene involved in hominoid placental physiology. Proceedings of the National Academy of Sciences, 101(6), 1731–1736. Manel, N., Kim, F. J., Kinet, S., Taylor, N., Sitbon, M., & Battini, J. L. (2003). The Ubiquitous Glucose Transporter GLUT-1 Is a Receptor for HTLV. Cell, 115(4), 449–459. Mangeney, M., & Heidmann, T. (1998). Tumor cells expressing a retroviral envelope escape immune rejection in vivo. Proceedings of the National Academy of Sciences, 95(25), 14920–5. Mangeney, M., Renard, M., Schlecht-Louf, G., Bouallaga, I., Heidmann, O., Letzelter, C., Richaud, A., Ducos, B., & Heidmann, T. (2007). Placental syncytins: Genetic disjunction between the fusogenic and immunosuppressive activity of retroviral envelope proteins. Pro- ceedings of the National Academy of Sciences, 104(51), 20534–9. Marchi, E., Kanapin, A., Magiorkinis, G., & Belshaw, R. (2014). Unfixed Endogenous Retroviral Insertions in the Human Population. Journal of Virology, 88(17), 9529–9537. Marder, W., Littlejohn, E. A., & Somers, E. C. (2016). Pregnancy and autoimmune connective tissue diseases. Best Practice and Research: Clinical Rheumatology, 30(1), 63–80. Maréchal, V., Clavel, F., Heard, J. M., & Schwartz, O. (1998). Cytosolic Gag p24 as an index of productive entry of human immunodeficiency virus type 1. Journal of Virology, 72(3), 2208–2212. Marin, M., Lavillette, D., Kelly, S. M., & Kabat, D. (2003). N-Linked Glycosylation and Sequence Changes in a Critical Negative Control Region of the ASCT1 and ASCT2 Neu- tral Amino Acid Transporters Determine Their Retroviral Receptor Functions. Journal of Virology, 77(5), 2936–2945. Marin, M., Tailor, C. S., Nouri, A. L. I., & Kabat, D. (2000). Sodium-Dependent Neutral Amino Acid Transporter Type 1 Is an Auxiliary Receptor for Baboon Endogenous Retrovirus. Journal of Virology, 74(17), 8085–8093. Matoušková, M., Blažková, J., Pajer, P., Pavlíček, A., & Hejnar, J. (2006). CpG methylation suppresses transcriptional activity of human syncytin-1 in non-placental tissues. Experimental Cell Research, 312(7), 1011–1020. Matsuoka, M., & Jeang, K.-T. (2007). Human T-cell leukaemia virus type 1 (HTLV-1) infectivity and cellular transformation. Nature Reviews Cancer, 7(4), 270–280. Mattei, S., Schur, F. K., & Briggs, J. A. (2016). Retrovirus maturation - An extraordinary structural transformation. Current Opinion in Virology, 18, 27–35. Mausfeld, P., Schmitz, A., Böhme, W., Misof, B. Y., Vrcibradic, D., & Rocha, C. F. D. D. (2002). Phylogenetic affinities of Mabuya atlantica Schmidt, 1945, endemic to the Atlantic ocean archipelago of Fernando de Noronha (Brazil): necessity of partitioning the genus Mabuya Fitzinger, 1826 (Scincidae: Lygosominae). Zoologischer Anzeiger, 241, 281–293. Mayhew, T. M., Wadrop, E., & Simpson, R. A. (1994). Proliferative versus hypertrophic growth in tissue subcompartments of human placental villi during gestation. Journal of Anatomy, 184(Pt 3), 535–43. Mayo, M. A. (2002). ICTV at the Paris ICV: Results of the plenary session and the binomial ballot. Archives of Virology, 147(11), 2254–2260. McClintock, B. (1950). The origin and behavior of mutable loci in maize. Proceedings of the National Academy of Sciences, 36(6), 344–355. McClintock, B. (1956). Controlling Elements and the Gene. Cold Spring Harbor Symposia on Quantitative Biology, 21, 197–216. McClure, M. a., Johnson, M. S., Feng, D. F., & Doolittle, R. F. (1988). Sequence comparisons

209 of retroviral proteins: relative rates of change and general phylogeny. Proceedings of the National Academy of Sciences, 85(8), 2469–73. Medstrand, P., & Mager, D. L. (1998). Human-specific integrations of the HERV-K endogenous retrovirus family. Journal of Virology, 72(12), 9782–7. Mendez-de la Cruz, F. R., Villagran-Santa Cruz, M., & Andrews, R. M. (1998). Constraints on the evolution of viviparity in the lizard genus Sceloporus. Herpetologica, 54(4), 521–532. Mertz, D., Geraci, J., Winkup, J., Gessner, B. D., Ortiz, J. R., & Loeb, M. (2017). Pregnancy as a risk factor for severe outcomes from influenza virus infection: A systematic review and meta-analysis of observational studies. Vaccine, 35(4), 521–528. Metallinou, M., Weinell, J. L., Karin, B. R., Conradie, W., Wagner, P., Schmitz, A., Jackman, T. R., & Bauer, A. M. (2016). A single origin of extreme matrotrophy in African mabuyine skinks. Biology Letters, 12(8), 20160430. Metz, J., Heinrich, D., & Forssmann, W. G. (1976). Gap junctions in hemodichorial and hemotrichorial placentae. Cell and Tissue Research, 171(3), 305–315. Mi, S., Lee, X., Li, X.-p., Veldman, G. M., Finnerty, H., Racie, L., LaVallie, E., Tang, X.-Y., Edouard, P., Howes, S., Keith, J. C., & McCoy, J. M. (2000). Syncytin is a captive retroviral envelope protein involved in human placental morphogenesis. Nature, 403(6771), 785–789. Miyaho, R. N., Nakagawa, S., Hashimoto-Gotoh, A., Nakaya, Y., Shimode, S., Sakaguchi, S., Yoshikawa, R., Takahashi, M. U., & Miyazawa, T. (2015). Susceptibility of domestic animals to a pseudotype virus bearing RD-114 virus envelope protein. Gene, 567(2), 189–195. Moffett, A., Chazara, O., & Colucci, F. (2017). Maternal allo-recognition of the fetus. Fertility and Sterility, 107(6), 1269–1272. Moffett, A., & Colucci, F. (2014). Uterine NK cells: Active regulators at the maternal-fetal interface. Journal of Clinical Investigation, 124(5), 1872–1879. Moffett, A., & Loke, C. (2006). Immunology of placentation in eutherian mammals. Nature Reviews Immunology, 6(8), 584–594. Møller, A. M. J., Delaissé, J. M., & Søe, K. (2017). Osteoclast Fusion: Time-Lapse Reveals Involvement of CD47 and Syncytin-1 at Different Stages of Nuclearity. Journal of Cellular Physiology, 232(6), 1396–1403. Mölling, K., Bolognesi, D. P., Bauer, H., Büsen, W., Plassmann, H. W., & Hausen, P. (1971). Association of viral reverse transcriptase with an enzyme degrading the rna moiety of RNA- DNA hybrids. Nature New Biology, 234(51), 240–243. Morgan, G., & Wooding, F. B. P. (1983). Cell migration in the ruminant placenta: A freeze- fracture study. Journal of Ultrasructure Research, 83(2), 148–160. Morozov, V. A., Dao Thi, V. L., & Denner, J. (2013). The Transmembrane Protein of the Human Endogenous Retrovirus - K (HERV-K) Modulates Cytokine Release and Gene Expression. PLoS ONE, 8(8), e70399. Morton, W. R. (1957). Placentation in the spotted hyena (Crocuta crocuta Erxleben). Journal of Anatomy, 91(3), 374–82. Mossman, H. (1937). Comparative morphogenesis of the fetal membranes and accessory struc- tures. Contributions to Embryology, 26, 129–246. Mossman, H. W. (1926). The rabbit placenta and the problem of placental transmission. American Journal of Anatomy, 37(3), 433–497. Muir, A., Lever, A. M., & Moffett, A. (2006). Human endogenous retrovirus-W envelope (syncytin) is expressed in both villous and extravillous trophoblast populations. Journal of General Virology, 87(7), 2067–2071.

210 Murakami, T., Ablan, S., Freed, E. O., & Tanaka, Y. (2004). Regulation of human immunod- eficiency virus type 1 Env-mediated membrane fusion by viral protease activity. Journal of Virology, 78(2), 1026–1031. Nakamura, Y., & Imakawa, K. (2011). Retroviral Endogenization and Its Role in the Genital Tract during Mammalian Evolution. Journal of Mammalian Ova Research, 28(4), 203–218. Nakaya, Y., Koshi, K., Nakagawa, S., Hashizume, K., & Miyazawa, T. (2013). Fematrin-1 Is Involved in Fetomaternal Cell-to-Cell Fusion in Bovinae Placenta and Has Contributed to Diversity of Ruminant Placentation. Journal of Virology, 87(19), 10563–10572. Nakaya, Y., & Miyazawa, T. (2015). The Roles of Syncytin-Like Proteins in Ruminant Placen- tation. Viruses, 7(6), 2928–2942. Neil, S. J. D., Zang, T., & Bieniasz, P. D. (2008). Tetherin inhibits retrovirus release and is antagonized by HIV-1 Vpu. Nature, 451(7177), 425–430. Nethe, M., Berkhout, B., & van der Kuyl, A. C. (2005). Retroviral superinfection resistance. Retrovirology, 2, 1–13. Nikolaitchik, O. A., Dilley, K. A., Fu, W., Gorelick, R. J., Tai, S.-H. S., Soheilian, F., Ptak, R. G., Nagashima, K., Pathak, V. K., & Hu, W.-S. (2013). Dimeric RNA Recognition Regulates HIV-1 Genome Packaging. PLoS Pathogens, 9(3), e1003249. Nowak, E., Miller, J. T., Bona, M. K., Studnicka, J., Szczepanowski, R. H., Jurkowski, J., Le Grice, S. F., & Nowotny, M. (2014). Ty3 reverse transcriptase complexed with an RNA-DNA hybrid shows structural and functional asymmetry. Nature Structural and Molecular Biology, 21(4), 389–396. Nowotny, M., Gaidamakov, S. A., Crouch, R. J., & Yang, W. (2005). Crystal structures of RNase H bound to an RNA/DNA hybrid: Substrate specificity and metal-dependent catalysis. Cell, 121(7), 1005–1016. Nyakatura, K., & Bininda-Emonds, O. R. (2012). Updating the evolutionary history of Carnivora (Mammalia): a new species-level supertree complete with divergence time estimates. BMC Biology, 10(1), 12. Nyathi, Y., Wilkinson, B. M., & Pool, M. R. (2013). Co-translational targeting and translocation of proteins to the endoplasmic reticulum. Biochimica et Biophysica Acta (BBA) - Molecular Cell Research, 1833(11), 2392–2402. O’Connell, C., O’Brien, S., Nash, W. G., & Cohen, M. (1984). ERV3, A full-length human endogenous provirus: Chromosomal localization and evolutionary relationships. Virology, 138(2), 225–235. Oduor Okelo, D., & Neaves, W. B. (1982). The chorioallantoic placenta of the spotted hyena (Crocuta crocuta Erxleben): An electron microscopic study. Anatomical Record, 204(July 1978), 215–222. Ono, A., Ablan, S. D., Lockett, S. J., Nagashima, K., & Freed, E. O. (2004). Phosphatidylinositol (4,5) bisphosphate regulates HIV-1 Gag targeting to the plasma membrane. Proceedings of the National Academy of Sciences, 101(41), 14889–14894. Ono, A., & Freed, E. O. (2004). Cell-type-dependent targeting of human immunodeficiency virus type 1 assembly to the plasma membrane and the multivesicular body. Journal of Virology, 78(0022-538X (Print)), 1552–1563. Orlinsky, K. J., Gu, J., Hoyt, M., Sandmeyer, S., & Menees, T. M. (1996). Mutations in the Ty3 major homology region affect multiple steps in Ty3 retrotransposition. Journal of virology, 70(6), 3440–8. Ostensen, M. (1999). Sex Hormones and Pregnancy in Rheumatoid Arthritis and Systemic Lupus Erythematosus. Annals of the New York Academy of Sciences, 876(1), 131–144.

211 Ostrovsky, A. N., Lidgard, S., Gordon, D. P., Schwaha, T., Genikhovich, G., & Ereskovsky, A. V. (2015). Matrotrophy and placentation in invertebrates: A new paradigm. Biological Reviews. Overbaugh, J., Miller, A. D., & Eiden, M. V. (2001). Receptors and Entry Cofactors for Retroviruses Include Single and Multiple Transmembrane-Spanning Proteins as well as Newly Described Glycophosphatidylinositol-Anchored and Secreted Proteins. Microbiology and Molecular Biology Reviews, 65(3), 371–389. Parham, P., & Moffett, A. (2013). Variable NK cell receptors and their MHC class I ligands in immunity, reproduction and human evolution. Nature Reviews Immunology, 13(2), 133–144. Passos, D. O., Li, M., Yang, R., Rebensburg, S. V., Ghirlando, R., Jeon, Y., Shkriabai, N., Kvaratskhelia, M., Craigie, R., & Lyumkis, D. (2017). Cryo-EM structures and atomic model of the HIV-1 strand transfer complex intasome. Science, 355(6320), 89–92. Peaston, A. E., Evsikov, A. V., Graber, J. H., de Vries, W. N., Holbrook, A. E., Solter, D., & Knowles, B. B. (2004). Retrotransposons Regulate Host Genes in Mouse Oocytes and Preimplantation Embryos. Developmental Cell, 7(4), 597–606. Pereira, A. G., & Schrago, C. G. (2017). Arrival and diversification of mabuyine skinks (Squamata: Scincidae) in the Neotropics based on a fossil-calibrated timetree. PeerJ, 5, e3194. Pereira, L. E., Clark, J., Grznarova, P., Wen, X., LaCasse, R., Ruml, T., Spearman, P., & Hunter, E. (2014). Direct evidence for intracellular anterograde co-transport of M-PMV Gag and Env on microtubules. Virology, 449(2014), 109–119. Perilla, J. R., & Gronenborn, A. M. (2016). Molecular Architecture of the Retroviral Capsid. Trends in Biochemical Sciences, 41(5), 410–420. Pérot, P., Mugnier, N., Montgiraud, C., Gimenez, J., Jaillard, M., Bonnaud, B., & Mallet, F. (2012). Microarray-Based Sketches of the HERV Transcriptome Landscape. PLoS ONE, 7(6), e40194. Peters, G., & Dahlberg, J. E. (1979). RNA-directed DNA synthesis in Moloney murine leukemia virus: interaction between the primer tRNA and the genome RNA. Journal of Virology, 31(2), 398–407. Philipson, L., Andersson, P., Olshevsky, U., Weinberg, R., Baltimore, D., & Gesteland, R. (1978). Translation of MuLV and MSV RNAs in nuclease-treated reticulocyte extracts: Enhancement of the gag-pol polypeptide with yeast suppressor tRNA. Cell, 13(1), 189–199. Pijnenborg, R., Bland, J. M., Robertson, W. B., & Brosens, I. (1983). Uteroplacental arterial changes related to interstitial trophoblast migration in early human pregnancy. Placenta, 4(4), 397–413. Pijnenborg, R., Dixon, G., Robertson, W. B., & Brosens, I. (1980). Trophoblastic invasion of human decidua from 8 to 18 weeks of pregnancy. Placenta, 1(1), 3–19. Pijnenborg, R., Robertson, W. B., Brosens, I., & Dixon, G. (1981). Review article: Trophoblast invasion and the establishment of haemochorial placentation in man and laboratory animals. Placenta, 2(1), 71–91. Pinter, A., Kopelman, R., Li, Z., Kayman, S. C., & Sanders, D. A. (1997). Localization of the labile disulfide bond between SU and TM of the murine leukemia virus envelope protein complex to a highly conserved CWLC motif in SU that resembles the active-site sequence of thiol-disulfide exchange enzymes. Journal of Virology, 71(10), 8073–8077. Pinto-Sánchez, N. R., Calderón-Espinosa, M. L., Miralles, A., Crawford, A. J., & Ramírez- Pinilla, M. P. (2015). Molecular phylogenetics and biogeography of the Neotropical skink genus Mabuya Fitzinger (Squamata: Scincidae) with emphasis on Colombian populations. Molecular Phylogenetics and Evolution, 93, 188–211.

212 Pizzato, M., Blair, E. D., Fling, M., Kopf, J., Tomassetti, A., Weiss, R. A., & Takeuchi, Y. (2001). Evidence for nonspecific adsorption of targeted retrovirus vector particles to cells. Gene Therapy, 8(14), 1088–1096. Pizzato, M., Marlow, S. a., Blair, E. D., & Takeuchi, Y. (1999). Initial binding of murine leukemia virus particles to cells does not require specific Env-receptor interaction. Journal of Virology, 73(10), 8599–8611. Planelles, V., Jowett, J. B. M., Li, Q.-X., Xie, Y., Hahn, B., & Chen, I. S. Y. (1996). Vpr-Induced Cell Cycle Arrest Is Conserved among Primate Lentiviruses. Journal of Virology, 70(4), 2516–2524. Poch, O., Sauvaget, I., Delarue, M., & Tordo, N. (1989). Identification of four conserved motifs among the RNA-dependent polymerase encoding elements. The EMBO Journal, 8(12), 3867– 74. Podbilewicz, B. (2014). Virus and cell fusion mechanisms. Annual review of cell and develop- mental biology, 30, 111–39. Pollard, V. W., & Malim, M. H. (1998). The HIV-1 Rev protein. Annual Review of Microbiology, 52(1), 491–532. Prabu-Jeyabalan, M., Nalivaika, E., & Schiffer, C. A. (2002). Substrate shape determines specificity of recognition for HIV-1 protease: Analysis of crystal structures of six substrate complexes. Structure, 10(3), 369–381. Quoy, J. R. C., & Gaimard, J. P. (1835). Zoologie. In Voyage de la corvette l’Astrolabe : exécuté par ordre du roi, pendant les années 1826-1827-1828-1829 / sous le commandement de J. Dumont d’Urville., chap. 3. Paris: J. Tastu. Rakyan, V. K., Blewitt, M. E., Druker, R., Preis, J. I., & Whitelaw, E. (2002). Metastable epialleles in mammals. Trends in Genetics, 18(7), 348–351. Ramírez-Pinilla, M. P., De Pérez, G., & Carreño-Escobar, J. F. (2006). Allantoplacental ultra- structure of an Andean population of Mabuya (Squamata, Scincidae). Journal of Morphology, 267(10), 1227–47. Ramírez-Pinilla, M. P., Rueda, E. D., & Stashenko, E. (2011). Transplacental nutrient trans- fer during gestation in the Andean lizard Mabuya sp. (Squamata, Scincidae). Journal of Comparative Physiology B, 181(2), 249–268. Rasaiyaah, J., Tan, C. P., Fletcher, A. J., Price, A. J., Blondeau, C., Hilditch, L., Jacques, D. A., Selwood, D. L., James, L. C., Noursadeghi, M., & Towers, G. J. (2013). HIV-1 evades innate immune recognition through specific cofactor recruitment. Nature, 503(7476), 402–405. Redelsperger, F., Cornelis, G., Vernochet, C., Tennant, B. C., Catzeflis, F., Mulot, B., Heidmann, O., Heidmann, T., & Dupressoir, A. (2014). Capture of syncytin-Mar1, a Fusogenic Endoge- nous Retroviral Envelope Gene Involved in Placentation in the Rodentia Squirrel-Related Clade. Journal of Virology, 88(14), 7915–7928. Redelsperger, F., Raddi, N., Bacquin, A., Vernochet, C., Mariot, V.,Gache, V.,Blanchard-Gutton, N., Charrin, S., Tiret, L., Dumonceaux, J., Dupressoir, A., & Heidmann, T. (2016). Genetic Evidence That Captured Retroviral Envelope syncytins Contribute to Myoblast Fusion and Muscle Sexual Dimorphism in Mice. PLoS Genetics, 12(9), 1–18. Rein, A., Mirro, J., Haynes, J. G., Ernst, S. M., & Nagashima, K. (1994). Function of the cytoplasmic domain of a retroviral transmembrane protein: p15E-p2E cleavage activates the membrane fusion capability of the murine leukemia virus Env protein. Journal of Virology, 68(3), 1773–1781. Rethwilm, A. (2010). Molecular biology of foamy viruses. Medical Microbiology and Im- munology, 199(3), 197–207.

213 Reznick, D. N., Mateos, M., & Springer, M. S. (2002). Independent Origins and Rapid Fish Genus Poeciliopsis. Science, 298, 1018–1020. Ribet, D., Dewannieux, M., & Heidmann, T. (2004). An active murine transposon family pair: retrotransposition of "master" MusD copies and ETn trans-mobilization. Genome research, 14(11), 2261–7. Ribet, D., Harper, F., Dupressoir, A., Dewannieux, M., Pierron, G., & Heidmann, T. (2008). An infectious progenitor for the murine IAP retrotransposon: emergence of an intracellular genetic parasite from an ancient retrovirus. Genome research, 18(4), 597–609. Riedel, C., Vasishtan, D., Siebert, C. A., Whittle, C., Lehmann, M. J., Mothes, W., & Grünewald, K. (2017). Native structure of a retroviral envelope protein and its conformational change upon interaction with the target cell. Journal of Structural Biology, 197(2), 172–180. Ristich, V., Liang, S., Zhang, W., Wu, J., & Horuzsko, A. (2005). Tolerization of dendritic cells by HLA-G. European Journal of Immunology, 35(4), 1133–1142. Ritson, A., & Bulmer, J. N. (1989). Isolation and functional studies of granulated lymphocytes in first trimester human decidua. Clinical and experimental immunology, 77(2), 263–8. Roberts, R. M., Green, J. A., & Schulz, L. C. (2016). The evolution of the placenta. Reproduction, 152(5), R179–R189. Roth, M. J., Tanese, N., & Goff, S. P. (1985). Purification and characterization of murine retroviral reverse transcriptase expressed in Escherichia coli. The Journal of biological chemistry, 260(16), 9326–35. Roubelakis, M. G., Martin-Rendon, E., Tsaknakis, G., Stavropoulos, A., & Watt, S. M. (2007). The murine ortholog of the SHP-2 binding molecule, PZR accelerates cell migration on fibronectin and is expressed in early embryo formation. Journal of Cellular Biochemistry, 102(4), 955–969. Rous, P. (1911). A sarcoma of the fowl transmissible by an agent separable from the tumor cells. The Journal of experimental medicine, 13(4), 397–411. Rovnak, J., Casey, R. N., Brewster, C. D., Casey, J. W., & Quackenbush, S. L. (2007). Estab- lishment of productively infected walleye dermal sarcoma explant cells. Journal of General Virology, 88(9), 2583–2589. Rovnak, J., & Quackenbush, S. L. (2010). Walleye dermal sarcoma virus: Molecular biology and oncogenesis. Viruses, 2(9), 1984–1999. Saad, J. S., Miller, J., Tai, J., Kim, A., Ghanam, R. H., & Summers, M. F. (2006). Structural basis for targeting HIV-1 Gag proteins to the plasma membrane for virus assembly. Proceedings of the National Academy of Sciences, 103(30), 11364–9. Sabapatha, A., Gercel-taylor, C., & Taylor, D. D. (2006). Specific isolation of placenta-derived exosomes from the circulation of pregnant women and their immunoregulatory consequences. American Journal of Reproductive Immunology, 56(5-6), 345–355. Saib, A., & Benkirane, M. (2009). Endogenous retroviruses: Thierry Heidmann wins the 2009 Retrovirology prize. Retrovirology, 6, 108. Salomon, C., Torres, M. J., Kobayashi, M., Scholz-Romero, K., Sobrevia, L., Dobierzewska, A., Illanes, S. E., Mitchell, M. D., & Rice, G. E. (2014). A Gestational Profile of Placental Exosomes in Maternal Plasma and Their Effects on Endothelial Cell Migration. PLoS ONE, 9(6), e98667. Samuel, C. A., Allen, W. R., & Steven, D. H. (1974). Studies on the Equine Placenta. Repro- duction, 41(2), 441–445. Santoni, F. A., Guerra, J., & Luban, J. (2012). HERV-H RNA is abundant in human embryonic stem cells and a precise marker for pluripotency. Retrovirology, 9, 1–15.

214 Sauter, D. (2014). Counteraction of the multifunctional restriction factor tetherin. Frontiers in Microbiology, 5, 1–14. Schlafke, S., & Enders, A. C. (1975). Cellular Trophoblast Basis of Interaction Uterus Implan- tation. Biology of Reproduction, 65(January), 41–65. Schlecht-Louf, G., Mangeney, M., El-Garch, H., Lacombe, V., Poulet, H., & Heidmann, T. (2014). A Targeted Mutation within the Feline Leukemia Virus (FeLV) Envelope Protein Immunosuppressive Domain To Improve a Canarypox Virus-Vectored FeLV Vaccine. Journal of Virology, 88(2), 992–1001. Schmitt, K., Heyne, K., Roemer, K., Meese, E., & Mayer, J. (2015). HERV-K(HML-2) Rec and Np9 transcripts not restricted to disease but present in many normal human tissues. Mobile DNA, 6(1), 1–13. Schöneberg, J., Lee, I. H., Iwasa, J. H., & Hurley, J. H. (2016). Reverse-topology membrane scission by the ESCRT proteins. Nature Reviews Molecular Cell Biology, 18(1), 5–17. Schreiber, J., Riethmacher-Sonnenberg, E., Riethmacher, D., Tuerk, E. E., Enderich, J., Bösl, M. R., & Wegner, M. (2000). Placental failure in mice lacking the mammalian homolog of glial cells missing, GCMa. Molecular and cellular biology, 20(7), 2466–74. Schröder, A. R., Shinn, P., Chen, H., Berry, C., Ecker, J. R., & Bushman, F. (2002). HIV-1 integration in the human genome favors active genes and local hotspots. Cell, 110(4), 521–529. Schubert, S. W., Lamoureux, N., Kilian, K., Klein-Hitpass, L., & Hashemolhosseini, S. (2008). Identification of integrin-α4, Rb1, and syncytin A as murine placental target genes of the transcription factor GCMa/Gcm. Journal of Biological Chemistry, 283(9), 5460–5465. Schur, F. K. M., Hagen, W. J. H., Rumlová, M., Ruml, T., Müller, B., Kraüsslich, H. G., & Briggs, J. A. G. (2015). Structure of the immature HIV-1 capsid in intact virus particles at 8.8 Å resolution. Nature, 517(7535), 505–508. Sedlacek, J., Fabry, M., Coward, J. E., Horejsi, M., Strop, P., & Luftig, R. B. (1993). Myeloblas- tosis Associated Virus (MAV) Proteinase Site-Mutated to Be HIV-like Has a Higher Activity and Allows Production of Infectious but Morphologically Altered Virus. Virology, 192(2), 667–672. Sheehy, A., Gaddis, N., Choi, J., & Malim, M. (2002). Isolation of a human gene that inhibits HIV-1 infection and is suppressed by the viral Vif protein. Nature, 418(6898), 646–650. Sheehy, A. M., Gaddis, N. C., & Malim, M. H. (2003). The antiretroviral enzyme APOBEC3G is degraded by the proteasome in response to HIV-1 Vif. Nature Medicine, 9(11), 1404–1407. Shine, J., Czernilofsky, A. P., Friedrich, R., Bishop, J. M., & Goodman, H. M. (1977). Nucleotide sequence at the 5’ terminus of the avian sarcoma virus genome. Proceedings of the National Academy of Sciences, 74(4), 1473–1477. Shinnick, T., Lerner, R., & Sutcliffe, J. G. (1981). Nucleotide Sequence of Moloney Murine Leukemia Virus. Nature, 293, 543–548. Shiroishi, M., Tsumoto, K., Amano, K., Shirakihara, Y., Colonna, M., Braud, V. M., Allan, D. S. J., Makadzange, A., Rowland-Jones, S., Willcox, B., Jones, E. Y., van der Merwe, P. A., Kumagai, I., & Maenaka, K. (2003). Human inhibitory receptors Ig-like transcript 2 (ILT2) and ILT4 compete with CD8 for MHC class I binding and bind preferentially to HLA-G. Proceedings of the National Academy of Sciences, 100(15), 8856–61. Simmons, D. G., Fortier, A. L., & Cross, J. C. (2007). Diverse subtypes and developmental origins of trophoblast giant cells in the mouse placenta. Developmental Biology, 304(2), 567–578. Simmons, D. G., Natale, D. R. C., Begay, V., Hughes, M., Leutz, A., & Cross, J. C. (2008). Early patterning of the chorion leads to the trilaminar trophoblast cell structure in the placental labyrinth. Development, 135(12), 2083–2091.

215 Simmons, G., Young, P., Hanger, J., Jones, K., Clarke, D., Mckee, J., & Meers, J. (2012). Prevalence of koala retrovirus in geographically diverse populations in Australia. Australian Veterinary Journal, 90(10), 404–409. Simpson, R. A., Mayhew, T. M., & Barnes, P. R. (1992). From 13 weeks to term, the trophoblast of human placenta grows by the continuous recruitment of new proliferative units: A study of nuclear number using the disector. Placenta, 13(5), 501–512. Sinha, A., & Johnson, W. E. (2017). Retroviruses of the RDR superinfection interference group: ancient origins and broad host distribution of a promiscuous Env gene. Current Opinion in Virology, 25, 105–112. Søe, K., Andersen, T. L., Hobolt-Pedersen, A. S., Bjerregaard Bolette, B., Larsson, L. I., & Delaissé, J. M. (2011). Involvement of human endogenous retroviral syncytin-1 in human osteoclast fusion. Bone, 48(4), 837–846. Sonigo, P., Barker, C., Hunter, E., & Wain-Hobson, S. (1986). Nucleotide sequence of Mason- Pfizer monkey virus: An immunosuppressive D-type retrovirus. Cell, 45(3), 375–385. Sorge, J., & Hughes, S. H. (1982). Polypurine tract adjacent to the U3 region of the Rous sarcoma virus genome provides a cis-acting function. Journal of Virology, 43(2), 482–8. Spencer, T. E., Mura, M., Gray, C. A., Griebel, P. J., & Palmarini, M. (2003). Receptor usage and fetal expression of ovine endogenous betaretroviruses: implications for coevolution of endogenous and exogenous retroviruses. Journal of Virology, 77(1), 749–753. Stankiewicz, P., & Lupski, J. R. (2002). Genome architecture, rearrangements and genomic disorders. Trends in Genetics, 18(2), 74–82. Strick, R., Ackermann, S., Langbein, M., Swiatek, J., Schubert, S. W., Hashemolhosseini, S., Koscheck, T., Fasching, P. A., Schild, R. L., Beckmann, M. W., & Strissel, P. L. (2007). Prolif- eration and cell-cell fusion of endometrial carcinoma are induced by the human endogenous retroviral Syncytin-1 and regulated by TGF-β. Journal of Molecular Medicine, 85(1), 23–38. Sugimoto, J., Sugimoto, M., Bernstein, H., Jinno, Y., & Schust, D. (2013). A novel human endogenous retroviral protein inhibits cell-cell fusion. Scientific Reports, 3, 1–9. Sutton, M. F. (1960). The Sexual Development of Salpa fusiformis (Cuvier): Part I. Journal of embryology and experimental morphology, 8(3), 268–290. Svarovskaia, E. S., Cheslock, S. S., Zhang, W.-H., Hu, W.-S., & Pathak, V. K. (2003). Retroviral mutation rates and reverse transcriptase fidelity. Frontiers in Bioscience, 8(4), 957. Szpakowski, S., Sun, X., Lage, J. M., Dyer, A., Rubinstein, J., Kowalski, D., Sasaki, C., Costa, J., & Lizardi, P. M. (2009). Loss of epigenetic silencing in tumors preferentially affects primate-specific retroelements. Gene, 448(2), 151–167. Tanaka, M., Robinson, B. A., Chutiraka, K., Geary, C. D., Reed, J. C., & Lingappa, J. R. (2016). Mutations of Conserved Residues in the Major Homology Region Arrest Assembling HIV-1 Gag as a Membrane-Targeted Intermediate Containing Genomic RNA and Cellular Proteins. Journal of Virology, 90(4), 1944–1963. Tarlinton, R. E., Meers, J., & Young, P. R. (2006). Retroviral invasion of the koala genome. Nature, 442(7098), 79–81. Taylor, G. M., Gao, Y., & Sanders, D. A. (2001). Fv-4: identification of the defect in Env and the mechanism of resistance to ecotropic murine leukemia virus. Journal of Virology, 75(22), 11244–8. Taylor, J. (1977). An analysis of the role of tRNA species as primers for the transcription into DNA of RNA tumor virus genomes. Biochimica et Biophysica Acta (BBA) - Reviews on Cancer, 473(1), 57–71. Temin, H. M. (1964). Nature of the provirus of Rous sarcoma. Journal of the National Cancer Institute Monographs, 17, 557–570.

216 Temin, H. M., & Mizutani, S. (1970). RNA-dependent DNA polymerase in virions of Rous sarcoma virus. Nature, 226, 1211–1213. Theze, J., Bezier, A., Periquet, G., Drezen, J.-M., & Herniou, E. A. (2011). Paleozoic origin of insect large dsDNA viruses. Proceedings of the National Academy of Sciences, 108(38), 15931–15935. Thomas, D. C., Voronin, Y. A., Nikolenko, G. N., Chen, J., Hu, W.-S., & Pathak, V. K. (2007). Determination of the ex vivo rates of human immunodeficiency virus type 1 reverse tran- scription by using novel strand-specific amplification analysis. Journal of Virology, 81(9), 4798–807. Thomas, J. H., & Schneider, S. (2011). Coevolution of retroelements and tandem zinc finger genes. Genome Research, 21(11), 1800–1812. Thompson, M. B., & Speake, B. K. (2006). A review of the evolution of viviparity in lizards: structure, function and physiology of the placenta. Journal of Comparative Physiology B, 176(3), 179–189. Thompson, P. J., Macfarlan, T. S., & Lorincz, M. C. (2016). Long Terminal Repeats: From Parasitic Elements to Building Blocks of the Transcriptional Regulatory Repertoire. Molecular Cell, 62(5), 766–776. Tilburgs, T., Roelen, D. L., van der Mast, B. J., de Groot-Swings, G. M., Kleijburg, C., Scher- jon, S. A., & Claas, F. H. (2008). Evidence for a Selective Migration of Fetus-Specific CD4+CD25bright Regulatory T Cells from the Peripheral Blood to the Decidua in Human Pregnancy. The Journal of Immunology, 180(8), 5737–5745. Tolosa, J. M., Schjenken, J. E., Clifton, V. L., Vargas, A., Barbeau, B., Lowry, P., Maiti, K., & Smith, R. (2012). The endogenous retroviral envelope protein syncytin-1 inhibits LPS/PHA- stimulated cytokine responses in human blood and is sorted into placental exosomes. Placenta, 33(11), 933–941. Toufaily, C., Lokossou, A. G., Vargas, A., Rassart, É., & Barbeau, B. (2015). A CRE/AP-1-like motif is essential for induced syncytin-2 expression and fusion in human trophoblast-like model. PLoS ONE, 10(3), 1–23. Toufaily, C., Vargas, A., Lemire, M., Lafond, J., Rassart, É., & Barbeau, B. (2013). MFSD2a, the Syncytin-2 receptor, is important for trophoblast fusion. Placenta, 34(1), 85–88. Trela, M., Nelson, P. N., & Rylance, P. B. (2016). The role of molecular mimicry and other factors in the association of Human Endogenous Retroviruses and autoimmunity. APMIS, 124(1-2), 88–104. Turner, G., Barbulescu, M., Su, M., Jensen-Seaman, M. I., Kidd, K. K., & Lenz, J. (2001). In- sertional polymorphisms of full-length endogenous retroviruses in humans. Current Biology, 11(19), 1531–1535. Ushizawa, K., Takahashi, T., Kaneyama, K., Hosoe, M., & Hashizume, K. (2006). Cloning of the bovine antiapoptotic regulator, BCL2-related protein A1, and its expression in trophoblastic binucleate cells of bovine placenta. Biology of Reproduction, 74(2), 344–51. Van Engelenburg, S. B., Shtengel, G., Sengupta, P., Waki, K., Jarnik, M., Ablan, S. D., Freed, E. O., Hess, H. F., & Lippincott-Schwartz, J. (2014). Distribution of ESCRT Machinery at HIV Assembly Sites Reveals Virus Scaffolding of ESCRT Subunits. Science, 343(6171), 653–656. Vargas, A., Moreau, J., Landry, S., LeBellego, F., Toufaily, C., Rassart, É., Lafond, J., & Barbeau, B. (2009). Syncytin-2 Plays an Important Role in the Fusion of Human Trophoblast Cells. Journal of Molecular Biology, 392(2), 301–318. Vargas, A., Zhou, S., Ethier-Chiasson, M., Flipo, D., Lafond, J., Gilbert, C., & Barbeau, B. (2014). Syncytin proteins incorporated in placenta exosomes are important for cell uptake

217 and show variation in abundance in serum exosomes from patients with preeclampsia. FASEB Journal, 28(8), 3703–3719. Vaughan, O. R., & Fowden, A. L. (2016). Placental metabolism: substrate requirements and the response to stress. Reproduction in Domestic Animals, 51, 25–35. Venables, P. J. W., Brookes, S. M., Griffith, D., Weiss, R., & Boyd, M. T. (1995). Abudnance of an endogenous retroviral envelope protein trophoblasts suggests a biological function. Virology, 211(2), 589–592. Vernochet, C., Heidmann, O., Dupressoir, a., Cornelis, G., Dessen, P., Catzeflis, F., & Heidmann, T. (2011). A syncytin-like endogenous retrovirus envelope gene of the guinea pig specifically expressed in the placenta junctional zone and conserved in Caviomorpha. Placenta, 32(11), 885–92. Vernochet, C., Redelsperger, F., Harper, F., Souquere, S., Catzeflis, F., Pierron, G., Nevo, E., Heidmann, T., & Dupressoir, A. (2014). The captured retroviral envelope syncytin-A and syncytin-B genes are conserved in the Spalacidae together with hemotrichorial placentation. Biology of Reproduction, 91(6), 148. Vieira, S., de Perez, G., & Ramírez-Pinilla, M. P. (2007). Invasive cells in the placentome of Andean populations of Mabuya: an endotheliochorial contribution to the placenta? Anatomical record (Hoboken, N.J. : 2007), 290(12), 1508–18. Villesen, P., Aagaard, L., Wiuf, C., & Pedersen, F. S. (2004). Identification of endogenous retroviral reading frames in the human genome. Retrovirology, 1, 1–13. Vishwanathan, S. A., & Hunter, E. (2008). Importance of the Membrane-Perturbing Properties of the Membrane-Proximal External Region of Human Immunodeficiency Virus Type 1 gp41 to Viral Fusion. Journal of Virology, 82(11), 5118–5126. Vitt, L. J., & Blackburn, D. G. (1983). Reproduction in the lizard Mabuya heathi (Scincidae): a commentary on viviparity in New World Mabuya. Canadian Journal of Zoology, 61, 2798–2806. Vogt, P. K. (2012). Retroviral oncogenes: a historical primer. Nature Reviews Cancer, 12(9), 639–648. von Chamisso, A. (1819). De animalibus quibusdam e classe vermium linnaeana in circumnav- igatione terrae auspicante Comite N. Romanzoff, duce Ottone de Kotzebue, annis 1815, 1816, 1817, 1818, peracta : observatis. Berlin, Germany: Dümmler, Ferdinand. von der Helm, K. (1977). Cleavage of Rous sarcoma viral polypeptide precursor into internal structural proteins in vitro involves viral protein p15. Proceedings of the National Academy of Sciences, 74(3), 911–915. Votteler, J., & Sundquist, W. I. (2013). Virus budding and the ESCRT pathway. Cell Host and Microbe, 14(3), 232–241. Wallin, M., Ekström, M., & Garoff, H. (2004). Isomerization of the intersubunit disulphide-bond in Env controls retrovirus fusion. EMBO Journal, 23(1), 54–65. Wang, J., Xie, G., Singh, M., Ghanbarian, A. T., Raskó, T., Szvetnik, A., Cai, H., Besser, D., Prigione, A., Fuchs, N. V., Schumann, G. G., Chen, W., Lorincz, M. C., Ivics, Z., Hurst, L. D., & Izsvák, Z. (2014). Primate-specific endogenous retrovirus-driven transcription defines naive-like stem cells. Nature, 516(7531), 405–409. Wang-Johanning, F., Liu, J., Rycaj, K., Huang, M., Tsai, K., Rosen, D. G., Chen, D. T., Lu, D. W., Barnhart, K. F., & Johanning, G. L. (2007). Expression of multiple human endogenous retrovirus surface envelope proteins in ovarian cancer. International Journal of Cancer, 120(1), 81–90. Wang-Johanning, F., Radvanyi, L., Rycaj, K., Plummer, J. B., Yan, P., Sastry, K. J., Piyathilake, C. J., Hunt, K. K., & Johanning, G. L. (2008). Human endogenous retrovirus K triggers

218 an antigen-specific immune response in breast cancer patients. Cancer Research, 68(14), 5869–5877. Wango, E. O., Wooding, F. B., & Heap, R. B. (1990). The role of trophoblastic binucleate cells in implantation in the goat: a morphological study. Journal of Anatomy, 171, 241–57. Waterston, R. H., Lindblad-Toh, K., Birney, E., Rogers, J., Abril, J. F., Agarwal, P., Agarwala, R., Ainscough, R., Alexandersson, M., An, P., Antonarakis, S. E., Attwood, J., Baertsch, R., Bailey, J., Barlow, K., Beck, S., Berry, E., Birren, B., Bloom, T., Bork, P., Botcherby, M., Bray, N., Brent, M. R., Brown, D. G., Brown, S. D., Bult, C., Burton, J., Butler, J., Campbell, R. D., Carninci, P., Cawley, S., Chiaromonte, F., Chinwalla, A. T., Church, D. M., Clamp, M., Clee, C., Collins, F. S., Cook, L. L., Copley, R. R., Coulson, A., Couronne, O., Cuff, J., Curwen, V., Cutts, T., Daly, M., David, R., Davies, J., Delehaunty, K. D., Deri, J., Dermitzakis, E. T., Dewey, C., Dickens, N. J., Diekhans, M., Dodge, S., Dubchak, I., Dunn, D. M., Eddy, S. R., Elnitski, L., Emes, R. D., Eswara, P., Eyras, E., Felsenfeld, A., Fewell, G. A., Flicek, P., Foley, K., Frankel, W. N., Fulton, L. A., Fulton, R. S., Furey, T. S., Gage, D., Gibbs, R. A., Glusman, G., Gnerre, S., Goldman, N., Goodstadt, L., Grafham, D., Graves, T. A., Green, E. D., Gregory, S., Guigó, R., Guyer, M., Hardison, R. C., Haussler, D., Hayashizaki, Y., LaHillier, D. W., Hinrichs, A., Hlavina, W., Holzer, T., Hsu, F., Hua, A., Hubbard, T., Hunt, A., Jackson, I., Jaffe, D. B., Johnson, L. S., Jones, M., Jones, T. A., Joy, A., Kamal, M., Karlsson, E. K., Karolchik, D., Kasprzyk, A., Kawai, J., Keibler, E., Kells, C., Kent, W. J., Kirby, A., Kolbe, D. L., Korf, I., Kucherlapati, R. S., Kulbokas, E. J., Kulp, D., Landers, T., Leger, J. P., Leonard, S., Letunic, I., Levine, R., Li, J., Li, M., Lloyd, C., Lucas, S., Ma, B., Maglott, D. R., Mardis, E. R., Matthews, L., Mauceli, E., Mayer, J. H., McCarthy, M., McCombie, W. R., McLaren, S., McLay, K., McPherson, J. D., Meldrim, J., Meredith, B., Mesirov, J. P., Miller, W., Miner, T. L., Mongin, E., Montgomery, K. T., Morgan, M., Mott, R., Mullikin, J. C., Muzny, D. M., Nash, W. E., Nelson, J. O., Nhan, M. N., Nicol, R., Ning, Z., Nusbaum, C., O’Connor, M. J., Okazaki, Y., Oliver, K., Overton-Larty, E., Pachter, L., Parra, G., Pepin, K. H., Peterson, J., Pevzner, P., Plumb, R., Pohl, C. S., Poliakov, A., Ponce, T. C., Ponting, C. P., Potter, S., Quail, M., Reymond, A., Roe, B. A., Roskin, K. M., Rubin, E. M., Rust, A. G., Santos, R., Sapojnikov, V., Schultz, B., Schultz, J., Schwartz, M. S., Schwartz, S., Scott, C., Seaman, S., Searle, S., Sharpe, T., Sheridan, A., Shownkeen, R., Sims, S., Singer, J. B., Slater, G., Smit, A., Smith, D. R., Spencer, B., Stabenau, A., Stange-Thomann, N., Sugnet, C., Suyama, M., Tesler, G., Thompson, J., Torrents, D., Trevaskis, E., Tromp, J., Ucla, C., Ureta-Vidal, A., Vinson, J. P., von Niederhausern, A. C., Wade, C. M., Wall, M., Weber, R. J., Weiss, R. B., Wendl, M. C., West, A. P., Wetterstrand, K., Wheeler, R., Whelan, S., Wierzbowski, J., Willey, D., Williams, S., Wilson, R. K., Winter, E., Worley, K. C., Wyman, D., Yang, S., Yang, S. P., Zdobnov, E. M., Zody, M. C., & Lander, E. S. (2002). Initial sequencing and comparative analysis of the mouse genome. Nature, 420(6915), 520–562. Wathes, D. C., & Wooding, F. B. P. (1980). An electron microscopic study of implantation in the cow. American Journal of Anatomy, 159(3), 285–306. Welsh, A. O., & Enders, A. C. (1991). Chorioallantoic placenta formation in the rat: I. Luminal epithelial cell death and extracellular matrix modifications in the mesometrial region of implantation chambers. American Journal of Anatomy, 192(3), 215–231. Whitelaw, E., & Martin, D. I. K. (2001). Retrotransposons as epigenetic mediators of phenotypic variation in mammals. Nature Genetics, 27(4), 361–365. Wildschutte, J. H., Williams, Z. H., Montesion, M., Subramanian, R. P., Kidd, J. M., & Cof- fin, J. M. (2016). Discovery of unfixed endogenous retrovirus insertions in diverse human populations. Proceedings of the National Academy of Sciences, 113(16), E2326–E2334.

219 Wooding, F. B. (1983). Frequency and localization of binucleate cells in the placentomes of ruminants. Placenta, 4 Spec No, 527–39. Wooding, F. B. (1984). Role of binucleate cells in fetomaternal cell fusion at implantation in the sheep. American Journal of Anatomy, 170(2), 233–250. Wooding, F. B., & Wathes, D. C. (1980). Binucleate cell migration in the bovine placentome. Journal of reproduction and Fertility, 59(2), 425–430. Wooding, F. B. P. (1992). The synepitheliochorial placenta of ruminants: Binucleate cell fusions and hormone production. Placenta, 13(2), 101–113. Wooding, F. B. P., Flint, A. P. F., Heap, R. B., & Hobbs, T. (1981). Autoradiographic evidence for migration and fusion of cells in the sheep placenta: Resolution of a problem in placental classification. Cell Biology International Reports, 5(8), 821–827. Wooding, F. B. P., Ramirez-Pinilla, M. P., & Forhead, A. S. (2010). Functional studies of the placenta of the lizard Mabuya sp. (Scincidae) using immunocytochemistry. Placenta, 31(8), 675–85. Wooding, P., & Burton, G. (2008). Comparative Placentation: Structures, Functions and Evo- lution. Berlin, Heidelberg: Springer Berlin Heidelberg. Woodward, C. L., Cheng, S. N., & Jensen, G. J. (2015). Electron Cryotomography Studies of Maturing HIV-1 Particles Reveal the Assembly Pathway of the Viral Core. Journal of Virology, 89(2), 1267–1277. Wootton, S. K., Halbert, C. L., & Miller, A. D. (2005). Sheep retrovirus structural protein induces lung tumours. Nature, 434(7035), 904–907. Wourms, J. P. (1981). Viviparity: The Maternal-Fetal Relationship in Fishes. American Zoolo- gist, 21(2), 473–515. Wourms, J. P. (1993). Maximization of evolutionary trends for placental viviparity in the spadenose shark, Scoliodon taticaudus. Environmental Biology of Fishes, 38(1-3), 269–294. Wu, X., Li, Y., Crise, B., Burgess, S. M., & Munroe, D. J. (2005). Weak Palindromic Consensus Sequences Are a Common Feature Found at the Integration Target Sites of Many Retroviruses Weak Palindromic Consensus Sequences Are a Common Feature Found at the Integration Target Sites of Many Retroviruses. Journal of Virology, 79(8), 5211–5214. Wyma, D. J., Jiang, J., Shi, J., Zhou, J., Lineberger, E., Miller, M. D., Aiken, C., & Lineberger, J. E. (2004). Coupling of Human Immunodeficiency Virus Type 1 Fusion to Virion Maturation : a Novel Role of the gp41 Cytoplasmic Tail Coupling of Human Immunodeficiency Virus Type 1 Fusion to Virion Maturation : a Novel Role of the gp41 Cytoplasmic Tail. Journal of Virology, 78(7), 3429–3435. Wynn, R. M., & Amoroso, E. (1964). Placentation in the Spotted Hyena (Crocuta crocuta), with Particular References to the Circulation. American Journal of Anatomy, 115, 327–362. Xu, J., Sivasubramaniyam, T., Yinon, Y., Tagliaferro, A., Ray, J., Nevo, O., Post, M., & Caniggia, I. (2016). Aberrant TGFβ signaling contributes to altered trophoblast differentiation in preeclampsia. Endocrinology, 157(2), 883–899. Yano, K. (1992). Comments on the Reproductive Mode of the False Cat Shark Pseudotriakis microdon. Copeia, 1992(2), 460. Yin, Z., Shi, K., Banerjee, S., Pandey, K. K., Bera, S., Grandgenett, D. P., & Aihara, H. (2016). Crystal structure of the Rous sarcoma virus intasome. Nature, 530(7590), 362–366. Yoshinaka, Y., & Luftig, R. B. (1977). Properties of a P70 proteolytic factor of murine leukemia viruses. Cell, 12(3), 709–720. Younis, I., & Green, P. L. (2005). The human T-cell leukemia virus Rex protein. Frontiers in bioscience : a journal and virtual library, 10(1), 431–45.

220 Yu, S. F., Sullivan, M. D., & Linial, M. L. (1999). Evidence that the human foamy virus genome is DNA. Journal of Virology, 73(2), 1565–1572. Zhao, R., Guerrah, A., Tang, H., & Zhao, J. Z. (2002). Cell surface glycoprotein PZR is a major mediator of concanavalin A-induced cell signaling. Journal of Biological Chemistry, 277(10), 7882–7888. Zhao, Z. J., & Zhao, R. (1998). Purification and cloning of PZR, a binding protein and putative physiological substrate of tyrosine phosphatase SHP-2. Journal of Biological Chemistry, 273(45), 29367–29372.

221 222 Part V

Résumé Français

223 Identification et caractérisation de deux nouveaux gènes d’enveloppes rétrovirales de type syncytine, capturées pour un possible rôle dans la structure atypique du placenta de hyène et l’émergence du placenta non-mammifère des lézards Mabuya

Introduction. Les syncytines, une famille de gènes d’enveloppe (env) de rétrovirus endogènes (ERV), jouent un rôle clé dans l’établissement des structures placentaires chez les mammifères. Les syncytines canoniques sont des gènes conservés au cours de l’évolution, exprimés dans le placenta et capables de faire fusionner les cellules en interagissant avec leur récepteur spécifique (fusiogénicité). Cette fusiogénicité est une propriété clé des syncytines qui sont responsables de la formation du syncytiotrophoblaste (ST), une couche de cellules placentaires fusionnées formant l’interface materno-fœtale. Leur présence a été démontrée dans de nombreuses familles de mammifères et leur inactivation chez des souris KO conduit à un défaut de formation des syncytia de l’interface materno-fœtale, provoquant un retard de croissance ou la mort des embryons. Les syncytines des différentes familles sont issues d’endogénisations indépendantes, ce qui pourrait contribuer à la diversité des placentas, l’organe le plus variable chez les mammifères. Un des grands critères de variabilité de la structure placentaire est le degré d’invasivité des tissus fœtaux. En effet on peut classer les placentas de mammifères en 3 grandes catégories : épithéliochorial/synépithéliochorial, endothéliochorial et hémochorial, du moins invasif au plus invasif. Chez les espèces hémochoriales, comme l’homme par exemple, l’épithélium utérin maternel est entièrement dégradé suite à l’invasion par le tissu placentaire fœtal, permettant un contact direct du ST fœtal avec le sang maternel, alors que chez les espèces épithéliochoriale comme le cheval, les tissus maternels et fœtaux sont simplement apposés.

Recherche de syncytine chez les lézards placentaires Mabuya. À la suite de l’identification de syncytines dans la plupart des clades de mammifères, nous avons étudié l’implication de ces gènes dans la placentation d’espèces non-mammifères. La placentation n’est pas un phénomène restreint aux mammifères, un certain nombre lézards par exemple présentent un placenta. Parmi ces derniers, le genre Mabuya présente un des placentas non-mammifères les plus complexes connus à ce jour. Dû à sa grande similarité structurelle aux placentas des mammifères et notamment à la présence d’une couche de cellule fusionnées à l’interface materno-fœtale, nous avons cherché à identifier un gène d’env domestiqué, fusiogène et à expression placentaire comme c’est le cas pour les syncytines de mammifère. Comme aucun génome de Mabuya n’a été séquencé, nous avons établi un transcriptome par séquençage à haut-débit d’ARN placentaire de Mabuya sp. IV. Un criblage in silico de ce transcriptome nous a permis d’identifier quatre phases ouvertes de lecture (ORFs) codant pour des env rétrovirales que nous avons appelées Mab-Env1 à 4. Ces nouvelles env peuvent être réparties en deux groupes, Mab-Env1 et 2 présentant de fortes similarités, tout comme Mab-

224 Env3 et 4. Les 4 séquences identifiées présentent tous les domaines classiques d’enveloppes rétrovirales et sont proches des enveloppes de type gamma sur un arbre phylogénétique des sous-unités transmembranaires (TM). Mab-Env3/4 sont proches de la TM du virus du sarcome de Rous (RSV) et présentent des caractéristiques d’enveloppes de type gamma aviaires : un peptide de fusion décalé et encadré par deux cystéines. Mab-Env1/2 présentent une séquence plus particulière qui associe la présence d’un domaine immunosupresseur (ISD) de type gamma et un motif CX7C de type beta. Leur position sur l’arbre phylogénétique est proche des TM de type gamma, suggérant que leur phénotype est principalement gamma. Une investigation plus approfondie de Mab-Env1 a permis l’identification du domaine conservé reverse transcriptase (RT) associé à ce gène et a permis de construire un arbre phylogénétique basé sur ce domaine qui place Mab-Env1 dans une position intermédiaire entre les gammarétrovirus et tous les autre genres orthorétroviraux. Des analyses de RT-qPCR ont montré que bien qu’il ne soit pas exprimé exclusivement dans le placenta, Mab-Env1 présente un niveau d’expression significatif dans ce tissu, alors que Mab-Env2 à 4 sont faiblement exprimées dans tous les tissus de notre panel. Pour étudier le profil d’expression de Mab-Env1 au niveau tissulaire, nous avons réalisé des expériences d’hybridation in situ ainsi que d’immunohistochimie (utilisant des anticorps que nous avons produits en souris). Dans le placentome, Mab-Env1 est exprimé dans les deux tissus formant l’interface materno-fœtale, mais son expression est plus importante dans le tissus maternel syncytialisé que le tissu fœtal, qui est constitué de cellules isolées. Dans le paraplacentome, qui ne présente pas de couche fusionnée, Mab-Env1 est fortement exprimé à l’interface entre les tissus maternels et fœtaux. Ce profil d’expression est en accord avec un potentiel rôle physiologique de Mab-Env1 à l’interface materno-fœtale. Pour étudier la conservation de Mab-Env1 au sein du genre Mabuya, nous avons essayé d’amplifier l’ORF entière ainsi qu’un fragment interne de 400 paires de bases par PCR au sein d’autres espèces de Scincidae. L’amplification de l’ORF entière n’a été possible qu’à partir de génomes de Mabuya, suggérant que Mab-Env1 n’a été conservée qu’au sein de ce clade. L’amplification du fragment interne a également été observée à partir des génomes de quatre espèces proches du genre Mabuya, suggérant qu’un provirus contenant Mab-Env1 a eu lieu, mais que l’enveloppe n’a pas été conservé dans ces espèces. Ayant établi que Mab-Env1 est exprimé dans le placenta et conservé chez toutes les espèces de Mabuya testées, nous avons ensuite analysé son pouvoir fusiogène, la troisième caractéristique des syncytines de mammifères. Des tests de fusiogénicité ex vivo ont montré que des virus MLV pseudotypés avec Mab-Env1 étaient hautement infectieux dans de nombreuses cellules mammifères, avec l’exception des cellules de rongeur. Dans des tests de fusion cellule-cellule, Mab-Env1 n’est fusiogène qu’après un choc acide, comme cela avait déjà été observé dans le cas de Syncytin-Rum1. Le tropisme conféré à des particules virales par Mab-Env1 nous a permis de chercher son récepteur spécifique en utilisant les clones d’hybrides d’irradiation humain/hamster Genbridge4, qui avaient déjà permis l’identification du récepteur spécifique de la Syncytin-2 humaine. Nous avons pu identifier MPZL1 (un transducteur de signal à un domaine transmembranaire) comme

225 récepteur de Mab-Env1. Après clonage du MPZL1 de Mabuya, nous avons pu montrer que son expression à la surface de cellules de rongeur permettait bien leur infection par des particules MLV pseudotypées avec Mab-Env1. L’expression de MPZL1 a été étudiée par RT-qPCR et hybridation in situ. De façon similaire à ce qui avait été observé pour Mab-Env1, MPZL1 montre un niveau d’expression significatif dans le placenta, mais également dans d’autre tissus. L’hybridation in situ montre que MPZL1 est fortement exprimé dans les couches maternelles de l”interface materno-fœtale, à la fois dans la couche fusionnée du placentome et la couche non-fusionnée du paraplacentome. Ce profil d’expression est en accord avec une interaction entre Mab-Env1 et MPZL1 in vivo, les deux étant exprimés soit dans la même couche cellulaire (placentome) ou dans deux couches adjacentes mais distinctes (paraplacentome). Des études précédentes ont montré que chez les mammifère MPZL1 était impliqué dans des phénomènes de migration et d’invasion cellulaires. Pour étudier si l’interaction entre Mab- Env1 et MPZL1 pouvait mener à l’activation de ce dernier, nous avons transduit des cellules non-transformées MCF10A afin qu’elles expriment Mab-Env1. Nous avons pu montrer que l’expression de l’enveloppe était suffisante pour induire la phosphorylation de MPZL1, une étape importante pour la transduction de signal par cette protéine. Pour conclure, Mab-Env1 est exprimée dans le placenta, conservée au cours de l’évolution et fusiogène, elle présente donc les trois caractéristiques des syncytines mammifère et a été nom- mée syncytin-Mab1. Syncytin-Mab1 est clairement différente de toutes les syncytine décrites précédemment et a été domestiquée dans un clade très éloigné de celui des mammifères. Cette domestication d’une enveloppe d’ERV pour un potentiel rôle dans le placenta est un exemple remarquable de convergence évolutive et suggère que ce phénomène pourrait avoir lieu dans de nombreux placentas non-mammifères. De plus, l’identification du récepteur de Syncytin- Mab1 suggère une implication de cette dernière dans l’invasion placentaire, un rôle qui a été précédemment suggéré pour les syncytines mammifères ainsi que les gènes de type syncytine.

Recherche d’env impliquées dans l’établissement du placenta atypique de hyène. Comme tous les autres carnivores, les hyènes expriment la syncytine Syncytin-Car1, mais leur placenta présente une structure très différente de celle de ses congénères. Le placenta des Hyaenidae est un placenta fortement invasif du type hémochorial, les tissus maternels étant entièrement dégradés et les tissus fœtaux baignant dans le sang maternel. Le placenta de tous les autres carnivores est du type endothéliochorial, dans lequel les vaisseaux fœtaux restent intact jusqu’à terme. La transition phénotypique du placenta des hyènes a dû avoir lieux relativement récemment, étant donné que cette famille s’est séparée du reste des carnivores il n’y a que 30 millions d’années. Pour étudier l’implication d’env d’ERV dans cette transition, et comme aucun génome de hyène n’a été publié, nous avons réalisé un séquençage haut débit des ARN placentaire de la hyène tachetée (Crocuta crocuta). Nous avons ensuite criblé ce transcriptome in silico et identifié trois ORFs d’env entières exprimées dans le placenta. Comme attendu, l’une d’elle s’est avéré être la

226 Syncytin-Car1 précédemment décrite. Les deux autres enveloppes, Hyena-Env2 et Hyena-Env3 sont différentes des syncytines décrites précédemment, bien que Hyena-Env3 ressemble à la non-syncytine Canis-Env3 du chien. L’analyse phylogénétique des deux enveloppes les identifie comme des enveloppes de type gamma. Comme la transition de structure placentaire investiguée n’a eu lieu qu’au sein de la famille de Hyaenidae, nous avons étudié l’état de conservation de ces deux nouvelles enveloppes au sein des carnivores. L’amplification par PCR de l’ORF ou d’une partie de l’ORF de Hyena-Env2 n’est possible qu’à partir du génome d’une des quatre espèces de hyène, indiquant que ce gène est bien spécifique des hyènes. De plus, la forte conservation de la séquence codante au sein des Hyaenidae (>98.4% d’identité de nucléotides) ainsi que le ratio de mutation non-synonyme sur synonyme faible suggère que Hyena-Env2 est sous sélection purifiante parmi les hyènes. Par opposition, Hyena-Env3 a pu être montrée comme présente dans de nombreux génome de Caniformia, par PCR ainsi que par analyse in silico. Une comparaison plus approfondie de Hyena-Env3 et Canis-Env3 a montré que cette dernière a perdu son peptide signal, un domaine essentiel pour la synthèse des enveloppes rétrovirale, elle peut donc être considérée comme non-fonctionnelle. Un provirus de chien contenant une copie mutée de Canis-Env3 présente deux LTR identiques à 100%, suggérant que le rétrovirus associé à Hyena-Env3/Canis-Env3 a été actif récemment ou est encore actif. Comme Hyena-Env2 est la seule enveloppe identifiée spécifique aux hyènes, nous l’avons privilégié dans notre étude. Nous avons pu identifier le locus d’insertion du provirus contenant Hyena-Env2 chez Crocuta crocuta et avons réussi à amplifier le provirus entier de 9 kb. La duplication de la cible est apparente des deux côtés du provirus et les LTRs présente une identité de séquence de 89%, en accord avec une intégration relativement récente. Les ORFs de gag et pol sont interrompues pas plusieurs mutations. Le domaine RT conservé associé à Hyena-Env2 a pu être identifié et suggère une origine gammarétrovirale. Des analyses in silico et PCR montrent que ce provirus est bien intégré à ce locus dans toutes les espèces de hyènes et que le locus est vide chez tous les autres carnivores testés. Le séquençage du transcrit placentaire exprimant Hyena-Env2 montre qu’il présente l’épissage canonique des enveloppes rétrovirales, mais qu’un autre épissage a lieu en amont de ce dernier et que l’initiation du transcrit a lieu dans un probable ilot CpG du génome hôte. L’orthologue de cet ilot CpG chez le chat est bien un site d’initiation de transcrits cellulaires. Le profil d’expression placentaire des trois env de hyène a été étudié par hybridation in situ et a montré que Syncytin-Car1 est bien exprimée de façon spécifique dans le ST, comme c’est le cas dans le chat et chien. Hyena-Env3 semble être également exprimée dans cette couche, mais le faible signal, potentiellement du au faible niveau d’expression de cette enveloppe, rend la conclusion difficile. Pour Hyena-Env2, la sonde sens contrôle présente un signal intense, indiquant la présence d’un transcrit antisens dans le placenta de hyène. Néanmoins, la sonde antisens produit un signal spécifique de la couche non-fusionnée sous-jacente au ST. Ce résultat a également été confirmé par des expériences d’immunohistochimie utilisant des anticorps

227 anti-Hyena-Env2 (que nous avons obtenus en inoculant des souris avec des peptides de SU de Hyena-Env2). Une expression spécifique de Hyena-Env2 dans les cytotrophoblastes est en accord avec un rôle placentaire de cette enveloppe. Finalement, nous avons investigué le pouvoir fusiogène des trois enveloppes de hyène. Syncytin-Car1 se comporte comme décrit précédemment, étant capable de rendre des particules MLV pseudotypées infectieuses, avec un fort tropisme carnivore. Hyena-Env2 n’a pas montré de pouvoir fusiogène dans nos expériences. Hyane-Env3 confère un tropisme carnivore strict aux particules pseudotypées. L’analyse de la séquence de la SU de Hyena-Env3 a révélé que cette dernière est proche de celle des SU du groupe d’interférence SLC1A5. Après clonage des SLC1A5 d’humain, de chat, de chien et de hyène, nous avons pu confirmer que les trois dernières rendaient des cellules résistantes vulnérables à l’infection par des virus pseudotypés, mais que ce n’était pas le cas avec SLC1A5 humain. L’utilisation de SLC1A5 comme récepteur est d’ordinaire lié à un tropisme très large et la question se pose de savoir pourquoi ce n’est pas le cas pour Hyena-Env3. Un alignement des séquences de SU du groupe d’interférence et de la SU de Hyena-Env3 montre que plusieurs domaines conservés au sein de ce groupe sont divergent chez Hyena-Env3, présentant souvent des insertions d’un ou deux acides aminés. Pour conclure, nous avons identifié un nouveau gène de type syncytine capturé et conservé de façon spécifique au sein de la famille des Hyaenidae. Le profil d’expression de Hyena-Env2 à l’interface materno-fœtale est en accord avec un potentiel rôle dans la transition d’un placenta endothélio- à un placenta hémochorial, qui a eu lieu de façon simultanée avec la capture de ce gène. Hyena-Env2 n’est pas fusiogène dans notre test, mais le rôle placentaire de cette enveloppe n’est pas nécessairement la fusion cellulaire, cette fonction étant déjà assuré par la Syncytin- Car1 qui semble se comporter de façon identique que chez les autres carnivores, n’étant donc probablement pas responsable de la transition structurelle du placenta d’hyène. Il est possible que Hyena-Env2 soit impliquée dans des phénomènes d’invasion ou de migration cellulaire et l’identification du récepteur spécifique de cette enveloppe pourrait mener à l’identification d’un potentiel mécanisme, comme cela a été le cas pour Syncytin-Mab1 et MPZL1. Pour le moment, Hyena-Env2 rejoint le rang des autres gènes de type syncytine non-fusiogènes mais conservés décrits précédemment. Nous avons également pu identifier une nouvelle enveloppe appartenant au groupe d’interférence SLC1A5, bien qu’elle présente un tropisme particulier qui n’est pas observé pour les autres membres de ce groupe. Les variations de séquence observées pour Hynea-Env3 dans les domaines conservés du groupe SLC1A5 pourraient fournir des indications concernant les interactions enveloppe/récepteur au sein de ce groupe.

Conclusion. Les nouvelles env caractérisées dans la cadre de ce travail renforcent l’hypothèse d’un rôle majeur des gènes syncytine et de type syncytine dans l’apparition et l’organisation structurelle des placentas, que ce soit au sein ou en dehors des mammifères. L’implication de Syncytin-Mab1 dans la convergence évolutive observée entre les placentas de différents clades de vertébrés et l’implication potentielle de Hyena-Env2 dans la diversification des placentas mam-

228 mifères suggère que la capture d’env est une force évolutive majeure dans ces deux processus. La description d’un nouveau gène de type non-fusiogène conservé au cours de l’évolution renforce la notion que le rôle des env capturées n’est pas uniquement fusiogène, mais qu’elles peuvent également être sélectionnées pour d’autres processus importants dans le cadre de la placentation, notamment l’invasion et la migration cellulaire. En plus de son rôle fusiogène, Syncytin-Mab1 pourrait également être impliquée dans ces processus, notamment par interaction avec son récep- teur MPZL1 que nous avons pu identifier, et qui a été impliqué précédemment dans la régulation de la migration et l’invasion cellulaire. Les récepteurs de la plupart des syncytines ne sont pas encore connus et leur identification pourrait, comme cela a été le cas ici, être une façon d’étudier un éventuel rôle non-fusiogène de ces gènes placentaires. En plus d’être la première syncytine non-mammifère décrite à ce jour, Syncytin-Mab1 est potentiellement la première syncytine fon- datrice identifiée. L’hypothétique syncytine associée à l’émergence du placenta des mammifères est probablement trop dénaturée pour être identifiée, cette émergence remontant à environ 175 millions d’années. L’émergence bien plus récente du placenta de Mabuya (25 millions d’années) nous a permis d’identifier une syncytine capturée de façon simultanée à cette émergence et qui pourrait donc avoir contribué directement à l’apparition du placenta dans ce clade particulier de lézards.

229

Titre : Identification et caractérisation de deux nouveaux gènes d'enveloppes rétrovirales de type syncytine, capturées pour un possible rôle dans la structure atypique du placenta de hyène et l'émergence du placenta non-mammifère des lézards Mabuya

Mots clés : rétrovirus endogène, syncytine, protéine d'enveloppe, placenta, récepteur

Résumé : Les syncytines sont des gènes d'enveloppes Ensuite, nous avons cherché des gènes syncytine dans le rétrovirales (env) capturés qui sont essentiels pour genre non-mammifère Mabuya, des lézards vivipares l'établissement du placenta chez les mammifères. Il a été présentant un type rare de placenta très complexe et proche proposé que la diversité des syncytines capturées explique de celui des mammifères. Nous avons identifié une env qui pourquoi le placenta est l'organe le plus variable chez les a été capturée et conservée dans ce genre depuis sa mammifères. Ici nous avons employé deux approches pour radiation, il y a 25 millions d'années. Ce gène, que nous étudier le lien entre la capture d'env et l'émergence et la avons appelé syncytin-Mab1, est capable d'induire la diversité des structures placentaires. D'abord, nous avons fusion cellule-cellule et est exprimé dans une couche de étudié la placentation des Hyaenidae, les seuls carnivores à cellules fusionnées à l'interface materno-fœtale du présenter un placenta très invasif hémochorial, comme placenta, deux propriétés canoniques de syncytine. Nous l'humain. Comme tous les carnivores, les hyènes avons aussi identifié le récepteur de syncytin-Mab1, expriment la syncytin-Car1 précédemment décrite, mais MPZL1, et avons montré que leur interaction induit son nous avons identifié une nouvelle env, capturée activation et sa phosphorylation. L'activation de MPZL1 a uniquement chez ces dernières, que nous avons nommée été liée à la migration et à l'invasion cellulaire, indiquant Hyena-Env2. Ce nouveau gène est présent au même locus que cette interaction env-récepteur pourrait jouer un rôle chez toutes les hyènes, ayant été capturé pendant la dans l'invasion placentaire du tissu maternel observée chez radiation de la famille. Il est non-fusiogène mais a les Mabuya. Pour conclure, la caractérisation de ces deux néanmoins été conservé pendant plus de 10 millions nouvelles env indique que les gènes de type syncytine ont d'années et est exprimé à l'interface materno-fœtale du pu jouer un rôle à la fois dans l'émergence du placenta de placenta, ce qui en fait un gène candidat pour expliquer le Mabuya et dans la structure atypique du placenta des passage à la placentation hémochoriale qui a eu lieu chez hyènes, supportant la notion que la capture d'env est une les Hyaenidae. force évolutive majeure.

Title: Identification and characterization of two novel syncytin-like retroviral envelope genes, captured for a possible role in the atypical structure of the hyena placenta and in the emergence of the non-mammalian Mabuya lizard placenta

Keywords : endogenous retrovirus, syncytin, envelope protein, placenta, receptor

Abstract: Syncytins are captured retroviral envelope Second, we searched for syncytin-like genes in the non- genes (env) that are essential for the establishment of mammalian Mabuya lizards, which are viviparous and placental structures in mammals. The syncytins present in present a rare type of highly complex placenta that is very different mammalian families are highly diverse, resulting reminiscent of mammalian placentas. We identified an from distinct capture events, and it has been suggested env gene that was captured and conserved in this genus that this might play a role in making the placenta the most since its radiation 25 million years ago. This gene, that we diverse structure in mammals. Here we used two different named syncytin-Mab1, is able to mediate cell-cell fusion approaches to investigate the links between env capture in vitro and is expressed in a fused cell layer at the and emergence and diversity of placental structures. First, materno-fetal interface of the placenta in vivo, we investigated placentation in Hyaenidae, the only characteristic features of canonical mammalian syncytin carnivorans that present a highly invasive hemochorial genes. We also identified the cellular gene MPZL1 as the placenta, as is also found in humans. Hyenas express the cognate receptor of syncytin-Mab1 and showed that their previously identified syncytin-Car1 gene, as do all interaction induces activation and phosphorylation of the carnivorans, but we identified a new hyena-specific former. MPZL1 activation has been linked with cell captured env that we named Hyena-Env2. This new gene migration and invasion, indicating that this env-receptor is present at the same locus in all hyenas, having been interaction could play a role in the placental invasion of captured during the radiation of this family. It is non- maternal tissues observed in Mabuya. In conclusion, the fusiogenic but still conserved over at least 10 million characterization of these two novel env genes indicates years of evolution and expressed at the materno-fetal that syncytin-like env might have played a role both in the interface in the hyena placenta, making it a candidate emergence of the Mabuya placenta and the atypical gene for explaining the endotheliochorial to hemochorial placental structure of hyenas, reinforcing the notion that placental transition that occurred in Hyeanidae. env capture is a major driving force in evolution.

Université Paris-Saclay Espace Technologique / Immeuble Discovery Route de l’Orme aux Merisiers RD 128 / 91190 Saint-Aubin, France