Molecular Evolution and in Silico Analysis of the Mitochondrial Genome of the Hermaphroditic Swan Mussel, Anodonta Cygnea (Bivalvia: Unionoida)

Molecular evolution and in silico analysis of the mitochondrial genome of the hermaphroditic swan mussel, Anodonta cygnea (Bivalvia: Unionoida)

Emily Erin Chase

B.Sc. Honours Acadia University 2015

Thesis submitted in partial fulfillment of the requirements for the Degree of Master of Science (Biology)

Acadia University Fall Graduation 2017

This thesis by Emily Erin Chase was defended successfully in an oral examination on 22 September 2017

The examining committee for the thesis was:

______Dr. Michael Robertson, Chair

______Dr. T. Rawlings, External Examiner

______Dr. M. Coombs, Internal Examiner

______Dr. D.T. Stewart, Co-Supervisor

______Dr. S. Breton, Co-Supervisor

______Dr. Brian Wilson, Head/Director

This thesis is accepted in its present form by the Division of Research and Graduate Studies as satisfying the thesis requirements for the degree Master of Science (Biology)

………………………………………….

I, Emily Erin Chase, grant permission to the University Librarian at Acadia University to archive, preserve, reproduce, loan or distribute copies of my thesis in microform, paper, or electronic formats on a non-profit basis. I undertake to submit my thesis, through my University, to Library and Archives Canada and to allow them to archive, preserve, reproduce, convert into any format, and to make available in print or online to the public for non-profit purposes. I, however, retain the copyright in my thesis.

______Author

______Supervisor

______Date

iii

TABLE OF CONTENTS

Title i Approval of thesis ii ii Library release form iii iii Table of contents iv List of tables vi List of figures vii Abstract viii Abbreviations and symbols ix Acknowledgements xiii Preface xiiv

Chapter 1: Mitochondrial DNA, doubly uniparental 1 inheritance and freshwater mussels: A literature review Overview 1 Freshwater mussels and study species: Anodonta cygnea 5 The conservation status of freshwater mussels 8 Hermaphroditism and sexual strategies of freshwater mussels 10 Doubly uniparental inheritance in freshwater mussels 17 Mitochondrial DNA and evolutionary relationships among 24 FWMs Trends in the analysis of complete genomes of freshwater 25 mussels Summary and objectives 26 Tables 28 References 29

iv Chapter 2: The complete H-type mitochondrial genome of 49 freshwater mussel Anodonta cygnea (Bivalvia: Unionidae) Abstract 49 Introduction 50 Materials and Methods 56 Results 60 Discussion 68 Figures 79 Tables 85 References 89

Chapter 3: Relative evolutionary rates of sex-associated 101 mtDNA genomes in the order Unionoida (Bivalvia): A comparison among male-, female- and hermaphrodite- transferred mitochondrial genomes Abstract 101 Introduction 102 Materials and Methods 106 Results and Discussion 111 Figures 123 Tables 126 References 131

LIST OF TABLES

Table Abbreviated Caption Page Chapter 1 1 Standard genes found in mtDNA 28 2 Shared unassigned regions of mtDNA 28 Chapter 2 1 Species specific primers used to amplify 85 Anodonta cygnea mitochondrial genomes 2 Hermaphroditic (H-type) mtDNA shared 85 unassigned regions S1 GenBank accessible complete or near 86 complete mtDNA of order Unionoida S2 GenBank accessible cox1 sequences used 87 S3 GenBank accessible F-, H- and M-ORFs 88 used Chapter 3 1 Cox1 nucleotide divergence of F-, H- and 126 M-type groups 2 Overall cox1 nucleotide divergence of F-, 126 H- and M-type mitochondrial DNA 3 Tajima’s D nucleotide relative rates test 127 of cox1 sequences

vi 4 Number of synonymous and 127 nonsynonymous substitutions between cox1 of F-, H- and M-type groups 5 Number of synonymous and 127 nonsynonymous substitutions between all protein coding genes of F-, H- and M- type groups 6 Purifying selection tests of cox1 within F- 128 and H-type mtDNA 7 Positive selection tests of cox1 within F- 128 and H-type mtDNA S1 GenBank accessible cox1 sequences used 129 S2 The ratio of nonsynonymous to 130 synonymous substitutions in F- and H- type mtDNA S3 Neutral selection tests of cox1 within F- 130 and H-type mtDNA

LIST OF FIGURES

Figure Abbreviated Caption Page Chapter 2 1 Gene order of Anodonta cygnea 79 mitochondrial DNA 2 Nucleotide and amino acid divergence of 80 Anodonta cygnea mitochondrial DNA 3 Example potential secondary structures in 81 shared unassigned region trnF-nad5 of Anodonta cygnea

vii 4 Amino acid properties of H-ORFs 81 5 Bayesian Inference tree of F- and H-type 82 cox1 with transmembrane descriptions of F- and H-ORFs 6 Bayesian Inference tree of M-type cox1 83 with transmembrane descriptions of M- ORFs 7 NAD5 protein alignment of three bivalve 83 species with translocated Anodonta cygnea partial NAD5 annotated S1 Predicted tRNA structures of H-type 84 mitochondrial DNA of Anodonta cygnea Chapter 3 1 Relative rates analysis group set up 123 methodology 2 Unionid cox1 maximum likelihood tree 124 3 Unionid cox1 maximum likelihood 125 unrooted trees of groups

ABSTRACT

viii Doubly uniparental inheritance is a fascinating phenomenon that still holds many mysteries. Two major components currently being explored are (1) novel proteins encoded by the mitochondrial DNA of male and female dioecious species (M- and F- type respectively) and of hermaphroditic species (H-type), and (2) the generally fast rate of evolution of these types of mitochondrial DNA and their different relative rates.

These two components are explored herein and prefaced with a thorough literature review of freshwater mussels (order Unionoida), which includes the study species,

Anodonta cygnea. Complete mitochondrial DNA of A. cygnea is sequenced and an in silico analysis is conducted on the subject and novel open reading frames (ORFs) of other members of order Unionoida. Relative rates of cox1 mitochondrial DNA within each mitotype are also explored. The complete mitochondrial DNA of A. cygnea consists of 13 protein coding genes, 2 rRNA, and 22 tRNA, and is 15,607 base pairs long, and contains a translocated portion of nad5. Based on in silico analysis, we conclude that the number and topology of transmembrane domains in F-ORFs are maintained across species, and that these contain signal peptides with corresponding signal cleavage sites. We suggest that H-type mitochondrial DNA is evolving at a slower rate than the F-type and discuss these results in the context of the arenas hypothesis.

ABBREVIATIONS AND SYMBOLS

ix ATP Adenosine atp6, ATP synthase

triphosphate ATP6 subunit 8 atp8, ATP synthase BI Bayesian

ATP8 subunit 8 Inference bp Base pairs C-TM-E Cellular-

transmembrane

domain-

extracellular

topology cox,1 Cytochrome cox2, Cytochrome

COX1 oxidase subunit I COX2 oxidase subunit II cox3, Cytochrome Cytb, Cytochrome b

COX3 oxidase subunit cob,

III COB

DNA Deoxyribonucleic DUI Doubly

acid uniparental

inheritance

E- Extracellular- f-orf, F- Female specific

TM-C transmembrane ORF open reading

domain-celluar frame

topology

x F-type Female mitotype FWM Freshwater

mussel h-orf, Hermaphroditic H-type Hermaphroditic

H- specific open mitotype

ORF reading frame

ITS Internal IUB International

transcribed Union of

spacer Biochemistry

LR- Long range LUR Longest

PCR polymerase chain unassigned region

reaction

m-orf, Male specific ML Maximum

M- open reading Likelihood

ORF frame

M- Male mitotype mtDNA Mitochondrial type deoxyribonucleic

acid mt Mitochondrial n non-synonymous

substitutions nd difference in Nd total difference in

number of non- number of non-

xi synonymous synonymous

substitutions substitutions nad2, NADH nad,1 NADH

NAD2 dehydrogenase NAD1 dehydrogenase

subunit 2 subunit 1 nad4, NADH nad3, NADH

NAD4 dehydrogenase NAD3 dehydrogenase

subunit 4 subunit 3 nad,5 NADH nad4L, NADH

NAD5 dehydrogenase NAD4L dehydrogenase

subunit 5 subunit 4L

ORF Open reading nad6, NADH

frame NAD6 dehydrogenase

subunit 6 rRNA Ribosomal ROS Reactive oxygen

ribonucleic acid species rrnS Small ribosomal rrnL Large ribosomal

subunit subunit ribosomal

ribosomal ribonucleic acid

ribonucleic acid s synonymous sd difference in

substitutions number of

xii synonymous

substitutions sd total difference in SMI Strictly maternal

number of inheritance

synonymous

substitutions

SCS Signal cleavage SUR Shared

site unassigned region

SP Signal peptide tRNA Transfer

ribonucleic acid

TM Transmembrane trnC Cysteine

domain trnA Alanine trnE Glutamic acid trnD Aspartic acid trnG Glycine trnF Phenylalanine trnI Isoleucine trnH Histidine trnL2 Leucine 2 trnL1 Leucine 1 trnN Asparagine trnM Methionine trnQ Glutamine trnP Proline trnS1 Serine 1 trnR Arginine trnT Threonine trnS2 Serine 2 trnW Trytophan trnV Valine trnY Tyrosine

xiii ACKNOWLEDGEMENTS

I would like to thank the following people and groups for their support, contribution, encouragement and effort towards this project and my experience at Acadia University:

Brent Robicheau

Dr. Anna Redden

Dr. Donald Stewart

Dr. David MacKinnon

Dr. Sophie Breton

Krista Mills

NSERC

Research and Graduate Studies Acadia University

Sarah Veinot

The Acadia Biology Department

xiv PREFACE

Chapters 2 and 3 of this thesis were written with the intent of being published in peer reviewed journals, thus, I have chosen to retain the more inclusive pronouns “we” and

“our” for those chapters. Co-authors on these publications are specified at the beginning of these chapters. Chapters 2 and 3 also include a brief introduction, which will have overlap with the literature review in Chapter 1.

Chapter 1

Mitochondrial DNA, doubly uniparental inheritance and freshwater mussels: A literature review

Chase E.E.1

1 Department of Biology, Acadia University, Wolfville, NS, Canada

OVERVIEW

Mitochondria (mt) are eukaryotic cell organelles whose primarily responsibility is energy metabolism via oxidative phosphorylation (Brand 1997). However, mitochondria are also involved in programmed cell death (apoptosis) (Kroemer et al. 1998), factor into various diseases (Graeber and Müller 1998) and are a key component in the process of aging (Wei 1998; Salvioli et al. 2001; Wei and Lee 2002; Balaban et al. 2005; Lee and Wei 2007; Weber and Reichert 2010; Cui et al. 2012). They also account for up to a quarter of cellular space or more (Ballard and Whitlock 2004). These factors make them an important part of the eukaryotic cell. Early observations by Altman (1890 cited in

Ernster and Schatz 1981) described mitochondria as an organism inside an organism and named them “bioblasts”. These observations would be addressed once again by

Margulis in 1981 through the Serial Endosymbiosis Theory (reviewed in Margulis

2004). This theory is now widely accepted to explain the origin of mitochondria in eukaryotic cells from a shared ancestor of alphaproteobacteria (Gray et al. 1999).

Mitochondrial DNA (mtDNA) was first described independently in the 1960s (Nass and

Nass 1963a; 1963b; Schatz et al. 1964). Eventually it would be established that animal

mtDNA typically houses thirty-seven genes: thirteen protein coding genes (PCG) related to ATP production, twenty-two transport RNA (tRNAs) and two ribosomal RNA

(rRNA). The study of mtDNA has played a key role in the formation of modern population genetics and phylogenetics (e.g., Brown et al. 1979; Wilson et al. 1985;

Avise et al. 1986;1987; Moritz et al. 1987; Avise et al. 1991; Rand 1994, 2001).

MtDNA sequences continue to provide genetic data used to test and identify species relationships (Saccone et al. 1999).

MtDNA, in general, has a high rate of evolution, although not all mitochondrial genes evolve at the same rate (Upholt and Dawid 1977; Brown et al. 1979; Rand et al. 2001).

Among the thirteen genes associated with ATP production, cytochrome c oxidase I

(cox1) is used as the molecular marker of choice for the metazoan Barcode of Life project (Stoeckle and Hebert 2008). Herbert et al. (2003a, 2003b) concluded that cox1 gene sequences reflect a relatively high rate of molecular evolution that can differentiate closely related species. Generally speaking, among mammals a divergence of >2% are considered separate species, and >3% for lepidopterans (Krishnamurthy and Francis).

Hebert et al. (2003b) demonstrated that 98% of species in the same genus showed more than 2% sequence divergence of cox1 among 13,320 total congeneric pairs tested, which included the following phyla: Annelida, Arthropoda, Chordata, Cnidaria,

Echinodermata, Mollusca, Nematoda and Platyhelminthes. Another benefit of using cox1, from a research perspective, is that amplification by cox1 universal primers can be successfully applied to most metazoan phyla. Consequently, Hebert et al. (2003a) proposed using cox1 as the foundation of a molecular systematic approach to species

identification. Although not without its challenges for some taxa (see Waugh 2007), cox1 barcoding is generally considered to be a relatively reliable method of inferring species evolutionary relationships and to be a useful tool for a wide range of types of biological enquiry.

MtDNA is transmitted generation to generation as is nuclear DNA, however, most metazoan mtDNA inheritance occurs only from mothers to their offspring. Thus, mtDNA transmission is referred to as an “uniparental mode of inheritance”, and also referred to as “strictly maternal inheritance” (SMI) in animals. This is because inheritance of mtDNA is exclusively from the oocyte portion of the developing embryo and typically no paternal mtDNA is passed on (Birky 1995; Sato and Sato 2013). An interesting exception to the rules of SMI is exhibited by many members of Bivalvia, including species of the orders Mytiloida, Veneroida, Nuculanoida, and Unionoida

(Breton et al. 2007; Theologidis et al. 2008; Bolye and Etter 2013). Order Unionoida, the freshwater mussels (FWMs) is the focus of this review and thesis research. This exceptional system of mtDNA transmission, named “doubly uniparental inheritance”

(DUI), is a mode of inheritance where both a so-called “male type” (M-type) and

“female type” (F-type) mtDNA genome are passed from parents to offspring (reviewed by Breton et al. 2007; Passamonti and Ghiselli 2009; Zouros 2000, 2013). Generally, adult bivalve somatic tissue will primarily consist of F-type mtDNA, whereas the dominant mtDNA in germ cells corresponds to the sex of the organism (Garrido-Ramos et al. 1998). Sperm cells contain almost exclusively M-type mtDNA and egg cells F- type mtDNA. Another interesting aspect of mtDNA inheritance associated with FWMs

that exhibit DUI is as follows. Most FWM species are dioecious, however, a number of hermaphroditic forms occur (van der Schalie 1970; Breton et al. 2011). The focus of this thesis is primarily, but not exclusively, concerned with obligate hermaphrodites (Avise

2011). In the case of the FWMs that will be examined herein, all of these species (or at least the populations of these species that have been the focus of recent molecular studies) have apparently lost the M-type mtDNA. Typically, male and female freshwater mussels possess M-type and F-type genomes that contain novel open reading frames (ORFs), termed the m-orf and f-orf genes in the M-type and F-type genomes, respectively. These orfs are unique to species exhibiting DUI. Hermaphrodites studied to date contain a highly modified f-orf gene called an h-orf (Breton et al. 2011). This h- orf, and the difference between an f-orf and an h-orf, may hold important answers to understanding both DUI in bivalve species and also SMI in metazoans in general. This literature review will further examine the natural history and reproductive biology of

FWMs and will briefly discuss some of the threats that are currently faced by FWMs.

This review will also discuss the use of mtDNA for inferring evolutionary relationships among FWM to date and summarize some recent findings regarding the mtDNA complete genomes of dioecious and hermaphroditic FWMs. The purpose of this review is to establish a foundation for understanding the research objectives of this thesis which focuses on analysis of the genes and complete mitochondrial genome of a hermaphroditic FWM mussel from the family Unionidae, the swan mussel, Anodonta cygnea, Linnaeus 1758.

LITERATURE REVIEW

Freshwater mussels and the study species; Anodonta cygnea

FWMs persist on all continents except Antarctica (Graf and Cummings 2007). As their name indicates, they inhabit lakes and river and appear in the fossil record as early as the Triassic (Bogan 1993). The genus Anodonta (family Unionidae) includes many dioecious species and also the hermaphroditic species that is the focus of this thesis, A. cygnea. Anodonta anatina has been reported to be a sister taxa to A. cygnea, however, internal transcribed spacer (ITS) analysis has suggested that Pseudanodonta complanata may be more closely related to A. cygnea (Källersjö et al. 2005). However, the complete mtDNA genome has yet to be sequenced for P. complanata, whereas A. anatina has several complete mitochondrial genomes available on GenBank, including both paternal

(Soroka and Burzyński 2016) and maternal (Soroka and Burzyński 2015) mitotypes.

The availability of complete mtDNA permits a comparison between relatively closely related hermaphroditic and dioecious species of genus Anodonta using A. anatina and A. cygnea (Nagel et al. 1996). An in depth comparative analysis of these genomes is the focus of Chapter 2.

FWMs exhibit a parasitic larval stage (Wood 1974) termed glochidia (Bauer and

Wächtler 2001). Their larval stages are non-specific or specific (depending on species) ectoparasites of fish, which typically attach to gills after being expelled from their female parent for dispersal to an appropriate habitat (Bogan 1993). Most host species are teleost fishes, and FWM species can possess generalist or specialist glochidia. For example, A. anatina has glochidia that can parasitise an array of freshwater fish hosts

and is considered to be a generalist (Anders and Wiese 1993). Bauer and Wächtler

(2001) stated that, generally, larger glochidia with larger hooks are generalists and that smaller types without hooks are more selective in their hosts. Glochidia were originally described in 1797 (Rathke) as bivalve parasites before being recognized as a larval stage

(Leydig 1866; Carus 1932). It is speculated that, in contrast to the planktonic stage of marine mussels, a parasitic larval stage reduces the possibility of larvae being dispersed into an uninhabitable environment within large river systems, alongside the added protection from predators during early development (Wood 1974). Fertilized eggs develop into glochidia, which are stored in structures in the mussel called marsupia

(Bauer 2001). The number of glochidia found in FWMs ranges from an estimated 9000-

16,000 in Unio crassus (Bednarczuk 1986; Engel 1990; Engel and Wächtler 1990;

Hochwald 1997) to 167 to 200 million in Sinanodonta woodiana (Bauer and Wächtler

2001). A. cygnea marsupia are estimated to contain between 310,000 to 370,000 glochidia, each of which is approximately 310 µm in diameter (Claes 1987; Niemeyer

1992). Glochidia persist in their host’s gills for between 3 days to 10 months (Seshaiya

1969). Glochidia of A. cygnea were described by Wood (1973) as possessing hooks, modified cilia on the mantle cells, and a thread (hypothesized to be for attachment to hosts). In A. cygnea, the glochidia typically mature by October, but are not released until May.

Early studies of FWM taxonomy and identification of the genus Anodonta and close relatives concentrated heavily on shell morphological features (conchology) and other physical characteristics of the shell and anatomy as well as on the substrate and depths

that the species were found at within lakes and rivers (e.g., Utterback 1916; Allen 1921;

Matteson 1955 as cited by Haag 2012). These studies played an important role in the progress of FWM research, but often did not consider significant intra-species variation in shell morphology linked with environmental conditions that occurs within FWMs

(Ball 1922). This practice resulted in incorrect species identification according to Shea et al. (2011). Furthermore, clear genetic differences have been demonstrated in species with extremely similar shell morphologies most likely due to environmental pressures and where multiple conspecific forms occur (Ortmann 1920; Ball 1922; Hoeh et al.

1995; Bailey and Green 1988; Hinch and Bailey 1988; Roper 1994; Hornbach et al.

2010), making species identification based on morphology alone highly questionable.

The subject of this thesis, A. cygnea, is found throughout Europe. It is a prime example of the historical challenges of FWM taxonomy as it has approximately 549 synonyms

(Graf and Cummings 2013)! It is also an excellent example of the challenges of describing life history strategies and reproductive patterns that occurs in FWMs, as A. cygnea has mistakenly been cited as an example of a species that exhibits sex protandry

(i.e., changing sex from male to female). This was thought to be the case for A. cygnea in a study by Bloomer (1934) but this was later corrected by several authors (e.g., Heard

1975; Smith 1979; Haag and Staton 2003). A. cygnea tissue samples used for research in this thesis were collected from Lake Konstanz, along the border of Germany and

Switzerland, a population that has been characterized by Gerhard Bauer, an expert on freshwater mussels and their reproductive strategies (see, for example, the excellent

book edited by Gerhard Bauer and Klaus Wächtler, Ecology and Evolution of the

Freshwater Mussels Unionoida).

The conservation status of freshwater mussels

Although many species of freshwater mussels are now endangered worldwide, their population declines may be most drastic and dramatic in North America, leading David

Strayer and colleagues to declare freshwater mussels to be the most imperiled group of animals on this continent (Strayer et al. 2004a). Substantial changes to North American freshwaters began at the time of European colonization (Haag 2012). These changes included construction of dams (Graf 1999), discharging of wastes into freshwater

(Trautman 1981; Anfinson 2003), clearing of erodible soils that caused fish kills

(Willoughby 1999), and habitat degradation caused by forest clearing and wetland drainage (Trautman 1981). All of the effects of European colonization in North America on FWM populations are not explicitly known (Haag 2012), however, the historical factors described above are being compounded by more modern documented events, including various agricultural practices, which are having a negative impact on FWM populations (e.g., Higgens 1858; Lewis 1868; Kunz 1898; Smith 1899; Stansbery 1970;

Tudorancea 1972; Watters 1996; Williams and Fradkin 1999). Anthropogenic influences on mussel populations increased with the Industrial Revolution (Haag 2012) as they were frequently locally extirpated due to industrial wastes (e.g., Rhoads 1899;

Simpson 1899; Smith 1899; Ortmann 1918; Remington and Clench 1925), particularly mining wastes (Ortmann 1909; Wilson and Clark 1914). Modern sewer systems also

began contributing to the decline of FWMs in the late 1800s as a consequence of large volumes of raw sewage being released into freshwater environments (Haag 2012).

Within North American, the pearl and button industry contributed significantly to the decline of FWM from 1850s onwards (Haag 2012). By the 1990s about 25 North

American FWM species were considered to have been driven to extinction by these factors (Martin 1997; Haag 2012). Many of these factors persist today, with the addition of other threats such as the introduction of more than 145 invasive species, frequently through ship ballast water (Ricciardi and MacIssac 2000). Two well documented examples are the introduction of Dreissena polymorpha, the zebra mussel and

Dreissena bugensis, the quagga mussel (e.g., Hebert et al. 1989; 1991; Strayer et al.

2004b; Naddafi and Rudstam 2013; Peñarrubia et al. 2016). A separate issue is the decline of FWMs without a clear reason, which is occurring in areas that do not appear to be obviously affected by the factors described above. These seemingly “random” declines may simply be the result of stochastic processes within isolated populations, although this is not yet clear (Haag 2012). These cases of random decline appear to impact most or all taxa included in some areas (Haag 2012). Present day pollution levels

(including pesticides and pharmaceuticals; Reis 2003), the spread of disease and the impact of fisheries on host fish populations have also contributed to the decline of

FWMs. However, the impact of changing fish populations may not be as critical as once believed considering that species with generalist glochidia do not appear to be better off than species with more narrow host fish affinities (Haag 2012). Haag (2012) has suggested that removal of dams and rehabilitation of streams could help remediate the decline of FWMs. In another review on the status of freshwater mussels, Strayer et al.

(2004a) suggest that laboratory propagation of species could be used to supplement natural populations and, not surprisingly, they also identify the need for further research on all aspects of the biology of FWMs to help provide the empirical data needed to develop effective recovery strategies for this entire group.

Hermaphroditism and sexual strategies of freshwater mussels

The sexual strategies of species from the family Unionidae are primarily either (1) dioecy or (2) hermaphroditism (Hinzmann et al. 2013). Although dioecy is by far the more common strategy (hermaphrodites represent less than 3% of North American species, Hoeh et al. 1995), and dioecy is apparently the ancestral system within the

Mollusca and Bivalvia (Collin 2013), hermaphroditism, and the evolutionary path to hermaphroditism, is of particular interest for this literature review. Hermaphrodites in general can be either simultaneous (synchronous; functional; Charnov 1979) or sequential (asynchronous; Lee et al. 2013). Furthermore, a simultaneous hermaphroditic

FWM can be an obligate (normal) hermaphrodite or an occasional (accidental) hermaphrodite. There are several potential benefits to the hermaphroditic sexual strategy (e.g., Dolgov 1992), which will be discussed below. Occasional hermaphrodites appear among populations of typically dioecious FWMs. While such individuals may provide the raw material for selection to act upon, which can facilitate the evolution of a population that will eventually become predominantly hermaphroditic, occasional hermaphrodites are not the focus of this thesis. The focus of this thesis is on simultaneous, obligate hermaphrodites, as appears to be illustrated by the Lake

Konstanz population of A. cygnea (Gerhard Bauer, personal communication; Stewart et al. 2013).

Generally, hermaphroditism, accompanied by a lowering of metabolism, was once thought to be a natural progression that arose from being sessile (Smith 1908). It was also thought to be the ancestral condition of early metazoans, which progressed to

“higher forms” from “lower forms”, in terms of hierarchy (Ghiselin 1969). Ghiselin proposed three possible models of the change from dioecy to hermaphroditism in animals including (1) the low density model, (2) the size advantage model, (3) and the gene dispersal model. Because the low-density model is the most likely to explain hermaphroditism in freshwater mussels, it will be reviewed here.

The low density model suggests that when a species has a low opportunity for encountering members of the opposite sex then hermaphroditism is a potential solution for reproduction. In this case, the environment may favour hermaphroditism, but hermaphroditism must exist in the species before selection of this sexual strategy can occur. A possible example of this occurred within Genus Anodonta (family Unionidae).

Weisensee (1916 cited in Stewart et al. 2013) proposed a hypothetical transformation of populations of Anodonta mussels from dioecy to hermaphroditism, which were originally from river populations and that had become established in isolated ponds.

Using examples associated with the Rhine River in Germany, Weisensee (1916) proposed that these populations had been isolated for upwards of three hundred years and over this time their sex ratio became skewed towards females and finally

hermaphrodites. Simultaneous hermaphroditism may be beneficial to an organism occurring in such a low density population because the chance of locating a mate or having your gametes reach a mate is extremely low (e.g., Heath 1977). The costs of self-fertilization (e.g., inbreeding depression) are outweighed by the benefit of producing any offspring at all (Ghiselin 1969). This may explain the change in some

Anodonta populations over time after their isolation in a relatively still pond environment.

It has been suggested that molluscs employing hermaphroditism as a sexual strategy do so to compensate for reproductive challenges associated with having two sexes (van der

Schalie 1970), e.g., locating a mate or complementary gametes in certain habitats.

Accordingly, the low density model could explain the transition from dioecy to hermaphroditism, which has occurred multiple times within the order Unionoida

(Breton et al. 2011; Stewart et al. 2013; Chase et al. in prep). The low density model can help explain this shift in reproductive mode specifically in isolated, relatively stagnant freshwater environments with a low population density and dispersed individuals.

Isolated sessile individuals in a pond or lake cannot easily access mates, either by movement of the individual or by broadcasting sperm, therefore selfing and hermaphroditism could overcome the reproductive challenges of stagnant freshwater environments.

Given the low density model explanation for hermaphrodites in FWM, we would expect that if a hermaphrodite spontaneously occurred in marine bivalves (with generally

greater access to mates), the population would be less likely to adopt a hermaphroditic sexual strategy and instead remain dioecious because of the advantage of greater fitness of outcrossed progeny. Saavedra et al. (1997) conducted crosses of a marine mussel,

Mytilus galloprovincialis, using male and female individuals, which resulted in hermaphroditic offspring alongside typical male and female offspring. These hermaphrodites contained a male and female “side” of the gonad, and both M- and F- type mtDNA. Despite the occurrence of such occasional hermaphrodites, the species remains dioecious. An assessment of 2313 cultivated M. galloprovincialis individuals from four locations along the Galicia coast, Spain, produced only three cases of hermaphrodites without significant deviations from a 1:1 sex ratio among all four sites

(Villalba 1995). Therefore, although occasional hermaphrodites occur within M. galloprovincialis in a marine environment they do not persist as an evolutionarily stable sexual strategy. Similarly, occasional hermaphrodites occur in other species such as

Perumytilus purpuratus (Villalobos et al. 2010), Perna viridis (Lee 1988) and Donax semigranosus (Wong 1989), but these species remain effectively dioecious. It is important to note that the distinction here is not between marine and freshwater environments, specifically, but between an isolated environment with isolated individuals in contrast to a high density population and an environment that permits greater opportunities for sperm exchange. To further exemplify this, within FWM, A. anatina is considered to be a dioecious species with a minimal frequency of hermaphrodites (Bauer and Wächtler 2001), but recent research has shown that A. anatina is mostly dioecious in moving waters and exhibits some degree of hermaphroditism in stagnant waters (Hinzmann et al. 2013). As will become clear after

reading the remainder of the thesis, such populations of hermaphroditic A. anatina will be of interest for further investigation into the process of the transition from dioecy to hermaphroditism, but these populations were not available for the present study. As another example, observations of a population of the FWM Elliptio complanata from the stagnant water environment of Lac de l’Achigan Québec, showed that the population was approximately 80% hermaphroditic with a clear variation in proportion of male and female gonad tissue within individuals (Downing et al. 1993). E. complanata were observed in the moving water environment of Nottoway River

Virginia, for which Henley (1997) reported that out of 57 mussels sexed, 47.4% were females, 36.8% males, and 5.3% were hermaphrodites (note: 10.5% of individuals were indeterminate). Bauer and Wächtler (2001) noted that Margaritifera auriclaria exhibits hermaphroditism in only small, low density, relatively isolated populations, and concluded that a third so called “cryptic microhermaphrodite” sexual strategy may occur in such populations. The term cryptic microhermaphrodite is not commonly used to describe hermaphrodites in Bivalvia literature. One possible example is that of

Hydriella depressa, a FWM from Australia that may exhibit a small amount of spermatogenic tissue within an otherwise egg producing gonad or “ovotestis” (Byrne

1998). The term microhermaphrodite was used to describe a similar gonad composure in both sexes of Margaritifera margaritifera collected from Esse river, Finland

(Hanstén et al. 1997). Although these terms are not widely used, the phenomena of possessing a bit of gametogenic tissue of the opposite sex in gonads is well established within mussels and has also been termed accidental hermaphrodite (van der Schalie

1970; Stewart et al. 2013). It is theorized that accidental hermaphrodites are the result of

a breakdown in the sex determining mechanism (van der Schalie 1970), but the differences between this breakdown and that of an obligate hermaphrodite may not be much different (Stewart et al. 2013). The work of van der Schalie (1970) concluded that four species of FWMs were obligate hermaphrodites and twenty-one were occasional hermaphrodites. A literature review conducted by Henley (2002) concluded that eight

FWM were obligate hermaphrodites and fifty were occasional hermaphrodites. The transition from the ancestral dioecious condition to obligate hermaphroditism in FWMs is seemingly the result of convergent evolution (Breton et al. 2011) with at least five independent transitions (Breton et al. 2011; Chase et al. in prep). A more detailed review of these transitions is addressed in Chapter 2. For further reviews on hermaphroditism see Charnov et al. 1976 and Charnov; Bull 1979; Morton 1991;

Valenzuela 2003.

Bauer and Wächtler (2001) state that species such as U. crassus, Alasmidonta heterodon, Cyclonaias tuberculate and Mutela bourguignati do not exhibit hermaphroditism even at small frequencies. It is possible that evidence of accidental hermaphrodites or occasional hermaphrodites has yet to be recorded in these species, but it is also possible that some species are exempt from this so called breakdown in the sex determination mechanism (van der Schalie 1969). It may be best to view dioecy and hermaphroditism as a spectrum among most FWMs given the number of terms describing the “level” of hermaphroditism occurring in FWM literature (explored above). There are even examples of so called obligate hermaphrodites whose hermaphrodite and dioecious population proportion may alter slightly, as well as the

internal proportion of male to female gonad tissue (Utterbackia imbecillis; Heard 1975;

Kat 1983; Johnson et al. 1998). Furthermore, there is evidence that different populations of U. imbecillis reproduce primarily through selfing and others through outcrossing

(Allen 1924; Johnson et al. 1998). In a dioecious species, M. margaritifera, there is evidence that transferring individuals from a high density environment to a low density environment resulted in a transition to hermaphroditism in individuals that were previously dioecious (Bauer 1987). However, Bauer (1987) also states that the claim of males switching to hermaphrodites in other studies may have been made due to the improper sexing of individuals. Because of this possibility, Bauer (1987) suggests that before an individual mussel can be judged to be true hermaphrodite, that gonad tissue be examined microscopically for the presence of spermatogonia (i.e., early stage, non- motile sperm cell) as opposed to simply the presence of mature sperm in an individual, as exogenous mature sperm may have entered the individual from a separate male. The transition from a female to a hermaphrodite has also been described for other species

(e.g., in the genus Margaritifera; Grande et al. 2001), but definitive confirmation of such a transition is clearly a difficult task. The same logic may apply to previous studies of the genus Anodonta, including studies of A. cygnea, the primary study species of this research (Hinzmann et al. 2013; Chojnacki et al. 2007; van der Schalie 1970). However,

Bauer and Wächtler (2001) suggest that some previous studies describing the reproductive status of A. cygnea need to be revisited in light of (1) the challenges of distinguishing A. anatina from A. cygnea and (2) the difficulties in distinguishing true hermaphrodites from females containing exogenous sperm. Bauer and Wächtler (2001) maintain that A. cygnea is mostly a “true” obligate hermaphroditic species with no more

than a few functional females within populations. Henley (2002) also identifies A. cygnea as an obligate hermaphrodite. While acknowledging the controversies associated with hermaphroditism in the genus, expert opinion suggests that A. cygnea represents one example of the transition from dioecy to hermaphroditism in the Unionidae and thus this species, and particularly the mussels examined herein, represent an appropriate system for studying the molecular consequences of the transition from dioecy to hermaphroditism in FWMs.

Doubly uniparental inheritance and molecular evolution of mtDNA in mussels

As previously mentioned, DUI occurs among FWMs, and this unusual system of cytoplasmic organelle inheritance is defined by separate evolutionary lineages of mtDNA between the sexes of dioecious species (reviewed in Breton et al. 2007; Zouros

2013). The age of this phenomenon is still the subject of study, but it is at least 200 million years old in FWMs (Hoeh et al. 2002; Doucet-Beaupré et al. 2010; Stewart et al.

2013). In terms of some general patterns of molecular evolution associated with the DUI system, studies of both marine and FWMs suggest that M-type mtDNA genomes typically evolve faster than F-type genomes (Hoeh et al. 1996; Zouros 2013). Breton et al. (2009) reviewed complete FWM mtDNA genomes from species exhibiting DUI and confirmed that the M- and F-type mtDNA evolve faster than other animal mtDNA and that the rate of evolution in the M-type exceeds that of the F-type. Sequence divergence between M- and F-types in marine mussels varies between 20 to 40% (reviewed in

Robicheau et al. 2017), and up to 50% in FWMs (Breton et al. 2011). Gene order is

often the same or highly similar within species of FWMs, but more variable in marine mussels (Breton et al. 2009).

It has been hypothesized that selective constraints on M-types are relaxed when compared to F-types, facilitating the higher rate of evolution within the M-types of both marine mussels and FWMs (Skibinski et al. 1994; Rawson and Hilbish 1995; Liu et al.

1996; Stewart et al. 1996). A comparison of the rates and patterns of molecular evolution of H-types compared to both the F- and M-type mtDNAs of dioecious species has not yet been extensively described. Other hypotheses to explain the difference in rates of molecular evolution between the M- and F-types have been explored by Stewart et al. (1996). These four hypotheses for the differential rates of evolution of the M- vs. the F-types are based on the four following ideas. (1) DNA replications occurring during spermatogenesis are much more frequent than replications occurring during oogenesis. A higher frequency of replications during spermatogenesis creates a higher probability of mutations and accumulation of mutations for the M-type genome. (2) A higher degree of free radical damage is likely experienced by M-types relative to F-

- types. Free radicals refer to reactive oxygen species (ROS) such as O2 (superoxide anion), which are byproducts of aerobic respiration (Halliwell 1991). Mitochondria are responsible for producing the majority of ROS in a cell (Chance et al. 1979). ROS can damage DNA, and although repair mechanisms are in place, these can only limit the damage and do not completely inhibit it (Krokan et al. 1997). It has been suggested that damage by ROS to mtDNA occurs more readily in sperm, when compared to eggs, due to their relatively higher degree of metabolic activity than is found in eggs (Lambert and

Battaglia 1993; Skibinski et al. 1994). (3) There may be differences in the nature and strength of selection acting on M- vs. F-type mtDNA as these genomes are functioning largely in different cellular environments. And finally, (4) the M-type mitochondrial genome has a smaller effective population size and this factor could also be responsible for faster rates of nucleotide substitution in M-types compared to F-types. There is a drastic difference in number of mitochondria contained within a single sperm vs. within a single egg (approximately 1:1000 in mammals; White et al. 2008), which creates unequal mtDNA diversity between the male and female germ lines. Based on this, slightly deleterious mutations produced within the male germ line may be more likely to persist (Ohta 2002). Stewart et al. (1996) focused on Mytilus edulis and Mytilus trossulus (marine mussel species), however, the study gives insight into species exhibiting DUI in general. Hypothesis (1) and (2) cannot account for the statistically equal number of synonymous substitutions between M- and F-type mtDNA while nonsynonymous substitutions are higher within M-type mtDNA, meanwhile this gives support to hypothesis (3) (Stewart et al. 1996). Stewart et al. (1996), ultimately suggest that the difference in rates of molecular evolution among the M- and F-types, and between mitochondrial genomes of DUI species vs. those of species with SMI can be understood in terms of “cellular arenas”, where the somatic line, female germ line and male germ line each count as one arena. Selective pressure placed on a mitochondrial genome functioning in all three arenas should be stronger than that of a mitochondrial genome functioning in only one or two arenas. Put another way, selective pressures should be more relaxed for an mtDNA genome function in one or two arenas in comparison to one functioning in all three. F-type mtDNA are present in the somatic

and female germ line, but not in the male germ line, combined with M-type mtDNA being rare in tissues outside the male gonad, and being excluded from females virtually completely, consequently excludes M-type mtDNA from the female tissue arena and nearly excludes it from the somatic tissue arena. This leaves it to function almost exclusively in one arena, the male tissue arena of sperm cells, thereby reducing the constraints placed on the M-type mtDNA in comparison to the F-type mtDNA, which is performing in all somatic tissue and in the female germ line. Conversely, an individual operating under SMI should be the most constrained under this arena hypothesis.

Stewart et al. (1996) noted that this was a working hypothesis and that the various factors hypothesized to affect rates of molecular evolution of the M- and F-types are not mutually exclusive. However, according to the “arenas” hypothesis, a dioecious FWM mussel species M-type mtDNA should be evolving faster than the F-type mtDNA, and both may be evolving faster than the mtDNA of a closely related hermaphroditic FWM species. This hypothesis, and how it relates to H-type mtDNA, will be explored further in Chapter 3.

Although DUI and the persistence of heritable M-type mtDNA within species of bivalves has been continuously assessed since its discovery in 1990 (Fisher and

Skibinski), the question still remains: what is the function of the M-type mtDNA? One explanation for the persistence of a male-transmitted mitochondrial genome is that it is advantageous for the organism possessing this distinct male-male line of cytoplasmic organelle transmission as selection can act directly on the M-type genome to optimize sperm function (Burt and Trivers 2006; Stewart et al. 2013). Optimization of sperm

related functions in mitochondrial DNA is not possible in species with SMI (Zouros

2013). Another explanation, which is not mutually exclusive of the first, is that M-type mtDNA plays a role in sex determination and this is the consequence of a selfish genetic element in the M-type mtDNA genome that increased in frequency in some ancestral bivalve as a consequence of adjusting the sex ratio in a way that benefitted its own proliferation throughout the population (Zouros 2000; Breton et al. 2007; Breton et al.

2017).

As previously mentioned novel ORFs termed m-orf and f-orf contained within M-type mtDNA and F-type mtDNA, respectively, exist in the mitochondrial genomes of FWM

(and marine mussel) species (Breton et al. 2011). The f-orf and m-orf genes occur in non-homologous regions of the M- and F-type mitochondrial genomes and in general f- orf genes are shorter than m-orf genes (Breton et al. 2011). Hermaphroditic species also house a modified f-orf, the h-orf. It is not yet known whether F-ORF proteins in both marine mussels and FWM are homologous (Breton et al. 2011). It is hypothesized that the source of both f-orf and m-orfs, and their corresponding proteins (F- and M-ORFs), could be from (1) a homologous ancestral bacterial gene, (2) a mtDNA gene that has been duplicated and diverged (Guerra et al. 2017), (3) a gene that has migrated from the nucleus to the mitochondria (Pont-Kingdon et al. 1998; Burger et al. 2003) or (4) a viral origin (Milani et al. 2013; Mitchell et al. 2016).

The presence of an m-orf in males/M-type mtDNA, an f-orf in females/F-type mtDNA and an h-orf in hermaphrodites/H-type mtDNA implies at least an association between

these novel ORFs and sex. Breton et al. (2011) demonstrated a link between the presence of an m-orf with dioecy and the absence of an m-orf with hermaphroditism in an examination of fourteen FWM species. This result demonstrated a tight correlation between these novel genes and the dioecious sexual strategy in FWM. Breton et al.

(2011) also showed that among these species, the transition from dioecy to hermaphroditism also accompanies macromutational modifications of the f-orf gene to the homologous h-orf gene in hermaphroditic species, thereby linking the change (either consequential or causational) in functional roles of these genes to a change in sexual strategy. This further implicates a link between these novel protein coding genes and sex determination, and fundamentally a link between hermaphroditism and the modification of the f-orf to an h-orf. An interesting finding is that through immunoelectron microscopy (immunostaining) Breton et al. (2011) also found that, in

Venustaconcha ellipsiformis eggs, the F-ORF protein is present not only in the mitochondria, but also associated with the nuclear membrane and nucleoplasm.

Furthermore, proteins coded by the mitochondria are associated with energy production

(oxidative phosphorylation), but the orfs are not homologous to any known genes associated with ATP synthesis (and are not readily identifiable as homologous to any known genes for that matter, however, see Guerra et al. 2016), this indicates that the

ORF proteins likely have functions other than energy production and adds supports to a possible link between sex determination and these novel proteins. A similar association has not yet been demonstrated for M-ORF proteins, but similar immunohistochemical studies of the M-ORF proteins have simply not yet been performed. Although the sex determination mechanism in bivalves possessing DUI are not yet defined, various

studies have given some clues into their possible properties (e.g. Milani et al. 2013;

Mitchell et al. 2016; reviewed by Breton et al. 2017).

Individuals developing into females exhibit a pattern in which sperm mitochondria are dispersed throughout the blastomere and then disappear, whereas in contrast, individuals destined to be male form an aggregate cluster of sperm mitochondria in a single blastomere (Cao et al. 2004; Cogswell et al. 2006). Sutherland et al. (1998) showed that virtually all eggs receive paternal mitochondria, and that within twenty-four hours they are eliminated in embryos destined to become females. It is worth noting that the mechanisms that work to eliminate maternal mitochondria in somatic tissue, and to aggregate maternal mitochondria in male gonads are not perfect (Garrido-Ramos et al.

1998; Zouros 2000; Passamonti and Scali 2001; Dalziel and Stewart 2002; Obata et al.

2006). These studies center around marine mussels, Mytilus, but are considered to be general characteristics of DUI in bivalves.

Among obligate hermaphrodites, the m-orf (and entire M-type genome) has been lost and these hermaphroditic populations effectively exhibit SMI, because the h-orf, as part of the hermaphroditic genome alone is passed down to all offspring (Breton et al. 2011).

H-ORF proteins, while highly variable, do share a few common patterns, as outlined by

Breton et al. (2011). (1) The H-ORFs are typically longer than F-ORF proteins, (2) they do not align well with F-ORFs from congeneric species, (3) they often contain repeat units that are not present in F-ORFs of dioecious species, (4) their amino acid sequences may not include the characteristic single transmembrane helix structure that has been

conserved in the 35 species of FWMs examined by Breton et al. (2011) and (5) their hydrophobicity profiles vary drastically from those of congeneric dioecious species’ F-

ORFs. These patterns will be revisited in an assessment of A. cygnea H-ORF in Chapter

Mitochondrial DNA and evolutionary relationships amongst FWMs

Recent phylogenetic analysis of family Unionidae, order Unionoida, provides a framework for studying patterns of molecular evolution of six subfamilies and eighteen tribes in this group (Breton et al. 2011; Lopes-Lima et al. 2017). Lopes-Lima et al.

(2017) proposed a phylogeny based on a combined mtDNA cox1 and nuclear DNA 28S data set for 70 FWMs. The earlier phylogenetic analysis conducted by Breton et al.

(2011) involved only mtDNA sequence data (cox1 and nad1). Lopes-Lima (2017) clarified this phylogeny and included species from the Eastern Palearctic and

Indotropical regions that were previously underrepresented. Their analysis, however, still had difficulties resolving certain relationships among species from the tribe

Anodontini, which includes A. cygnea, A. anatina and P. complanata. They highlight the importance of phylogenetic studies in the conservation of FWMs and for further systematic and ecological studies.

These phylogenetic studies have highlighted the challenges of resolving FWM species relationships. Several other studies have also suggested that partial sequences from the cox1 gene, which is routinely used as part of the Barcode of Life initiative (e.g., Hebert

et al. 2003a), are insufficient to resolve FWM families (Graf and Cummings 2006;

Hoeh et al. 2009; Whelan et al. 2011; Graf 2013). Nonetheless, due to data availability, and for the purpose of comparative studies, this thesis will include an analysis of patterns of molecular evolution of cox1 sequence data. However, the limits of cox1 are acknowledged and will be discussed.

Trends in the analysis of complete genomes of freshwater mussels

The circular nature of mtDNA creates two strands of DNA, termed the heavy and light strand. The strand that consists of a higher proportion of purines (adenine and guanine) is considered the “heavy strand” because purines are heavier than pyrimidines (cytosine and thymine). Heavy or light strand, and alternatively the direction of translation, are used to describe the location of protein coding genes (PCG), tRNA and rRNA along mtDNA. The complete list of PCG, tRNA and rRNA genes located in FWM mtDNA is outlined in Table 1, alongside designations used in this thesis.

A total of 35 FWM species (including A. cygnea) have complete mtDNA accessible on

GenBank (see Chapter 2, Table S1). Both male and female complete mtDNA are available for the following species: A. anatina (a close relative of A. cygnea),

Cumberlandia monodonta, Echyridella menziesii, Hyriopsis cumingii, Hyriopsis schlegelii, Lamprotula leai, Lampsilis powellii, Lampsilis siliquiodea, Pyganodon grandis, Quadrula quadrula, S. woodiana, Solenaia carinatus, Unio delphinus, Unio tumidulu, Utterbackia peninsularis and V. ellipsiformis. There are five complete mtDNA of hermaphroditic individuals available: T. parvum, U. imbecillis, Lasmigona

compressa, Margaritifera falcata and (from this study) Anodonta cygnea. A near complete genome is also available for the hermaphrodite Lasmigona subviridis.

Transcription of mtDNA is governed by the nucleus through the mitochondrial control region, a typically non-coding region of the mtDNA genome (Shadel and Clayton

1997). In addition to the standard mitochondrial genes, freshwater mussel mtDNA usually contains three common non coding regions (i.e., shared unassigned regions or

SURs; Breton et al. 2011), which differ slightly between the M- and F-type mtDNA

(Table 2).

Shared unassigned regions (SURs) are located where the direction of transcription changes along the mt genome (Breton et al. 2009). The control region of mtDNA is contained within non coding regions, thus these SURs are candidates for control regions

(e.g., Breton et al. 2009; 2011a). Both SURs and potential control regions are examined in Chapter 2.

SUMMARY AND OBJECTIVES

Prior to this research, the A. cygnea mt genome had yet to be sequenced and uploaded to

GenBank. The objectives of this thesis are (1) to add to the available complete mitochondrial genome data of FWMs by the addition of the sequence of the complete A. cygnea mitochondrial genome and (2) to investigate the molecular characteristics of the

H-ORF in A. cygnea, and (3) to compare evolutionary rates among M-, F- and H-type

mtDNA for the first time. Objectives 1 and 2 will be presented in Chapter 2, and objective 3 in Chapter 3.

TABLES

Table 2. Shared unassigned regions within F- and M-type mitochondrial DNA of Order Unionoida. Designations reflect the location on the mitochondrial genome. Mitotype Shared Unassigned Regions F-Type F nd5- F trnF- F nd3- trnQ nd5 trnH+trnA M-Type M nd5- M trnF- M nd3- trnQ+trnH nd5 trnA

Table 1. Standard genes found in mitochondrial DNA: 13 protein coding, 2 ribosomal RNA and 22 tRNA. Modified from Boore 1999. Protein Designation Figure Designation Cytochrome oxidase subunit I, II, cox1, cox2, cox1, cox2, III cox3 cox3 Cytochrome b cytb cob NADH dehydrogenase subunits 1- nad1-6, nad4L nad1-6, nad4L 6, 4L ATP synthase subunit 6,8 atp6, atp8 atp6, atp8 Large ribosomal subunit RNA rrnL rrnL Small ribosomal subunit RNA rrnS rrnS tRNA alanine trnA A tRNA cysteine trnC C tRNA aspartic acid trnD D tRNA glutamic acid trnE E tRNA phenylalanine trnF F tRNA glycine trnG G tRNA histidine trnH H tRNA isoleucine trnI I tRNA lysine trnK K tRNA leucine 1 TRNL1 L1 tRNA leucine 2 TRNL2 L2 tRNA methionine trnM M tRNA asparagine trnN N tRNA proline trnP P tRNA glutamine trnQ Q tRNA arginine trnR R tRNA serine 1 TRNS1 S1 tRNA serine 2 TRNS2 S2 tRNA threonine trnT T tRNA valine trnV V tRNA tryptophan trnW W tRNA tyrosine trnY Y

REFERENCES

Allen, E., 1924. The existence of a short reproductive cycle in Anodonta imbecillis. The

Biological Bulletin 46, 88–94.

Allen, W.R., 1921. Studies of the biology of freshwater mussels. III. Distribution and

movements of Winona Lake mussels, in: Proceedings of the Indiana Academy of

Science. pp. 227–238.

Altmann, R., 1890. Die elementarorganismen und ihre beziehungen zu den zellen. Veit,

Leipzig, Germany.

Anders, K., Wiese, V., 1993. Glochidia of the freshwater mussel, Anodonta anatina,

affecting the anadromous European smelt (Osmerus eperlanus) from the Eider

estuary, Germany. Journal of Fish Biology 42, 411–419.

Anfinson, J.O., 2003. The River We Have Wrought: A History of the Upper

Mississippi. University of Minnesota Press.

Avise, J.C., 1991. Ten unorthodox perspectives on evolution prompted by comparative

population genetic findings on mitochondrial DNA. Annual Review of Genetics

25, 45–69.

Avise, J.C., Arnold, J., Ball, R.M., Bermingham, E., Lamb, T., Neigel, J.E., Reeb, C.A.,

Saunders, N.C., 1987. Intraspecific phylogeography: the mitochondrial dna

bridge between population genetics and systematics. Annual Review of Ecology

and Systematics 18, 489–522.

Avise, J.C., Ellis, D., 1986. Mitochondrial DNA and the evolutionary genetics of higher

animals and Discussion. Philosophical Transactions of the Royal Society of

London. Series B, Biological Sciences 312, 325–342.

Bailey, R.C., Green, R.H., 1988. Within-basin variation in the shell morphology and

growth rate of a freshwater mussel. Canadian Journal of Zoology 66, 1704–

1708.

Balaban, R.S., Nemoto, S., Finkel, T., 2005. Mitochondria, oxidants, and aging. Cell

120, 483–495.

Ball, G.H., 1922. Variation in fresh-water mussels. Ecology 3, 93–121.

Ballard, J.W.O., Whitlock, M.C., 2004. The incomplete natural history of mitochondria.

Molecular Ecology 13, 729–744.

Bauer, G., 1987. Reproductive strategy of the freshwater pearl mussel Margaritifera

margaritifera. Journal of Animal Ecology 56, 691–704.

Bauer, G., Wächtler, K., 2001. Ecology and Evolution of the Freshwater Mussels

Unionoida. Springer, Verlag, Berlin, Germany.

Bednarczuk, J., 1986. Untersuchungen zu wirtsfischspektrum und entwicklung der

bachmuschel Unio crassus. Thesis, Tierärztliche Hochschule Hannover.

Birky, C.W., 1995. Uniparental inheritance of mitochondrial and chloroplast genes:

mechanisms and evolution. PNAS 92, 11331–11338.

Bloomer, H.H., 1934. On the sex, and sex-modification of the gill, of Anodonta cygnea.

Journal of Molluscan Studies 21, 21–28.

Bogan, A.E., 1993. Freshwater bivalve extinctions (Mollusca: Unionoida): a search for

causes. Integrative and Comparative Biology 33, 599–609.

Boyle, E.E., Etter, R.J., 2013. Heteroplasmy in a deep-sea protobranch bivalve suggests

an ancient origin of doubly uniparental inheritance of mitochondria in Bivalvia.

Marine Biology 160, 413–422.

Brand, M.D., 1997. Regulation analysis of energy metabolism. Journal of Experimental

Biology 200, 193–202.

Breton, S., Beaupré, H.D., Stewart, D.T., Hoeh, W.R., Blier, P.U., 2007. The unusual

system of doubly uniparental inheritance of mtDNA: isn’t one enough? Trends

in Genetics 23, 465–474.

Breton, S., Beaupré, H.D., Stewart, D.T., Piontkivska, H., Karmakar, M., Bogan, A.E.,

Blier, P.U., Hoeh, W.R., 2009. Comparative mitochondrial genomics of

freshwater mussels (Bivalvia: Unionoida) with doubly uniparental inheritance of

mtDNA: gender-specific open reading frames and putative origins of replication.

Genetics 183, 1575–1589.

Breton, S., Capt, C., Guerra, D., Stewart, D., 2017. Sex Determining Mechanisms in

Bivalves. Preprints, 2017060127.

Breton, S., Stewart, D.T., Shepardson, S., Trdan, R.J., Bogan, A.E., Chapman, E.G.,

Ruminas, A.J., Piontkivska, H., Hoeh, W.R., 2011. Novel protein genes in

animal mtDNA: A new sex determination system in freshwater mussels

(Bivalvia: Unionoida)? Molecular Biology and Evolution 28, 1645–1659.

Brown, W.M., George, M., Wilson, A.C., 1979. Rapid evolution of animal

mitochondrial DNA. PNAS 76, 1967–1971.

Burger, G., 2003. The enigmatic mitochondrial ORF ymf39 codes for ATP synthase

chain b. Nucleic Acids Research 31, 2353–2360.

Burt, A., Trivers, R., 2006. Genes in conflict: the biology of selfish genetic elements.

Harvard University Press.

Byrne, M., 1998. Reproduction of river and lake populations of Hyridella depressa

(Unionacea: Hyriidae) in New South Wales: implications for their conservation.

Hydrobiologia 389, 29–43.

Cao, L., Kenchington, E., Zouros, E., 2004. Differential segregation patterns of sperm

mitochondria in embryos of the blue mussel (Mytilus edulis). Genetics 166, 883–

894.

Carus, C., 1832. Neue Untersuchungen über die entwicklungsgeschichte unserer

flußmuschel. Verh K Leopoldinisch-Carolinischen Akad Naturforsch 1–87.

Carus, C.G., 1832. Neue untersuchungen über die entwickelungsgeschichte unserer

flussmushel. Weber.

Chance, B., Sies, H., Boveris, A., 1979. Hydroperoxide metabolism in mammalian

organs. Physiological reviews 59, 527–605.

Charnov, E.L., 1979. Simultaneous hermaphroditism and sexual selection. Proceedings

of the National Academy of Sciences 76, 2480–2484.

Charnov, E.L., Bull, J., 1977. When is sex environmentally determined? Nature 266,

828–830.

Charnov, E.L., Bull, J.J., Maynard Smith, J., 1976. Why be an hermaphrodite? Nature

263, 125–126.

Chase, E., Robicheau, B., Veinot, S., Bauer, G., Breton, S., Stewart, D.T., n.d. The

complete H-type mitochondrial genome of freshwater mussel Anodonta cygnea

(Bivalvia: Unionidae). Unpublished.

Chojnacki, J.C., Lewandowska, A., Rosińska, B., 2007. Biometrics of the mussel

Anodonta cygnea (L.) inhabiting in 2005 the Binowo and Bobolin Lakes near

Szczecin. Oceanological and Hydrobiological Studies 21–37.

Claes, M., 1987. Untersuchungen zur entwichlungsbiologie der teichmuschel Anodonta

cygnea. Dipl Thesis, Universität Hannover.

Cogswell, A.T., Kenchington, E.L.R., Zouros, E., 2006. Segregation of sperm

mitochondria in two- and four-cell embryos of the blue mussel Mytilus edulis:

implications for the mechanism of doubly uniparental inheritance of

mitochondrial DNA. Genome 49, 799–807.

Collin, R., 2013. Phylogenetic patterns and phenotypic plasticity of molluscan sexual

systems. Integrative and Comparative Biology 53, 723–735.

Cui, H., Kong, Y., Zhang, H., 2012. Oxidative stress, mitochondrial dysfunction, and

aging. Journal of Signal Transduction 2012, 1–13.

Dalziel, A.C., Stewart, D.T., 2002. Tissue-specific expression of male-transmitted

mitochondrial DNA and its implications for rates of molecular evolution in

Mytilus mussels (Bivalvia: Mytilidae). Genome 45, 348–355.

Dolgov, L.V., 1992. Sexual structure of a Tridacna squamosa population: relative

advantages of sequential and simultaneous hermaphroditism. Journal of

Molluscan Studies 58, 21–27.

Doucet-Beaupré, H., Breton, S., Chapman, E.G., Blier, P.U., Bogan, A.E., Stewart,

D.T., Hoeh, W.R., 2010. Mitochondrial phylogenomics of the Bivalvia

(Mollusca): searching for the origin and mitogenomic correlates of doubly

uniparental inheritance of mtDNA. BMC Evolutionary Biology 10, 50.

Engel, H., 1990. Untersuchungen zur autökologie von Unio crassus (Philipsson) in

Norddeutschland. Thesis, Unisersität Hannover.

Engel, H., Wächtler, K., 1990. Folgen von bachentkrautungsmaßnahmen auf einen

Süßwassermuschelbestand am beispiel eines kleinen fließgewässers des

südlichen drawehn (Lüchow-Dannenberg). Natur Landschaft 63–65.

Ernster, L., Schatz, G., 1981. Mitochondria: a historical review. J Cell Biol 91, 227–

255.

Fisher, C., Skibinski, D.O.F., 1990. Sex-Biased Mitochondrial DNA Heteroplasmy in

the Marine Mussel Mytilus. Proceedings of the Royal Society of London B:

Biological Sciences 242, 149–156.

Garrido-Ramos, M.A., Stewart, D.T., Sutherland, B.W., Zouros, E., 1998. The

distribution of male-transmitted and female-transmitted mitochondrial DNA

types in somatic tissues of blue mussels: Implications for the operation of doubly

uniparental inheritance of mitochondrial DNA. Genome 41, 818–824.

Ghiselin, M.T., 1969. The evolution of hermaphroditism among animals. The Quarterly

Review of Biology 44, 189–208.

Graeber, M.B., Müller, U., 1998. Recent developments in the molecular genetics of

mitochondrial disorders. Journal of the Neurological Sciences 153, 251–263.

Graf, D.L., 2013. Patterns of freshwater bivalve global diversity and the state of

phylogenetic studies on the Unionoida, Sphaeriidae, and Cyrenidae. American

Malacological Bulletin 31, 135–153.

Graf, D.L., Cummings., K.S., 2013. The Freshwater Mussels (Unionoida) of the World

(and other less consequential bivalves). MUSSEL Project Web Site,

http://www.mussel-project.net/.

Graf, D.L., Cummings, K.S., 2007. Review of the systematics and global diversity of

freshwater mussel species (Bivalvia: Unionoida). Journal of Molluscan Studies

73, 291–314.

Graf, D.L., Cummings, K.S., 2006. Palaeoheterodont diversity (Mollusca: Trigonioida+

Unionoida): what we know and what we wish we knew about freshwater mussel

evolution. Zoological Journal of the Linnean Society 148, 343–394.

Graf, W.L., 1999. Dam nation: A geographic census of American dams and their large-

scale hydrologic impacts. Water Resources Research 35, 1305–1311.

Grande, C., Araujo, R., Ramos, M.A., 2001. The gonads of Margaritifera auricularia

(Spengler, 1793) And M. margaritifera (Linnaeus, 1758) (Bivalvia:

Unionoidea). Journal of Molluscan Studies 67, 27–35.

Gray, M.W., Burger, G., Lang, B.F., 1999. Mitochondrial evolution. Science 283,

1476–1481.

Guerra, D., Plazzi, F., Stewart, D.T., Bogan, A.E., Hoeh, W.R., Breton, S., 2017.

Evolution of sex-dependent mtDNA transmission in freshwater mussels

(Bivalvia: Unionida). Scientific Reports 7.

Haag, W.R., 2012. North American freshwater mussels: natural history, ecology, and

conservation. Cambridge University Press.

Haag, W.R., Leann Staton, J., 2003. Variation in fecundity and other reproductive traits

in freshwater mussels. Freshwater Biology 48, 2118–2130.

Halliwell, B., 1992. Iron and damage to biomolecules. Iron and human disease 209–236.

Hansten, C., Pekkarinen, M., Valovirta, I., 1997. Effect of transplantation on the gonad

development of the freshwater pearl mussel, Margaritifera margaritifera (L.).

Boreal Environment Research 2, 247–256.

Heard, W.H., 1975. Sexuality and other aspects of reproduction in Anodonta

(Pelecypoda: Unionidae). Malacologia 15, 81–103.

Heath, D.J., 1977. Simultaneous hermaphroditism; cost and benefit. Journal of

Theoretical Biology 64, 363–373.

Hebert, P.D., Wilson, C.C., Murdoch, M.H., Lazar, R., 1991. Demography and

ecological impacts of the invading mollusc Dreissena polymorpha. Canadian

Journal of Zoology 69, 405–409.

Hebert, P.D.N., Cywinska, A., Ball, S.L., deWaard, J.R., 2003a. Biological

identifications through DNA barcodes. Proceedings of the Royal Society of

London B: Biological Sciences 270, 313–321.

Hebert, P.D.N., Muncaster, B.W., Mackie, G.L., 1989. Ecological and genetic studies

on Dreissena polymorpha (Pallas): a new mollusc in the Great Lakes. Canadian

Journal of Fisheries and Aquatic Sciences 46, 1587–1591.

Hebert, P.D.N., Ratnasingham, S., deWaard, J.R., 2003b. Barcoding animal life:

cytochrome c oxidase subunit 1 divergences among closely related species.

Proceedings of the Royal Society of London B: Biological Sciences 270, S96–

S99.

Henley, W.F., 2002. Evaluation of Diet, Gametogenesis, and Hermaphroditism in

Freshwater Mussels (Bivalvia: Unionidae). Virginia Tech Doctoral Dissertation.

Higgins, F., 1858. A catalogue of the shell-bearing species of Mollusca, inhabiting the

vicinity of Columbus Ohio with some remarks thereon. Ohio State Board of

Agriculture with an Abstract in the Proceedings of the County Agricultural

Societies 12 Annual Report for 1857.

Hinch, S.G., Bailey, R.C., 1988. Within-and among-lake variation in shell morphology

of the freshwater clam Elliptio complanata (Bivalvia: Unionidae) from south-

central Ontario lakes. Hydrobiologia 157, 27–32.

Hinzmann, M., Lopes-Lima, M., Teixeira, A., Varandas, S., Sousa, R., Lopes, A.,

Froufe, E., Machado, J., 2013. Reproductive cycle and strategy of Anodonta

anatina (L., 1758): Notes on Hermaphroditism. Journal of Experimental

Zoology 319, 378–390.

Hochwald, S., 1997. Populationsökologie der bachmuschel (Unio crassus). Bayreuther

Forum Ökologie, Band 50, Bayreuth.

Hoeh, W.R., Bogan, A.E., Heard, W.H., Chapman, E.G., 2009. Palaeoheterodont

phylogeny, character evolution, diversity and phylogenetic classification: A

reflection on methods of analysis. Malacologia 51, 307–317.

Hoeh, W.R., Stewart, D.T., Sutherland, B.W., Zouros, E., 1996. multiple origins of

gender-associated mitochondrial DNA lineages in Bivalves (mollusca: Bivalvia).

Evolution 50, 2276–2286.

Hoeh, W.R., Frazer, K.S., Naranjo-Garcia, E., Trdan, R.R., 1995. A phylogenetic

perspective on the evolution of simultaneous hermaphroditism in a freshwater

mussel clade (Bivalvia Unionidae Utterbackia). Malacological Review 25–42.

Hoeh, W.R., Stewart, D.T., Guttman, S.I., 2002. High fidelity of mitochondrial genome

transmission under the doubly uniparental mode of inheritance in freshwater

mussels (Bivalvia: Unionoidea). Evolution 56, 2252–2261.

Hornbach, D.J., Kurth, V.J., Hove, M.C., 2010. Variation in freshwater mussel shell

sculpture and shape along a river gradient. The American Midland Naturalist

164, 22–36.

Johnston, M.O., Das, B., Hoeh, W.R., 1998. Negative correlation between male

allocation and rate of self-fertilization in a hermaphroditic animal. Proceedings

of the National Academy of Sciences 95, 617–620.

Kallersjo, M., von Proschwitz, T., Lundberg, S., Eldenas, P., Erseus, C., 2005.

Evaluation of ITS rDNA as a complement to mitochondrial gene sequences for

phylogenetic studies in freshwater mussels: an example using Unionidae from

north-western Europe. Zoologica Scripta 34, 415–424.

Kat, P.W., 1983. Sexual selection and simultaneous hermaphroditism among the

Unionidae (Bivalvia: Mollusca). Journal of Zoology 201, 395–416.

Krishnamurthy, P.K., Francis, R.A., 2012. A critical review on the utility of DNA

barcoding in biodiversity conservation. Biodiversity and Conservation 21, 1901–

1919.

Kroemer, G., Dallaporta, B., Resche-Rigon, M., 1998. The mitochondrial death/life

regulator in apoptosis and necrosis. Annual review of physiology 60, 619–642.

Krokan, H.E., Standal, R., Slupphaug, G., 1997. DNA glycosylases in the base excision

repair of DNA. Biochemical Journal 325, 1–16.

Kunz, G.F., 1898. A brief history of the gathering of fresh-water pearls in the United

States. US Fish Commission Bulletin 17, 321–330.

Lambert, C.C., Battaglia, D.E., 1993. Loss of the paternal mitochondrion during

fertilization. Zoological Science 31–37.

Lavett Smith, C., 1967. Contribution to a theory of hermaphroditism. Journal of

Theoretical Biology 17, 76–90.

Lee, H.-C., Wei, Y.-H., 2007. Oxidative stress, mitochondrial DNA mutation, and

apoptosis in aging. Experimental biology and medicine 232, 592–606.

Lee, J.S., Park, J.S., Shin, Y.K., Lee, Y.G., Park, J.J., 2013. Sequential hermaphroditism

in Manila clam Ruditapes philippinarum (Bivalvia: Veneridae). Invertebrate

Reproduction & Development 57, 185–188.

Lee, S.Y., 1988. The reproductive cycle and sexuality of the green mussel Perna viridis

(Bivalvia: Mytilacea) in Victoria Harbour, Hong Kong. Journal of Molluscan

Studies 54, 317–323.

Lewis, J, 1868. Remarks on some of the Mollusca of the valley of the Mohawk.

American Journal of Conchology 341–245.

Leydig, F., 1866. Mittheilung über den parasitismus junger Unioniden an Fischen in

Noll. Tübingen, Inaugural-Dissertation, Frankfurt-am-Main.

Liu, H.-P., Mitton, J.B., Wu, S.-K., 1996. Paternal mitochondrial DNA differentiation

far exceeds maternal mitochondrial DNA and allozyme differentiation in the

freshwater mussel, Anodonta grandis grandis. Evolution 50, 952.

Lopes-Lima, M., Froufe, E., Do, V.T., Ghamizi, M., Mock, K.E., Kebapçı, Ü., Klishko,

O., Kovitvadhi, S., Kovitvadhi, U., Paulo, O.S., Pfeiffer, J.M., Raley, M.,

Riccardi, N., Şereflişan, H., Sousa, R., Teixeira, A., Varandas, S., Wu, X.,

Zanatta, D.T., Zieritz, A., Bogan, A.E., 2017. Phylogeny of the most species-

rich freshwater bivalve family (Bivalvia: Unionida: Unionidae): Defining

modern subfamilies and tribes. Molecular Phylogenetics and Evolution 106,

174–191.

Margulis, L., 2004. Serial endosymbiotic theory (SET) and composite individuality.

Microbiology Today 31, 172–175.

Martin, S.M., 1997. Freshwater Mussels (Bivalvia: Unionoida) of Maine. Northeastern

Naturalist 4, 1.

Matteson, M.R., 1955. Studies on the Natural History of the Unionidae. The American

Midland Naturalist 53, 126–145.

Milani, L., Ghiselli, F., Guerra, D., Breton, S., Passamonti, M., 2013. A comparative

analysis of mitochondrial ORFans: new clues on their origin and role in species

with doubly uniparental inheritance of mitochondria. Genome Biology and

Evolution 5, 1408–1434.

Mitchell, A., Guerra, D., Stewart, D., Breton, S., 2016. In silico analyses of

mitochondrial ORFans in freshwater mussels (Bivalvia: Unionoida) provide a

framework for future studies of their origin and function. BMC Genomics 17,

597.

Moritz, C., Dowling, T.., Brown, W.., 1987. evolution of animal mitochondrial DNA:

Relevance for population biology and systematics. Annual Review of Ecology

and Systematics 18, 269–292.

Morton, B., 1991. Do the Bivalvia demonstrate environment-specific sexual strategies?

A Hong Kong model. Journal of Zoology 223, 131–142.

Naddafi, R., Rudstam, L.G., 2014. Predation on invasive zebra mussel, Dreissena

polymorpha, by pumpkinseed sunfish, rusty crayfish, and round goby.

Hydrobiologia 721, 107–115.

Nagel, K.O., Badino, G., Alessandria, B. 1996. Population genetics of European

Anodontinae (Bivalvia: Unionidae). Journal of Molluscan Studies 62, 343–357.

Nass, M.M.K., Nass, S., 1963a. Intramitochondrial Fibers with Dna Characteristics. The

Journal of Cell Biology 19, 593–611.

Nass, S., Nass, M.M.K., 1963b. Intramitochondrial fibers with DNA characteristics. The

Journal of Cell Biology 19, 613–629.

Niemeyer, B., 1992. Vergleichende utersuchungen zur bionomischen strategie der

teichmuscherlarten Anodonta cygnea L. und Anodonta anatina L. Thesis,

Universität Hannover.

Obata, M., Kamiya, C., Kawamura, K., Komaru, A., 2006. Sperm mitochondrial DNA

transmission to both male and female offspring in the blue mussel Mytilus

galloprovincialis. Development, Growth and Differentiation 48, 253–261.

Ohta, T., 2002. Near-neutrality in evolution of genes and gene regulation. PNAS 99,

16134–16137.

Ortmann, A.E., 1920. Correlation of shape and station in fresh-water mussels (Naiades).

Proceedings of the American Philosophical Society 59, 269–312.

Ortmann, A.E., 1918. The Nayades (Freshwater Mussels) of the upper Tennessee

Drainage. With Notes on Synonymy and Distribution. Proceedings of the

American Philosophical Society 57, 521–626.

Ortmann, A.E., 1909. The Destruction of the Fresh-Water Fauna in Western

Pennsylvania. Proceedings of the American Philosophical Society 48, 90–110.

Passamonti, M., Ghiselli, F., 2009. Doubly uniparental inheritance: two mitochondrial

genomes, one precious model for organelle DNA inheritance and evolution.

DNA and Cell Biology. 28, 79–89.

Passamonti, M., Scali, V., 2001. Gender-associated mitochondrial DNA heteroplasmy

in the venerid clam Tapes philippinarum (Mollusca Bivalvia). Current Genetics

39, 117–124.

Peñarrubia, L., Vidal, O., Viñas, J., Pla, C., Sanz, N., 2016. Genetic characterization of

the invasive zebra mussel (Dreissena polymorpha) in the Iberian Peninsula.

Hydrobiologia 779, 227–242.

Rand, D.M., 2001. The units of selection of mitochondrial DNA. Annual Review of

Ecology and Systematics 32, 415–448.

Rand, D.M., 1994. Thermal habit, metabolic rate and the evolution of mitochondrial

DNA. Trends in Ecology & Evolution 9, 125–131.

Rathke, J., 1797. Om dam-muslingen. Naturhist Selskabets Skr Kjöbenhavn 139–179.

Rawson, P.D., Hilbish, T.J., 1995. Evolutionary relationships among the male and

female mitochondrial DNA lineages in the Mytilus edulis species complex. Mol.

Biol. Evol. 12, 893–901.

Reis, J., 2003. The freshwater pearl mussel Margaritifera margaritifera (L.) (Bivalvia,

Unionoida) rediscovered in Portugal and threats to its survival. Biological

Conservation 114, 447–452.

Remington, P.S., Jr., Clench, W.J., 1925. Vagabonding for shells. Nautilus 127–143.

Rhoads, S.N., 1899. On a recent collection of Pennsylvanian mollusks from the Ohio

River system below Pittsburg. Nautilus 12, 133–138.

Ricciardi, A., MacIsaac, H.J., 2000. Recent mass invasion of the North American Great

Lakes by Ponto–Caspian species. Trends in Ecology & Evolution 15, 62–65.

Roper, D.S., Hickey, C.W., 1994. Population structure, shell morphology, age and

condition of the freshwater mussel Hyridella menziesi (Unionacea: Hyriidae)

from seven lake and river sites in the Waikato River system. Hydrobiologia 284,

205–217.

Saavedra, C., Reyero, M.-I., Zouros, E., 1997. Male-dependent doubly uniparental

inheritance of mitochondrial DNA and female-dependent sex-ratio in the mussel

Mytilus galloprovincialis. Genetics 145, 1073–1082.

Saccone, C., De Giorgi, C., Gissi, C., Pesole, G., Reyes, A., 1999. Evolutionary

genomics in Metazoa: the mitochondrial DNA as a model system. Gene 238,

195–209.

Salvioli, S., Bonafè, M., Capri, M., Monti, D., Franceschi, C., 2001. Mitochondria,

aging and longevity–a new perspective. FEBS letters 492, 9–13.

Sato, M., Sato, K., 2013. Maternal inheritance of mitochondrial DNA by diverse

mechanisms to eliminate paternal mitochondrial DNA. BBA - Molecular Cell

Research 8, 1979–1984.

Seshaiya, R., 1969. Some observations on the life histories of South Indian freshwater

mussels. Malacologia 286–287.

Shadel, G.S., Clayton, D.A., 1997. Mitochondrial DNA maintenance in vertebrates.

Annual Review of Biochemistry 66, 409–435.

Shea, C.P., Peterson, J.T., Wisniewski, J.M., Johnson, N.A., 2011. Misidentification of

freshwater mussel species (Bivalvia:Unionidae): contributing factors,

management implications, and potential solutions. Journal of the North

American Benthological Society 30, 446–458.

Simpson, C.T., 1899. The pearly fresh-water mussels of the United States: their habits,

enemies, and diseases: with suggestions for their protection. US Government

Printing Office.

Skibinski, D.O., Gallagher, C., Beynon, C.M., 1994. Sex-limited mitochondrial DNA

transmission in the marine mussel Mytilus edulis. Genetics 138, 801–809.

Smith, D., 1979. Sexual characteristics of Margaritifera margaritifera (Linnaeus)

populations in central new-england. The veliger. 21, 381–383.

Smith, G., 1908. Sex in the Crustacea, with special reference to the origin and nature of

hermaphroditism. Report of the British Association for the Advancement of

Science 543–544.

Smith, H.M., 1899. The mussel fishery and pearl-button industry of the Mississippi

River. Bulletin of the U.S. Fish Commission 18, 289–314.

Soroka, M., Burzyński, A., 2015. Complete female mitochondrial genome of Anodonta

anatina (Mollusca: Unionidae): confirmation of a novel protein-coding gene (F

ORF). Mitochondrial DNA 26, 267–269.

Soroka, M., Burzyński, A., 2014. Complete male mitochondrial genome of Anodonta

anatina (Mollusca: Unionidae). Mitochondrial DNA 1–2.

Stansbery, D.H., 1970. Eastern freshwater mollusks. (I) The Mississippi and St.

Lawrence river systems. Malacologia 9–22.

Stewart, D.T., Hoeh, W.R., Bauer, G., Breton, S., 2013. Mitochondrial genes, sex

determination and hermaphroditism in freshwater mussels (Bivalvia:

Unionoida), in: Evolutionary Biology: Exobiology and Evolutionary

mechanisms. Springer, Berlin, Heidelberg, pp. 245–255.

Stewart, D.T., Kenchington, E.R., Singh, R.K., Zouros, E., 1996. Degree of selective

constraint as an explanation of the different rates of evolution of gender-specific

mitochondrial DNA lineages in the mussel Mytilus. Genetics 143, 1349–1357.

Stoeckle, M.Y., Hebert, P.D.N., 2008. Barcode of Life. Scientific American 299, 82–88.

Strayer, D.L., Downing, J.A., Haag, W.R., King, T.L., Layzer, J.B., Newton, T.J.,

Nichols, S.J., 2004a. Changing perspectives on pearly mussels, North America’s

most imperiled animals. BioScience 54, 429–439.

Strayer, D.L., Hattala, K.A., Kahnle, A.W., 2004b. Effects of an invasive bivalve

(Dreissena polymorpha) on fish in the Hudson River estuary. Canadian Journal

of Fisheries and Aquatic Sciences 61, 924–941.

Sutherland, B., Stewart, D., Kenchington, E.R., Zouros, E., 1998. The fate of paternal

mitochondrial DNA in developing female mussels, Mytilus edulis: implications

for the mechanism of doubly uniparental inheritance of mitochondrial DNA.

Genetics 148, 341–347.

Theologidis, I., Fodelianakis, S., Gaspar, M.B., Zouros, E., 2008. Doubly uniparental

inheritance (DUI) of mitochondrial DNA in Donax trunculus (Bivalvia:

Donacidae) and the problem of its sporadic detection in Bivalvia: DUI

distribution in Bivalves. Evolution 62, 959–970.

Trautman, M.B., 1981. The fishes of Ohio: with illustrated keys. Ohio State University

Press.

Tudorancea, C., 1972. Studies on Unionidae populations from the Crapina-Jijila

complex of pools (Danube zone liable to inundation). Hydrobiologia 39, 527–

561.

Upholt, W.B., Dawid, I.B., 1977. Mapping of mitochondrial DNA of individual sheep

and goats: Rapid evolution in the D loop region. Cell 11, 571–583.

Utterback, W.I., 1915. The Naiades of Missouri. American Midland Naturalist 4, 41.

Valenzuela, N., Adams, D.C., Janzen, F.J., 2003. Pattern does not equal process: exactly

when is sex environmentally determined? The American Naturalist 161, 676–

683.

Van der Schalie, H., 1970. Hermaphroditism among North American freshwater

mussels. Malacologia 93–112.

Van der Schalie, H., 1969. Two unusual unionid hermaphrodites. Science 163, 1333–

1334.

Villalba, A., 1995. Gametogenic cycle of cultured mussle, Mytilus galloprovincialis, in

the bays of Galicia (N.W. Spain). Aquaculture 2–3, 269–277.

Villalobos, M.D., Paz, A.O., González, M.T., 2010. Hermafroditismo en el Chorito

Perumytilus purpuratus (Lamarck, 1819), (Mollusca: Mytilidae). International

Journal of Morphology 28, 569–573.

Watters, G.T., 1996. Small dams as barriers to freshwater mussels (Bivalvia,

Unionoida) and their hosts. Biological Conservation 75, 79–85.

Waugh, J., 2007. DNA barcoding in animal species: progress, potential and pitfalls.

Bioessays 29, 188–197.

Weber, T.A., Reichert, A.S., 2010. Impaired quality control of mitochondria: aging

from a new perspective. Experimental gerontology 45, 503–511.

Wei, Y.-H., 1998. Oxidative Stress and Mitochondrial DNA Mutations in Human

Aging. Proceedings of the Society for Experimental Biology and Medicine 217,

53–63.

Wei, Y.-H., Lee, H.-C., 2002. Oxidative Stress, Mitochondrial DNA Mutation, and

Impairment of Antioxidant Enzymes in Aging 1. Experimental biology and

medicine 227, 671–682.

Weisensee, H., n.d. Die Geschlechtsverhaltnisse und der Geschlechtsapparat bei

Anodonta. Zeitschrift für wissenschaftliche Zoologie 262–335.

Whelan, N.V., Geneva, A.J., Graf, D.L., 2011. Molecular phylogenetic analysis of

tropical freshwater mussels (Mollusca: Bivalvia: Unionoida) resolves the

position of Coelatura and supports a monophyletic Unionidae. Molecular

Phylogenetics and Evolution 61, 504–514.

White, D.J., Wolff, J.N., Pierson, M., Gemmell, N.J., 2008. Revealing the hidden

complexities of mtDNA inheritance. Molecular Ecology 17, 4925–4942.

Williams, J.D., Fradkin, A., 1999. Fusconaia apalachicola, a new species of freshwater

mussel (Bivalvia: Unionidae) fom precolumbian archaeological sites in the

Apalachicola Basin of Alabama, Florida, and Georgia. Tulane Studies in

Zoology and Botany 31, 51–62.

Willoughby, L., 2012. Flowing Through Time: A History of the Lower Chattahoochee

River. University of Alabama Press.

Wilson, A.C., Cann, R.L., Carr, S.M., George, M., Gyllensten, U.B., Helm-Bychowski,

K.M., Higuchi, R.G., Palumbi, S.R., Prager, E.M., Sage, R.D., Stoneking, M.,

1985. Mitochondrial DNA and two perspectives on evolutionary genetics.

Biological Journal of the Linnean Society 26, 375–400.

Wilson, C.B., Clark, H.W., 1914. The mussels of the Cumberland River and its

tributaries. United States Government Publishing Office.

Wong, C.-K., 1989. The ecology of exposed sandy beaches in Hong Kong with

particular reference to Donax semigranosus and D. cuneatus (Bivalvia:

Donacidae). HKU Theses Online (HKUTO).

Wood, E.M., 1974. Development and morphology of the glochidium larva of Anodonta

cygnea (Mollusca: Bivalvia). Journal of Zoology 173, 1–13.

Zouros, E., 2013. Biparental Inheritance through uniparental transmission: The doubly

uniparental inheritance (DUI) of mitochondrial DNA. Evolutionary Biology 40,

1–31.

Zouros, E., 2000. The exceptional mitochondrial DNA system of the mussel family

Mytilidae. Genes and Genetic Systems 75, 313–318.

Chapter 2

The complete mitochondrial genome of the hermaphroditic freshwater mussel Anodonta cygnea (Bivalvia: Unionidae)

Chase E.E.1, Robicheau B.M.2, Veinot S.1, Bauer G., Breton S.4, Stewart D.T.1

1 Department of Biology, Acadia University, Wolfville, NS, Canada

2 Department of Biology, Dalhousie University, Halifax, NS, Canada

3 Institut für Biologie, Universität Freiburg, Freiburg, Germany

4 Département de Sciences Biologiques, Université de Montréal, QC, Canada

ABSTRACT

Doubly uniparental inheritance of mitochondrial DNA (mtDNA) in bivalves remains a fascinating and active area of research. There are still many questions left unanswered, including ones relating to the role of the mitochondrial genome in sex determination or sexual development in freshwater mussels (order Unionoida). In this study, one complete mitochondrial genome of the swan mussel, Anodonta cygnea, is sequenced, analyzed and compared to the complete mitochondrial genome of the duck mussel,

Anodonta anatina, a dioecious species. An in silico assessment of novel proteins found within bivalve species (known as the F-, H-, and M-open reading frame or ORFs) is conducted, with special attention to putative transmembrane domains, signal peptides, signal cleavage sites, subcellular localization and potential control regions. The number and topology of putative transmembrane domains appears to be maintained among both

F- and H-ORFs, however, this is not the case for M-ORFs. The in silico analyses suggested the presence of signal peptides and signal cleavage sites and provided some insights into possible function of these novel ORFs, but confidence in these structures and functions was highly variable, which may be due to the extreme novelty of these proteins. In situ analysis will be necessary to further explore the intracellular localizations and potential functions of these ORFs in the process of sex determination in freshwater mussels.

INTRODUCTION

Order Unionoida is composed of freshwater mussels (FWM) that inhabit rivers, lakes and ponds. FWMs exhibit several fascinating reproductive and life-history strategies.

For example, their larval stage (glochidia) are parasitic on one or more species of fish.

Glochidia attach to fish, most commonly their gills, after being expelled from their parent, which facilitates dispersal to new habitats (Bogan 1993). A fascinating characteristic of dioecious FWMs is that they do not exhibit strictly maternal inheritance

(SMI) of mitochondrial (mt) DNA. In animals with SMI, maternal mtDNA is exclusively passed on to all offspring (Birky 1995). In contrast, many species of FWMs, marine mussels (order Mytiloida), nut shells (order Nuculanoida), and marine clams

(order Veneroida), mtDNA is transmitted via an unusual system called doubly uniparental inheritance or DUI (see Breton et al. 2007; Zouros 2013; Gusman et al.

2016 for reviews). Under DUI, female offspring inherit and maintain maternal mtDNA

(called female-transmitted or F-type mtDNA), while male offspring inherit and maintain

both maternal mtDNA and paternal mtDNA (called male-transmitted or M-type mtDNA). Consequently, with respect to mitotype, males are heteroplasmic and females are homoplasmic (Stewart et al. 2013). In male offspring, F-type mtDNA is found throughout their somatic tissue, whereas M-type mtDNA is contained predominantly within gametic cells (Garrido-Ramos et al. 1998; Dalziel and Stewart 2002). The respective M-type and F-type mtDNA genomes are also unusual, in contrast to typical animal mtDNA gene content, as each of the sex-associated mtDNA genomes contains their own novel open reading frames (ORFs, Breton et al. 2009). These are referred to as the m-orf and f-orf in the male and female mt genomes, respectively, and these two regions code for proteins that do not show evidence of homology to each other or to any other known proteins (Breton et al. 2009; but also see Guerra et al. 2017). As a result of the paternal and maternal lineages participating in their own independent forms of uniparental inheritance, M-types and F-types evolve on separate evolutionary trajectories. This also frees M-types from being evolutionary dead-ends, as would otherwise typically be the case under an SMI mode of transmission (Passamonti and

Giselli 2009).

Recent work has shown that hermaphroditic species of FWM do not transmit, or even possess, M-type mtDNA (Breton et al. 2011; Stewart et al. 2013). Instead hermaphrodites have reverted to SMI and transmit only a modified F-type mt genome, termed the hermaphroditic or H-type (Breton et al. 2011). In several hermaphroditic species examined to date, H-type mtDNA carries a modified f-orf gene that has experienced considerable mutational divergence into what is termed the h-orf gene

(Breton et al. 2011). Hermaphroditism has evolved independently multiple times in

FWMs, therefore the h-orfs documented in multiple species of FWM are not homologous (Breton et al. 2011; Stewart et al. 2013). Notably, Breton et al. (2011) identified four independent transitions from a dioecious to hermaphroditic sexual strategy in the following genera: Lasmigona, Margaritifera, Toxolasma and

Utterbackia. Interspecific comparisons of hermaphroditic FWM species show that h- orfs have highly divergent nucleotide sequences and variable amino acid hydrophobicity profiles, whereas f-orfs are much more conserved in both respects (Breton et al. 2011;

Stewart et al. 2013). Comparing amino acid sequences and hydrophobicity profiles of f- orfs and closely related h-orfs suggested that this apparent molecular modification often includes: (1) a lengthening of the sequence, possibly due to relatively relaxed selection compared to closely related f-orfs, (2) the introduction of repeating motifs of varying lengths, and (3) the presence of transmembrane (TM) protein domains (Breton et al.

2011).

We anticipate that the analysis of additional FWM h-orfs will provide insight into the transition from dioecy to monoecy, the role of these novel protein-coding genes (m-orf, f-orf, and h-orf) and mt inheritance (DUI and SMI) in general. Thirty-four FWM species have at least one complete mt genome sequenced and are available on GenBank for comparative analyses. To this list, we are adding the complete sequence of the mt genome of a hermaphroditic population of another species, the swan mussel, Anodonta cygnea (order Unionoida; family Unionidae) (Table S1). A. cygnea is found across the

British and Irish Isles, Europe, Siberia and northern Africa (Pourang et al. 2009). This

species reportedly exhibits hermaphroditism in standing or slow moving waters, however, dioecious individuals have been reported in rivers (Chojnacki et al. 2007).

Both the M-type and F-type mt genomes are available for A. anatina (a close relative of

A. cygnea), as well as for other FWMs listed in Table S1. Complete H-type mtDNA sequences are now available for five species (including the new H-type genome described herein): A. cygnea, Toxolasma parvum, Utterbackia imbecillis, Lasmigona compressa and Lasmigona subviridis. A partial sequence is available for H-type

Lasmigona subviridis, which excludes cob, nad5 and neighboring transfer RNAs

(tRNAs).

Given the novelty of this unusual system of mtDNA inheritance in bivalves, there is considerable interest in characterizing the structure and function of the various elements of the bivalve F- and M-type mt genomes (e.g., Breton et al. 2014; Passamonti and

Ghiselli 2009; Zouros 2013). The mitochondrial control region of marine mussels (order

Mytiloida) has been identified as the longest unassigned region (LUR) (see Cao et al.

2004). Within FWMs, Breton et al. (2009) suggest that the region between nad5 and trnQ (nad5-trnQ) is the control region. Mt control regions are located in non coding stretches of DNA, and include secondary structures that act as the origin of replication.

DUI permits the existence and inheritance of F-, H- and M-type mtDNA, therefore it is important to confirm whether or not there is a difference in control regions (or initiating replication in general) among these mitotypes (Cao et al. 2004). We cannot assume each mitotype functions in the same way, especially given the presence of mitotype specific

ORFs. The findings of Breton et al. (2009) reference typical properties of control

regions such as a high A-T percentage, repeating units, and stemloops/hairpin structures, which are all present in the nad5-trnQ region of the F-type mt genomes examined. A main control region or multiple control regions have yet to be suggested for H-type mtDNA of FWM. In addition, control regions of marine mussels have been shown to contain unique ORFs and/or DNA sequence motifs thought to play a role in sexual development or sequestration of the M-type mitochondria into the spermatogonia during sexual maturation (e.g., Kyriakou et al. 2015; Robicheau et al. 2017). One objective of the present study is to evaluate and characterize potential control regions of the A. cygnea mt genome. F- and M-type mtDNA of FWMs share three sizable regions of non-coding DNA; these shared unassigned regions (SURs) include the putative FWM nad5-trnQ control region as well as two areas flanking nad3 termed trnF-nad3 and nad3-trnH+A in F-type mtDNA. Accordingly, the properties of these regions in A. cygnea will be subjected to a thorough in silico comparative analysis, which consists of an attempt to gain insights into possible functions of complex molecular and cellular processes using computer-based analyses of nucleotide and amino acid sequence data

(e.g., Mitchell et al. 2016).

Immunostaining localized the F-ORF protein in not only the mitochondria of mature egg cells, but also on the nuclear membrane and in the nucleoplasm of the ellipse,

Venustaconcha ellipsiformis (Bivalvia: Unionidae; Breton et al. 2011). The movement of the F-ORF protein from the mitochondria to the nucleus infers a function outside of the primary mitochondrial function of oxidative phosphorylation (Breton et al. 2011).

This finding, along with the transition of sexual strategies accompanied by a loss of the

m-orf and degeneration of the f-orf to a h-orf, suggests a link between sex determination and these novel ORFs in species exhibiting DUI (Breton et al. 2011). Previous in silico analysis of M-ORFs and F-ORFs has been conducted by Milani et al. (2013) in marine mussels, and of M-ORFs, F-ORFs and H-ORFs in freshwater mussels by Mitchell et al.

(2016). Both studies explored protein secondary structures in these novel ORFs, including potential transmembrane domains (TMs) and signal peptides (SPs). They similarly concluded a potential viral origin of these novel ORFs, although Mitchell and colleagues also proposed that the m-orf and f-orf could also have arisen via duplication of the nad2 and atp8 genes, respectively, and they also identified several potential functions for these proteins.

For comparative purposes, A. anatina, the most closely related species for which complete mtDNA is available, will be compared to A. cygnea herein. Another pair of closely related species, Utterbackia imbecillis (hermaphroditic) and U. peninsularis

(dioecious), will be included in an analysis of molecular evolutionary patterns. These two Utterbackia species were previously characterized by Breton et al. (2009). In addition, an in silico analysis of H-ORFs, F-ORFs, and M-ORFs, will be performed to further explore the transition from dioecy to monoecy in FWMs with a focus on the potential movement of these proteins through subcellular localization within the cell.

Finally, an assessment of the location of putative control region or control regions within SURs of complete FWM mt genomes will be conducted using the data from

Breton et al. (2009) and the novel H-type mtDNA sequence presented herein.

MATERIALS AND METHODS

Sequencing the complete mitochondrial genome of Anodonta cygnea

Anodonta cygnea samples were obtained from Lake Konstanz, along the border of

Germany and Switzerland, and stored in AllProtect® (QIAGENTM). Genomic DNA was extracted using a saturated NaCl protocol similar to Appendix B in Sambrook et al.

(1989), except Queen’s lysis buffer was used (Seutin et al. 1991). Long-range polymerase chain reactions (LR-PCRs) were used to obtain DNA from the entire circular mtDNA molecule. This was accomplished by amplifying two segments of mtDNA using species specific primers targeting two genes: cox1 and rrnL (Table 1).

LR-PCRs were conducted using a Phusionâ High-Fidelity DNA Polymerase Kit (New

England BioLabs Inc.) and its associated protocol. Thermocycling conditions were: activation at 98°C for 30 seconds followed by 35 cycles of denaturation at 98°C for 15 seconds, annealing at 64°C for 40 seconds, extension at 72°C for 7 minutes, and a final extension at 72°C for 10 minutes. LR-PCR products were sequenced using PacBio RSII sequencing technology and reads were assembled with de novo assembly (see Rhoads and Au 2015) at the McGill University and Génome Québec Innovation Centre,

Montréal, Canada. The resulting circular mtDNA was annotated for protein coding and

RNA genes using MITOS WebServer revision 917 (Bernt et al. 2013), and validated using MFannot (Beck and Lang 2015). Putative tRNA genes were further confirmed and their structure assessed using tRNAScan-SE 2.0 (Lowe and Eddy 1997). The gene order of the mt genome was visualized using Geneious version 9.1.7 (Kearse et al.

2012). Images of tRNA secondary structures were also generated in the MITOS

WebServer revision 917 (Bernt et al. 2013).

Determining potential control regions of Anodonta cygnea mtDNA

Shared unassigned regions (SURs) of the various Anodonta genomes were assessed for control regions as follows. Secondary structures were identified using mfold webserver

(Zuker 2003) and results were visualized using VARNA version 3.93 (Darty et al.

2009). SURs were ran through REPFIND Webserver 4.09 (Benson 1999) to check for tandem repeats, which may provide variation in function for multiple protein products transcribed and translated from a single gene following post-transcriptional or post- translational modification (Fuchs 2013). A-T nucleotide content was checked with

Geneious (Kearse et al. 2012) and a comparison of SURs in A. cygnea, A. anatina, U. imbecillis and U. peninsularis was conducted in Geneious to calculate sequence similarity.

Mapping of predicted transmembrane domains on phylogenetic trees of the F- types, H-types and M-types

For the purpose of mapping properties of the predicted TMs of the F-type, H-type and

M-type open reading frames in a historical context, phylogenetic trees were produced based on cox1 sequences of FWMs obtained from GenBank (Table S2). Basing these trees on cox1 sequence data (rather than on complete genomes or multiple genes) allowed us to maximize the number of F-, H- and M-ORF proteins that could be included in the analysis because (1) cox1 sequence data are relatively abundant in

comparison to complete mtDNA and (2) as including other genes that are less readily available would also reduce the number of species that could be included in the analysis. F- and H-type cox1 sequences were aligned using ClustalW (Thompson et al.

1994) with default parameters (i.e., Interational Union of Biochemistry [IUB] cost matrix, a gap open cost of 15 and a gap extend cost of 6.66). M-type sequences were aligned separately in the same way. A Tamura-Nei model of evolution (Tamura and Nei

1993) was used, as per the recommendation from the model test function in MEGA

7.0.16 (Kumar et al. 2016). Separate Bayesian inference (BI) F-/H-type and M-type trees were produced using BEAUti and BEAST version 2.4.6 with a Yule speciation process (Steel and McKenzie 2001) and 100 million Markov chain Monte Carlos

(MCMC; Drummond 2002) steps with samples taken every 1000 steps. A burn in of

10% was performed on the resulting trees. Tree samples were visualized using

DensiTree included with BEAST version 2.4.6 package (Bouckaert 2010), and compiled into one “best” topology using TreeAnnotator version 1.4 (Rambaut and

Drummond 2002) and viewed using FigTree version 1.4.3 (Rambaut 2007). TM domain predictions for the F-, H- and M-ORFs were run using the transmembrane Hidden

Markov Model (TMHMM) version 0.9 in a Geneious plugin and using the TMHMM

Server version 2.0 (Krogh et al. 2001). These programs determine potential TM locations, number of potential TMs, and their orientation through the membrane (i.e., topology). Mesquite 3.2 (Maddison and Maddison 2017) was used to assess the ancestral state and most parsimonious reconstruction of transitions for characters such as different TM topology and numbers of TMs present in an ORF.

Calculating nucleotide and amino acid divergence

Nucleotide and amino acid divergence values between A. cygnea and A. anatina, U. imbecillis and U. peninsularis, respectively, were calculated for each gene with MEGA

7.0.16 (Kumar et al. 2016) using the Tamura-Nei model of evolution and a gamma distribution. Results were visualized with statistical program R 3.2.2 using the package

“lattice” (Figure 2).

Exploring in silico analysis of H-ORF, F-ORF and M-ORF properties

An alignment of all available unionid H-ORF and F-ORF proteins (Table S3) was conducted using ClustalW (Thompson et al. 1994) with default settings (e.g., a

BLOSUM cost matrix and a gap open cost of 10) as implemented by the package

Geneious. H-ORFs were assessed for SPs using the TargetP 1.1 (Guda 2005) server with default settings. If a putative SP was detected using TargetP, then the SignalP 4.1 server (Emanuelsson et al. 2007) was used to further characterize the validity of a potential SP using the “sensitive” D-cutoff value. As an alternative strategy for maximizing the detection of SPs, H-ORFs were run through SignalP a second time with the method set to exclude the possibility of TMs. Subcellular locations of H-ORFs were predicted using iLoc-Animal (Lin et al. 2013) and Euk-mPLoc 2.0 (Cell-PLoc 2.0 package; Chou and Shen 2010). The program iLoc-Animal checks for probable localizations to the following 20 subcellular sites: acrosome, cell membrane, centriole, centrosome, cell cortex, cytoplasm, cytoskeleton, endoplasmic reticulum, endosome, extracellular, Golgi apparatus, lysosome, mitochondrion, melanosome, microsome, nucleus, peroxisome, plasma membrane, spindle and synapse. Euk-mPLoc searches for

the same localizations, plus additional sites specific to plant cells. Both programs can assess multiple localizations of a protein (i.e., multiplex proteins). H-ORFs were also assessed for amino acid property composition using EMBOSS Pepstats (Rice et al.

2000). M-ORFs were assessed for SPs and subcellular localization. All ORFs were also assessed for signal cleavage sites (SCS) using Geneious and TargetP.

RESULTS

A description of the complete mtDNA genome of a hermaphroditic Anodonta cygnea

The complete mtDNA of Anodonta cygnea consists of all 13 protein coding genes, the two rRNA genes, and 22 tRNAs typically found in FWM mtDNAs (Figure 1) and other animal mtDNA (Boore 1999). This genome has the same gene order as the F-type of A. anatina. Similar to A. anatina (Soroka and Burzyński 2015), both trnS1 and trnS2 are missing a dihydrouridine arm (Figure S1), which is also the case in other FWMs (e.g.,

Lampsilis ornata; Serb and Lydeard 2003) and members of order Mytiloida (Breton et al. 2006). A fourteenth putative gene, the h-orf, is located between nad2 and trnE, the same location as the f-orf in A. anatina. The size of the genome is also typical of FWM mtDNAs (15,607 bp), and it does not have any obvious duplication events. However, the h-orf does not possess the molecular features associated with most h-orfs as described by Breton et al. (2011). For example, the A. cygnea h-orf is shorter than most h-orfs and it does not possess any repetitive DNA sequence motifs that are found in several other h-orfs. Another distinguishing feature is that the hydrophobicity profile of the h-orf is not strongly similar to that of A. anatina (see Veinot et al. in prep).

Methionine (nucleotide triplet ATG) was the most frequent start codon (atp6, atp8, cox2, and cox3), but several alternative start codons were found: isoleucine (ATC) in cob and nad1, isoleucine (ATT) in nad3, nad4, nad6, leucine (TTG) and valine (GTG) in nad2, and nad5, respectively. Valine also acts as a start codon in Lampsilis ornata

(Serb and Lydread 2003). Stop codons included TAG (atp6, atp8, cox1, cox3, nad2, nad3, nad4), TGA (cox2) and TAA (nad1, nad5 and nad6). The overall G-C content is

35.6%, making the A-T content relatively high at 64.4%. The most frequent amino acid is leucine at 10.3%, followed by serine at 8.7% and lysine at 8.3%. These results are similar to A. anatina, which has a G-C content of 34.0%, leucine at 15.4%, serine at

10.6%, and an overall size of 15,653bp. Although leucine and serine are the top two most frequent amino acids appearing in A. anatina, lysine only accounts for 2.5% of amino acids.

Nucleotide and amino acid divergences are smaller between A. cygnea and A. anatina than between A. cygnea and either U. imbecillis or U. peninsularis (Figure 2), and unsurprisingly A. cynea and A. anatina were sister species based on a phylogeny of unionid partial cox1 sequences (Figure 5). Thus, the mt genome of this hermaphroditic

A. cygnea individual is more similar to that of a dioecious female than to the mitochondrial genome of other hermaphroditic species. This is in agreement with the hypothesis that hermaphroditism in A. cygnea is the result of convergent evolution, as was also inferred for multiple cases of hermaphroditism in other unionids (Breton et al.

2011). Within A. cygnea, cox1, cox2 and cox3 are among the least divergent genes and atp8 is the most divergent, which is typical of FWM (Breton et al. 2011) and bivalves in

general (Breton et al. 2010). The remaining mtDNA genes had divergence values intermediately between atp8 and the cox genes.

Characterization of the shared unassigned regions in H-type mtDNA and a comparison of molecular divergence patterns between H-type and F-type mtDNA of closely related species

All available H-type mt genomes have the same shared unassigned regions (SURs) as found in other FWM F-type mt genomes. The percentage of identical bases between A. anatina and A. cygnea within SURs nad3-trnH+A, nad5-trnQ and trnF-nad5 were 83.9,

63.8 and 56.6, respectively. Between U. imbecillis and U. peninsularis the percentage of identical bases for these three SURs were 71.1, 60.0 and 68.8, respectively. The smallest SUR was trnF-nad5 in A. anatina, U. imbecillis and U. peninsularis, however, this is the largest SUR in A. cygnea. At 442bp, the trnF-nad5 SUR in A. cygnea is larger than the next largest SUR in these four species (i.e., the nad3-trnH+A SUR of U. imbecillis at 282bp). Due to its size, the nad3-trnH+A SUR of A. cygnea was analyzed further, and an ORF was detected using Geneious. The resulting ORF of 62 amino acids was then assessed for signal cleavage sites (SCS), SPs, and presence of TM. The topologies of the TM domains were also reconstructed. Geneious predicted two potential TM domains, however, the predicted topology of the TM structure (i.e., extracellular-TM-cytoplasmic-TM-extracellular) was not particularly strong (less than

100% likelihood). Both TargetP and SignalP detected an SP in this sequence. TargetP suggested a secretory SP, with a reliability class (RC) of 4 (where 1 is the strongest prediction and 5 is the least). SignalP was run under sensitive parameters including the

potential of finding a TM. A nucleotide BLAST search (Altschul et al. 1990; Wheeler et al. 2007) was conducted using both the nucleotide and amino acid sequence generated from the ORF. The nucleotide search had one alignment hit with 93% identity and 35% coverage of the query sequence matching U. peninsularis (HM856636) within its nad5 gene (E-value = 6×10-16). The protein search had 49 alignment hits, with Lasmigona compressa (another hermaphroditic FWM) being the top hit with 100% query coverage, and a 68% identity similarity (E-value = 6×10-16). Among the 49 hits, query coverage ranged from 90%-100% and the percent identity ranged from 46-68. The above analysis indicated that the identified ORF was putatively related to nad5. As such, the annotated nad5 from A. cygnea and A. anatina were compared to elucidate how the ORF is related to the nad5 gene. An alignment of the NAD5 protein from A. cygnea and A. anatina showed a percent identity of 68.8 overall, with a major gap in A. cygnea beyond amino acid 420 of A. anatina. NAD5 in A. cygnea is 448 amino acids in length, whereas that of

A. anatina is 578 amino acids long. The predicted ORF protein from the A. cygnea trnF- nad5 SUR has a percent identity of 61.3 when aligned with NAD5 of A. anatina. The alignment places this ORF starting at amino acid 453 of A. anatina. Among all species, the trnF-nad5 SUR was highly variable; the largest percent similarity between A. cygnea and any other species was with T. parvum at 63.2. Among the other two SURs, i.e. the nad3-trnH+A SUR and the nad5-trnQ SUR, the highest percent similarity was with L. compressa in both cases at 73.5 and 65.8, respectively. At 1344 bp, nad5 of A. cygnea is smaller than other FWM nad5 genes; for example, the F-type A. anatina nad5 is 1737bp, F-type U. peninsularis is 1752bp and H-type U. imbecillis is 1743 bp. The

implication of the potential translocation of a part of nad5 within the A. cygnea mtDNA genome is discussed below.

An assessment of shared unassigned regions for control region properties.

Based on an assessment of all SURs in available H-type mtDNA (Table 2), the trnF- nad5 SUR is less likely to be the main control region compared to the other SURs due to its small size; however, in all instances this SUR does exhibit putative secondary structures. H-type SURs appear to rarely contain repeat units, but do tend to have high

A-T content. The percent sequence divergence between the proposed FWM F-type and

M-type mtDNA proposed control region (i.e., the region between nad5-trnQ; Breton et al. 2009) ranges from 43-50. Between F-type mt genomes of A. anatina and U. peninsularis, and the corresponding closely related H-type mt genomes, A. cygnea and

U. imbecillis, the nad5-trnQ region have similar percent sequence divergences (i.e.,

63.8 and 60.0, respectively). The trnF-nad3 SUR of A. cygnea possesses many secondary structures including one with strong similarity to a typical metazoan control region origin of replication (Figure 3).

In silico analysis of H-ORFs compared to F-ORFs

The amino acid sequence length of H-ORFs varies considerably with the smallest reported to date being A. cygnea at 34 amino acids long and the largest, U. imbecillis, with 286 amino acids. TM predictions for all H-ORFs reached 100% with the exception of A. cygnea; however, there was some evidence that the A. cygnea has a TM. The topology for all the TMs included a cytoplasmic to extracellular (C-TM-E) crossing of

the membrane, although this orientation was less certain for A. cygnea and M. falcata.

TargetP detected an SP in all H-ORFs (RC = 1: A. cygnea, L. compressa, and L. subviridus, RC = 4: M. falcata, T. parvum, and RC = 5: U. imbecillis). TargetP putatively assigned all H-ORF SPs as part of a secretory pathway except for M. falcata, which it identified as a mt SP. With SignalP parameters set to “sensitive” and with a potential TM presence, only L. subviridis came back positive for an SP. When SignalP parameters were set to exclude the possibility of a TM being present all H-ORFs came back as positive for SP presence except M. falcata. Subcellular localization analysis predicted the following: (1) that the T. parvum H-ORF was localized to the cytoplasm and nucleus (iLoc-Animal) or extracellular and nucleus (EuK-mPLoc), (2) that the L. compressa H-ORF was localized to nucleus (iLoc-Animal, EuK-mPLoc), (3) that the L. subviridis H-ORF was localized to cellular membrane and plasma membrane (iLoc-

Animal) or extracellular and nucleus (EuK-mPLoc), (4) that the M. falcata H-ORF was localized to cellular membrane and plasma membrane (iLoc-Animal) or extracellular

(EuK-mPLoc), (5) and finally, that the U. imbecillis H-ORF was localized to cytoplasm

(iLoc-Animal) or cellular membrane (EuK-mPLoc). EuK-mPLoc could not assess the potential localization of the A. cygnea H-ORF due to its small size, and iLoc-Animal did not localize it to any area, again possibly because of its small size. While H-ORFs did not have a clear subcellular localization, the F-ORF analysis using EuK-mPLoc predicted that all F-ORFs, except for C. monodonta and U. peninsularis, have an extracellular localization. Some other hits among the F-ORFs included localization to the cytoplasm, nucleus, and mitochondria; in addition, U. peninsularis exhibited a unique localization to endoplasmic reticulum and cell membrane. Subcellular

localization analysis of F-ORFs using iLoc-Animal did not provide any conclusive results and many did not produce any localization hits. No clear pattern of subcellular localization hits resulted from EuK-mPLoc or iLoc-Animal analysis of M-ORFs.

The amino acid properties were similar among all H-ORFs (Figure 4), although for A. cygnea, the non-polar amino acid composition was noticeably higher than other H-

ORFs, and its polar, “small”, and “tiny” amino acid composition was noticeably lower than for other H-ORFs. Geneious predicted an SCS in the H-ORFs of T. parvum, M. falcata, L. compressa and L. subviridis, but neither A. cygnea nor U. imbecillis produced a hit. TargetP predicted an SCS in all hermaphroditic species except M. falcata. In all cases that both Geneious and TargetP predicted an SCS they were in the same location along the H-ORFs. TargetP predicted an SCS within A. cygnea, although the pre-sequence length is 33 out of 34 amino acids in the proposed H-ORF. Geneious predicted that 25 out of 35 F-ORF sequences examined contained an SCS. For all 10 F-

ORFs that Geneious did not predict a cleavage site for, TargetP predicted a cleavage site in all of them. Each of these ORFs contained a predicted SCS within the potential TM hydrophobic region except T. glans, V. ellipsiformis, and Potamilus metnecktayi.

In silico analysis of M-ORFs

All M-ORFs contained at least one potential TM with variable topologies. All M-ORFs possessed an SCS predicted by either or both Geneious and TargetP. All predicted SCS were within the potential TM hydrophobic stretch of amino acids.

Determination of evolutionary pattern(s) of TM predictions and their associated topologies among the sex-associated mt ORF proteins

The following results are based on constructing BI trees of F-/H-type or M-type cox1 sequences and examining the trees for evolutionary patterns of F-, H- and M-ORF TM predictions and topologies as identified in the previous section. Almost all female dioecious and hermaphroditic species have one F-/H-ORF TM prediction with a cytoplasmic-TM-extracellular (C-TM-E) topology (Figure 5), although some predictions are not at 100% likelihood for either orientation or presence of a TM. The only major exception was the F-ORF of Ellipsaria lineolata, which potentially possesses two TMs, but the topology of these TMs was not strongly supported. The initial orientation of the TM protein is unclear (either cytoplasmic or extracellular localizations were equally likely). There is no clear link between phylogenetic placement on the cox1 tree and the absence of an SCS predicted within the F-ORFs. The corresponding cox1 tree groups all members of subfamily Anodontinae together, with tribes Anodontini and Cristariini positioned separately. Both members of subfamily

Unioninae and tribe Unionini are grouped together. All members of tribe Lampsilini

(subfamily Ambleminae) cluster together, except for a member of tribe Amblemini,

Reginaia ebena, which has a weakly supported BI grouping with Echyridella menziesii, a member of family Hyriidae. Members of genus Margaritifera group together. Overall the posterior probabilities were strong within the BI tree. The M-type cox1 tree (Figure

6) also featured strong posterior probability support and resolved the FWMs into their respective subfamilies. In contrast to the F-/H-type cox1 tree the number of M-ORF

TMs and their topology differs among subfamilies. A parsimonious reconstruction of

both number of TMs and their topologies suggested that one TM and an extracellular-

TM-cellular (E-TM-C) topology is the ancestral state of all M-type FWMs included

(Figure 6).

DISCUSSION

The complete mtDNA of A. cygnea is similar to other F-/H-type FWM mtDNA with respect to location of protein coding genes, rRNAs, tRNAs and the H-ORF. An interesting exception is a translocation of a portion of nad5, which is highlighted below.

The SURs of A. cygnea and other H-type mtDNA share some, but not all, common features of F-type SURs. Sequence comparisons among members of the same tribe show a closer sequence similarity between hermaphroditic species and a dioecious species. For example, in Anodontini, which includes A. cygnea, A. anatina, U. peninuslaris and U. imbecillis, there is a closer sequence similarity between A. cygnea and F-type A. anatina than there is between A. cygnea and other sequences in this group. As A. anatina is dioecious, and U. imbecillis, a hermaphroditic species, is on a separate branch, the close relationship between A. cygnea and A. anatina represents a fifth transition from a dioecy to hermaphroditism in FWMs together with the four such transitions previously identified by Breton et al. (2011). This supports the hypothesis that obligate hermaphroditism among FWMs is an example of repeated convergent evolution in this group. As expected, the H-ORF of A. cygnea shows little similarity to the F-ORF of A. anatina, however, somewhat unexpectedly, it does not contain the other common features of H-ORFs as outlined by Breton et al. (2011). Also, as

expected, there was no M-ORF or any evidence of an M-type mt genome detected in this population of hermaphroditic A. cygnea (see Stewart et al. 2013). Further insights into the H-ORF of A. cygnea, and F-ORF/H-ORFs of FWMs are discussed below.

The translocation of partial nad5 of Anodonta cygnea to shared unassigned region trnF-nad5

Our BLAST search of the SUR trnF-nad5 region suggested that there is some degree of similarity of this region to the nad5 gene of U. peninsularis. A similar strategy was used by Serb and Lydeard (2003) to identify the translocation of a portion of the atp8 gene to between trnD and nad4L in the F-type genome of Unio japanensis. Although nad5 does not appear to be highly modified among A. anatina, U. peninsularis, and U. imbecillis it is possible that the middle region of nad5 is relatively conserved compared to the two outer regions. If an outer part of nad5 can translocate (or the parts of nad5 flanking it) then it is possible that selective pressures are relatively unequal across the gene and that oxidative phosphorylation functions are not interrupted with the movement and loss of this piece or, alternatively, with this piece being transcribed and translated separately from the remainder of the gene. It is possible that the specific amino acids that play an essential role in the active sites of the protein’s tertiary structure can have stronger selective pressures on them than neighboring amino acids that play a more structural role (e.g., McClellan et al. 2004). NAD5 of A. cygnea may function adequately with this component being “moved”, or else with the areas flanking it being translocated. It is also possible that the gene coding for NAD5 has migrated to the nucleus and the mtDNA nad5 is no longer being transcribed or a functional protein after translation; this

type of transfer is documented in some eukaryotes (e.g., Adams and Palmer 2003).

However, it appears more likely that the portion is simply not required as the amino acid composition of this portion within other FWM species ranges from only a percent identity of 45.2 to 56.6, except in the unusual case of L. compressa (another hermaphrodite), which is 83% similar to this portion in A. cygnea. BLAST searches of this stretch of NAD5 with other FWM species resulted in similar percent identities among species, whereas the percent identity of NAD5 overall ranged from 81.7 to 88.6 overall within the same set of FWM species. It is unlikely that a nucleotide BLAST of a portion of the A. cygnea trnF-nad5 SUR would produce a hit in a FWM (U. peninsularis) within a neighbouring gene by chance alone. An alignment of bivalve cox1 sequences, with the approximate location of the translocated NAD5 portion identified, suggests that this region is not especially well conserved (Figure 7). Given these results, we argue that the area of nad5 that is translocated in A. cygnea is likely a relatively unconstrained part of the nad5 gene of FWM mtDNA in general.

Control region(s) location(s) are inconclusive in FWM.

Although secondary structures were located within all SURs, and within the longest unassigned region (LUR) of all complete H-type mtDNA available, it is difficult to pinpoint a main control region similar to other metazoan mt genomes. The proposition that nad5-trnQ SUR (Breton et al. 2009) is the main control region of F-type mtDNA is also possible for the H-type mtDNA, however, nad3-trnH+trnA also appears to be a potential candidate for the primary control region. Curiously, only three SURs within five H-type complete mtDNAs (15 SURs total) assessed contain repeat unit. The

majority of SURs did not possess repeats, which is notably different than the pattern described in Breton et al. (2009) in which all SUR types of the seven F-type complete mtDNA genomes assessed (35 SURs total) contained at least one repeat, and at most 33.

It is unclear why a shift from a dioecious sexual strategy to a hermaphroditic sexual strategy would accompany a loss in repeat units within SURs. The transition from a dioecious to hermaphroditic sexual strategy occurs independently, and may involve different genetic signatures (i.e., different modifications to cellular processes to produce both types of gametes), thereby examining these multiple transitions in the future will provide us with a better understanding of the diversity of these developmental systems.

Within F-type mtDNA nad5-trnQ ranges from 202-450 bp (excluding the exceptional case of U. japanensis at 1196 bp), whereas this region ranges from 152-291 bp in H- type mt genomes. There does not appear to be a reduction in complete mtDNA length overall with the transition from dioecy to hermaphroditism, thus there may be something specific about these SURs that is no longer required after this shift from dioecy to monoecy. Figure 3 provides an example of secondary structures, and thereby potential nucleotide motifs such as an origin of replication or transcription, occurring within the A. cygnea trnF-nad3 SUR. However, similar structures occur within SURs from all other H-type complete mtDNA analyzed and this feature is therefore not unique to A. cygnea. As noted in Breton et al. (2009) these SURs are associated with a change in the direction of transcription within the mt genome, and an origin of transcription occurring within each would seem to make sense.

Putative TM and topology of F-/H-ORF and M-ORFs exemplifies the variable selective constraints placed on F-, H- and M-type mtDNA

The F-type and H-type cox1 tree (Figure 5) closely reflects the groupings suggested in

Lopes-Lima et al. (2017), with the exception of the position of Echyridella menziesii.

However, there is not strong support for the placement of this particular species on the tree (i.e., <75% posterior probability). It is possible that including additional genes would resolve this tree topology to more closely reflect the BI tree presented in Breton et al. (2011), which was based on both cox1 and nad1 sequences. The E. lineolata F-

ORF was scored as potentially having two TMs, however, this most likely occurs due to stretches of hydrophobic amino acids that do not constitute a true TM. Stretches of hydrophobic amino acids can trigger programs to report potential TMs, although with a low likelihood percentage. The topology of the putative TM was also somewhat uncertain, further reducing the likelihood of this region forming a TM. Consequently, it is likely that all the FWM species included in the F-/H-type cox1 tree possess an ORF with a single TM domain and a C-TM-E orientation. Subcellular localization analysis did suggest that F-ORFs are localized as “extracellular”, and although this hit is consistent within the analysis it does not agree with the immunostaining results of

Breton et al. (2011). In that study, an in vivo assessment of the localization of the F-

ORF in V. ellipsiformis suggested that this protein was localized to the mitochondria, nuclear membrane and nucleoplasm. It may be that the programs iLoc-Animal and

EuK-mPLoc cannot provide fully realistic results for localization of the H-, F- and M-

ORFs because both of these programs rely on the Swiss-Prot Database (Bairoch and

Apweiler 2000; Shen et al. 2007; Chou and Shen 2010; Lin et al. 2013), which may not

be optimized to predict localization for novel mt proteins that are exported to other cellular structure(s). While subcellular localization results remain unclear, these trees

(Figure 5 and 6) suggest that a stronger selective pressure is placed on the F-ORFs and potentially the H-ORFs to maintain a hydrophobic region that could be a TM. An analysis of a total of 40 FWM species (either female or hermaphroditic) provides a clear indication that this TM and its topology is being maintained across several subfamilies.

In contrast, a total of 10 males suggest that the number of TMs and their topology is not maintained across several subfamilies, but may be maintained within some subfamilies.

Based on the consistent presence of one TM in the female tree, as well as patchy presence through the male tree, it is likely that one TM is the ancestral condition, although the topology is unclear based on these trees alone as the M-type tree contains both an E-TM-C and C-TM-E topology with E-TM-C appearing twice. A maximum parsimony analysis in Mesquite agreed with the ancestral state of one TM, and predicted the ancestral topology to be E-TM-C. Interestingly, among closely related subfamilies on the M-type tree for the families Anodontinae and Unioninae (this study; Fonseca et al. 2016), a drastic change from one TM and an E-TM-C topology to two TMs and a C-

TM-E-TM-C topology occurs. The ancestral topology of F-/H-ORFs is opposite to the

M-ORFs, suggesting a potential switch in topology coinciding with the evolution of

DUI. The m-orf and f-orf are likely not homologous and are derived from separate evolutionary events (Breton et al. 2011; Milani et al. 2013), it is not surprising that they would have different ancestral topologies. However, it is unclear if DUI or these novel

ORFs appeared first as it has yet to be determined precisely what function the ORFs

have, if they all have similar functions, or if some are functioning while others have lost functionality.

A comparison among these two trees exemplifies the selective constraints placed on the

F-ORFs compared to the M-ORFs and similarly the selective constraints placed on F- type and M-type mtDNA of species with DUI. Although it has been shown that both

FWM ORFs have a high amino acids substitution rate (Breton et al. 2009), it has also been shown that M-type mtDNA typically evolves much more quickly than F-type mtDNA in bivalves exhibiting DUI (Hoeh et al. 1996; Stewart et al. 1996; Hoeh et al.

2002). It is possible that selective constraints placed on F-ORFs are much higher than those placed on M-ORFs, which is illustrated by the multiple character states in this relatively small sample of FWM M-ORFs. It is also interesting to note that across the F-

/H-type tree, H-ORFs have maintained a relatively similar amino acid composition.

Although patterns have emerged from the comparison of H-ORFs and closely related F-

ORFs (Breton et al. 2011), H-ORF comparisons show a high level of sequence divergence and their hydrophobicity profiles vary greatly (Breton et al. 2011). Despite their many differences, the amino acid composition properties appear quite similar

(Figure 4). This is particularly interesting given that the hermaphroditic reproductive mode has evolved independently multiple times within order Unionoida.

In silico analyses of H-, F-, and M-ORFs

The in silico analyses presented here agree with the findings of Mitchell et al. (2016) in that (1) TMs were predicted in all ORFs (although their likelihood varied) and (2) SPs

were predicted in almost all H-ORFs and all M-ORFs, but again the certainty in these predictions remains somewhat less certain for the H-ORFs as “sensitive” parameters were used for this analysis. Nonetheless, SCS were predicted in almost all H-ORFs. SPs possess a tripartite structure that includes a hydrophilic region, a hydrophobic region and an SCS (Kapp et al. 2009). The hydrophilic region portion usually consists of positively charged amino acids, varies in length (Martoglio and Dobberstein 1998), and generally has the highest variability of all three regions (Martoglio 2003). In contrast, the hydrophobic region is considered the most essential feature of an SP (Hegde and

Bernstein 2006). Although some SP predictions were uncertain, the confirmation of both a hydrophobic region (TM) and an SCS leave only the variable hydrophilic region to be confirmed; theoretically this variability (especially in these novel ORFs with a high substitution rate) could simply result in under-identification of the region in question. Although SPs share these characteristics they do not possess a single characteristic motif and instead can tolerate mutations (Gierasch 1989; Hegde and

Bernstein 2006). SPs target proteins to a membrane and can reach their destination through different pathways. The topology of a protein during translocation through a protein conducting channel is C-TM-E, which interestingly is the orientation consistently predicted for F- and H-ORFs (Figure 5). M-ORFs adopt this orientation less often (Figure 6). If the SCSs of SPs leave a significant hydrophobic region after cleavage then this could function as an anchor for the SP at its target destination (Roy et al. 1993; Chen and Kendall 1995; Martoglio and Dobberstein 1998). The SCS locations across these novel ORFs vary in relation to their placement along the hydrophobic stretch of amino acids, therefore the number of hydrophobic amino acids remaining

after cleavage is variable; consequently, it is unclear if these proteins will be anchored or not at their final destination. Interestingly, all known examples of TMs formed from the hydrophobic region of SPs have a viral origin (Kapp et al. 2009). Both Milani et al.

(2013) and Mitchell et al. (2016) suggest that these ORFs could have a viral origin, although Mitchell et al. (2016) and Guerra et al. (2017) also suggested that the process of gene duplication and subsequent modification of a mitochondrial gene may have been responsible; determining whether or not these orfs produce anchored proteins is relevant for testing this hypothesis.

For a number of FWMs Mitchell et al. (2016) did identify motifs and domains of F-

ORFs and M-ORFs related to cell membrane and surface anchoring, but independent in situ verification would be useful. In the case of V. ellipsiformis, the F-ORF was associated with both cytoplasm (mitochondria and nucleoplasm) and membrane (plasma membrane), but this does not necessarily imply that these structures are the final destinations for these proteins. Multiple cleavage sites were predicted for many F-

ORFs, consequently it may be possible to have both anchored and non-anchored

“versions” of one protein. However, it should be noted that for the F-ORF of V. ellipsiformis only one SCS was predicted using TargetP, which was located near the end of the hydrophobic stretch leaving few hydrophobic amino acids to act as an anchor. It has also been suggested that SP fragments, which have been cleaved at the SCS, may have post-targeting functions and that the cleaved proteins themselves might undergo additional processing (Martoglio 2003). This further complicates the potential of these novel ORFs given their (putative) SPs, and SCSs. However, this could explain the

potential cleavage within the H-ORF of A. cygnea, as almost all of the protein would be cleaved off. To add another layer of complexity, not all SPs are cleaved, specifically in the case of some viral proteins (Kapp et al. 2009). Considering these factors may shed some light on the variety of subcellular localization hits. For example, the V. ellipsiformis F-ORF was only localized as extracellular based on the in silico analysis and not predicted to occur in any of the locations identified through immunostaining

(Breton et al. 2011). It is possible that because these ORFs are (1) not homologous to other proteins, (2) may be participating in unique functions associated with a unique phenomenon (i.e., DUI), and (3) are produced in the mitochondria and migrating to the nucleus (also a unique process), that standard in silico analysis programs are not optimized to characterize them. Although in silico analysis is useful, it will be beneficial to compare it with experimental evidence such as the immunostaining conducted by

Breton et al. (2011) to create a more complete image of what these fascinating proteins are doing.

Conclusion and suggestions for future work

As with Mitchell et al. (2015), the in silico analyses performed herein have provided a valuable framework for future studies of the origin and function of orfs associated with

DUI in FWMs, as we provided further evidence of SPs indicative of mitochondria to nucleus transportation of putative ORF proteins. In situ analysis of F-, H- and M-ORFs will complement the in silico analysis conducted in this study. Future studies targeted to understanding subcellular localizations of these novel proteins may clarify their function, verify findings of this study and answer new questions (e.g., are M-ORFs

being translated and transported out of the mitochondria given the diversity in number of TMs and their topology across subfamilies?). Accessing samples of obligate hermaphroditic species that have “reverted” back to dioecy (e.g., A. cygnea in rivers) will (1) confirm this phenomenon and (2) serve as an interesting investigation of the

“reappearing” M-ORF, when or if males reappear in a population of previously hermaphroditic individuals, and these reversions exhibit DUI as opposed to SMI in hermaphrodites.

FIGURES

Figure 1. Gene order and relative size of hermaphroditic freshwater mussel (order Unionid) Anodonta cygnea H-type mitochondrial genome, with annotated tRNAs and RNAs. Image generated in Geneious 9.1.7.

Figure 2. (A) Nucleotide and (B) Amino acid sequence divergence between Anodonta cygnea, a hermaphroditic freshwater mussel (order Unionidae), and a closely related dioecious female freshwater mussel: Anodonta anatina. For comparative purposes, another set of hermaphroditic (Utterbackia imbecillis) and a dioecious female (Utterbackia peninsularis) sequence divergence from A. cygnea are also compared. Nucleotide divergence was calculated using (A) the Tamura-Nei model and (B) p- distance. Please note the difference in scale.

Figure 3. Example excerpt sequence from a potential control region of Anodonta cygnea within a non-coding segment between trnF and nad5. Potential secondary structures are shown, a star indicates a potential origin of replication due based on structural similarities to the classic origin of replication in metazoan control regions. Note that not all secondary structures within this region are pictured.

Figure 4. Amino acid property composition of all available H-ORFs determined using EMBOSS Pepstats. Amino acids included by each property are as follows: tiny (A+C+G+S+T), small (A+B+C+D+G+N+P+S+T+V), aliphatic (A+I+L+V), aromatic (F+H+W+Y), non-polar (A+C+F+G+I+L+M+P+V+W+Y), polar (D+E+H+K+N+Q+R+S+T+Z), charged (B+D+E+H+K+R+Z), basic (H+K+R) and acidic (B+D+E+Z).

Figure 5. F-type and H-type BI phylogenetic tree based on cox1 DNA sequences. Posterior probabilities >0.75 are marked in red. A (*) denotes a species that has a slight potential of possessing two TMs. Number of predicted transmembrane domains (TM) and their topology are shown (E = extracellular, C = cytoplasmic). Subfamilies are denoted by colour, pink; Anodontinae, purple; Unioninae, blue; Ambleminae and green; Margaritiferdae. A single species represents subfamily Hyriidae (Echyridella menziesii). H-types include: Anodonta cygnea, Lasmigona compressa, Utterbackia imbecillis, Toxolasma parvus, and Margaritifera margaritifera, all other taxa are F-types.

Figure 6. M-type BI phylogenetic tree of based on cox1 DNA sequences. Posterior probabilities > 0.75 are marked in red. Number of predicted transmembrane domains (TM) and their topology are shown (E = extracellular, C = cytoplasmic). Subfamilies are denoted by colour: pink; Anodontinae, purple; Unioninae, blue; Ambleminae. A single species represents subfamily Margaritiferdae (Cumberlandia monodonta), Hyriidae (Echyridella menziesii), and Gonideinae (Solenaia carinata).

Figure 7. NAD5 protein alignment from members of family Unionoida (Anodonta anatina; KF030964), Mytiloida (Modiolus modiolus; KX821782) and Veneroida (Meretrix lusoria; ACV92129). The approximate location of the translocated segment of A. cygnea NAD5 (blue) is aligned with A. anatina. A sliding window size of 25 was used. Image generated in Geneious version 9.1.7.

Figure S1. Predicted tRNA structures for all 22 tRNA of H-type mitochondrial DNA in A. cygnea. Left-to-right and top-to-bottom order follows their placement within the annotated complete mitochondrial genome beginning at base 2,668.

TABLES

REFERENCES

Adams, K., 2003. Evolution of mitochondrial gene content: gene loss and transfer to the

nucleus. Molecular Phylogenetics and Evolution 29, 380–395.

Altschul, S.F., Gish, W., Miller, W., Myers, E.W., Lipman, D.J., 1990. Basic local

alignment search tool. Journal of Molecular Biology 215, 403–410.

An, C., Ouyang, S., Zhou, C.-H., Wu, X.-P., 2016. The complete F-type mitochondrial

genome of Chinese Anodonta arcaeformis (Bivalvia: Unionidae: Anodontinae).

Mitochondrial DNA 27, 1552–1553.

Bairoch, A., Apweiler, R., 2000. The SWISS-PROT protein sequence database and its

supplement TrEMBL in 2000. Nucleic Acids Res 28, 45–48.

Beck, N., Lang, B.F., 2010. MFannot, organelle genome annotation webserver.

Université de Montréal QC, Canada.

Benson, G., 1999. Tandem repeats finder: a program to analyze DNA sequences.

Nucleic Acids Research 27, 573-580.

Bernt, M., Donath, A., Jühling, F., Externbrink, F., Florentz, C., Fritzsch, G., Pütz, J.,

Middendorf, M., Stadler, P.F., 2013. MITOS: Improved de novo metazoan

mitochondrial genome annotation. Molecular Phylogenetics and Evolution,

Mitogenomics and Metazoan Evolution 69, 313–319.

Birky, C.W., 1995. Uniparental inheritance of mitochondrial and chloroplast genes:

mechanisms and evolution. PNAS 92, 11331–11338.

Bogan, A.E., 1993. Freshwater bivalve extinctions (Mollusca: Unionoida): A search for

causes. Integrative and Comparative Biology 33, 599–609.

Boore, J.L., 1999. Animal mitochondrial genomes. Nucleic Acids Research 27, 1767–

1780.

Bouckaert, R.R., 2010. DensiTree: making sense of sets of phylogenetic trees.

Bioinformatics 26, 1372–1373.

Breton, S., Beaupré, H.D., Stewart, D.T., Hoeh, W.R., Blier, P.U., 2007. The unusual

system of doubly uniparental inheritance of mtDNA: isn’t one enough? Trends

in Genetics 23, 465–474.

Breton, S., Beaupré, H.D., Stewart, D.T., Piontkivska, H., Karmakar, M., Bogan, A.E.,

Blier, P.U., Hoeh, W.R., 2009. Comparative mitochondrial genomics of

freshwater mussels (Bivalvia: Unionoida) with doubly uniparental inheritance of

mtDNA: gender-specific open reading frames and putative origins of replication.

Genetics 183, 1575–1589.

Breton, S., Burger, G., Stewart, D.T., Blier, P.U., 2006. Comparative analysis of

gender-associated complete mitochondrial genomes in marine mussels (Mytilus

spp.). Genetics 172, 1107–1119.

Breton, S., Milani, L., Ghiselli, F., Guerra, D., Stewart, D.T., Passamonti, M., 2014. A

resourceful genome: updating the functional repertoire and evolutionary role of

animal mitochondrial DNAs. Trends in Genetics 30, 555–564.

Breton, S., Stewart, D.T., Hoeh, W.R., 2010. Characterization of a mitochondrial ORF

from the gender-associated mtDNAs of Mytilus spp. (Bivalvia: Mytilidae):

Identification of the “missing” ATPase 8 gene. Marine Genomics 3, 11–18.

Breton, S., Stewart, D.T., Shepardson, S., Trdan, R.J., Bogan, A.E., Chapman, E.G.,

Ruminas, A.J., Piontkivska, H., Hoeh, W.R., 2011. Novel protein genes in

animal mtDNA: A new sex determination system in freshwater mussels

(Bivalvia: Unionoida)? Molecular Biology and Evolution 28, 1645–1659.

Cao, L., 2004. Evidence that the large noncoding sequence is the main control region of

maternally and paternally transmitted mitochondrial genomes of the marine

mussel (Mytilus spp.). Genetics 167, 835–850.

Chen, H., Kendall, D.A., 1995. artificial transmembrane segments: requirements for

stop transfer and polypeptide orientation. Journal of Biological Chemistry 270,

14115–14122.

Chen, L., Wang, G., Li, J., 2012. Analysis on complete F type of mitochondrial genome

in Lamprotula leai. Shengtai Xuebao/Acta Ecologica Sinica.

Chojnacki, J.C., Lewandowska, A., Rosińska, B., 2007. Biometrics of the mussel

Anodonta cygnea (L.) inhabiting in 2005 the Binowo and Bobolin Lakes near

Szczecin. Oceanological and Hydrobiological Studies 21–37.

Chou, K.-C., Shen, H.-B., 2010. Cell-PLoc 2.0: An improved package of web-servers

for predicting subcellular localization of proteins in various organisms. Natural

Science 2, 1090–1103.

Dalziel, A.C., Stewart, D.T., 2002. Tissue-specific expression of male-transmitted

mitochondrial DNA and its implications for rates of molecular evolution in

Mytilus mussels (Bivalvia: Mytilidae). Genome 45, 348–355.

Darty, K., Denise, A., Ponty, Y., 2009. VARNA: Interactive drawing and editing of the

RNA secondary structure. Bioinformatics 25, 1974–1975.

Doucet-Beaupré, H., Breton, S., Chapman, E.G., Blier, P.U., Bogan, A.E., Stewart,

D.T., Hoeh, W.R., 2010. Mitochondrial phylogenomics of the Bivalvia

(Mollusca): Searching for the origin and mitogenomic correlates of doubly

uniparental inheritance of mtDNA. BMC Evolutionary Biology 10, 50.

Drummond, A.J., Nicholls, G.K., Rodrigo, A.G., Solomon, W., 2002. Estimating

mutation parameters, population history and genealogy simultaneously from

temporally spaced sequence data. Genetics 161, 1307–1320.

Emanuelsson, O., Brunak, S., von Heijne, G., Nielsen, H., 2007. Locating proteins in

the cell using TargetP, SignalP and related tools. Nature Protocols 2, 953–971.

Fonseca, M.M., Harris, D.J., Posada, D., 2014. The inversion of the Control Region in

three mitogenomes provides further evidence for an asymmetric model of

vertebrate mtDNA replication. PLoS ONE 9, e106654.

Fuchs, S.M., 2013. Chemically modified tandem repeats in proteins: natural

combinatorial peptide libraries. ACS Chemical Biology 8, 275–282.

Garrido-Ramos, M.A., Stewart, D.T., Sutherland, B.W., Zouros, E., 1998. The

distribution of male-transmitted and female-transmitted mitochondrial DNA

types in somatic tissues of blue mussels: Implications for the operation of doubly

uniparental inheritance of mitochondrial DNA. Genome 41, 818–824.

Gierasch, L.M., 1989. Signal sequences. Biochemistry 28, 923–930.

Guda, C., Subramaniam, S., 2005. TARGET: a new method for predicting protein

subcellular localization in eukaryotes. Bioinformatics 21, 3963–3969.

Guerra, D., Plazzi, F., Stewart, D.T., Bogan, A.E., Hoeh, W.R., Breton, S., 2017.

Evolution of sex-dependent mtDNA transmission in freshwater mussels

(Bivalvia: Unionida). Scientific Report 7.

Gusman, A., Lecomte, S., Stewart, D.T., Passamonti, M., Breton, S., 2016. Pursuing the

quest for better understanding the taxonomic distribution of the system of

doubly uniparental inheritance of mtDNA. PeerJ 4, e2760.

Han, Z., Wang, G., Xue, T., Chen, Y., Li, J., 2016. The F-type complete mitochondrial

genome of Chinese freshwater mussels Cuneopsis pisciculus. Mitochondrial

DNA Part A 27, 3376–3377.

He, F., Wang, G. ling, Li, J., 2016. Complete F-type mitochondrial genome of Chinese

freshwater mussels Lamprotula gottschei. Mitochondrial DNA Part A 27, 246–

247.

Hegde, R.S., Bernstein, H.D., 2006. The surprising complexity of signal sequences.

Trends in Biochemical Sciences 31, 563–571.

Hinzmann, M., Lopes-Lima, M., Teixeira, A., Varandas, S., Sousa, R., Lopes, A.,

Froufe, E., Machado, J., 2013. Reproductive cycle and strategy of Anodonta

anatina (L., 1758): notes on hermaphroditism. Journal of Experimental Zoology

319, 378–390.

Hoeh, W.R., Stewart, D.T., Guttman, S.I., 2002. High fidelity of mitochondrial genome

transmission under the doubly uniparental mode of inheritance in freshwater

mussels (Bivalvia: Unionoidea). Evolution 56, 2252–2261.

Hoeh, W.R., Stewart, D.T., Sutherland, B.W., Zouros, E., 1996. Cytochrome c oxidase

sequence comparisons suggest an unusually high rate of mitochondrial DNA

evolution in Mytilus (Mollusca: Bivalvia). Molecular Biology and Evolution 13,

418–421.

Huang, X.-C., Zhou, C.-H., Ouyang, S., Wu, X.-P., 2015. The complete F-type

mitochondrial genome of threatened Chinese freshwater mussel Solenaia

oleivora (Bivalvia: Unionidae: Gonideinae). Mitochondrial DNA 26, 263–264.

Jiang, W.., Li, J.., Zheng, R.., Wang, G.., 2010. Analysis of complete mitochondrial

genome of Cristaria plicata. Yi Chuan 32, 153–162.

Kapp, K., Schrempf, S., Lemberg, M.K., Dobberstein, B., 2013. Post-targeting functions

of signal peptides. Landes Bioscience.

Kearse, M., Moir, R., Wilson, A., Stones-Havas, S., Cheung, M., Sturrock, S., Buxton,

S., Cooper, A., Markowitz, S., Duran, C., Thierer, T., Ashton, B., Meintjes, P.,

Drummond, A., 2012. Geneious Basic: An integrated and extendable desktop

software platform for the organization and analysis of sequence data.

Bioinformatics 28, 1647–1649.

Krogh, A., Larsson, B., von Heijne, G., Sonnhammer, E.L., 2001. Predicting

transmembrane protein topology with a hidden markov model: application to

complete genomes. Journal of Molecular Biology 305, 567–580.

Kumar, S., Stecher, G., Tamura, K., 2016. MEGA7: molecular evolutionary genetics

analysis version 7.0 for bigger datasets. Molecular Biology and Evolution 33,

1870–1874.

Lee, J.H., Choi, E.H., Kim, S.K., Ryu, S.H., Hwang, U.W., 2012. Mitochondrial

genome of the cockscomb pearl mussel Cristaria plicata (Bivalvia, Unionoida,

Unionidae). Mitochondrial DNA 23, 39–41.

Lee, J.H., Kim, S.K., Hwang, U.W., 2016. Complete mitochondrial genome of

Lamprotula coreana (Unionidae, Unionoida, Bivalvia). Mitochondrial DNA Part

A 27, 1026–1027.

Lin, W.-Z., Fang, J.-A., Xiao, X., Chou, K.-C., 2013. iLoc-Animal: a multi-label

learning classifier for predicting subcellular localization of animal proteins.

Molecular BioSystems 9, 634.

Lopes-Lima, M., Froufe, E., Do, V.T., Ghamizi, M., Mock, K.E., Kebapçı, Ü., Klishko,

O., Kovitvadhi, S., Kovitvadhi, U., Paulo, O.S., Pfeiffer, J.M., Raley, M.,

Riccardi, N., Şereflişan, H., Sousa, R., Teixeira, A., Varandas, S., Wu, X.,

Zanatta, D.T., Zieritz, A., Bogan, A.E., 2017. Phylogeny of the most species-

rich freshwater bivalve family (Bivalvia: Unionida: Unionidae): Defining

modern subfamilies and tribes. Molecular Phylogenetics and Evolution 106,

174–191.

Lowe, T.M., Eddy, S.R., 1997. tRNAscan-SE: a program for improved detection of

transfer RNA genes in genomic sequence. Nucleic Acids Res. 25, 955–964.

Maddison, W.., Maddison, D.., 2017. Mesquite: a modular system for evolutionary

analysis Version 3.2.

Martoglio, B., 2003. Intramembrane proteolysis and post-targeting functions of signal

peptides. Biochemical Society Transactions 31, 1243–1247.

Martoglio, B., Dobberstein, B., 1998. Signal sequences: more than just greasy peptides.

Trends in Cell Biology 8, 410–415.

McClellan, D.A., Palfreyman, E.J., Smith, M.J., Moss, J.L., Christensen, R.G.,

Sailsbery, J.K., 2005. physicochemical evolution and molecular adaptation of

the cetacean and artiodactyl cytochrome b proteins. Molecular Biology and

Evolution 22, 437–455.

Milani, L., Ghiselli, F., Guerra, D., Breton, S., Passamonti, M., 2013. A comparative

analysis of mitochondrial ORFans: new clues on their origin and role in species

with doubly uniparental inheritance of mitochondria. Genome Biology and

Evolution 5, 1408–1434.

Mitchell, A., Guerra, D., Stewart, D., Breton, S., 2016. In silico analyses of

mitochondrial ORFans in freshwater mussels (Bivalvia: Unionoida) provide a

framework for future studies of their origin and function. BMC Genomics 17,

597.

Passamonti, M., Ghiselli, F., 2009. Doubly uniparental inheritance: two mitochondrial

genomes, one precious model for organelle DNA inheritance and evolution.

DNA and Cell Biology 28, 79–89.

Pourang, N., Richardson, C.A., Mortazavi, M.S., 2010. Heavy metal concentrations in

the soft tissues of swan mussel (Anodonta cygnea) and surficial sediments from

Anzali wetland, Iran. Environmental Monitoring and Assessment 163, 195–213.

Rambaut, A., Drummond, A., 2002. Tree Annotator version 1.4 part of the BEAST

package.

Rambaut, A., 2007. FigTree, a graphical viewer of phylogenetic trees.

Rhoads, A., Au, K.F., 2015. PacBio Sequencing and Its Applications. Genomics,

Proteomics & Bioinformatics 13, 278–289.

Rice, P., Longden, I., Bleasby, A., 2000. EMBOSS: The European molecular biology

open software suite. Elsevier Current Trends.

Robicheau, B.M., Breton, S., Stewart, D.T., 2017. Sequence motifs associated with

paternal transmission of mitochondrial DNA in the horse mussel, Modiolus

modiolus (Bivalvia: Mytilidae). Gene 605, 32–42.

Roy, P., Chatellard, C., Lemay, G., Crine, P., Boileau, G., 1993. Transformation of the

signal peptide/membrane anchor domain of a type II transmembrane protein into

a cleavable signal peptide. Journal of Biological Chemistry 268, 2699–2704.

Sambrook, J., Fritsch, E.F., Maniatis, T., 1989. Molecular cloning: a laboratory manual.

Cold Spring Harbor Laboratory Press.

Serb, J.M., Lydeard, C., 2003. Complete mtDNA sequence of the North American

freshwater mussel, Lampsilis ornata (Unionidae): an examination of the

evolution and phylogenetic utility of mitochondrial genome organization in

Bivalvia (Mollusca). Molecular Biology and Evolution 20, 1854–1866.

Seutin, G., White, B.N., Boag, P.T., 1991. Preservation of avian blood and tissue

samples for DNA analyses. Canadian Journal of Zoology 69, 82–90.

Shen, H.-B., Yang, J., Chou, K.-C., 2007. Euk-PLoc: an ensemble classifier for large-

scale eukaryotic protein subcellular location prediction. Amino Acids 33, 57–67.

Song, X.-L., Ouyang, S., Zhou, C.-H., Wu, X.-P., 2016. Complete maternal

mitochondrial genome of freshwater mussel Anodonta lucida (Bivalvia:

Unionidae: Anodontinae). Mitochondrial DNA Part A 27, 549–550.

Soroka, M., Burzyński, A., 2015. Complete female mitochondrial genome of Anodonta

anatina (Mollusca: Unionidae): confirmation of a novel protein-coding gene (F-

ORF). Mitochondrial DNA 26, 267–269.

Soroka, M., Burzyński, A., 2017. Doubly uniparental inheritance and highly divergent

mitochondrial genomes of the freshwater mussel Unio tumidus (Bivalvia:

Unionidae). Hydrobiologia 1–16.

Soroka, M., 2010. Characteristics of mitochondrial DNA of unionid bivalves (Mollusca:

Bivalvia: Unionidae). II. Comparison of complete sequences of maternally

inherited mitochondrial genomes of Sinanodonta woodiana and Unio pictorum.

Folia Malacologica 18, 189–209.

Steel, M., Mckenzie, A., 2001. Properties of phylogenetic trees generated by Yule-type

speciation models. Mathematical Biosciences 170, 91–112.

Stewart, D.T., Hoeh, W.R., Bauer, G., Breton, S., 2013. mitochondrial genes, sex

determination and hermaphroditism in freshwater mussels (Bivalvia:

Unionoida), in: evolutionary biology: exobiology and evolutionary mechanisms.

Springer, Berlin, Heidelberg, pp. 245–255.

Stewart, D.T., Kenchington, E.R., Singh, R.K., Zouros, E., 1996. Degree of selective

constraint as an explanation of the different rates of evolution of gender-specific

mitochondrial DNA lineages in the mussel Mytilus. Genetics 143, 1349–1357.

Tamura, K., Nei, M., 1993. Estimation of the number of nucleotide substitutions in the

control region of mitochondrial DNA in humans and chimpanzees. Molecular

Biology and Evolution 10, 512–526.

Thompson, J.D., Higgins, D.G., Gibson, T.J., 1994. CLUSTAL W: improving the

sensitivity of progressive multiple sequence alignment through sequence

weighting, position-specific gap penalties and weight matrix choice. Nucleic

acids research 22, 4673–4680.

Veinot, S., Chase, E.E., Robicheau, B., Breton, S., Stewart, D.T., n.d. The H-ORF of

Anodonta cygnea (Title TBD). In prep.

Wang, G., Cao, X., Li, J., 2013. Complete F-type mitochondrial genome of Chinese

freshwater mussel Lamprotula tortuosa. Mitochondrial DNA 24, 513–515.

Wei, M., Yang, S., Yu, P., Wan, Q., 2014. The complete mitochondrial genome of

Hyriopsis cumingii (Unionoida: Unionidae): genome description and related

phylogenetic analyses. Mitochondrial DNA 1–2.

Wheeler, D.L., Barrett, T., Benson, D.A., Bryant, S.H., Canese, K., Chetvernin, V.,

Church, D.M., DiCuccio, M., Edgar, R., Federhen, S., Feolo, M., Geer, L.Y.,

Helmberg, W., Kapustin, Y., Khovayko, O., Landsman, D., Lipman, D.J.,

Madden, T.L., Maglott, D.R., Miller, V., Ostell, J., Pruitt, K.D., Schuler, G.D.,

Shumway, M., Sequeira, E., Sherry, S.T., Sirotkin, K., Souvorov, A.,

Starchenko, G., Tatusov, R.L., Tatusova, T.A., Wagner, L., Yaschenko, E.,

2007. Database resources of the National Center for Biotechnology Information.

Nucleic Acids Research 36, D13–D21.

Xue, T., Chen, M., Wang, G., Han, Z., Li, J., 2016. The complete F-type mitochondrial

genome of Chinese freshwater mussel Anodonta euscaphys. Mitochondrial DNA

Part A 27, 2698–2699.

Zhou, C.-H., Ouyang, S., Wu, X.-P., Ding, M.-H., 2016. The complete maternal

mitochondrial genome of rare Chinese freshwater mussel Lepidodesma

languilati (Bivalvia: Unionidae: Unioninae). Mitochondrial DNA Part A 27,

4615–4616.

Zouros, E., 2013. Biparental inheritance through uniparental transmission: the doubly

uniparental inheritance (DUI) of mitochondrial DNA. Evolutionary Biology 40,

1–31.

Zuker, M., 2003. Mfold web server for nucleic acid folding and hybridization

prediction. Nucleic Acids Research 31, 3406–3415.

100

Chapter 3

Relative evolutionary rates of sex-associated mitochondrial genomes in the order Unionoida (Bilvalvia): A comparison among male-, female-, and hermaphrodite-transferred mitochondrial genomes.

Chase E.E.1, Robicheau B.M.2, Breton S.3, Stewart D.T.1

1 Department of Biology, Acadia University, Wolfville, NS, Canada

2 Department of Biology, Dalhousie University, Halifax, NS, Canada

3 Département de Sciences Biologiques, Université de Montréal, QC, Canada

ABSTRACT

An extensive examination of the relative rates of F-, H- and M-type mitochondrial DNA within species exhibiting doubly uniparental inheritance has yet to be conducted, although the relative rates of F- vs M- has been explored many times previously. There is substantial support for faster rate of evolution of M-type mitochondrial DNA compared to F-type mitochondrial DNA. Herein, the relative rates of F- vs H-type mtDNA is tested in the context of the “cellular arenas” hypothesis. According to this hypothesis, mitochondrial DNA operating in more cellular arenas should be under stronger selective constraints than mitochondrial DNA operating in fewer arenas.

Relatively speaking, M-type mitochondrial operates in a smaller number of arenas

(primarily sperm) than either F- or H-type mitochondrial DNA, and F-type mitochondrial DNA operates in fewer arenas (eggs and somatic tissues) than H-type

101

mitochondrial DNA (sperm, eggs, and somatic tissues). This study tests differences in these relative rates of molecular evolution using a variety of statistical tests. Based on the results of these comparisons, we conclude that there is evidence of differential rates and patterns of molecular evolution that is consistent with the cellular arenas hypothesis.

INTRODUCTION

The unique system of doubly uniparental inheritance (DUI) of mitochondrial DNA

(mtDNA) occurs in members of the order Veneroida, Mytiloida, Nuculanoida, and

Unionoida (Bivalvia) (Skibinski et al. 1994; Zouros 1994; Breton et al 2007, 2009,

2011; Boyle and Etter 2013; Zouros 2013). DUI is defined by the inheritance of two mitotypes (i.e. a maternally inherited or F-type and a female inherited or M-type) in contrast to the system of strictly maternal inheritance (SMI) that occurs among all other metazoans (Birky 1995). Under DUI, female offspring contain F-type mtDNA in both their somatic and gonad tissues, whereas male offspring contain exclusively M-type mtDNA in their germline and F-type mtDNA predominates within somatic tissues

(Garrido-Ramos et al. 1998, Venetis et al. 2006). While SMI causes mtDNA in males to come to an evolutionary dead end, DUI results in the evolution of separate F- and M- type lineages (Hurst and Hoekstra 1994). The transition from a dioecious sexual strategy to an obligate hermaphroditic strategy has occurred at least five times within separate freshwater mussel species (FWM; order Unionoida) (Breton et al. 2011; Chase et al. in prep). Associated with this transition is the loss of DUI, as hermaphrodites only

102

contain and pass on one type of mtDNA, known as the H-type. Consequently, hermaphrodites follow a SMI model of mtDNA transmission.

A fascinating property of DUI species is that F- and M-type mtDNA evolve at different rates, with the M-type typically evolving substantially faster (Stewart et al. 1996;

Quesada et al. 1998; Ort and Pogson 2007; Robicheau et al. 2016) although both the M- and the F-type have been shown to evolve at an exceptionally high rate compared to other metazoans (Hoeh et al. 1996). It is currently hypothesized that the high rate of molecular evolution of both sex-associated mitochondrial genomes in bivalves and the particularly high rate of molecular evolution of the M-types can be explained as a consequence of the different cellular “arenas” in which these mitotypes function compared to typical animal mtDNA under a system of SMI (Stewart et al. 1996).

Sequence divergence between conspecific F- and M-type have been estimated as upwards of 52% in FWM species (Doucet-Beaupré et al 2010), and 22-37% in marine mussel species (Mizi et al. 2005; Breton et al. 2006; Ort and Pogson 2007; Zbawicka et al. 2007; Robicheau et al. 2017) depending on the gene or molecular marker assessed and whether divergence is estimated using nucleotide or amino acid sequences. M-type mtDNA evolves separately from the F-type (given that recombination does not appear to have occurred in freshwater mussel genomes for >200 million years; Hoeh et al.

2002) and the M-type appears to be under relatively relaxed selection compared to the

F-type (Stewart et at. 1995; Hoeh et al. 1996, 1997; Breton et al. 2011). This separation into two sex-associated lineages may enable selection to act on the M-type mtDNA to enhance sperm function, which is one of the main hypotheses to explain the origin and

103

maintenance of this unusual system of mitochondrial DNA inheritance (Burt and

Trivers 2006; Breton et al. 2007; Zouros 2013). DUI has also been hypothesized to be the result of a “selfish” genetic element (Crow 2006) appearing in a mitochondrial genome in sperm that was able to secure its transition to the next generation through sperm (Burt and Trivers 2006; Milani et al. 2011; Stewart et al. 2013). This hypothesis suggests that elements of the M and F genomes play a role in sex determination in

Bivalves (Breton et al. 2011; Milani et al. 2011; Zouros 2013; Breton et al. 2017).

Regardless of which of these hypotheses is true (and they are not mutually exclusive), the existence of two distinct lineages sets the stage for different patterns of molecular evolution in the M-type vs. the F-type mtDNAs. For example, it has been proposed that there may be stronger selective constraints on F-type mtDNA than M-type mtDNA because the F-type operates in more cellular “arenas” compared to the M-type mtDNA

(Stewart et al. 1996). The F-type mtDNA must operate in all somatic tissue and reproductive tissues arenas, whereas the M-type primarily only operates in sperm

(Garrido-Ramos et al. 1998). Oogenesis also represents a greater energy investment per gamete than spermatogenesis (Hayward and Gillooly 2011) and consequently

“mistakes” have a greater energetic consequence. Because fewer eggs are produced compared to sperm, survival and success of the comparatively few eggs is crucial for reproduction (Stewart et al. 1995; Rawson and Hilbish 1995). Another non-exclusive explanation for higher rates of nucleotide substitution in the M type compared to the F type is that the M type is expected to experience elevated levels of oxidative damage because of the higher metabolic activity of sperm compared to eggs (Skibinski et al.

1994; Quesada et al. 1996). This higher rate of evolution of DNA in male gametes than

104

female gametes is even evident in plants (e.g. Whittle and Johnston 2002) which likely explains, in part, the higher rate of mutation within the M-type mtDNA than the F-type.

However, in species with DUI, both the M and F-types evolve faster than species with

SMI (e.g. Hoeh et al. 1996; Doucet-Beaupré et al. 2010). One possibility for this pattern of molecular evolution is that the existence of two sex-associated lineages leads to a relaxation of selection on the F type compared to that experienced by the single maternally transmitted genome present in species with SMI (Stewart et al. 1996; Zouros

2013).

While sequence divergence and evolutionary rates have been examined multiple times among F- and M-type mtDNA (e.g., Stewart et al. 1996; Quesada et al. 1998; Ort and

Pogson 2007; Robicheau et al. 2016), little work has been done comparing patterns of molecular evolution of the H-type mtDNA compared to F- and M-types. The aim of this study is to examine rates of molecular evolution and selective constraints on H-type mtDNA relative to both F- and M-type mtDNA. In theory, H-type mt genomes are operating within more arenas than either the M- or F-type mtDNA in species with DUI species. Therefore, H-type mtDNA should evolve in a fashion more consistent with the maternally transmitted genome of species with SMI. This is because the H-type is the only mitochondrial genome being transmitted from generation to generation, and as with species with SMI, even though there is mtDNA present in the sperm of hermaphrodites, it is not likely transmitted from parent to offspring. We therefore predict that the H-type mtDNA will evolve more slowly and be under stronger selective constraints than the F-type of closely related dioecious species. Because F-type mtDNA

105

has been shown to generally evolve more slowly than M-type mtDNA, we also predict that H-type mtDNA will evolve more slowly than M-type mtDNA.

MATERIALS AND METHODS

Quantifying nucleotide divergence of unionid M-, F-, and H-type cox1 sequence data in a phylogenetic context

For comparative purposes and to examine evolutionary patterns in a phylogenetic context, M-, F-, and H-type cox1 sequences were downloaded from GenBank for the following taxa in the family Unionidae: Anodonta cygnea, Anodonta anatina,

Sinanodonta woodiana, Utterbackia peninsularis, Utterbackia imbecillis, Pyganodon grandis, Lasmigona compressa, Margaritifera margarifera, Margaritifera falcata,

Cumberlandia monodonta, Toxolasma glans, Toxolasma parvum and Potamilus alatus

(Table S1). Because taxonomy of freshwater mussels is frequently challenging and the subject of considerable debate (e.g., Stanton et al. 2012), we have used some principles from the Barcode of Life initiative (http://www.barcodeoflife.org) to help select appropriate sequences for analysis. Species barcode thresholds are usually around >2% nucleotide divergence in mammals (Krishnamurthy and Francis 2012) and >3% in lepidopterans (Hebert et al. 2003), but previous work on Pygandon species (Doucet-

Beaupré et al. 2012) used a difference of 1.8% between M-types and 0.43% of F-types partial cox1 and cox2 to infer two separate species. To be conservative, sequences that were similar across separate studies were used, others that were above a 2% divergence threshold difference were removed. For this reason, some sequence data available from

106

GenBank were not included (e.g. in species S. woodiana and P. grandis), and a sixth available H-type sequence (L. subviridis) was removed as it is unclear what sequence data is the most likely correct identification. Sequences were aligned in MEGA 7.0.16

(Kumar et al 2016) with ClustalW (Thompson et al 1994) using default parameters (i.e.

International Union of Biochemistry [IUB] cost matrix, a gap open cost of 15 and a gap extend cost of 6.66). A maximum likelihood (ML) tree was produced in MEGA 7.0.16

(Kumar et al 2016) using representative sequences from each species included in this study, with 1000 bootstrap replicates and a Tamura-Nei (Tamura and Nei 1993) evolutionary model. The tree was then visualized using FigTree version 1.4.3 (Rambaut

2007). For comparative purposes, five groups of F- and H-type sequences were identified using the phylogeny presented in Breton et al (2011) and Lopes-Lima et al.

(2017). Each group includes two dioecious species, for which both F- and M-type cox1 sequences were available, and for which a closely related H-type cox1 sequence was also available. Specific comparisons consisted of (1) M-types for both dioecious species in the groups identified, and (2) F-types of both dioecious species and the closely related H-type for each group identified. The formation of a group is examined in

Figure 1. These groupings are set up in such a way that the distance between female 1 and the hermaphrodite can be compared with the distance between female 2 and the same hermaphrodite. It should be noted that although one female and the hermaphrodite share a “closer” common ancestor (C2), it is not clear if the initial divergence was between a dioecious species and a hermaphrodite. In other words, the obligate hermaphrodite may have been diverged from the common ancestor between itself and its closely related dioecious species as a dioecious species before transitioning into an

107

obligate hermaphrodite at a later date. As discussed below, it is not known at what point along the branch to a hermaphroditic sequence (i.e., distance “b” in Figure 1), that the species actually evolved hermaphroditism. The five groups are the following: (1) A. anatina, S. woodiana and A. cygnea (2) U. peninsularis, P. grandis and U. imbecillis (3)

U. peninsularis, P. grandis and L. compressa (4) M. margaritifera, C. monodonta and

M. falcata (5) T. glans, P. alatus and T. parvum.

For each group, the nucleotide p-distance and Tamura-Nei distance (Tamura and Nei

1993) with standard error (1000 bootstrap replicates) were calculated using MEGA

7.0.16 (Kumar et al. 2016). Overall nucleotide sequence divergence values between the

F- and H-types within a dioecious/hermaphroditic grouping were calculated as per

Table 1. For comparative purposes, the M- vs. M-type distance values were also calculated. Species chosen for these comparisons were limited to sections of the family

Unionidae for which there was phylogenetic evidence that a dioecious species had speciated into a hermaphroditic species and for which with both F- and M-type cox1 sequences were available from GenBank.

For each of the five established groups an ML tree was produced using the same parameters as the “all species ML tree”. Groupings also included any additional sequence data used for comparative purposes (i.e., an outgroup). These trees used representative cox1 sequence from each taxon.

108

Nucleotide divergence between five hermaphrodites and an F-type cox1 outgroup were compared to the nucleotide divergence between closely F-type cox1 sequences.

Outgroups diverged before the most recent common ancestor of the H-type and F-type cox1 being compared. Nucleotide divergences were calculated as described above, except that the representative species from the group ML trees were used.

Testing relative rates of divergence of F- vs. H-type unionid cox1 sequences

A relative rates test was performed on nucleotide sequences between H-type cox1 sequences and cox1 sequences of a closely related F-type using an outgroup F-type sequence as the reference, which diverged earlier than the divergence between the F- and H-type being assessed, using MEGA 7.0.16 (Kumar et al 2016) and the Tajima’s D method (Tajima 1993). Complete deletion was used and all codon positions were assessed. This test provides evidence of whether or not the H-type is relatively more constrained than the F-type.

Comparing synonymous and nonsynonymous substitutions, respectively, between pairs of H-type cox1 and F-type cox1 of closely related species

As for the first relative rates test described above, sequences included in this comparison consisted of an H-type cox1 sequence, a closely related F-type cox1 sequence, and a homologous cox1 sequence from an outgroup F-type sequence, which diverged earlier than the divergence between the F- and H-type being assessed. All outgroup species branched off earlier than the split between the hermaphrodite and its dioecious sister species. MEGA 7.0.16 (Kumar et al 2016) was used to determine the

109

number of synonymous and nonsynonymous substitutions, respectively, between all available H-type cox1 and F-type cox1 sequences compared with the outgroup. The difference between these comparisons was calculated to determine if the number of synonymous and nonsynonymous substitutions, respectively, between the H- and F-type cox1 sequences and the outgroup were approximately equal or not. This analysis provides insight into the change in amino acids over time to infer if selection is relatively relaxed or constrained.

Testing unionid cox1 sequence data for statistical evidence of purifying, positive and neutral evolution

For each of the five H-type species and eight F-type species included in our comparisons, 422 cox1 sequences were aligned in MEGA 7.0.16 (Kumar et al 2016) with ClustalW (Thompson et al 1994) with default parameters. Each species was tested for evidence of purifying selection in MEGA 7.0.16 (Kumar et al 2016) using both the

Nei-Gojobori (Jukes-Cantor) (Zhang et al. 1998) and Kumar methods (Nei and Kumar

1998) with 1000 bootstrap replicates and pairwise deletion. In cases where tests of purifying selection were rejected, a test of positive selection was performed using the same parameters and models. In any cases where neither purifying or positive selection is supported by statistical tests, then a test of neutral selection was conducted. These tests provide insight into what selection is occurring within cox1 of FWMs.

110

RESULTS AND DISCUSSION

Phylogenetic relationships among the M-, F- and H-types of unionid species

The relationships among all species within this collection of unionid M-, F-, and H- types are shown in a maximum likelihood (ML) tree based on cox1 sequence data

(Figure 2). This tree does not completely reflect relationships established by other FWM trees (e.g., Breton et al. 2011; Lopes-Lima et al. 2017) possibly because only one gene was included in this analysis (vs. two or more genes in the other studies), and because different species were included in each study. However, the tree presented in Figure 2 serves the purpose of showing, generally, how these species are related to each other, and indicating which F-type cox1 sequences are most closely related to which H-type cox1 sequences. As expected, the nucleotide divergence values (Table 1) of the five focal groups coincide well with the observed branch lengths on the ML tree of all species (Figure 2), which group F- and H-type species with a relatively small cox1 sequence divergence together, although not all of these relationships have strong bootstrap supports (e.g. >51). An ML tree of each group (groups 1-5; Figure 3) depicts the relationships among sequences that was assumed for later analysis (i.e., establishing the relationships used for conducting relative rates tests). The complete tree (Figure 2) clearly shows that there are multiple examples of the transition from dioecy to hermaphroditism among the collection of F- and H-type sequences, and that there is a clear distinction between the F- and H-type lineage and the M-type lineage.

111

As expected, both Tamura-Nei and p-distance nucleotide divergence values suggest a larger sequence divergence between M-type cox1 of closely related species than between the corresponding F-type cox1 sequences for the same species (Table 1A).

Both Tamura-Nei and p-distance nucleotide divergence values also suggest that H-type mtDNA is more similar to one F-type than the two F-types are to each other (based on cox1 alone). An exception to this pattern is evident in group 3, which contained the dioecious species, U. peninsularis and P. grandis, and a hermaphroditic species, L. compressa. This exception is explored in more detail below. The groupings we chose to examine were limited by the availability of finding two dioecious, closely related species that also have M-type cox1 sequence data available; although the most closely related species were chosen based on this requirement, some groups had a more closely related alternative F-type cox1 sequences available for comparison with the relevant H- type sequence. In the case of group 3 the addition of one of these more closely related

F-type cox1 sequences (from A. heterodon) provided an F-type that had a smaller sequence divergence to the H-type than the nucleotide divergence between F-types. A. heterodon, has an extremely small divergence value from the H-type of group 3 (i.e., L. compressa). When nucleotide divergence was calculated between a closely related F- type of a dioecious species and an H-type (hermaphrodite) to an F-type outgroup, the difference in these nucleotide divergences were greater between the F-type and the outgroup than between the H-type and the outgroup in three out of five comparisons

(Table 1B). For one comparison, the nucleotide divergences between the F-type and the reference outgroup species and between the H-type and the reference outgroup species were equal. In the case of A. cygnea and A. anatina, the nucleotide divergence values

112

were larger between the hermaphrodite and the outgroup compared to the F-type and the outgroup. Curiously, this pattern, i.e., greater divergence between the A. anatina F-type and the outgroup than between the A. cygnea iH-type and the outgroup is an exception to the trend observed for these same comparisons for all 13 mitochondrial genes (see below). It is possible that Pseudanodonta complanata may be a more appropriate dioecious F-type for the comparison between A. cygnea and outgroup P. grandis because it has been suggested by Källersjö et al. (2005) that A. cygnea is more closely related to P. complanata than it is to A. anatina. Nonetheless, three out of five comparisons suggest that the H-type cox1 is evolving at a slower rate than the F-type cox1 between closely related species.

Overall, these findings are consistent with the hypothesis that H-types are more slowly evolving and possibly under more selective constraints than F-types, however, these results must be considered in the context of when exactly the ancestor of the hermaphroditic species under examination made the transition from dioecy to hermaphroditism (explored further below). A lack of significant nucleotide divergence can be explained by some hermaphroditic species simply being too “new”, and not having experienced selective pressures associate with hermaphroditism for very long.

This complicates tests of molecular clocks (such as Tajima’s D) as ideally the comparison should be between an obligate hermaphrodite and a dioecious female, and not include a very recent, and possibly even an “accidental” hermaphrodite or a dioecious population on its way to transitioning into an obligate hermaphroditic species.

113

If selective constraints placed on H-type mtDNA are larger than those placed on F- types, which would follow the arenas hypothesis, then H-type cox1 sequence should be diverging at a slower rate than F-type sequences. Tajima’s D (Tajima 1989) relative rates test can be used to determine if sequences are evolving at unequal rates. When

Tajima’s D (Table 3) was preformed between F- and H-type comparisons among groups that had a nucleotide divergence smaller than F- vs F-types only one comparison rejected the null hypothesis of equal evolutionary rates between the two species. Taken as a collective, these results do not support the arenas hypothesis of unequal selective constraints between H- and F-type sequences within cox1 specifically. However,

Tajima’s D has difficulty detecting a significant difference in relatively short sequence data (<1000bp in length; Robinson et al. 1998), such as the cox1 sequences used in this study (approximately 600bp in length). Therefore, a statistical evaluation of unequal rates between lineages is currently left unresolved for the cox1 sequences themselves, however, using complete mtDNA of an F- and H-type ingroup and an appropriate F- type outgroup should provide enough sequence data to perform a Tajima’s D test.

Preforming relative rates test on A. anatina and A. cygnea (ingroups) and P. grandis

(outgroup) complete mtDNA genomes is currently being pursued (Chase et al., in prep.).

When the overall level of nucleotide divergence is teased apart into numbers of synonymous and nonsynonymous substitutions, there is no clear difference in the number of non-synonymous substitutions between H- and F-type cox1 sequences compared to corresponding outgroup species (Table 4), with the possible exception of T.

114

parvum and T. glans compared to the outgroup V. ellipsiformis. Overall, this suggests that selection on the H-type mtDNA is not any “better” at removing non-synonymous substitutions, in the specific case of cox1, for the H-types compared to the F-types.

However, the total number of nucleotide substitutions, and the number of nonsynonymous, or amino acid replacement substitutions, in particular is extremely small.

The non-synonymous substitutions present in the sequences may not cause particularly important changes in the COX1 protein. They may, for example, potentially just impact structural components of the protein rather than occur in the active sites of the enzyme.

Nonetheless, if more substitutions of this kind occurred within F-type cox1 sequence this would provide support for stronger selective constraints on H-type compared to F- type. The same assessment of all other protein coding genes with a complete mitochondrial genome of A. cygnea, A. anatina and outgroup P. grandis do show differences in the number of non-synonymous corresponding to the sequence divergence between A. cygnea and A. anatina from Chase et al. (in prep; see Chapter 2)

(Table 5). In other words, genes that have a higher nucleotide divergence between two species would have a greater difference in number of non-synonymous substitutions in comparison to an outgroup, most likely because there are relatively lower selective constraints on those particular genes. For example, atp8 is known to be highly divergent among bivalves generally (Breton et al. 2010; Śmietanka et al. 2010), and in the gene by gene comparison conducted herein, atp8 has the greatest difference in number of nonsynonymous substitutions. Another example is with nad2 comparisons, which have the second largest divergence values. However, in the case of atp8, this gene has fewer nonsynonymous substitutions in A. cygnea than in A. anatina when compared to P.

115

grandis and the opposite is true for nad2. It is worth noting that when considering standard error for these measures that these differences in the number of nonsynonymous substitutions are not clear in the cox1 (Table 4) comparisons. To summarize this portion of the analysis, there is not strong support for a difference n the number of nonsynonymous substitutions between F- and H-type mtDNAs based on cox1 alone. It is worth noting that background mutation rate as measured by the number of synonymous substitutions is about equal for cox1 from different species.

Consequently, the cox1 sequence analysis does not provide support for the arenas hypothesis when applied to H- and F-type mtDNA. However, when a complete gene by gene comparison is made (Table 5), there is good evidence that there is an overall difference between H- and F-type mtDNA that is consistent with the arenas hypothesis.

The specific comparisons between different genes varies up or down, however, overall the total non-synonymous substitutions indicate that the H-type mtDNA may be more constrained, and evolving more slowly, than the F-type mtDNA. Therefore, this observation provides support for the arenas hypothesis when applied to the H- and F- type mtDNA. The addition of the A. cygnea mtDNA genome has enabled this complete genome comparison and has provided valuable insights into the relative rates of molecular evolution of the F- and H-type mtDNAs in this genus of freshwater mussels.

The transition from a dioecious to obligate hermaphroditic sexual strategy in unionids may be coupled with the progression of mitochondrial DNA towards purifying selection

116

Purifying selection (or background selection) is characterized by the removal of deleterious mutations from a coding sequence to maintain protein structure and function

(Charlesworth et al. 1995). Testing for purifying selection indicates that with few exceptions F-type cox1 fall into two extremes: a strong indication of purifying selection, or an indication of the absence of purifying selection (Table 6). Tests of selection are based on a number of factors, including the ratio of non-synonymous (dN) to synonymous substitutions (dS), dN / dS (Table S2) and the length of sequence being analyzed. Within H-type cox1 sequences three out of five sets of gene sequences examined show clear support for purifying selection. In contrast, both U. imbecillis and

L. compressa do not show such clear evidence, however, statistical support for the rejection of purifying selection is not as strong as in the majority of F-type cases (e.g., p values were 0.391 and 0.132 for this test for U. imbecillis and L. compressa, respectively). While findings are either statistically significant or not, it is interesting to explore why H-type results do not follow the almost clear-cut pattern of p-values either being 1.0 or 0.0 among most F-type tests. Work conducted by Hoeh et al. (1995) suggests that U. imbecillis is a relatively recent hermaphrodite. Given this, it may be possible that U. imbecillis mtDNA has yet to demonstrate purifying selection, but may be on its way to it. If this is the case, then U. imbecillis would be expected to have a smaller nucleotide divergence to a closely related F-type cox1 sequence than other comparisons that have statistically significant support for purifying selection, which is the case (Table 1). Given that L. compressa cox1 also did not provide statistically significant support for purifying selection, and if all H-types progress to purifying selection this may indicate that L. compressa is also a very recent dioecious to

117

hermaphroditic transition (newly diverged H-type from an F-type mtDNA). Within H- type cox1 sequences, species with a nucleotide divergence of 0.117 or higher had statistically significant support for purifying selection. Consequently, it is possible that all H-type mtDNA sequences may eventually progress towards demonstrating purifying selection.

Among the eight FWM species F-type cox1 sequences tested for purifying selection, four have statistically significant support for purifying selection and four do not. It is not entirely clear why, for the majority of dioecious F-types, purifying test results are so clear cut as either having strong support for rejecting the null hypothesis of dN = dS in favour or purifying selection or strong support for accepting dN = dS. For example, S. woodiana has a p-value of 1.000, thereby rejecting the null hypothesis, and A. anatina has a p-value of 0.0000, thereby accepting the null hypothesis. Purifying selection also does not appear to persist among closely related species (e.g., both C. monodonta and

M. margaritifera, of the family Margaritiferidae, are not under purifying selection, whereas M. falcata is). In other words, phylogenetic relationships do not appear to be a factor in whether or not cox1 is under purifying selection within a species. Purifying selection of mtDNA is logical given that mtDNA codes for proteins playing highly specific cellular roles in oxidative phosphorylation that most likely benefit from removal of deleterious mutations. In contrast, some nuclear genes code for proteins for which changes in amino acid sequences can be highly beneficial (e.g., MHC genes that play a role in the immune response; Rand 2001). Maintaining the current functions of mtDNA coded proteins appears to be beneficial – “if it is not broken then why fix it”?

118

In fact, the majority of mitochondrial genes seem to evolve under purifying selection

(Rand 2001; Ballard and Whitlock 2003; Stewart et al. 2008), and the role of energy metabolism has been credited with the presence of purifying selection of mtDNA (Shen et al. 2009; Feng et al. 2015). More specifically, work conducted on mice by Stewart et al. (2008) found that the strongest support for purifying selection in mtDNA was on cox1 and cox2. Previous work has also indicated that overall cox1 has a smaller sequence divergence compared to cob, implying stronger selective constraints on cox1 than cob generally (Meiklejohn et al. 2007). Bazin et al. (2006) challenged the idea that mtDNA is almost certainly under purifying selection by exploring polymorphism data across many vertebrate and invertebrate species. They conclude that adaptive selection is likely occurring on mtDNA, and that patterns suggest that selective sweeps are a potential factor in this adaptation. Selective sweeps include both “hard sweeps”, in which a novel adaptation is quickly brought to fixation, and “soft sweeps”, in which previously neutral mutation becomes favourable as the environment changes over time

(Ferrer-Admetlla et al. 2014). Our results raise the following questions: (1) Is it possible that hard and/or soft selective sweeps are influencing this seemingly all-or-nothing, clear cut pattern or lack thereof among F-type mtDNA within unionids? And (2) Are novel or previously neutral alleles permitting specific FWMs to excel in new or changing environments?

Among the six species (two hermaphrodites and four dioecious female) that did not test positive for purifying selection, three of those did test positive for positive selection and the remaining three (which included both H-types that did not test positive for purifying

119

selection; U. imbecillis and L. compressa) did not test positive for neutral evolution

(Table 7, Table S3). These two H-type cox1 do not appear to have clear statistically significant support for any of the three types of selection tested for here, which may strengthen the proposal that they are progressing towards showing evidence of purifying selection and that all H-type cox1 sequences may eventually do the same. The major exception to this is that T. glans also appears to potentially trend towards purifying selection. This raises the question of whether hermaphroditism may be an example of a soft selective sweep in which the transition from a dioecious to a hermaphroditic sexual strategy is exceptionally favourable in a particular environment, and the mitochondrial genome contains a combination of synonymous and non-synonymous substitutions that show a mix of bursts of rapid adaptation to the new reproductive mode coupled with gradual tweaking of gene mutations to become more optimal over time. It is difficult to conceptualize (although not impossible) why a change in cox1 per se would permit such a sweep, however, genetic hitchhiking (Gillespie 2000, 2001) may be occurring, which permits “nearby” polymorphisms to also change in frequencies alongside mutations that are the focus of the selective sweep.

Cox1 sequence data alone may not provide a clear picture or substantial support for an extension of the arenas hypothesis with respect to F- and H-type mtDNA

In conclusion, our tests do not provide strong support that, based on the arenas hypothesis, H-type mtDNA in unionids is under more constrained selection than F-type mtDNA based on cox1 sequences. It is possible that examination of another single gene

120

may provide a different picture as there are clear differences in the number of nonsynonymous substitutions among other mitochondrial genes overall, however, these differences do not exclusively show that H-type mtDNA possesses fewer nonsynonymous substitutions than F-type mtDNA (in the case of A. cygnea vs A. anatina). It may be possible that more conserved mitochondrial genes (such as cox1), which are frequently under purifying selection, do not have a clear difference in divergence between H- and F-type. It may also be possible that mitochondrial genes that are under relatively relaxed selection may provide a clearer picture of different selective constraints among H- and F-type mtDNA. Most importantly complete mtDNA overall

(as indicated by Table 5) may provide the clearest picture of whether or not H-type mtDNA is more (or less) constrained than F-type mtDNA. The results of this study suggest that M-type cox1 sequences is evolving faster than F-type cox1 sequences, which agrees with previous studies. Furthermore, there is evidence that F-type complete mtDNA evolves faster than complete H-type mtDNA. Sequencing of unionid mtDNA is an ongoing process, consequently additional data sets for comparing patterns of molecular evolution of M-, F- and H-type mtDNA will increase in time.

Conclusions and suggestions for future work

By assessing cox1 sequence data alone among F-, H- and M-type mtDNA and complete mtDNA of A. anatina, A. cygnea and P. grandis we have provided evidence for the arenas hypothesis, specifically for a higher level of selective constraints on F-type mtDNA compared to H-type mtDNA. We have demonstrated that (1) three out of five

121

comparisons of two dioecious (M- and F-type) and one hermaphrodite (H-type) cox1 sequences suggest that H-type mtDNA is evolving more slowly than F-type mtDNA and

(2) among complete mtDNA of A. cygnea, A. anatina and P. grandis H-type evolves more slowly than F-type within most genes and overall. Alongside the fact that H-type mtDNA is operating in more cellular arenas than F-type mtDNA, it is also possible that

H-type mtDNA would be under stronger selective constraints because, in the case of obligate hermaphrodites, reproducing solely through selfing there would be no offspring produced if sperms (which will also contain and be driven by the H-type mtDNA) do not function. This is in contrast to other animals exhibiting SMI in which non- functioning sperm can be passed on from mother to son, and then passed to further generations through daughters (Gemmell et al. 2004; Pal et al. 2017). In hermaphrodites, SMI cannot mediate for non-functional sperm, the individual will simply be unable to reproduce. As a consequence, it is logical to assume that the H-type will be under stronger selection to produce fully functional eggs and sperm compared to a dioecious female that exists in a population of other individuals who can provide sperm. It is important to note that this is under the assumption that obligate hermaphrodites reproduce through selfing alone, which currently is difficult to detect and not always fully discussed in the FWM literature. Further assessments of the reproductive biology of various FWM species and specific populations will need to occur when more complete mtDNA of all mitotypes become available.

122

FIGURES

Figure 1. Relative rates analysis group set up. “C” denotes a common ancestor, “F” denotes a female of a dioecious species, “H” denotes an obligate hermaphrodite. The distance from common ancestor two to female one is represented by “a”, likewise the distance to the hermaphrodite is “b”. The distance from common ancestor two to the female two is represented by “c”. The two dioecious species and one obligate hermaphrodite form one group. The male of both dioecious species is not pictured here, but fall on a different clade.

123

Figure 2. Maximum likelihood tree, with 1000 bootstrap replicates, of unionid cox1 sequences used in this study. Red branches indicate a high bootstrap value (>70), blue covers the F-type lineage and purple covers the M-type lineage. A star (*) denotes H- type cox1.

124

Figure 3. Maximum likelihood tree, with 1000 bootstrap replicates, of unionid cox1 sequences that have been grouped across analysis in this study, with additional outgroup species. Where “F”, “M” and “H” denote F-type, M-type and H-type mtDNA respectively. A, B, C, D and E represent groups 1, 2, 3, 4 and 5 respectively.

125

TABLES

126

127

128

129

130

REFERENCES

Ballard, J.W.O., Whitlock, M.C., 2004. The incomplete natural history of mitochondria.

Molecular Ecology 13, 729–744.

Bazin, E., 2006. Population size does not influence mitochondrial genetic diversity in

animals. Science 312, 570–572.

Birky, C.W., 1995. Uniparental inheritance of mitochondrial and chloroplast genes:

mechanisms and evolution. PNAS 92, 11331–11338.

Boyle, E.E., Etter, R.J., 2013. Heteroplasmy in a deep-sea protobranch bivalve suggests

an ancient origin of doubly uniparental inheritance of mitochondria in Bivalvia.

Marine Biology 160, 413–422.

Breton, S., Beaupré, H.D., Stewart, D.T., Hoeh, W.R., Blier, P.U., 2007. The unusual

system of doubly uniparental inheritance of mtDNA: isn’t one enough? Trends

in Genetics 23, 465–474.

Breton, S., Beaupré, H.D., Stewart, D.T., Piontkivska, H., Karmakar, M., Bogan, A.E.,

Blier, P.U., Hoeh, W.R., 2009. Comparative mitochondrial genomics of

freshwater mussels (Bivalvia: Unionoida) with doubly uniparental inheritance of

mtDNA: gender-specific open reading frames and putative origins of replication.

Genetics 183, 1575–1589.

Breton, S., Burger, G., Stewart, D.T., Blier, P.U., 2006. Comparative analysis of

gender-associated complete mitochondrial genomes in marine mussels (Mytilus

spp.). Genetics 172, 1107–1119.

Breton, S., Capt, C., Guerra, D., Stewart, D., 2017. Sex determining mechanisms in

bivalves. Preprints, 2017060127.

131

Breton, S., Stewart, D.T., Hoeh, W.R., 2010. Characterization of a mitochondrial ORF

from the gender-associated mtDNAs of Mytilus spp. (Bivalvia: Mytilidae):

identification of the “missing” ATPase 8 gene. Marine Genomics 3, 11–18.

Breton, S., Stewart, D.T., Shepardson, S., Trdan, R.J., Bogan, A.E., Chapman, E.G.,

Ruminas, A.J., Piontkivska, H., Hoeh, W.R., 2011. Novel protein genes in

animal mtDNA: A new sex determination system in freshwater mussels

(Bivalvia: Unionoida)? Molecular Biology and Evolution 28, 1645–1659.

Burt, A., Trivers, R., 2006. Genes in conflict: the biology of selfish genetic elements.

Harvard University Press.

Charlesworth, D., Charlesworth, B., Morgan, M.T., 1995. The pattern of neutral

molecular variation under the background selection model. Genetics 141, 1619–

1632.

Crow, J.F., 2006. A selfish element. Nature 440, 609–609.

Dalziel, A.C., Stewart, D.T., 2002. Tissue-specific expression of male-transmitted

mitochondrial DNA and its implications for rates of molecular evolution in

Mytilus mussels (Bivalvia: Mytilidae). Genome 45, 348–355.

Doucet-Beaupré, H., Blier, P.U., Chapman, E.G., Piontkivska, H., Dufresne, F.,

Sietman, B.E., Mulcrone, R.S., Hoeh, W.R., 2012. Pyganodon (Bivalvia:

Unionoida: Unionidae) phylogenetics: A male- and female-transmitted

mitochondrial DNA perspective. Molecular Phylogenetics and Evolution 63,

430–444.

Doucet-Beaupré, H., Breton, S., Chapman, E.G., Blier, P.U., Bogan, A.E., Stewart,

D.T., Hoeh, W.R., 2010. Mitochondrial phylogenomics of the Bivalvia

132

(Mollusca): searching for the origin and mitogenomic correlates of doubly

uniparental inheritance of mtDNA. BMC Evolutionary Biology 10, 50.

Feng, P., Zhao, H., Lu, X., 2015. Evolution of mitochondrial DNA and its relation to

basal metabolic rate. Mitochondrial DNA: The Journal of DNA Mapping,

Sequencing & Analysis 26, 566–571.

Ferrer-Admetlla, A., Liang, M., Korneliussen, T., Nielsen, R., 2014. On detecting

incomplete soft or hard selective sweeps using haplotype structure. Molecular

Biology and Evolution. 31, 1275–1291.

Garrido-Ramos, M.A., Stewart, D.T., Sutherland, B.W., Zouros, E., 1998. The

distribution of male-transmitted and female-transmitted mitochondrial DNA

types in somatic tissues of blue mussels: Implications for the operation of doubly

uniparental inheritance of mitochondrial DNA. Genome 41, 818–824.

Gemmell, N.J., Metcalf, V.J., Allendorf, F.W., 2004. Mother’s curse: the effect of

mtDNA on individual fitness and population viability. Trends in Ecology &

Evolution 19, 238–244.

Gillespie, J.H., 2001. Is the Population Size of a Species Relevant to Its Evolution?

Evolution 55, 2161–2169.

Gillespie, J.H., 2000. Genetic drift in an infinite population: The pseudohitchhiking

model. Genetics 155, 909–919.

Hayward, A., Gillooly, J.F., 2011. The cost of sex: quantifying energetic investment in

gamete production by males and females. PLoS ONE 6, e16557.

133

Hebert, P.D.N., Cywinska, A., Ball, S.L., deWaard, J.R., 2003. Biological

identifications through DNA barcodes. Proceedings of the Royal Society of

London B: Biological Sciences 270, 313–321.

Hoeh, W.R., Stewart, D.T., Saavedra, C., Sutherland, B.W., Zouros, E., 1997.

Phylogenetic evidence for role-reversals of gender-associated mitochondrial

DNA in Mytilus (Bivalvia: Mytilidae). Molecular Biology and Evolution 14,

959–967.

Hoeh, W.R., Stewart, D.T., Sutherland, B.W., Zouros, E., 1996. multiple origins of

gender-associated mitochondrial DNA lineages in Bivalves (Mollusca:

Bivalvia). Evolution 50, 2276.

Hurst, L.D., Hoekstra, R.F., 1994. Evolutionary genetics: Shellfish genes kept in line.

Nature 811–812.

Krishna Krishnamurthy, P., Francis, R.A., 2012. A critical review on the utility of DNA

barcoding in biodiversity conservation. Biodiversity and Conservation 21, 1901–

1919.

Kumar, S., Stecher, G., Tamura, K., 2016. MEGA7: Molecular evolutionary genetics

analysis version 7.0 for bigger datasets. Molecular Biology and Evolution 33,

1870–1874.

Lopes-Lima, M., Froufe, E., Do, V.T., Ghamizi, M., Mock, K.E., Kebapçı, Ü., Klishko,

O., Kovitvadhi, S., Kovitvadhi, U., Paulo, O.S., Pfeiffer, J.M., Raley, M.,

Riccardi, N., Şereflişan, H., Sousa, R., Teixeira, A., Varandas, S., Wu, X.,

Zanatta, D.T., Zieritz, A., Bogan, A.E., 2017. Phylogeny of the most species-

rich freshwater bivalve family (Bivalvia: Unionida: Unionidae): Defining

134

modern subfamilies and tribes. Molecular Phylogenetics and Evolution 106,

174–191.

Meiklejohn, C.D., Montooth, K.L., Rand, D.M., 2007. Positive and negative selection

on the mitochondrial genome. Trends in Genetics 23, 259–263.

Milani, L., Ghiselli, F., Guerra, D., Breton, S., Passamonti, M., 2013. A comparative

analysis of mitochondrial ORFans: New clues on their origin and role in species

with doubly uniparental inheritance of mitochondria. Genome Biology and

Evolution 5, 1408–1434.

Mizi, A., Zouros, E., Moschonas, N., Rodakis, G.C., 2005. The complete maternal and

paternal mitochondrial genomes of the Mediterranean mussel Mytilus

galloprovincialis: implications for the doubly uniparental inheritance mode of

mtDNA. Molecular Biology and Evolution. 22, 952–967.

Nei, M., Kumar, S., 2000. Molecular Evolution and Phylogenetics. Oxford University

Press.

Ort, B.S., Pogson, G.H., 2007. Molecular population genetics of the male and female

mitochondrial DNA molecules of the California sea mussel, Mytilus

californianus. Genetics 177, 1087–1099.

Pal, A.K., Ambulkar, P.S., Sontakke, B.R., Waghmare, J.E., Shende, M.R., Tarnekar,

A.M., 2017. Role of nuclear and mitochondrial genes in human male infertility:

a review. Nucleus 60, 209–220.

Quesada, H., Gallagher, C., Skibinski, D. a. G., Skibinski, D.O.F., 1998. Patterns of

polymorphism and gene flow of gender-associated mitochondrial DNA lineages

in European mussel populations. Molecular Ecology 7, 1041–1051.

135

Rambaut, A., 2007. FigTree, a graphical viewer of phylogenetic trees. Available from

http://tree.bio.ed.ac.uk/software/figtree/.

Rand, D.M., 2001. The units of selection on mitochondrial DNA. Annual Review of

Ecology and Systematics 32, 415–448.

Rawson, P.D., Hilbish, T.J., 1995. Evolutionary relationships among the male and

female mitochondrial DNA lineages in the Mytilus edulis species complex.

Molecular Biology and Evolution 12, 893–901.

Robicheau, B.M., Powell, A.E., Del Bel, L., Breton, S., Stewart, D.T., 2017. Evidence

for extreme sequence divergence between the male- and female-transmitted

mitochondrial genomes in the bivalve mollusc, Modiolus modiolus (Mytilidae).

Journal of Zoological Systematics and Evolutionary Research 55, 89–97.

Robinson, M., Gouy, M., Gautier, C., Mouchiroud, D., 1998. Sensitivity of the relative-

rate test to taxonomic sampling. Molecular Biology and Evolution 15, 1091–

1098.

Shen, Y.-Y., Shi, P., Sun, Y.-B., Zhang, Y.-P., 2009. Relaxation of selective constraints

on avian mitochondrial DNA following the degeneration of flight ability.

Genome Research 19, 1760–1765.

Skibinski, D., Gallagher, C., Beynon, C.M., 1994. Sex-limited mitochondrial DNA

transmission in the marine mussel Mytilus Edulis. Genetics 138, 801–809.

Śmietanka, B., Burzyński, A., Wenne, R., 2010. Comparative genomics of marine

mussels (Mytilus spp.) gender associated mtDNA: rapidly evolving atp8. Journal

of Molecular Evolution 71, 385–400.

136

Stanton, L.M., Hoeh, W.R., McAlpine, D.F., Hebda, A., Stewart, D.T., 2012. mtDNA

and AFLP markers demonstrate limited genetic differentiation within the

Pyganodon cataracta–Pyganodon fragilis freshwater mussel complex in

Atlantic Canada. Canadian Journal of Zoology. 90, 1307–1319.

Stewart, D.T., Hoeh, W.R., Bauer, G., Breton, S., 2013. Mitochondrial genes, sex

determination and hermaphroditism in freshwater mussels (Bivalvia:

Unionoida), in: Pontarotti, P. (Ed.), Evolutionary Biology: Exobiology and

Evolutionary Mechanisms. Springer Berlin Heidelberg, Berlin, Heidelberg, pp.

245–255.

Stewart, D.T., Kenchington, E.R., Singh, R.K., Zouros, E., 1996. Degree of selective

constraint as an explanation of the different rates of evolution of gender-specific

mitochondrial DNA lineages in the mussel Mytilus. Genetics 143, 1349–1357.

Stewart, D.T., Saavedra, C., Stanwood, R.R., Ball, A.O., Zouros, E., 1995. Male and

female mitochondrial DNA lineages in the blue mussel (Mytilus edulis) species

group. Molecular Biology and Evolution 12, 735–747.

Stewart, J.B., Freyer, C., Elson, J.L., Wredenberg, A., Cansu, Z., Trifunovic, A.,

Larsson, N.-G., 2008. Strong purifying selection in transmission of mammalian

mitochondrial DNA. PLoS Biol 6.

Tajima, F., 1993. Simple methods for testing the molecular evolutionary clock

hypothesis. Genetics 135, 599–607.

Tamura, K., Nei, M., 1993. Estimation of the number of nucleotide substitutions in the

control region of mitochondrial DNA in humans and chimpanzees. Molecular

Biology and Evolution 10, 512–526.

137

Thompson, J.D., Higgins, D.G., Gibson, T.J., 1994. CLUSTAL W: improving the

sensitivity of progressive multiple sequence alignment through sequence

weighting, position-specific gap penalties and weight matrix choice. Nucleic

Acids Research 22, 4673–4680.

Venetis, C., Theologidis, I., Zouros, E., Rodakis, G.C., 2006. No evidence for presence

of maternal mitochondrial DNA in the sperm of Mytilus galloprovincialis males.

Proceedings of the Royal Society of London B: Biological Sciences 273, 2483–

2489.

Whittle, C.-A., Johnston, M.O., 2002. Male-driven evolution of mitochondrial and

chloroplastidial DNA sequences in plants. Molecular Biology and Evolution 19,

938–949.

Zbawicka, M., Burzyński, A., Wenne, R., 2007. Complete sequences of mitochondrial

genomes from the Baltic mussel Mytilus trossulus. Gene 406, 191–198.

Zhang, J., Rosenberg, H.F., Nei, M., 1998. Positive Darwinian selection after gene

duplication in primate ribonuclease genes. PNAS 95, 3708–3713.

Zouros, E., 2013. Biparental inheritance through uniparental transmission: the doubly

uniparental inheritance (DUI) of mitochondrial DNA. Evolutionary Biology 40,

1–31.

Zouros E, Oberhauser Ball A, Saavedra C, Freeman Kr, 1994. An unusual type of

mitochondrial DNA inheritance in the blue mussel Mytilus. Proceedings of the

National Academy of Sciences USA 91, 7463–7467.

138