Tunicata 4 Alberto Stolfi and Federico D. Brown

Chapter vignette artwork by Brigitte Baldrian. © Brigitte Baldrian and Andreas Wanninger. A. Stolfi Department of Biology , Center for Developmental Genetics, New York University , New York , NY , USA F. D. Brown (*) EvoDevo Laboratory, Departamento de Zoologia , Instituto de Biociências, Universidade de São Paulo , São Paulo , SP , Brazil Evolutionary Developmental Biology Laboratory, Department of Biological Sciences, Universidad de los Andes, Bogotá , Colombia Centro Nacional de Acuicultura e Investigaciones Marinas (CENAIM) , Escuela Superior Politécnica del Litoral (ESPOL), San Pedro , Santa Elena , Ecuador e-mail: [email protected]

A. Wanninger (ed.), Evolutionary Developmental Biology of Invertebrates 6: Deuterostomia 135 DOI 10.1007/978-3-7091-1856-6_4, © Springer-Verlag Wien 2015

[email protected] 136 A. Stolfi and F.D. Brown

Above all , perhaps , I am indebted to a decidedly the phylogenetic relationships between the three vegetative , often beautiful , and generally obscure classes and many orders and families have yet to group of marine animals , both for their intrinsic interest and for the enjoyment I have had in search- be satisfactorily settled. Appendicularia, ing for them . N. J. Berrill (1955) Thaliacea, and Ascidiacea remain broadly used in textbooks and scientifi c literature as the three classes of tunicates; however, recent molecular INTRODUCTION phylogenies have provided support for the mono- phyly of only Appendicularia and Thaliacea, but Tunicates are a group of marine fi lter-feeding not of Ascidiacea (Swalla et al. 2000 ; animals1 that have been traditionally divided into Tsagkogeorga et al. 2009 ; Wada 1998 ). A para- three classes: (1) Appendicularia, also known as phyletic Ascidiacea calls for a reevaluation of larvaceans because their free-swimming and tunicate relationships. The most up-to-date phy- pelagic adult stage resembles a larva; (2) logeny with a comprehensive taxonomic sam- Thaliacea, which includes three orders of free- pling of tunicates using 18S rRNA supports three swimming and pelagic adult forms with complex clades for the Tunicata (Fig. 4.1 ; T s a g k o g e o r g a life cycles (Salpida, Pyrosomida, and Doliolida); et al. 2009 ): (1) Appendicularia, (2) Stolidobranch and (3) Ascidiacea, colloquially referred to as sea ascidians, and (3) Aplousobranch ascidians+ squirts, 2 which is comprised of diverse sessile Phlebobranch ascidians+Thaliacea. All three solitary and colonial species and includes some groups of the latter clade show resemblance in of the most extensively studied tunicates. It was association of the gonads to the gut in concor- mainly the ascidians that served as the inspiration dance with the Enterogona classifi cation (Perrier for seminal studies in developmental biology and 1898 ). However, the precise position of the evolution conducted last century by British zool- Appendicularia and the Thaliacea remains unre- ogist and prolifi c writer N. J. Berrill, whose quote solved. Appendicularia has traditionally been appears above. considered to be at the base of the Tunicata (Wada 1998 ; Swalla et al. 2000 ), but recent molecular phylogenies place them as sister group to Evolutionary Relationships Stolidobranchia (Zeng et al. 2006 ; T s a g k o g e o r g a et al. 2009 ). On the other hand, Thaliacea gener- Although Tunicata is well established as a mono- ally groups closer to Phlebobranchia than to phyletic group and united by their indisputably Aplousobranchia using ribosomal molecular synapomorphic ability to synthesize cellulose, phylogenies (Zeng et al. 2006 ; T s a g k o g e o r g a et al. 2009 ), but the low taxonomic sampling and 1 Actually, there are specialized predatory ascidians in the a high 18S evolutionary rate of thaliaceans and Molgulidae and Octacnemidae families that do not fi lter aplousobranchs bring some uncertainty to this feed but rather swallow whole crustaceans and other grouping. In an attempt to provide an evolution- invertebrates (Tatián et al. 2011 ). ary framework of the developmental mechanisms 2 The term “sea squirt” has been adopted for these ani- in Tunicata, developmental studies of several mals, as seawater can be strongly propelled through the oral siphon upon physical perturbation. Other colloquial species of tunicates are reviewed here in a com- terms apply to certain species or genera (“sea peach” for parative manner with respect to our understand- Halocynthia aurantium, “sea grape” for Molgula spp.). ing of tunicate relationships. The literate translation for “sea squirt” is used in Spanish ( chorro de mar), Finnish ( meritupet), Swedish ( Sjöpungar), or Polish (Ż achwy); more externally descrip- Relationships to Other Chordates tive terms are also used, such as “sea bag” ( Sekkedyr in For several centuries after Linnaeus fi rst estab- Norwegian or Søpunge in Danish), “sack pipe” ( Zakpijpen lished the classifi cation system for living organ- in Dutch), or “sea sheath” or “sea vagina” ( Seescheide in isms, ascidians were grouped together based on German). Humorous, if crude, terms abound especially in Portuguese, like mija- mija (“piss-piss”), Maria Mijona adult morphological features. This group con- (“Pissing Mary”), or mijão (“big pisser”). tained a stereotypical adult body plan including

[email protected] 4 Tunicata 137

Fig. 4.1 Current understanding of tunicate relationships gray ). Descriptions of the most parsimonious ancestral within the deuterostomes. Dashed lines represent possible characteristics are shown at every node representing pos- relationships in uncertain branches (see text for details). sible tunicate ancestors. Cases of multiple larval or adult Larval (below ) and adult (above ) forms have been ancestral forms are separated by “or”, direct development included at each node to facilitate understanding of dis- (no larva) is shown by “juvenile,” and unknown forms are tinct arguments discussed in the text of the major evolu- presented as a question mark (?) in a circle tionary transitions of extant clades and their ancestors (in

(1) a U-shaped gut delimited by an incurrent and tury, tunicates were grouped together with an excurrent siphon, (2) a large branchial cavity ectoprocts and brachiopods within the mol- for fi lter feeding, and (3) the tunic, a characteris- lusks (Fig. 4.2A ) (Milne-Edwards 1843 ; Haeckel tic mantle later discovered to be made of animal 1866 ). A unifi ed Tunicata as we consider them cellulose. Already by the turn of the nineteenth today was fi rst proposed by Huxley (1851 ), who century, Jean-Baptiste de Lamarck realized that included Appendicularia in the Tunicata pro- ascidians and thaliaceans could be grouped as posed by Lamarck. Tunicata (Lamarck 1816 ). However, based Only after Kowalevsky had described the strictly on the adult morphological features of tailed larval form of ascidians (Kowalewski solitary or colonial forms, these animals remained 1866), containing a dorsal neural tube, a typical diffi cult to place in broader evolutionary animal chordate notochord, and lateral muscle cells, did groups; for example, in the mid-nineteenth cen- zoologists begin to realize that tunicates should be

[email protected] 138 A. Stolfi and F.D. Brown

A

C

B

Fig. 4.2 Historic overview of the position of Tunicata twentieth century with the lophophorates (Hyman 1959 ; within the Metazoa. ( A ) View of animals relationships Zimmer and Larwood 1973; Nielsen 2002 ) basal to the through the nineteenth century (Milne-Edwards 1843 ; tunicates. ( C ) Current view using molecular phylogenies Haeckel 1866 ), highlighting the position of the Tunicata (Field et al. 1988 ; Cameron et al. 2000 ; Swalla et al. 2000 ; as a chordate due to the presence of the notochord (in red ). Winchell et al. 2002 ) ( B ) Popular view of animal relationships throughout the placed within the chordates. In the late nineteenth were placed at the base of Chordata. Because of century, zoologists such as Haeckel (1874 ) , the unique body plan of adult tunicates, it was Bateson (1884 , 1 8 8 6 ) , B r o o k s (1893 ) , W i l l e y even proposed that they be raised to phylum status ( 1894 ) , a n d G a r s t a n g (1894 ) supported Chordata (Kozloff 1990 ; C a m e r o n e t a l . 2000 ; Z e n g a n d and Hemichordata as sister phyla based on noto- Swalla 2005 ) . chord-like structures as the main synapomorphy. At the turn of the twentieth century, zoologists Within Chordata three subphyla were recognized, realized that chordates, hemichordates, and echi- Vertebrata, Cephalochordata, and Tunicata, which noderms share a common embryological trait: the Lankester (1877 ) attempted to rename as blastopore, the fi rst opening that forms in the gas- Urochordata, to highlight the evolutionary impor- trulating embryo, developed into the anus of the tance of this character within the phylum. Because adult. Thus, they were classifi ed within the super- cephalochordates shared more morphological phylum Deuterostomia (Grobben 1908 ) , o p p o s i t e features with the vertebrates, including a meta- to Protostomia, or animals whose blastopores meric muscle segmentation pattern, the tunicates developed into the mouth instead. A morphological

[email protected] 4 Tunicata 139 feature that generated confusion for almost the monophyly of Deuterostomia and Ambulacraria entire twentieth century but that was relevant to the (Echinodermata + Hemichordata) and also resulted general debate of early deuterostome evolution in new and previously unexplored hypothetical rela- was the lophophore of brachiopods, ectoprocts, tionships such as the placement of Xenoturbella and phoronids (see Vol. 2, Chapters 1 0 , 1 1 , and together with acoel worms as the sister group to the 1 2 ), a crown-like feeding apparatus composed of Ambulacraria (see Vol. 1, Chapter 9 for discussion) numerous tentacles. In spite of convincing argu- or the inversion of Tunicata and Cephalochordata ments for nearly half a century that supported a positions within Chordata (Delsuc et al. 2006 ; monophyletic Lophophorata (those animals bear- Telford and Copley 2011 ). The latter proposed rela- ing a lophophore) as being part of or closely tionship for Chordata suggests that Tunicata is the related to Deuterostomia (Fig. 4.2B ; H y m a n 1959 ; sister group to Craniata/Vertebrata within a group Zimmer and Larwood 1973 ; N i e l s e n 2002 ), named Olfactores (reviewed in Delsuc et al. 2006 ), molecular phylogenetic work contested their deu- to the exclusion of Cephalochordata (Fig. 4.2C ). terostome affi nities and instead grouped them Some of the implications of this new rearrangement together with annelids, mollusks, and other diverse for our understanding of chordate evolution will be phyla in Lophotrochozoa (Halanych 2004 ; discussed in the following. Helmkampf et al. 2008 ). Conserved gene sequences, such as large and Tunicate and Chordate Origins small subunit ribosomal rDNAs, were the fi rst For nearly 200 years, zoologists have wondered molecular characters used for testing phylogenies about the origins of our own phylum, and the lit- of deuterostomes and other phyla (Field et al. 1988 ; erature is rife with heated debates on the origins Cameron et al. 2000 ; S w a l l a e t a l . 2000 ; Winchell of chordates (Fig. 4.3 ). In this section the main et al. 2002 ). Additional protein-coding gene hypotheses that have been proposed and how sequences (both mitochondrial and nuclear), intron- more recent views of deuterostome relationships exon boundaries, miRNAs, and other empirical have affected our understanding of the chordate evidence rendered additional support for the ancestor are reviewed.

AB

C

Fig. 4.3 Chordate origins hypotheses. (A ) Garstang’s ancestor eventually transitioned to a free-living life his- hypothesis suggested that chordates evolved from larval tory and modifi ed the tentacles into gills. (C ) Wada’s pelagic forms; a scenario representing the evolutionary hypothesis was formulated according to the phylogenetic transition from a dipleurula-like ancestor that modifi ed understanding of chordate relationships by the turn of the the ciliary band into the nerve cord of the chordate larva is century; the basal position of the larvaceans among the shown. (B ) Romer hypothesized the evolution of chor- tunicates and the paraphyly of ascidians suggested an dates from a colonial pterobranch-like ancestor; the ancestor with both a pelagic larva and adult

[email protected] 140 A. Stolfi and F.D. Brown

Garstang Hypothesis (1928 ): Walter Garstang larities of the tornaria and auricularia larvae. proposed that chordates evolved through a series Ideas of recapitulation prevailed in the nine- of modifi cations to the ancestral larva, indepen- teenth century (Haeckel 1869 ) and set the stage dently from changes in the adult form (Fig. 4.3A ), for twentieth century ideas of heterochrony in and therefore retraced the evolutionary steps development as an important factor in the evolu- back in time starting from (1) a permanently free- tion of animals. swimming chordate larva to (2) a protochordate- Romer Hypothesis (1967 ): Alfred Romer was like larva (i.e., cephalochordate or ascidian), to a vertebrate paleontologist and comparative (3) a tornaria-like larva (i.e., hemichordate), to morphologist who came to address the question (4) an auricularia-like larva (i.e., echinoderm). of chordate evolution from a vertebrate point of From this ancestral reconstruction, it may be view (Fig. 4.3B ). In his hypothesis, two major extrapolated that the free-swimming larval and evolutionary novelties were necessary for the adult stages of Craniata/Vertebrata as well as evolution of the chordate body plan: (1) the evo- Cephalochordata evolved from an ascidian-like lution of pharyngeal slits from feeding tentacles ancestor that suppressed metamorphosis. The lat- (lophophores) and (2) the evolution of a free- ter, in turn, evolved from plankton-feeding ben- living stage in the life cycle of an ancestral thic organisms (similar to extant hemichordates s e s s i l e , fi lter-feeding organism. As Romer and echinoderms) with external ciliated tentacles explained in his departing American Association and food groves that contained pelagic larvae for the Advancement of Science (AAAS) presi- (tornaria and auricularia, respectively) (reviewed dential address published in Science ( 1967 ): in Berrill 1955 ; Brown et al. 2008 ). Garstang also “Instead of passively waiting for food to come observed similarities between the development of to it, the animal could go in search of food and the neural tube and endostyle (i.e., the thyroid could explore new areas or new habitats in gland precursor; see Metamorphosis section which it might exist,” thus leading to the evolu- below for more details) of chordate larvae and the tion of the cephalochordates and the vertebrates. formation of the circumoral and adoral ciliary A recurrent idea previously postulated by bands of echinoderm larvae. He assumed these Garstang was that of a pedomorphic event in an processes were homologous, which would sup- ancestral ascidian-like larva that evolved into a port his hypothesis. However, expression patterns free-living form in the cephalochordate and cra- of several key developmental genes in the echi- niate/vertebrate ancestor. noderm larva do not resemble the expected The Romer hypothesis of chordate evolution expression patterns of either invertebrate or ver- became popular in the twentieth century and pre- tebrate embryos and generally show echinoderm- vailed in many Zoology textbooks. Phylogenies at specifi c patterns. These results have caused some the time supported the pterobranchs and crinoids diffi culties in inferring homologies between at the base of Deuterostomia, with the lophopho- chordate and echinoderm larvae and have been rates as the sister group to the deuterostomes. used to support the view of a secondarily acquired However, phylogenetic data no longer support indirect-developing larva in the Ambulacraria this hypothesis (Fig. 4.3B ), and the sessile ptero- (Echinodermata + Hemichordata) (reviewed in branchs are likely derived from a free- living Swalla 2006 ; Brown et al. 2008 ). wormlike ancestor within the hemichordates Garstang was highly infl uenced at the time (Fig. 4.1 ; B r o w n e t a l . 2008 ) . by work and ideas previously developed by Wada Hypothesis (Wada 1998 ) : B a s e d o n p h y - Arthur Willey ( 1894 ). Willey had already noted logenetic understanding of Tunicata relationships that larval forms could recapitulate ancestral using 18S rDNA, Hiroshi Wada proposed a forms since the discovery of the chordate nature pelagic free-living ancestor of the chordates. The of the ascidian larva (Kowalewski 1866 ) and relationship generally accepted at the time was as had also proposed the close relationship of the follows: Cephalochordates were the sister taxon hemichordates and echinoderms due to the simi- to the Craniata/Vertebrata; larvaceans were at the

[email protected] 4 Tunicata 141 base of the tunicates; ascidians were shown to be termed the “notoneuron,” was discussed by paraphyletic (Fig. 4.3C ). Therefore, the most par- Nielsen (1995 ) in his Trochaea hypothesis. simonious reconstruction scenario of the chor- However, the evolutionary sequence that gave date ancestor was that of a pelagic noncolonial rise to the notoneuron assumed the existence of and indirect-developing free-swimming form ancestral holopelagic forms, which very much resembling the life cycles of lampreys, cephalo- resembles the recapitulation hypothesis of the chordates, and larvaceans that were placed at the nineteenth century (Fig. 4.3A ; Haeckel 1869 ; base of the Tunicata. Since then, phylogenetic reviewed in Garstang 1928 ; Collins and Valentine relationships of the chordates have gained from 2001 ). new analysis methods, additional gene sequences, In contrast to the previous hypotheses, the and new genomes and transcriptomes that have most complete and parsimonious scenario based substantially changed our view of tunicate rela- on current phylogenetic evidence (Fig. 4.1 ) is tionships. In particular, a new placement of presented here. Within the Chordata, three hypo- Tunicata as sister group of the Craniata/Vertebrata thetical ancestors are discussed: chordate, olfac- and genome-wide sequence analyses and devel- torean (tunicate + vertebrate), and tunicate. opmental studies in larvaceans have begun to The chordate and olfactorean ancestors may change our views of chordate ancestry. These be assumed as being very similar in development will be discussed next. and form: indirect-developing vermiform organ- The Wormlike or Vermiform Deuterostome isms with gill slits, dorsal neural tube, notochord, Ancestor Hypothesis : Although references to a and a pigmented sensory organ throughout all wormlike ancestor of the chordates date back to stages of their life cycle (Fig. 4.1 ). Whether the A. Willey (1894 ),3 later work by morphologists adult of the olfactorean ancestor was pelagic or focused mostly on macroevolutionary recon- benthic is diffi cult to predict with confi dence, but structions based solely on ancestral pelagic larval the chordate ancestor was very likely benthic, forms (Garstang 1928 ; Nielsen 1995 ). Only with a lifestyle resembling that of the cephalo- recently has this hypothesis been revisited, thanks chordate adult. This view can be supported as to recent developmental studies and a new under- cephalochordates are now placed basal to standing of deuterostome relationships. More Olfactores. The parsimonious conclusion is that specifi cally, gene expression patterns have the olfactorean ancestor was also benthic. emerged as new characters that can be evaluated The tunicate ancestor was likely an indirect as shared novelties (synapomorphies) among the developer that had a pelagic free-swimming larva various groups within the deuterostomes (Nielsen with a notochord, a dorsal neural tube, and asym- 1995 ; Cameron et al. 2000 ; Wicht and Lacalli metric pigmented sensory organs, traits that are 2005 ; Gudo and Syed 2008 ). found in most tunicate clades (Fig. 4.1 ). This The vermiform ancestor hypothesis postulates ancestral larva probably did not have any gills, as that chordates and deuterostomes evolved from a most larvae in extant tunicates do not develop or solitary, sexually reproducing, free-living worm- utilize their gills until after metamorphosis. Due like ancestor with pharyngeal slits, i.e., similar to to the unresolved phylogenetic positions of an enteropneust worm (Fig. 4.1 ; Cameron et al. Appendicularia, Thaliacea, and several species of 2000 ; Gerhart et al. 2005 ; Brown et al. 2008 ). phlebobranch and aplousobranch ascidians, there Initially, a benthic vermiform deuterostome are few confi dent predictions we can make on the ancestor with a pelagic dipleurula-like larva, adult form of the tunicate ancestor, other than to say it was likely a fi lter feeder. While the fi rst 3 A. Willey wrote in his treaty on “Amphioxus and the tunicate ancestor was likely very similar to its Ancestry of the Vertebrates” (1894 ): “The ultimate or pri- vermiform olfactorean forebear, we cannot rule mordial ancestor of the vertebrates would, on the contrary, out a scenario in which the transition to sessility be a wormlike animal whose organization was approxi- mately on a level with that of the bilateral ancestors of the occurred in the stem leading to the last common Echinoderms.” ancestor of all extant tunicates.

[email protected] 142 A. Stolfi and F.D. Brown

Historical Milestones of Tunicate thus held up in support of a recapitulatory view of Developmental Biology evolution. Although such views have been largely superseded, it is imperative to contextualize The chordate nature of ascidians was fi rst recog- them. At that time, these studies were among the nized by Russian zoologist Alexander fi rst to suggest that the development of an organ- Kowalevsky (1866 ). He generated detailed ism could reveal information about its evolution- descriptions of embryogenesis and the anatomy ary origins. A modern-day analogy would be the and general organization of the chordate larva in “hidden” evolutionary history that is stored in two species of ascidians, Phallusia mammillata morphogenesis as well as the genes and genomes and Ciona intestinalis . As mentioned above, his of organisms, which has revolutionized our original descriptions suggested that these animals understanding of phylogeny and evolution in the were closer to the vertebrates rather than to their last decades. previous classifi cation as mollusks. His discov- Near the end of the nineteenth century, the ery set the stage for renewed discussions of the French embryologist Laurent Chabry, working on centuries-old questions about the evolution of Ascidiella aspersa , p e r f o r m e d t h e fi rst experiments animal form (Raff and Love 2004 ). on animal embryos at the single-cell level, manipu- Kowalevsky’s careful anatomical descriptions lating individual blastomeres with the microma- of the ascidian embryo and larva served as one of nipulator he had invented and built himself the inspirations for an attempt to unravel the evo- (Fig. 4.4A ; C h a b r y 1887 ; reviewed in Fischer lutionary history of animals by analyzing the fea- 1992 ; Sander and Fischer 1992 ) . T h i s w a s a r a d i - tures that appeared transiently during particular cally new approach at the time and generated much developmental stages. The ascidian embryo was discussion in those days of wildly differing theories

ABC

Fig. 4.4 Classic works of ascidian embryology. ( A ) Top : or “organ-forming substance” (Conklin 1905a ), and the Laurent Chabry’s micromanipulator, of his own invention invariant cell lineages of the early embryo (Conklin and construction. Bottom row : some of the “monsters” 1905b ). (C ) Conklin’s illustrations of the embryos whose (damaged or naturally defective embryos) whose abnor- blastomeres he managed to damage through swirling or mal development he characterized. The illustrations are suction. His documentation of these half or quarter from his thesis (Chabry 1887 ). ( B ) Illustrations by embryos helped establish a role for mosaic, as opposed to Conklin of a Styela canopus embryo, showcasing two of regulative, mechanisms of cell fate specifi cation (Conklin his greatest discoveries: the posterior vegetal myoplasm, 1905c )

[email protected] 4 Tunicata 143 on the very essence of embryogenesis. Chabry’s “mosaic.” He and others also noted similar pat- work was eagerly followed up by other luminaries terns of embryogenesis and highly stereotypical of those early days of experimental embryology, cell numbers and lineages in embryos of other such as Roux, Driesch, Spemann, and Mangold. species of solitary ascidians. This was followed In the history of ascidian embryology, we by several decades of comparative works describ- skip some 20 years forward to describe the work ing the diversity and evolution of developmental of American embryologist Edwin G. Conklin, modes found in ascidians and tunicates, most who conducted both detailed descriptive and notably by Berrill (1930 , 1931 , 1932 , 1935a , b , ingenious experimental research on the ascidian 1936 , 1947a , b , c , 1948a , b , c , d , e , 1950a , b , embryo. At the Marine Biological Laboratories 1951 ). at Woods Hole, Conklin used a local species, A s t h e b r o a d e r fi eld of developmental genet- Styela canopus ,4 to ask basic questions about ics matured and incorporated the techniques and cell fates during embryonic development. This approaches of molecular biology and genomics, species was particularly suited for this line of ascidian embryogenesis quickly came to be inquiry because its eggs contained a naturally understood and interpreted in the light of gene occurring yellow-orange pigment in the cyto- expression and function. Gone were the sectar- plasm that became progressively restricted after ian squabbles of regulative vs. mosaic develop- each round of cell division until it was ultimately ment: conserved intercellular signaling pathways inherited by the larval muscles (Fig. 4.4B ) . were found to regulate the induction of the Conklin therefore hypothesized, correctly, that majority of cell fates in the ascidian embryo. A this cytoplasm contained an “organ-forming new picture emerged, in which mechanisms of substance” that was suffi cient and necessary for “mosaic” development (e.g., localized maternal larval muscle specifi cation and differentiation determinants) critically set the stage for the intri- (Conklin 1905a ) . cate and invariant unfolding of events that none- The cellular simplicity of the ascidian embryo theless require the constant instructions and also allowed Conklin to follow embryonic cells affi rmations of cell-cell communication, until all during development up to their fi nal fate. The cell progenitors are specifi ed and their derivatives lineages he described for Styela became the most fully differentiated. complete embryonic lineage mapping ever done Many of the fi rst discoveries in this new for any animal at the time (Conklin 1905b ). molecular age of ascidian biology were achieved Moreover, he could manipulate early embryos en by researchers working on Halocynthia roretzi , masse by spurting and shaking and assess the but Ciona intestinalis quickly rose to promi- results of injuring particular blastomeres in this nence5 when its genome was sequenced in 2002 way (Conklin 1905c ). This allowed him to dem- (Dehal et al. 2002 ). Researchers working on onstrate the highly deterministic mode of devel- ascidians were privileged to be among the fi rst to opment of ascidians, which contrasted abruptly lay eyes on the genomic underpinnings of animal with the highly regulative mode of development development, as C. intestinalis was only the sev- that was discovered in sea urchins around the enth animal to have its genome sequenced at the same time (Fig. 4.4C ). He called this highly time. The adaptation of an electroporation-based deterministic mode of development in ascidians high-throughput transgenesis protocol also boosted the popularity of C. intestinalis among 4 There appears to be some confusion as to which species developmental biologists (Corbo et al. 1997 ). was originally used in Conklin’s experiments. Although in his original studies the species was named as Cynthia ( Styela) partita , this name is no longer valid, having been 5 There are approximately 70 laboratories in the world cur- synonymized with Styela canopus . However, the ecologi- rently using Ciona as a model species; for further infor- cally dominant Styela species around Woods Hole nowa- mation on Ciona and other ascidian online resources, visit days is the invasive S. clava. http://www.aniseed.cnrs.fr/aniseed/ .

[email protected] 144 A. Stolfi and F.D. Brown

Tunicata: Experimental Models for EvoDevo number of cells as Ciona a n d o t h e r s m a l l e r Research solitary species. Many of the modern-day There has been a variety of work on many dif- classics in ascidian developmental biology ferent species representing all orders in the were performed on H. roretzi b y H i r o k i Tunicata, and it is with some injustice that the Nishida in Japan. However, its limited distri- four species below have been highlighted above bution has prevented its adoption as a model all the others, solely based on the relative organism by researchers outside East Asia. importance attached to them by a handful of Other obstacles to widespread use as a model bipedal terrestrial olfactoreans. However, they organism include its inaptitude for the elec- represent a good sampling of different traits, troporation protocol so useful for transient historical backgrounds, and other points of transfections in Ciona , and deadly outbreaks interest to the experimentalist. of the insidious soft tunic syndrome disease Ciona intestinalis ( Linnaeus , 1767 ), like its (caused by a species of kinetoplastid fl agel- close relative C. savignyi , is a rather unspe- late) among sea pineapples under intense cul- cialized solitary phlebobranch ascidian. A tivation. However, important insights into common invasive species in temperate waters developmental biology continue to be regu- around the globe, it is easily obtainable from larly generated by work on this venerable the wild. In fact, this was the major reason C. model, and the community eagerly awaits the intestinalis gained traction with developmen- results of the ongoing efforts to sequence and tal biologists (culminating in the sequencing annotate its genome and that of its North of its genome in 2002): researchers in Japan, American close relative, the sea peach North America, and Northern and Southern Halocynthia aurantium . Europe could work on the same species… or Botryllus schlosseri ( Pallas , 1766 ) is the so they thought! It is now understood that the leading model colonial species. Research same name has been used for two closely in Botryllus schlosseri has focused mainly related but genetically distinct and reproduc- in the study of budding, i.e., blastogenesis, tively isolated species, currently referred to as and in studies of self- and nonself recogni- “Type A” and “Type B.” The Ciona commu- tion of adult colonies. Also known as the nity awaits offi cial publication of the bomb- star tunicate due to the star-shaped arrange- shell revelation (Lucia Manni, personal ment of individual zooids within the colony, communication) that the specimen originally B. schlosseri develops synchronously and described by Linnaeus, and thus the specifi c undergoes weekly developmental cycles that epithet intestinalis , most likely belonged to generate new zooids at the same time as old the “Type B” species of Northern Europe. zooids regress. A relatively fast turnover of This would mean that the “Type A,” studied zooids and simplicity in the maintenance of by the majority of labs around the world, actu- colonies in the laboratory have made this ally corresponds to Ciona robusta , currently a species a suitable model organism for stud- junior synonym that nonetheless describes the ies on stem cells, differentiation, apoptosis, type specimen for “Type A.” and sexual vs. asexual development. Stages Halocynthia roretzi ( Drasche , 1884 ) , a l s o of blastogenic development were described known as the sea pineapple, is a prized culi- by Sabbadin as well as Watanabe, and its nary delicacy in parts of Japan and Korea. A complete genome was fi nally published in large solitary stolidobranch ascidian, its eggs 2013. Upon contact of two adult colonies, and larvae are roughly twice as large as those an allorecognition reaction occurs, which of Ciona spp., but develop according to the results in either fusion of the colonies form- same cell lineages, with roughly the same ing a larger chimeric colony (parabiont) or

[email protected] 4 Tunicata 145

rejection of the two colonies, revealing a strains, enabling classical genetic analyses. histocompatibility locus with important evo- Furthermore, the female germ line and ovary lutionary implications for the evolution of develop as a single, multinucleated syncytium adaptive immunity of vertebrates. called a coenocyst, allowing for targeted Oikopleura dioica (Fol , 1872 ) is an emerg- genetic manipulations in ovo by injecting the ing model organism for larvacean biology. coenocyst. Larvaceans hold many of the keys Their small size and rapid generation cycle to understanding the evolution of tunicates mean they have to be cultured in suspensions and chordates, and work on O. dioica will fi g- of large numbers of individuals. Nonetheless, ure heavily into understanding the biology of this allows for inland culturing of different these amazing creatures.

Soon thereafter the genome of the related Ciona Thanks to the Ciona intestinalis sequencing savignyi was sequenced (Vinson et al. 2005 ; Small efforts, a large amount of gene expression data et al. 2007 ). This was followed by the genome has been generated for this species, including sequence of the larvacean Oikopleura dioica spatiotemporal expression pattern profi les for the (Denoeud et al. 2010 ) and the colonial stolido- vast majority of the approx. 700 transcriptional branch ascidian Botryllus schlosseri (Voskoboynik regulators (i.e., sequence-specifi c transcription et al. 2013 ), which was already the established factors and major cell signaling molecules) dur- colonial tunicate model for studies on blastogene- ing embryonic development (Satou et al. 2001 , sis and histocompatibility. Other tunicate genomes 2002 , 2003 ; Kusakabe et al. 2002 ; Mochizuki currently in the sequencing pipeline include those et al. 2003 ; Imai et al. 2004 ). Concurrent to the of the phlebobranch ascidians Phallusia mammil- genomic revolution, the adaptation and refi ne- lata , Phallusia fumigata ( T a s s y e t a l . 2010 ; Roure ment of techniques for molecular manipulation et al. 2014 ) , a n d Didemnum vexillum (Abbott et al. of ascidian embryos (Corbo et al. 1997 ; reviewed 2011 ; Stefaniak et al. 2012 ), as well as the stolido- in Stolfi and Christiaen 2012 ; Treen et al. 2014 ) branch ascidians Halocynthia roretzi , Halocynthia have allowed researchers to characterize the aurantium ( T a s s y e t a l . 2010 ) , Molgula occulta , functions of many developmentally important Molgula oculata , a n d Molgula occidentalis ( S t o l fi genes in ascidians in quite some detail. The ease et al. 2014 ). of manipulation of great numbers of synchro- Tunicate genomes are rapidly evolving nized embryos allows for large-scale analyses of (Denoeud et al. 2010 ; Tsagkogeorga et al. 2012 ) gene function, a prime example of which was the and quite compact relative to the genomes of construction of a provisional “blueprint” of regu- their sister taxa, the vertebrates. The Ciona latory networks at single-cell resolution for every genomes are approx. 150–170 Mb in length, cell of the C. intestinalis early embryo and larval while the Oikopleura dioica g e n o m e i s t h e nervous system (Fig. 4.5 ; Imai et al. 2006 , 2009 ). smallest chordate genome sequenced to date, at 70 Mb. The Botryllus schlosseri g e n o m e i s t h e largest tunicate genome sequenced thus far, at an EARLY DEVELOPMENT estimated 580 Mb. The sizes of the tunicate genomes currently being sequenced appear to Tunicate development is marked by the contrast fall somewhere in the range of those already between the simple and the complex and between sequenced. Protein-coding genes are estimated the conservative and the innovative. We use the to number between 15.000 in Ciona a n d ascidian embryo as a starting point for under- 27.000 in Botryllus . standing the development of tunicates as a group,

[email protected] 146 A. Stolfi and F.D. Brown

A

B

Fig. 4.5 Example of cell-specifi c gene regulatory net- specifi c circuit diagrams of regulatory connections oper- works in Ciona intestinalis . ( A ) Circuit diagram of overall ating in selected cells of the posterior sensory vesicle regulatory connections between a subset of regulatory (A11.64), neck (A11.62), motor ganglion (A11.118), and genes (transcription factors and intercellular signaling nerve cord (A11.116) (Figures adapted with permission of pathway components) expressed in the developing central the authors from Imai et al. (2009 )) nervous system of the Ciona intestinalis larva. (B ) Cell- as phylogenetic and embryological evidence sug- the early development of ascidians, which have gest all extant tunicates descend from an ascidian- been the intense focus of over 100 years’ worth like common ancestor. Much is known about of developmental genetic studies. The vast

[email protected] 4 Tunicata 147

knowledge on the subject generated over the in Lemaire 2009 , 2011 ). In all species, the years can only be done justice by an extended bilaterally symmetric embryos develop in an format review, as has been elegantly done most invariant manner, and cell division rates are recently by Satoh (2013 ). Here, we attempt to tightly regulated. Unlike certain species of summarize a summary of what is known. nematodes in which apoptosis, or programmed The ascidian life cycle may be separated into cell death, plays a large role in development by three or four important developmental periods: culling unused cells, there is no evidence that (1) embryogenesis (i.e., the formation of the apoptosis is involved in normal embryogenesis larva), (2) larval development (i.e., settlement in ascidians. All the cells that are born during and metamorphosis), (3) growth and maturation the development of a typical ascidian embryo (i.e., transition from juvenile to sexually mature seem to be accounted for upon hatching (Hotta adult), and (4), only for colonial species, blasto- et al. 2007 ). genesis (i.e., development of asexually propa- Quite remarkably, embryogenesis of the vari- gated clonal generations). Considerable ous solitary ascidians is perfectly conserved at interspecifi c variation is found in egg/embryo least up until the neurula stage (and probably size, reproductive strategies, and adult zooidal well beyond that) as far as cell lineages, divi- size (Fig. 4.6 ). The relative timing of the different sions, and arrangements are concerned. The stages of the ascidian life cycle also varies sub- embryos may differ in size, developmental rate, stantially from species to species, which high- and even gene expression, but there is a 1:1 cor- lights the importance of heterochrony and respondence between each and every cell in modularity in the evolution of developmental Ciona intestinalis and Molgula occidentalis g a s - processes. trula embryos, in spite of the estimated 500 mil- lion years of evolutionary divergence between the two species (Fig. 4.7 ). The Ascidian Embryo That species on opposite branches of the tuni- cate phylogeny have the same exact cell lineages Although ascidian embryos are greatly reduced would seem to imply that genetically they must be in cell number and highly derived relative to ver- very conservative as well and that the cis- tebrates in strategies for cell fate specifi cation, regulatory elements (or “enhancers”) orchestrat- they are profoundly steeped in their chordate ing the gene regulatory networks responsible for pedigree, as evidenced by highly conserved embryonic development must be highly con- chordate- specifi c embryonic fate maps and gene served. In fact, the exact opposite is true. There is regulatory networks. Within the ascidians, one virtually no conservation between Ciona intesti- fi nds great variation in the mode of development: nalis a n d Halocynthia roretzi n o n c o d i n g viviparity, asexual reproduction, direct develop- sequences (Oda-Ishii et al. 2005 ) n o r b e t w e e n ment, and adultation, i.e., the precocious differ- certain congeneric species, like Molgula occiden- entiation of adult structures prior to hatching of talis a n d Molgula oculata ( S t o l fi et al. 2014 ) . T h i s the swimming larva. However, even distantly is consistent with the observations that the related ascidians share the same streamlined, genomes of ascidians, and tunicates as a whole, invariant embryonic cell lineages. Therefore, are rapidly evolving (Seo et al. 2001 ; T s a g k o g e o r g a phylogenetic evidence suggests that the simple, et al. 2012 ). In some cases, a lack of sequence solitary ascidian embryo likely represents the conservation does not necessarily imply divergent ancestral condition and that all the other strik- regulatory mechanisms, as some enhancers are ingly different modes are variations on this basic fully compatible in cross-species transgenic theme. assays (Johnson et al. 2004 ; Oda-Ishii et al. 2005 ; Embryonic development has been character- Roure et al. 2014 ). However, there is mounting ized in great detail in a number of indirectly evidence for great variation in regulatory strate- developing, solitary ascidian species (reviewed gies controlling identical developmental processes

[email protected] 148 A. Stolfi and F.D. Brown

AB

CD

EF

Fig. 4.6 Two styelid ascidians with different life histo- light and gravity sensing, i.e., photolith (ph ) , a n d a n t e r i o r ries: solitary Boltenia villosa ( A , B ) a n d c o l o n i a l protrusions known as ampullae (am ) . (D ) A n t e r i o r m a g n i - Botrylloides violaceus (C –F ) . (A ) B. villosa o r a n g e s o l i - fi ed view of B. violaceus larval head, which shows the tary larva shows chordate features including the notochord prominent ampullae and some pigmented cells before set- (no ) and the dorsal nerve cord (nc ); these larvae present tlement of the larva. (E ) The oozoid with open oral (os ) light and gravity sensing organs, i.e., ocellus (oc ) a n d o t o - and atrial (as ) siphons after settlement shows lateral buds lith (ot ). Palps or adhesive papillae can be seen in the ante- that will develop into the next asexual generation of blasto- rior (left). (B ) A n a d u l t Boltenia villosa “hitchin’ a ride” on zooids; note extended ampullae around the edges of the a Chlamys scallop at Friday Harbor Laboratories. ( C ) B. newly settled colony. ( F ) M a g n i fi cation of a B. violaceus violaceus larva also shows the characteristic chordate fea- colony shows the oral openings of each zooid and the com- tures in the tail, i.e., notochord ( no ) a n d n e r v e c o r d (nc ) , a s mon cloacal aperture ( c ) that allows internal visualization well as a single sensory organ in the head that serves for of branchial sacs (Courtesy of J. Greer)

[email protected] 4 Tunicata 149

Fig. 4.7 Comparison of Molgula occidentalis (Stolido- ans that share near-identical cell lineages and fate maps, branchia) and Ciona intestinalis (Phlebobranchia) in spite of negligible conservation of noncoding genomic embryogenesis. Comparative diagram of embryonic sequences. Embryos were stained with DAPI and fl uores- stages of Molgula occidentalis (left ) and Ciona intestina- cent phalloidin conjugates and are all at roughly the same lis ( right ), two distantly related species of solitary ascidi- scale (Images were adapted from Stolfi et al. (2014 )) in distantly related ascidians6 ( T a k a h a s h i e t a l . “typical” solitary ascidian embryo that will serve 1999 ; H u d s o n e t a l . 2007 , 2013 ; T o k u o k a e t a l . heavily as a reference for the general description 2007 ; T a k a t o r i e t a l . 2010 ) . of embryogenesis in this chapter, as it has been This remarkably minimalist embryo must the workhorse of most genetic and molecular have evolved at a bottleneck – the last common studies in the tunicates. ancestor of all extant ascidians and possibly all extant tunicates. However, the incredible deriva- tions seen in the more unusual ascidian embryos The Typical Ascidian Larva demonstrate that, far from being an evolutionary dead end, the simplifi ed ascidian embryo has Before describing the embryonic development of been elaborated and reconfi gured many times ascidians, it is crucial to explain its end product: over. Later we will discuss the exceptional exam- the typical solitary ascidian larva. The chordate ples of derivation from this typical embryo and nature of ascidians is most obvious during the their impact on our understanding of the evolu- larval stage (Figs. 4.6 , 4.7 , and 4.8 ). The ascidian tion of tunicate development. For now, it is this larva is roughly divided into two parts, frequently but misleadingly termed the “trunk” and the 6 Both examples given above are referred to as develop- “tail.” The former is mostly composed of the mental system drift or DSD (see Vol. 1, Chapter 1 ) This anterior central nervous system (CNS), periph- refers to the divergence in molecular mechanisms govern- eral nervous system (PNS), and undifferentiated ing the development of identical homologous characters mesoderm and endoderm. From this mesodermal in different species (True and Haag 2001). As outlined above, this can refer to the divergence in primary nucleo- component are derived the adult branchiomeric tide sequence of orthologous enhancers that are function- muscles, heart, blood, tunic cells, and other cell ally interchangeable in cross-species transgenic assays but types. The endoderm gives rise to the pharyngeal also to the functional divergence of orthologous enhancers gill slits (later, the branchial basket), the endo- that do not drive identical gene expression patterns in their respective species of origin but are not interchangeable in style, and the anterior digestive tract. Given the cross-species assays. homology of these tissues and progenitor types to

[email protected] 150 A. Stolfi and F.D. Brown

ABC D

Symplegma brakenhielmi H E F G

K L I J

M N O

P

Q R Eudistoma sp. S

T U

Aplidium sp. V WX Y

Fig. 4.8 Diversity of colonial forms and budding modes the colony reveals the cloacal canals (P ). ( Q – S ) Eudistoma of Ecuadorean ascidians. ( A – D ) Symplegma brakenhielmi sp. larva ( Q ) and zooids ( R ) bud by abdominal strobila- larva (A ), zooid (B ), and live colony (C ) buds mainly by tion ( S ). ( T , U ) Aplidium sp . larva (T ) and zooids (U ) bud vascular budding (D ). ( E – H ) Polyandrocarpa zorritensis by strobilation ( S ). (V – Y ) Clavelina oblonga larva (V ), larva (E ), zooid (F ), and live colony (G ) buds mainly by zooid ( W ), and colony (X ) bud by stolonial budding from stolonial budding (H ). ( I – L ) Cystodytes dellechiajei larva the abdominal region ( Y ). ( D , H , L , S , Y ) represent only ( I ), brooding zooids (J ), and live colony (K ) buds mainly symbolic/generic drawings of the distinct budding modes by pyloric budding ( L ). ( M – P ) Trididemnum sp. pyloric ( capital and bold letters ) in ascidians, but are not meant to bud during the formation of a new abdomen (M ), lateral illustrate the specifi c modes of budding observed for the view of zooids and buds (N ); dorsal view of the colony species shown (Modifi ed from Brown and Swalla (2012 ), shows branchial regions of the zooids in blue and a clear and photo courtesy from G. Agurto (CENAIM-Ecuador)) common protruding cloaca ( O ), whereas a ventral view of

[email protected] 4 Tunicata 151

A

B

Fig. 4.9 Comparison of ascidian and vertebrate embryo yellow , endoderm + yolk sac; red, trunk (notochord + par- anatomies. ( A ) Ciona intestinalis mid-tailbud stage axial muscles); green , postanal tail. Comparison between embryo, with major tissues/territories color coded: blue , the two suggests the “tail” of ascidian larvae is homolo- sensory vesicle + brain; yellow , endoderm; red , axial + gous to the trunk + tail of vertebrates, and the ascidian paraxial mesoderm (notochord and muscle, respectively); larval “head” (formerly known as the “trunk”) contains green , tailbud (© Alberto Stolfi , 2015. All Rights tissues and precursor cells homologous to those located in Reserved). ( B ) Zebrafi sh embryo with tissues color-coded the vertebrate head (brain, endostyle/thyroid, pharyngeal according to their homologs in Ciona : blue , forebrain; endoderm). Zebrafi sh image is from Okuda et al. (2010 ) their counterparts in vertebrates, mostly located but rather to the trunk (cervical + thoracic in the head or just posterior to it, it would be more regions) of vertebrates, with perhaps a vestigial accurate to refer to this region as a “head.” postanal tail at its very posterior end (Fig. 4.9 ). Therefore, this term is preferentially used herein. Although it would be more appropriate to refer to The posterior half of the larva is composed this structure as a “trunk-tail,” it is here called mostly of caudal CNS and PNS, notochord (axial “tail” for simplicity. mesoderm), larval muscles (paraxial mesoderm), Thus, although the larvae of ascidians and and a strand of undifferentiated endoderm that amphibians indeed share a basic chordate body extends posteriorly almost to the very end of the plan, as defi ned by the unequivocal presence of larva. The presence of this gut rudiment, which chordate-specifi c body structures such as the dor- gives rise to the adult intestine (Nakazawa et al. sal hollow neural tube and a notochord, their 2013 ), suggests that this “tail” is not entirely “tadpole” characteristics may only be superfi - equivalent to the defi ning chordate postanal tail, cially similar.

[email protected] 152 A. Stolfi and F.D. Brown

A

B C

Fig. 4.10 Anural development in the Molgulidae. (A ) dele species (© Alberto Stolfi , 2015. All Rights Reserved). Illustrations by Berrill ( 1931 ) of the development, hatch- ( C ) Confocal image of a hatched larva of the urodele ing, and metamorphosis of Molgula arenata , an indirect- (tailed) species Molgula oculata. M. occulta and M. ocu- developing tailless (anural ) species. Amp. epidermal lata are closely related and sympatric off the coast around ampulla. ( B ) Confocal microscopy image of a hatching Roscoff, in northern Brittany (France). Embryos were larva of the anural species Molgula occulta . Note the pres- stained with a fl uorescent phalloidin conjugate. Anterior ence of notochord cells that have failed to proliferate and is to the right in ( B , C ). Scale bars = 50 μm (© Alberto intercalate like their counterparts in the embryos of uro- Stolfi , 2015. All Rights Reserved)

Anural and Direct Development by Jeffery and Swalla (1992 ). The best-studied Within the ascidians, there are some curious case is the loss of the familiar tadpole form by departures from the typical tadpole-shaped larva. larvae of diverse stolidobranchs, notably in the The incredible diversity of developmental modes Molgulidae and Styelidae families (Fig. 4.10 ). In was studied in detail by Berrill (1928 , 1930 , many of these cases, the embryo develops like a 1931 , 1935a , 1936 ) and more recently reviewed typical solitary ascidian embryo, but morphogen-

[email protected] 4 Tunicata 153 esis of the tail structures – the notochord, larval Table 4.1 Berrill’s survey of development in the muscles, and nerve chord – is severely impaired Molgulidae (Fig. 4.10A ). These embryos are said to be Depth found (in “anural” or “tailless,” which are misnomers if Species fathoms) one considers that these lost or degenerate struc- Oviparous, urodele Attached tures correspond mostly to the axial and paraxial Molgula tubifera 0–3 mesoderm of the vertebrate thoracic or cervical Molgula manhattensis 0–3 segments. However, in this chapter we will con- Molgula simplex 20 tinue to use these terms and their opposites, “uro- Molgula socialis 20 dele” or “tailed,” for lack of better words. Unattached Anural development may occur due to Molgula oculata 40 improper cell fate specifi cation, arrested differen- Oviparous, anural tiation, impaired morphogenetic movements, Attached acceleration of apoptosis, or any combination of Molgula retortiformis 2–50 these. In the anural species Molgula occulta , Unattached there are fewer notochord cells (Swalla and Molgula occulta 0–70 Jeffery 1990 ), which remain relatively amor- Molgula solenata 0–5 phous and do not extend a functional notochord Molgula provisionalis 15 through convergence and intercalation Molgula arenata 8 (Fig. 4.10B, C ; Berrill 1931 ; Swalla and Jeffery Bostrichobranchus pilularis 8 1990 ). Furthermore, the larval muscles of M. Eugyra arenosa 5–20 occulta fail to differentiate, in part due to recent Ovoviviparous, urodele Attached loss of certain larval muscle-specifi c “structural” Molgula complanata 2–10 genes such as muscle actin (Kusakabe et al. Molgula cooperi 0–1 1996 ). The anural condition has independently Molgula verrucifera 0–5 evolved in different clades (Berrill 1931 ; Hadfi eld Molgula citrina 2–5 et al. 1995 ; Huber et al. 2000 ), although the Molgula platei – developmental causes underlying the condition Ovoviviparous, anural in these other species have not been extensively Attached studied. Vestigial acetylcholinesterase activity in Molgula bleizi 0–1 the muscles of some larvae but not others (Bates Berrill ( 1931 ) lists a detailed account of the development and Mallett 1991 ; Swalla and Jeffery 1992 ; Bates (urodele, anural), reproductive (oviparous, ovovivipa- 1995 ; Tagawa et al. 1997 ) and distinct inactivat- rous), or ecological (range of depth of occurrence) strate- ing mutations in larval muscle actin pseudogenes gies of each species in the Molgulidae family. Adapted from Berrill (1931 ) in different anural species (Jeffery et al. 1999 ) suggest that the tailless conditions of different anural species are not identical. the need for hatching prior to settlement. The Berrill associated the anural condition with extreme tidal changes of high latitudes have also mud- or sand-fl at habitat in which larval motility been proposed as a factor that could make swim- may be dispensible (offering no adaptive value) ming at best redundant as a dispersal mechanism and, as a result, easily lost (Berrill 1931 ). (Huber et al. 2000 ). It is likely that a combination However, the fi nding of anural Molgula species of such biogeographical and ecological factors that preferentially attach to rocky substrates may have been permissive for the multiple exam- along rough coastlines challenged this view ples of evolution of tailless larvae in the (Table 4.1 ; Young et al. 1988 ; Bates and Mallett Molgulidae. Whether the molgulids are geneti- 1991 ; Nishikawa 1991 ; Bates 1995 ). In these cally predisposed to generate anural larvae is a cases, there is even perhaps positive selection for tantalizing hypothesis that remains to be direct development from adhesive eggs, without answered.

[email protected] 154 A. Stolfi and F.D. Brown

A closely related condition found in some the cER and mitochondria toward a presumptive ascidians is direct development (Young et al. posterior pole, giving rise to a posterior/vegetal 1988 ; Bates and Mallett 1991 ; Swalla and Jeffery cytoplasmic complex (Sardet et al. 1989 , 2007; 1992 ; Tagawa et al. 1997 ). So similar are the two Roegiers et al. 1999 ). Thus, the anterior-posterior conditions that in different studies the distinction (AP) axis is established roughly perpendicular to is made using different criteria and often the the AV axis. terms made interchangeable. In this review we distinguish direct development from anural Early Cleavages development only by the fact that direct develop- The fi rst cleavage occurs along the AP axis, ers will not hatch or break through the chorion equally partitioning posterior/vegetal cytoplasm prior to the extension of the ampullae. In direct into bilaterally symmetric left and right blasto- developers, it is these epidermal protrusions that meres. The second cleavage occurs perpendicular penetrate the chorion and allow the juvenile to to the fi rst, and the third cleavage occurs perpen- “hatch.” In many of these species, the eggs them- dicular to the second, resulting in a bilaterally selves are adhesive, allowing for settlement prior symmetric eight-cell embryo in which each half to hatching (Young et al. 1988 ). In both anural is partitioned into four major quadrants: anterior and direct development, individuals are presumed animal (“a-line,” pronounced “small a-line”), to follow the normal course of embryogenesis, posterior animal (“b-line” or “small b-line”), with specifi cation of vestigial larval structures anterior vegetal (“A-line” or “big/large A-line”), that may or may not differentiate. Other criteria and posterior vegetal (“B-line” or “big/large for distinguishing direct vs. indirect development B-line”) (Fig. 4.11 ). Starting from these eight of anural species (e.g., differentiation of larval major octants (two halves divided into four quad- muscles) are diffi cult to assay and may be rants each), the cell lineages of the embryo can be ambiguous depending on the assay. Further traced throughout development using Conklin’s detailed characterization of the embryogenesis supremely elegant nomenclature system that and morphogenesis in the Molgulidae will be allows for easy inference of the mitotic history of needed to classify the potentially multiple grada- any given cell7 (Conklin 1905b ). tions of anural and indirect development.

7 Conklin’s nomenclature system hinges on a tripartite name for each and every cell. This name is composed of a From Fertilization to Gastrulation letter and two integers. Let us consider an example, the B 7.5 cell. The letter indicates it is derived from the poste- Fertilization and Embryonic Axis rior vegetal (“B”) blastomere of the right half (indicated Determination by underlining) of the eight-cell embryo. The fi rst number (“7”) indicates the mitotic generation to which the cell During oogenesis, the egg is polarized along an belongs (the seventh generation, defi ning the fi rst genera- animal-vegetal (AV) axis. Unfertilized eggs are tion as the single-cell zygote). The second number is the arrested in metaphase of meiosis I, with the mei- cell’s unique identifi er, calculated by simple arithmetic otic spindle localized at the animal pole, where from the identifi er of its mother cell, B 6.3. To derive the identifi ers of its daughter cells, the mother cell’s identifi er the polar bodies will form following fertilization. (“3”) is fi rst multiplied by two (3 × 2 = 6). The daughter Endoplasmic reticulum-rich cortex (cER), asso- cell closest to the original sperm-entry point receives this ciated maternal “postplasmic” mRNAs, and even number (“6”) as its identifi er ( B 7.6). The other mitochondria are largely excluded from the ani- daughter cell is then given this number minus 1 (6 − 1 = 5). Hence, its name is B 7.5. There are some limitations to mal pole in unfertilized eggs, but only after sperm applying this nomenclature to later development. For entry the trigger of a calcium wave leads to acto- instance, sperm-entry point becomes an untenable land- myosin contraction and concentration of these mark as cells rearrange and alter their positioning relative maternal determinants at the vegetal pole (Prodon to one another. Furthermore, all lineages eventually cease to develop in a stereotypic manner. Therefore, it is not et al. 2008 ). A second phase of ooplasmic segre- advisable to insist upon Conklin’s nomenclature beyond a gation occurs after meiosis is complete, shunting certain time point in some lineages.

[email protected] 4 Tunicata 155

Fig. 4.11 The major cell lineages of the ascidian embryo. with each blastomere labeled with its unique identifi er and Top left : diagram of an ascidian one-cell stage embryo color-coded according to the quadrants of the one-cell (zygote) viewed laterally, showing the major embryonic stage embryo. These are the founder cells of the major axes and the polar bodies at the animal pole. Bottom left : lineages of the ascidian embryo: a-line (anterior/animal), division of this zygote into the major quadrants of the b-line (posterior/animal), A-line (anterior/vegetal), and embryo, as defi ned by the major axes, animal-vegetal and B-line (posterior/vegetal). The cells on the right hemi- anterior- posterior. Top right : lateral view of an eight-cell sphere have the same identifi ers, which are underlined to stage embryo, showing left and right hemispheres of the indicate their origin from the right hemisphere (e.g., a4.2 , bilaterally symmetric embryo. Bottom right : lateral view b4.2 , A4.2 , B4.2 ) (© Alberto Stolfi , 2015. All Rights of only the left hemisphere of the embryo shown at top, Reserved)

Each quadrant gives rise to a multitude of cell ric cleavage patterns of the early embryos. The lineages that develop according to distinct, but molecular nature of several such determinants invariant, cleavage patterns and orientations. The was soon to be revealed, heralded initially by the only symmetry is bilateral, with right-side and discovery of Macho1 , the maternal gene at the left-side lineages being identical until neurula- top of the regulatory cascade for the autonomous tion is complete. The observations and classic specifi cation of primary larval muscles (Nishida experiments of Conklin ( 1905a , c ), Whittaker and Sawada 2001 ). ( 1973 , 1977 , 1982 ), and Nishida (1992 , 1993 , Additional proteins were subsequently identi- 1994a , b , 1996 ) indicated the presence of mater- fi ed whose mRNAs are localized to the cER/ nal determinants that patterned the ascidian centrosome-attracting body (CAB), comprising embryo not only molecularly, through the speci- the postplasmic/PEM (posterior end mark) genes fi cation of the various cell lineages and their gen- (Satou and Satoh 1997 ; Satou 1999 ; Sardet et al. eration of the various differentiated larval tissues, 2003 ). These encode a wide variety of proteins, but also physically, by controlling the asymmet- most of which have not been extensively studied

[email protected] 156 A. Stolfi and F.D. Brown in ascidians. One exception is the original PEM includes the A-line neural/notochord progenitors. protein, Pem1, a fascinating molecule that exe- The vegetal mesendoderm in turn is largely fated cutes multiple functions critical for early devel- to give rise to endoderm, except for A6.3, which opment in multiple ascidian species. First, Pem1 will also give rise to the A7.6 mesenchymal is involved in the control of unequal cell cleav- lineage. ages through asymmetric mitotic spindle posi- In Ciona intestinalis , binary fate choice tioning (Negishi et al. 2007 ). Remarkably, Pem1 between vegetal mesendoderm and marginal is also a transcriptional regulator, globally zone involves sustained nuclear Beta-catenin repressing transcription in early posterior blasto- accumulation. While nuclear accumulation of meres and in the germ line (Kumano et al. 2011 ). maternally deposited Beta-catenin drives specifi - While the fi rst function requires Pem1 localiza- cation of the entire vegetal half of the embryo tion to the CAB (independently of Pem1 mRNA between the eight- and 16-cell stages, Beta- localization to the same structure), the second catenin nuclear localization is rapidly downregu- depends on Pem1 nuclear localization. These dis- lated in the marginal zone, being restricted to tinct functions also depend on different domains their more vegetal sister cells. This differential within Pem1. Furthermore, Pem1 mRNA local- Beta-catenin signaling is suffi cient and necessary ization to the CAB is another important property, to distinguish the two layers (Hudson et al. 2013 ). as it ensures an AP “gradient” of cell size and Surprisingly, this mechanism is not conserved transcriptional activity, with more posterior cells in Halocynthia roretzi , in which marginal zone being progressively smaller and transcriptionally cells also show sustained nuclear Beta-catenin silenced for longer, due to this asymmetric inher- staining. Instead, a peculiar mechanism for asym- itance of Pem1 mRNA at each round of cell divi- metric inheritance of mRNA of the marginal sion. The stepwise release from transcriptional zone determinant Not seems to drive mesendo- silencing constitutes a “timing mechanism” that derm patterning. This mechanism involves is critical for the patterning of the early ascidian nuclear migration, retention of locally transcribed embryo along the AP axis (Kumano and Nishida Not mRNA in a Wnt5-dependent manner, and 2009 ). later repositioning of the nucleus at mitosis, resulting in one daughter cell inheriting all the Germ Layer Specifi cation Not mRNAs, which are suffi cient and necessary At the vegetal pole, nuclear accumulation of for marginal zone specifi cation (Takatori et al. maternally provided Catenin Beta-1 (Beta- 2010 ). This is a remarkable example of develop- catenin) helps these cells maintain mesendoderm mental system drift between Ciona and potential and suppresses their specifi cation as Halocynthia . ectoderm (Imai et al. 2000 ), although the posi- tional cue for the accumulation itself is not Gastrulation known. In Ciona intestinalis , s uppression of By the 110-cell stage, the germ layers have segre- ectoderm fates involves antagonism of maternal gated, the fate map has been established, and gas- Gata4/5/6 function, which is required for “naïve” trulation commences. The endoderm cells drive ectoderm specifi cation and later for neural induc- gastrulation by invagination, which occurs in a tion (Rothbächer et al. 2007 ). two-step process in which the cells undergo fi rst The vegetal half of the embryo at the 16-cell apical constriction then basolateral shortening stage, having repressed an ectodermal fate, is (Sherrard et al. 2010 ). This is immediately fol- then further divided into vegetal mesendoderm lowed by the involution of the cells of the mar- and a “marginal zone” by the 32-cell stage. The ginal zone (mesoderm). Since this marginal zone marginal zone refers to a ring of cells surround- is only one to two cell diameters wide, its ingres- ing the mesendoderm proper and abutting the sion is swift and nearly concurrent with epiboly animal pole (ectoderm). This marginal zone does of ectodermal cells. Although this epiboly was not correspond to a specifi c germ layer, as it hypothesized to push the marginal zone cells as

[email protected] 4 Tunicata 157 they ingressed, experiments revealed that the restricted Fgf9/16/20 ligand. This induction is ectoderm cells were not required for proper gas- dependent on direct contact of inducing and trulation (Sherrard et al. 2010 ). Once ingressed, induced cells, as is the case for almost every other mesodermal cells do not immediately move great documented FGF-dependent induction event in distances, the most dramatic being the inversion the C. intestinalis embryo. of the larval muscle precursors along the AP axis. Induction of CNS from A-line progenitors Subsequent neurulation and epiboly of the epi- also involves FGF signaling. Surprisingly, in this dermis effectively seal in the endoderm, meso- lineage FGF signaling inhibits CNS specifi ca- derm, and neural tube. tion. A-line neural precursors are specifi ed by the Less is known about how the different germ absence of FGF signaling, while their sister cells layers and tissues remain physically compart- are induced by FGF signaling to become noto- mentalized, though evidence from other organ- chord precursors (Minokawa et al. 2001 ; Yasuo isms suggests that differential contractility and and Hudson 2007 ). The availability of FGF cortical tensions may be involved (Krieg et al. ligands is not limiting, as these cells are the 2008 ). Additionally, cadherin family gene expres- source of Fgf9/16/20 and are thus surrounded by sion patterns are dynamically regulated during ligand. Instead, the localized cue for differential Ciona intestinalis development and may play a activation cells was found to be Ephrin, which is role in the adhesive properties of the different cell expressed in a-line cells and antagonizes FGF- lineages of the embryo (Noda and Satoh 2008 ). It dependent ERK signaling in A-line neural pro- remains to be seen how the general principles of genitors in a contact-dependent manner (Picco germ layer organization and cell sorting in larger et al. 2007 ). It was shown that this occurs through embryos apply to the much smaller, invariant Ephrin/Eph-dependent localized recruitment of embryos of solitary ascidians. Rasa1 (p120 RasGAP) protein, which is an inhib- itor of the Ras/ERK pathway (Haupaix et al. 2013 ). Development of the Ectoderm Given the requirement for FGF in induction of a-line neural tissue and the conserved role for Neural Induction FGF in neural induction in vertebrates, this inver- The CNS is derived from two of the major quad- sion in the outcomes of induction seems quite rants of the embryo: anterior/animal (a-line) and puzzling. However, FGF has since been shown to anterior/vegetal (A-line). Either lineage gives be repeatedly employed as a simple binary switch rise to both neural and nonneural lineages. The to decide between two opposing cell fates in sis- basic mechanisms by which this induction occurs ter cell pairs throughout ascidian embryogenesis in ascidian embryos are known and differ slightly (Davidson et al. 2006 ; Hudson et al. 2007 ; Shi between the two lineages. In Ciona intestinalis , and Levine 2008 ; Squarzoni et al. 2011 ; Stolfi neural induction of a-line occurs at the 32-cell et al. 2011 ; Wagner and Levine 2012 ). stage and was shown to be carried out by FGF Accordingly, the role of FGF in the A-line noto- signaling. More specifi cally, Fgf9/16/20 ligand chord/neural decision is likely to be an ascidian- expressed in vegetal-pole cells is suffi cient and specifi c regulatory motif, although the induction necessary to induce neurogenic ectoderm specifi - of a-line neural progenitors may indeed be a cally in those a-line cells in contact with the manifestation of an ancestral FGF-dependent Fgf9/16/20-expressing cells (Hudson and mechanism for neural induction. Lemaire 2001 ; Bertrand et al. 2003 ). Those cells that do not receive this signal go on to become Neurulation epidermis and components of the PNS. While Neurulation refers to the chordate-specifi c mor- this induction still requires Gata4/5/6 activity, phogenesis of a dorsal, hollow neural tube start- like earlier induction of basic ectoderm, the ing from a fl at neural plate. In the typical ascidian instructive cue is nonetheless the spatially larva, the neural plate is a fl at epithelium of

[email protected] 158 A. Stolfi and F.D. Brown

Fig. 4.12 Neural plate patterning in Ciona . Cartoon dia- of the neural plate (A-line, a-line, and b-line). Dashed line gram summarizing the fi ndings of Hudson et al. ( 2007 ) on indicates the dorsal midline. Color coding in the neural the grid-like patterning of the C. intestinalis neural plate plate represents unique cell identities as assayed by gene by FGF/MAPK(ERK), Nodal, and Delta-Notch signaling expression profi les. Cell numbers and identities are accu- pathways. Boxed cells are under the direct infl uence of the rately and realistically portrayed, but the relative sizes and indicated signals. Further input into the regulatory mecha- positions of the cells are slightly modifi ed for clarity. nisms for cell fate specifi cation come from the cell lineage Anterior is to the bottom left (© Alberto Stolfi , 2015. All histories of each cell, divided into the three major lineages Rights Reserved)

neurogenic a- and A-line-derived ectodermal tube roof plate, formed by the lateral- most cells of cells that stretch over the dorsal surface of the the neural plate), one ventral column (the fl oor embryo by epiboly. It forms a highly unusual grid plate, formed by the medial-most columns of the of rectangular cells, an organization that is criti- neural plate), and two bilaterally symmetric lat- cal for its proper patterning into rows and col- eral columns, formed from the remainder of neu- umns (Hudson et al. 2007 ). In Ciona intestinalis , ral plate cells) (Nicol and Meinertzhagen 1988a , this plate was shown to be under the infl uence of b ; Cole and Meinertzhagen 2004 ) . T h u s , t h e p a t - several inductive cues that pattern it into rows terning of the neural plate is critical not only for and columns of differently fated cells (Fig. 4.12 ). later compartmentalization of the CNS but for the Namely, Delta and Nodal signals emanate from actual morphogenesis of the neural tube. regions just lateral to the neural plate, while an Neurulation is incomplete at the anterior terminus intricate interplay of FGF and ephrin signals reit- of the neural tube where it remains open, giving eratively determine the binary cell fate choices of rise to the neurogenic portion of the larval stomo- anterior/posterior sister cell pairs, by their oppos- deum (Veeman et al. 2010 ). The neural tube is ing actions on ERK pathway activation (Hudson eventually internalized, after being completely et al. 2007 ; Haupaix et al. 2014 ). covered dorsally by the epidermis. Upon neurulation (Fig. 4.13 ), the rows and col- Like in most of the embryo, the cell lineages umns of the neural plate are precisely reorganized of the neural tube are invariant through neurula- into the simplest of neural tubes: four AP columns tion, except for the exact AP order of intercala- or stacks of cells (one dorsal column (the neural tion of the dorsal and ventral cells. Neural tube

[email protected] 4 Tunicata 159

Fig. 4.13 Neurulation. Cartoon diagram depicting the the neural tube “roof.” The cells of the medial columns mode of neurulation in solitary ascidian embryos. The fl at, also intercalate, forming a single column of cells dorsal neural plate is a grid-like sheet of cells. The lateral (the “fl oor”). Thus, the ascidian neural tube is a simple and medial columns are displaced in opposite directions bundle of four single-cell columns (one dorsal, one ven- (dorsal and ventral, respectively) by as of yet unknown tral and two lateral). A small lumen forms along the inside forces. The cells of the lateral columns meet at the dorsal of the bundle, from the apical surfaces of the neural tube midline and intercalate to form a single column of cells, cells (© Alberto Stolfi , 2015. All Rights Reserved) progenitors proliferate during neurulation, but it anterior, the neural tube lumen is engorged was shown that a prolonged G2 phase in dorsal (Marino et al. 2007 ) and eventually becomes the epidermal midline cells is essential for their sensory vesicle (SV), which houses the larval intercalation (Ogura et al. 2011 ). Thus, there is a melanocytes and associated sensory (light- and tight coordination between cell division and mor- acceleration-sensing cells). Distinct rates of pro- phogenesis of the neural tube. liferation and differentiation occur along the Further elaboration of the neural tube is uneven length of the AP axis, resulting in an irregularly along the AP axis, foreshadowing the regionaliza- shaped neural tube and discontinuous neural tube tion of the larval central nervous system. At the lumen (Cole and Meinertzhagen 2004 ) .

[email protected] 160 A. Stolfi and F.D. Brown

Fig. 4.14 The ascidian larval central nervous system subdivisions of the CNS are demarcated, from anterior to (CNS). Ciona intestinalis larva electroporated with the posterior: sensory vesicle, posterior sensory vesicle (PSV , Vesicular acetylcholinesterase transporter (Vacht )> unc - also known as the brain), neck ( N ), motor ganglion (MG ), 76 :: eGFP plasmid, fl uorescently labeling the major neu- and nerve cord ( NvC ) (Adapted with permission from ronal centers of the larval CNS. The fi ve major anatomical Stolfi et al. (2011 ))

Central Nervous System Patterning expression characterizes the remainder (Ikuta and At the end of larval development, the central ner- Saiga 2007 ). However, a tripartite organization vous system (CNS) contains approximately 100 emerges when one looks at several genes expressed neurons (Imai and Meinertzhagen 2007a ) a n d i s in the neck. Among these, Pax2 /5 /8 a n d Phox2 divided into at least fi ve anatomical compart- suggest homology of the neck to the vertebrate ments along the AP axis (Fig. 4.14 ) . A t t h e a n t e - hindbrain, a claim strongly supported by the fact rior end lies the sensory vesicle (SV). Immediately that these progenitors give rise to Tbx20 + motor posterior to that is the larval “brain” or posterior neurons innervating the pharyngeal muscles of the sensory vesicle (PSV): a tightly bundled cluster of adult (Dufour et al. 2006 ). The motor ganglion neurons, photoreceptors, and other putative sen- (MG) progenitors express several sequence- sory cells associated with the SV. Posterior to the specifi c transcription factor (TF) genes involved in brain is the neck, an undifferentiated mass of neu- spinal cord motor neuron specifi cation in verte- ral progenitors that will give rise to certain bran- brates such as Pax6 , Lhx3 / 4 , Islet , Onecut , and chiomeric neurons of the adult (Dufour et al. Nk6 , such that homology of this region to the ver- 2006 ). Next is the motor ganglion (MG), com- tebrate spinal cord is hard to refute. Not enough is prised of motor neurons and interneurons that known about the adult anterior CNS progenitors, drive the swimming behavior of the larva (Bone and whether they constitute a molecularly or phys- 1992 ; H o r i e e t a l . 2010 ). Most posteriorly, along ically distinct compartment apart from the larval the length of the tail lies the nerve cord. Other brain, leaving us three functionally distinct than a few scattered neurons of uncertain func- regions: larval/adult brain, adult branchiomeric tion, the nerve cord is composed chiefl y of elon- motor neurons, and larval tail motor neurons. gated ependymal cells and the axons of MG The correspondences to the major divisions of neurons that project posteriorly toward the tail- the vertebrate CNS are somewhat tenuous and bud. The clonal subdivisions of the neural tube controversial. There are at least three major fac- (a-line and A-line) together with cell signaling are tors that make these comparisons a dicey propo- responsible for the establishment of regionalized sition. First, there is a drastic alteration in the patterns of gene expression. These molecular positioning of CNS neural precursors and differ- domains are assumed to then specify and drive the entiated neurons. For instance, the motor gan- differentiation of the anatomical subdivisions. glion, argued to be homologous to the vertebrate Gene expression patterns support a fundamen- spinal cord based on molecular, embryological, tally bipartite organization of the ascidian CNS: and functional grounds, is arrayed as only a sin- Otx expression marks the larval brain and adult gle cluster of several bilateral pairs of cells. There anterior CNS progenitors, while absence of Otx is no dorsoventral axis to speak of and no serial

[email protected] 4 Tunicata 161 repetition of this structure, as there is only one That said, it seems that the embryological and muscle to innervate on either side of the larva. molecular evidence thus far presented best sup- Posterior to the motor ganglion, the nerve cord ports a tripartite nature of the ascidian CNS (Wada over the majority of its length contains no neu- et al. 1998). When comparing to vertebrates, the rons of its own, supporting instead the axon bun- motor ganglion and nerve cord may correspond to dles of the motor ganglion neurons. This drastic the spinal cord, the neck may be homologous to reconfi guration is no doubt a result of the extreme the hindbrain, and the anterior Otx - e x p r e s s i n g reduction in cell numbers and size of the larva. domain would encompass everything anterior to Another issue complicating comparative stud- the hindbrain. The fi ner subdivisions seen both in ies is the heterochrony of larval and adult CNS ascidians and vertebrates (midbrain, midbrain/ differentiation. Popularly, the ascidian larval hindbrain boundary, etc.) would be lineage-spe- CNS is thought to simply degenerate and the cifi c elaborations of this basic frame, adapted to adult CNS thought to be formed de novo from an their respective specialized needs. unorganized mass of naïve progenitors. This view has been convincingly refuted, with lineage Motor Neuron Specifi cation tracing experiments revealing the contribution to The gene regulatory networks and inductive the adult CNS of progenitors from specifi c events governing neuronal specifi cation in the regions of the embryonic neural tube (Dufour ascidian CNS have best been studied in motor et al. 2006 ; Horie et al. 2011 ). The different larval ganglion neurons (MGNs). In the Ciona intestina- and adult-differentiating portions of the neural lis M G , f o u r o f t h e fi ve pairs of MGNs are speci- tube will naturally be in different genetic regula- fi ed from the A9.30 blastomere (Fig. 4.15A ; C o l e tory states, compromising the simple one-to-one and Meinertzhagen 2004 ). Each pair corresponds pairing of static “snapshots” of gene expression to a molecularly and morphologically distinct patterns that tend to be emphasized. subtype (Stolfi and Levine 2011 ) a n d i s b o r n a n d Related to this is a third complication, namely, specifi ed in an invariant manner by a series of that many of the genes used in comparing the short-range signaling events (Fig. 4.15B, C ; S t o l fi regionalization of the CNS are reiteratively used et al. 2011). This is in stark contrast to subtype in different cell types at different stages, carrying specifi cation in the vertebrate spinal cord, which out different functions. For example, although is patterned along the DV axis by opposing BMP Fgf8/17/18 was shown to act as an organizing and Shh morphogen gradients. It was shown that molecule in the motor ganglion and neck regions in C. intestinalis , although the neural tube roof as well as at the vertebrate midbrain/hindbrain plate expresses BMP and the fl oor plate expresses boundary (Imai et al. 2009 ), the repeated use of Hedgehog, these molecules have no bearing on FGF signaling for cell fate decisions in Ciona patterning the MG (Hudson et al. 2011 ) . T h i s i s leaves open the question as to whether this orga- not entirely surprising, given that all MGNs are at nizer is a conserved or convergent feature. the same DV position and thus cannot be easily Similarly, Otx is also a critical regulator of poste- patterned by DV morphogen gradients. rior PNS fate (Roure et al. 2014 ), muddling its The A9.29 lineage, just posterior to the A9.30 intimate association with anterior CNS specifi ca- lineage, gives rise to the fi fth MGN, the A10.57 tion. Finally, given the rapid pace of development motor neuron, and to the decussating GABAergic of ascidian embryos, where every cell division is interneurons that lie at the base of the tail and potentially a binary fate choice between two presumably play a role in the left-right alterna- completely different lineages, gene expression tion of muscle contraction that drive swimming patterns are highly dynamic and rapidly changing behavior (Nishino et al. 2010 ; Nishitsuji et al. over the course of embryogenesis (Ikuta and 2012 ). Although physically removed from the Saiga 2007 ). Great care must be taken to ascer- rest of the MGNs, these neurons are likely an tain the temporal and cellular context in which integral component of the ascidian larval motor gene expression is taking place before seizing system and should be considered as belonging to these to support claims of homology. the MG.

[email protected] 162 A. Stolfi and F.D. Brown

AB

C

Fig. 4.15 The cell lineages, patterning, and neuronal in the posterior half of the lineage signals to A11.119 cell, subtypes of the Ciona m o t o r g a n g l i o n . (A ) The cell lin- limiting the expression of Pax3 / 7 t o t h e a n t e r i o r - m o s t c e l l eage of the motor ganglion. Cells in red are those of the lineage, which gives rise to the Dmbx + interneuron descended from the A9.30 blastomeres of the neural plate, (A12.239). The Mnx +/ Nk6 + motor neuron (A11.118) is which gives rise to four of the fi ve pairs of motor ganglion specifi ed by later activation of ERK signaling, and neuro- neurons (indicated by asterisks ). Anterior is to the left nal differentiation of A12.239 is driven by stereotyped (Diagram adapted from Stolfi and Levine (2011 ). Lineage downregulation of ERK and Notch signaling in this cell information is from Cole and Meinertzhagen ( 2004 )). ( B ) by unknown means. Anterior is to the top. ( C ) Summary Summary diagram adapted from Stolfi et al. (2011 ) diagram from Stolfi and Levine (2011 ) of the different depicting the major cell signaling events that pattern the neuronal subtypes of the motor ganglion of C. intestinalis , motor ganglion into specifi c progenitor and neuronal sub- depicting their most salient morphological features, such types. Ephrin-A signals from the adjacent A9.29 lineage as decussating axons of A12.239, frondose motor end suppress MAPK(ERK) signaling in the posterior half of plates of A11.118, dendritic arborizations of A11.117, and the A9.30 lineage and then in the A11.117 cell, which is elongated cell body and en passant motor end plates of specifi ed as a Vsx + interneuron. Delta2 ligand expressed A10.57. Anterior is to the left

Surprisingly, the Hox cluster genes appear to found in C. intestinalis and Oikopleura dioica , play limited functions in the larval CNS of Ciona with a nearly complete set of anterior and poste- intestinalis , with no obvious CNS patterning rior Hox genes, but lacking a number of central defects observed in loss-of-function experiments Hox genes (Fig. 4.16 ) found in vertebrates and (Ikuta et al. 2010 ). Nine Hox genes have been cephalochordates (Seo et al. 2004 ; Brown et al.

[email protected] 4 Tunicata 163

AB

Fig. 4.16 Chordate Hox and ParaHox genes. ( A ) and green boxes ; and posterior genes, blue boxes. Evx Evolutionary reconstruction of the Hox gene complex in duplications and transpositions are shown in gray boxes. chordates and cephalochordates. The complete Hox and Dashed lines indicate large gaps between Hox genes that ParaHox clusters are shown for the cephalochordate remain on the same chromosome but have been dispersed. (Branchiostoma fl oridae ). A disintegration in the order of (B ) Hox gene expression pattern relative to the anterior- the clusters and the lack of central Hox genes can be seen in posterior anatomical position of the larva of O. dioica . ( C ). the tunicates ( Oikopleura dioica a n d Ciona intestinalis ). Hox gene expression pattern relative to the anterior-poste- The coelacanth (Latimeria menadoensis ) Hox and ParaHox rior anatomical position of the larva of C. intestinalis clusters are also shown (dark colored circles ). Hox genes (Figures are based on data from Ikuta et al. (2004 ), Seo are color-coded by their relative position of expression: et al. (2004 ) , B r o w n e t a l . (2008 ), Amemiya et al. (2010 ), anterior genes, red a n d orange boxes ; c e n t r a l g e n e s , yellow David and Mooi (2014 ), Mulley and Holland (2014))

2008 ; Amemiya et al. 2010 ; Ikuta et al. 2004 ; the basic larva of species covering diverse fami- Ikuta 2011 ; David and Mooi 2014 ). The Hox lies, there are only two pigment cells. One is an complex itself is fragmented and eroded in the otolith (also known as the statocyst or statocyte) genomes of tunicates (Spagnuolo et al. 2003 ; that is required for geotaxis of the larva, presum- Ikuta et al. 2004 ; Seo et al. 2004 ), with partial ably by acting as a weight that can activate asso- loss of collinearity of gene expression and cluster ciated mechanosensory cells, depending on the organization (Fig. 4.16 ). The same is observed relative orientation of the larva in the water col- for the ParaHox genes (Ferrier and Holland umn (Tsuda et al. 2003; Jiang et al. 2005 ). The 2002 ). Those Hox cluster genes that remain are other is an ocellus, associated with a photorecep- expressed in the nerve cord, notochord, and mus- tor complex (Tsuda et al. 2003 ; Horie et al. 2008 ). cle of the larva (Ikuta et al. 2004 ; Seo et al. 2004 ) In some species, the otolith or the ocellus may be and undoubtedly play indispensable biological missing. In other species (e.g., Botryllus schlos- roles that have yet to be fully elucidated, but it is seri and other botryllids; Fig. 4.6C ), both have nonetheless stunning to see the rock stars of been combined into a dual function pigment cell bilaterian embryonic patterning remain behind termed the “photolith” (Sorrentino et al. 2000 ). the scenes in the tunicates. In species with separate otolith and ocellus, both cell types are derived from left-right equiva- Neural Plate Borders, Placodes, lent cells in the neural plate, and their differentia- and Neural Crest tion into one or the other occurs mainly through From the lateral borders of the neural plate arise relative positioning along the anterior-posterior a handful of different cell types, mostly associ- axis after intercalation at the dorsal midline ated with the PNS. The most conspicuous of (Darras and Nishida 2001 ; Abitua et al. 2012 ). these are the melanin-containing pigment cells The cell that ends up more anterior invariably associated with the sensory vesicle (Fig. 4.17 ). In becomes the otolith, and the one that ends up

[email protected] 164 A. Stolfi and F.D. Brown

Fig. 4.17 The pigment cells of ascidian larvae. ( A ) A Larva of Molgula occidentalis , showing a single melanocyte in the sensory vesicle, an otolith (also known as a statocyte or statocyst) associated with putative geotactic mechanoreceptory cells. (B ) Larva of Ciona intestinalis , showing two melanocytes in the sensory vesicle: one otolith and one ocellus pigment cell (© Alberto Stolfi , 2015. All Rights Reserved)

B

more posterior is specifi ed to become an ocellus. expression of common neural plate border speci- Both subsequently undergo an epithelial-to- fying genes like the ones listed above further sup- mesenchymal transition as they delaminate from ports the claims for homology. Later, neural plate the neural tube epithelium and enter the sensory border descendants express genes associated vesicle lumen. with more advanced neural crest developmental Wada was the fi rst to note that the specifi ca- processes, including Foxd and Mitf orthologs tion of otolith/ocellus precursors from the neural (Abitua et al. 2012 ). plate borders closely resembles the development Now, are the NPBCs in fact neural crest cells? of neural crest-derived melanocytes in verte- That is a trickier question to answer. Since neural brates (Wada et al. 1997 ; Wada 2001 ). crest cells were initially identifi ed in vertebrates, Interestingly, neural plate border cells (NPBCs) they were defi ned by those criteria that described express combinations of transcription factors that them in vertebrates. And there are more than a in vertebrates participate in the specifi cation of few such criteria, given the myriad embryonic neural crest cells. These include orthologs of origins, molecular signatures, and differentiated Pax3 / 7 , Msx , Dlx , Zic , Snail , Id , and others (Wada fates of all those cells collectively referred to as et al. 1997 ; Abitua et al. 2012 ). Molecularly, the the “neural crest.” As such, it is a tall order to

[email protected] 4 Tunicata 165 expect to fi nd cells outside the vertebrates that Epidermis and Epidermal Sensory strictly fi t those multiple criteria. 8 Taking this Neurons into consideration, we venture that the answer is The entire larva is covered by an epithelium com- then ascidian NPBCs are homologous to neural posed of epidermal cells and related cell types crest cells, but are not neural crest cells. that share a common clonal origin. Neurons com- We have seen that ascidian NPBCs share an prising the PNS are embedded throughout the embryological origin and certain regulatory sig- epidermis in both the head and tail regions, and natures with vertebrate neural crest cells. other specialized structures arise at the border Furthermore, like their craniate counterparts, the regions between the epidermis and neurogenic NPBCs give rise to melanocytes and sensory ectoderm. All these epidermis derivatives are neurons. What are obviously missing are two descended from the four animal-pole octants of things: (1) the capacity/propensity for long-range the eight-cell embryo. The major roles of the lar- migration and (2) the ability to give rise to val epidermis are to seal in the different tissues mesoderm- like differentiated cell types. and organ primordia and to synthesize the tunic. Along the descent of vertebrates, a rudimen- The tunic, from which the tunicates derive their tary neural crest similar to the ascidian NPBCs name, is made of cellulose. Tunicates are the only must have come to express genes that conferred metazoans capable of synthesizing cellulose. on them the entire suite of neural crest properties This is made possible by epidermal expression of we have come to recognize. In tunicates, we fi nd the cellulose synthase enzyme, thought to be other cells that display some of these neural derived from a gene that was horizontally trans- crest-like properties. Pigment cell precursors ferred from an unidentifi ed prokaryotic organism capable of migrating long distances were identi- to the last common ancestor of all tunicates fi ed in the complex adultative larvae of (Nakashima et al. 2004 ; Matthysse et al. 2004 ; Ecteinascidia turbinata (Jeffery et al. 2004 ). Sasakura et al. 2005 ). These were proposed as homologous to the A7.6- In the head, PNS neurons (including those derived mesenchymal cells of Ciona intestinalis derived from the neural plate borders) are dis- and Halocynthia roretzi , which do express some persed along the anterior and dorsal regions, neural crest factors such as Twist - related genes being absent from the ventral side of the epider- (Jeffery et al. 2008 ). However, the embryological mis (Imai and Meinertzhagen 2007b ). In the tail, origin of the A7.6 lineage is not from the neural regularly interspaced epidermal cells at the dor- plate borders, being of clear endomesodermal sal and ventral midlines differentiate into caudal origin. Interestingly, the forced expression of epidermal sensory neurons (CESNs) that form Twist - related genes in neural plate border cells of thin projections into the tunic (Fig. 4.18 ) and are C. intestinalis was found to confer migratory thought to serve a mechanosensory role (Pasini properties to them, resulting in a long-range et al. 2006 ; Imai and Meinertzhagen 2007b ; migration of pigment cell precursors, mimicking Terakubo et al. 2010 ; Yokoyama et al. 2014 ). the behavior of E. turbinata pigment cell precur- Caudal CESNs are induced from neurogenic sors (Abitua et al. 2012 ). Such an experiment epidermis of the tail, which is induced at the dor- may represent a very simplifi ed reenactment of sal and ventral midlines. The dorsal midline neu- the evolutionary co-option that gave rise to bona rogenic epidermis is induced by FGF signaling, fi de vertebrate neural crest cells. while the ventral midline neurogenic epidermis is induced by ADMP/BMP signaling (Pasini et al. 2006 ). Within these median stripes of neurogenic 8 If the forelimbs had been initially identifi ed in birds and epidermis, putative mechanosensory neurons are been strictly defi ned as feathery appendages that bear specifi ed through Delta-Notch lateral inhibitory three digits, then one would be infallibly correct in claim- signaling (Pasini et al. 2006 ), which controls ing that only birds have forelimbs and that the homolo- gous structures found in all non-avian vertebrates are like the deployment of a core PNS neurogenic code forelimbs, but not forelimbs. consisting of the transcription factors Mytf,

[email protected] 166 A. Stolfi and F.D. Brown

Fig. 4.18 Caudal epidermal sensory neurons A of Ciona . ( A ) The thin ciliary projections of the caudal epidermal sensory neurons into the tunic fi n revealed by beta-tubulin immunostaining. (B ) Higher magnifi cation view of tunic fi n sensory projections (Figure is adapted from Pasini et al. (2006 )) B

NeuroD-related, Pou4, and Atoh and the micro- While part of the incurrent siphon primordium RNA miR - 124 (Chen et al. 2011 ; Tang et al. is determined in a lineage-dependent manner 2013 ). Further AP patterning of these sensory from cells forming the neuropore9 ( V e e m a n e t a l . neurons is achieved through antagonistic retinoic 2010 ), the excurrent siphon primordia are acid and FGF signals emanating from the ante- induced de novo from naïve epidermis by reti- rior larval muscles and tailbud, respectively noic acid-dependent expression of Hox1 in com- (Pasini et al. 2012 ). bination with later induction by FGF/ERK signaling (Kourakis and Smith 2007 ; Sasakura Adult Siphon Primordia et al. 2012 ). Also arising from the head epidermis are the siphon primordia that give rise to the oral and Adhesive Organ aboral (or incurrent and excurrent) openings, or At the most anterior end of the head is the adhe- siphons, of the adult. The excurrent siphon is sive organ, which in the majority of ascidians (one also known as the “atrial” siphon, due to its asso- notable exception being the Molgulidae) consists ciation with the peribranchial atrium or chamber. of three protruding clusters of putative sensory These are observed late in embryogenesis as and adhesive-secreting cells. These cell clusters rosettes of morphologically distinct epidermal are commonly referred to as “palps.” The palps cells. In phlebobranchs, one incurrent and two are remarkably elaborated in phlebobranch ascid- excurrent siphon primordia are specifi ed, and ians that undergo adultation (discussed below), initially the juveniles, or young adults, have perhaps because their larger, heavier larvae need three siphons each. Later in adult life, the two to be more securely fastened to the substrate. excurrent siphons fuse into a single siphon in a The adhesive organ primordium arises from poorly studied process. In contrast, the excurrent the anterior borders of the neural plate and is siphon of stolidobranch ascidians arises from a single primordium in the larva or young 9 The opening formed by the incomplete closure of the juvenile. neural tube

[email protected] 4 Tunicata 167 specifi ed in part by Foxc (Wagner and Levine 2002 , 2007 ; Imai et al. 2006 ). Mrf, Tbx6, and 2012 ). Later adhesive organ development appears other transcription factors then cooperate to acti- to require FGF/ERK signaling and an intricate vate the expression of a battery of “differentia- interplay between the transcription factors Sp8, tion” genes such as myosins, muscle actins, Emx, and Islet to specify the different cell types troponins, tropomyosins, etc. (Johnson et al. composing the palps and to regulate their mor- 2004 , 2005 ; Brown et al. 2007 ; Izzi et al. 2013 ). phogenesis (Wagner et al. 2014 ). In contrast, secondary muscle cells are induced by extracellular signals from A- and b-line progenitors that do not inherit any mater- Development of the Mesoderm nal Macho1 . Cell-cell signaling results in activa- tion of Tbx6 - r and Mrf independent of Macho1 Larval Muscles (Kim and Nishida 2001 ; Hudson et al. 2007 ; The ascidian larva is specialized for dispersal, Tokuoka et al. 2007 ). However, the exact identity this being in fact its only role in the life cycle. of the signaling pathways used for the same Indispensable to this function are the larval mus- inductive event differs between Ciona intestina- cles, arranged as bands of striated, mononucle- lis and Halocynthia roretzi , representing another ated cells on either side of the tail. These muscle example of developmental system drift in ascid- cells are divided into primary and secondary ian evolution (Hudson et al. 2007 ; Tokuoka et al. muscle lineages, according to their mode of spec- 2007 ; Hudson and Yasuo 2008 ). ifi cation. In Ciona intestinalis , there are 18 mus- The variation in the number of secondary cle cells per side and 21 cells per side in the muscle cells being specifi ed between Ciona and slightly larger Halocynthia roretzi embryo. This Halocynthia appears to be due to slight differ- difference in muscle cell number is entirely due ences in cell positioning. In C. intestinalis , to variation in the secondary lineage, while the muscle- inducing cells are in contact with only primary lineage is invariant between the two spe- the b7.9 blastomere, whereas in H. roretzi contact cies. It does not appear that these numbers devi- is made with both b7.9 and b7.10, resulting in ate substantially among the larvae of the various induction of extra muscle progenitors (Kim and solitary species. However, in those species show- Nishida 2001 ; Hudson et al. 2007 ; Tokuoka et al. ing extreme adultation and caudalization 2007 ). (explained later), larval muscles are greatly elab- orated, both in terms of number of rows of cells Cardiopharyngeal Mesoderm and total numbers of cells ( Cavey and Cloney The term “cardiopharyngeal mesoderm” refers to 1976 ; Cavey 1982 ). multipotent progenitors that give rise to heart Primary muscle cells are autonomously speci- precursors and muscles of the peribranchial fi ed, through inheritance of posterior/vegetal chamber and excurrent siphon of the adult (col- mitochondria-rich cytoplasm (Conklin’s famous lectively referred to as “pharyngeal muscles”). In orange-yellow myogenic cytoplasm or “myo- the embryo, four cardiopharyngeal progenitors plasm”), shown by Whittaker to be suffi cient and (CPPs), two on either side of the embryo, are necessary for larval muscle formation (Whittaker specifi ed from the B7.5/ B 7.5 pair of blastomeres, 1973 , 1982 ). The critical molecular component which also give rise to anterior larval muscle of the myoplasm was identifi ed as maternally cells (Fig. 4.19 ). deposited Macho1 mRNA, which is localized to Investigations into marginal zone specifi ca- the cER and translated into Macho1 protein in the tion have largely ignored the B7.5 pair of blasto- primary larval muscle precursors (Nishida and meres, which are situated at the posterior end of Sawada 2001 ). Macho1 protein in turn activates the marginal zone. Embryologically, the B7.5 its targets, among them the Tbx6 - related family cells are a confl uence of vegetal and posterior of genes (Yagi et al. 2005 ). Tbx6-related proteins endoderm and mesoderm, and this “hybrid” then activate the myogenic regulatory factor gene nature holds also true at the molecular level. In Mrf , also known as Myod (Meedel et al. 1997 , Ciona intestinalis , the B7.5 cells are delineated

[email protected] 168 A. Stolfi and F.D. Brown

Fig. 4.19 The B7.5 lineage and cardiopharyngeal meso- sors, ASMFs atrial siphon muscle founder cells, ASMPs derm. Diagram of the cell division and morphogenetic atrial siphon muscle precursors, HPs heart precursors, events of the B7.5 lineage, which gives rise to the cardio- OSM oral siphon muscles, ASM atrial siphon muscles, pharyngeal progenitor cells ( CPPs ) that will generate the LoM longitudinal body wall muscles, Myh.c Myosin heart and atrial siphon muscles ( ASM ) of the juvenile/ heavy chain.c (also known as MHC3 ), Myh.b Myosin adult. Refer to text for details. ATMs anterior tail muscles, heavy chain.b (also known as MHC2 ) (Figure is adapted FHPs fi rst heart precursors, SHPs second heart precur- from Wang et al. (2013 )) by expression of the Mesogenin / MesP homolog Experiments carried out in Ciona intestinalis Mesp , which is exclusively expressed in these and later extended to other ascidian species cells and is necessary for the subsequent onto- revealed that the CPPs are specifi ed from the genesis of lineage and its derivatives including B7.5 lineage by FGF signaling (Davidson et al. the adult heart and pharyngeal muscles (Davidson 2006 ; Cooley et al. 2011 ). FGF-dependent acti- et al. 2005 ; Satou et al. 2004). In Ciona intestina- vation of the ERK pathway and its effector, the lis , Mesp is activated by Tbx6.b and Lhx3/4, transcription factor Ets.b, results in upregulation zygotically expressed transcription factors under of a core of cardiac regulatory network genes and the direct regulatory control of beta-catenin and cell migration effector genes (Davidson and Macho1 maternal determinants, respectively. The Levine 2003 ; Davidson et al. 2006 ; Beh et al. expression territories of Tbx6.b (posterior) and 2007 ; Christiaen et al. 2008 ). In contrast, their Lhx3/4 (vegetal) overlap precisely in the B7.5, sister cells do not activate the ERK pathway and resulting in synergistic activation of Mesp in the cardiopharyngeal program and are instead these cells alone (Christiaen et al. 2009 ). A specifi ed as anterior primary larval muscle cells, Pem1-dependent timing mechanism has been as they inherit some myoplasm. proposed to help coordinate the precise overlap The CPPs detach from their sister cells and in time of these transcriptional cascades in the migrate anteriorly into the head (Hirano and B7.5 (Tolkin and Christiaen 2012 ). Nishida 1997 ). In Ciona intestinalis , they migrate

[email protected] 4 Tunicata 169 along a ventral path and are thus called trunk ven- cells, occupying the posterior tip of the noto- tral cells (TVCs), but in other species such as chord, are derived from the B8.6 blastomere. Molgula spp., these cells undertake a more lateral The primary notochord precursors are induced migration (Stolfi et al. 2014 ). by an FGF signal and activate transcription of In the head, the CPPs undergo two rounds of Brachyury ( Bra ) . Bra i s s u f fi cient and necessary for asymmetric cell division to give rise to ventral/ notochord fate in Ciona a n d Halocynthia (Yasuo medial heart progenitors and dorsal/lateral pha- and Satoh 1998 ; Takahashi et al. 1999 ). Later, the ryngeal muscle progenitors. The pharyngeal secondary notochord precursors are induced by muscle progenitors are specifi ed by a Tbx1/10- Ebf Delta-Notch signaling to also activate Bra (Hudson cascade (Stolfi et al. 2010 ; Wang et al. 2013 ; and Yasuo 2006 ). Upstream factors also regulating Razy-Krajka et al. 2014 ), which is antagonized Bra activation in both lineages include Foxa and by Nk4 in the heart precursors. The heart pro- Foxd, while Zic-related appears to be required for genitors remain quiescent in the larva until Bra only in the primary notochord precursors (Imai needed for heart organogenesis during metamor- et al. 2002a , b , 2006 ; W a d a a n d S a i g a 2002 ; Yagi phosis. On the other hand, the adult muscle pro- et al. 2004 ; K u m a n o e t a l . 2006 ). genitors migrate to dorsal regions and surround At the moment of gastrulation, notochord pre- the excurrent siphon primordia while the larva is cursors are fate restricted and begin to ingress, still swimming (Fig. 4.19 ). prior to dividing twice to give rise to the fi nal During metamorphosis, cardiac and pharyngeal number of notochord cells. The cells form a noto- muscles differentiate, each expressing a unique chord plate, which undergoes convergent exten- suite of muscle structural genes (Stolfi et al. 2010 ; sion, contributing to the elongation of the tail Razy-Krajka et al. 2014 ). The pharyngeal muscle (reviewed in Jiang and Smith 2007 ). Final differ- progenitors will further migrate and proliferate to entiation of the notochord involves the deposition elaborate the musculatures of the excurrent siphon of extracellular luminal matrix by each notochord and the peribranchial chamber, which comprises cell. These lumens then fuse with each other to essentially the entire body wall of the adult (Hirano generate a single tube extending along the length and Nishida 1997; Razy- Krajka et al. 2014 ). Given of tail (Dong et al. 2009 ). the striking parallels in ontogeny and clonal topol- Downstream of Bra, approximately 450 genes ogy between the cardiopharyngeal mesoderm in have been identifi ed as being upregulated in the ascidians and vertebrates, it has been proposed notochord (Hotta et al. 1999 ; Takahashi et al. that ascidian pharyngeal muscles are homologous 1999 ). Of these, some are direct targets of Bra, to certain craniofacial muscles in vertebrates with earlier-activated targets containing more (Stolfi et al. 2010 ; Tzahor and Evans 2011 ; Tolkin Bra binding sites in upstream cis-regulatory and Christiaen 2012 ). sequences than targets activated later, which typi- cally contain only one binding site (Katikala Notochord Formation et al. 2013 ). There are also indirect targets of Bra, The notochord, one of the defi ning chordate mediated by transcription factors that are direct traits, is the most salient feature of the ascidian Bra targets (Katikala et al. 2013 ). All in all, the larva and serves primarily as a hydrostatic skele- notochord genes constitute a diverse group of ton required for the biomechanics of swimming proteins involved in several steps of notochord (Jiang and Smith 2007 ). It is a single column of morphogenesis and differentiation. 40 cells, which are vacuolated late in larval devel- opment to form a hollow tube. Embryologically, Mesenchyme the notochord has two origins (Fig. 4.20 ). The 32 With the exception of the anterior portion of the primary notochord cells are derived from the notochord, mesodermal cells in the head of the A-line, specifi cally the A7.3 and A7.7 blasto- larva are undifferentiated and have been collec- meres. The remaining eight secondary notochord tively referred to as “mesenchyme” due to their

[email protected] 170 A. Stolfi and F.D. Brown

A B

C D

EF

Fig. 4.20 The ascidian notochord. (A ) 110-cell stage eage colored in orange . ( D ) Magnifi ed view of boxed area Ciona intestinalis embryo. The primary notochord precur- in ( C ). ( E ) Notochord cells in late tailbud stage embryo, sors are false colored in green ; the secondary notochord already elongated and undergoing vacuolization. ( F ) precursors are colored in orange . ( B ) Early tailbud stage Vacuolization of notochord cells is complete in the larva, embryo, with the cluster of intercalating notochord cells with the unifi ed lumens resulting in one long hollow tube outlined by the dotted line . ( C ) Notochord in a mid-tail- running the entire length of the notochord. Scale bars in bud stage embryo, showing a single row of cells in the all panels equal 50 μm (© Alberto Stolfi , 2015. All Rights “stack of coins” confi guration. Secondary notochord lin- Reserved) loose, seemingly disorganized distribution. These the B7.5-derived cardiopharyngeal progenitors and cells undergo an epithelial-to-mesenchymal transi- mesenchyme derived from the A7.6, B7.7, and tion after gastrulation and engage in proliferation B8.5 pairs of blastomeres (Tokuoka et al. 2004). and migration to varying degrees. Among these, The most anterior mesodermal lineage, the there is usually a strong distinction made between A7.6, has been reported to give rise to diverse

[email protected] 4 Tunicata 171 adult tissues including hematopoietic (blood) laid out in the larva and are not signifi cantly rear- cells, gill slit lining, part of the stomach and peri- ranged during metamorphosis. However, there branchial chamber, and incurrent (oral) siphon was variation among individuals in the contribu- muscles (Hirano and Nishida 1997 ; T o k u o k a e t a l . tions of each blastomere to the fi nal differentiated 2005). There were discrepancies between A7.6 organs, suggesting that the endoderm is patterned tracing in different species, and new tools like at the larval stage in a position-dependent manner photoconvertible intracellular fl uorescent proteins by extrinsic cues, and not according to determin- may be needed to ascertain whether these repre- istic, lineage-dependent processes. Late in larval sent true species-specifi c differences. The remain- development, the digestive tract rudiment can be ing mesenchymal lineages, B7.7 and B8.5, have seen to undergo tubular morphogenesis been reported in Ciona as giving rise to blood (Nakazawa et al. 2013 ), while in those species cells as well as diverse cells that come to populate that show adultation, these and other organs are the adult tunic, whereas in Halocynthia roretzi fully formed at the moment of larval hatching. they are described as giving rise to tunic cells only (Hirano and Nishida 1997 ; T o k u o k a e t a l . 2005 ) . FGF signaling and the basic helix-loop-helix Primordial Germ Cells (bHLH) factor Twist-related were shown to be required for specifi cation of mesenchyme (Imai The highly asymmetric divisions of the posterior et al. 2002a , 2003 ; Tokuoka et al. 2005 ), but little vegetal end of the embryo eventually give rise to else is known about the developmental events the B7.6/B7.6 pair of blastomeres, the smallest and processes involved in specifi cation and dif- and most posterior cells in the pre-gastrula ferentiation of A7.6, B7.7, and B8.5 mesenchy- embryo. They are kept transcriptionally silent mal derivatives of the adult. through inheritance of Pem1 mRNAs and protein. They also inherit other postplasm components including the homolog of the major metazoan Development of the Endoderm germ cell marker Vasa (Fig. 4.21 ; Takamura et al. 2002 ; Brown and Swalla 2007 ; Brown et al. Endoderm specifi cation occurs immediately 2009 ; Shirae-Kurabayashi et al. 2006 ). downstream of vegetally localized, stabilized After gastrulation, these cells end up in the ven- maternal Beta-catenin, in part by activation of troposterior portion of the larval tail. Each B7.6 Lhx3/4 (Imai et al. 2000 ; Satou et al. 2001 ). In cell divides and gives rise to two daughter cells: the Ciona intestinalis larva, there are approx. 500 B8.11 and B8.12. Vasa mRNAs and protein are endodermal cells, the majority of these residing asymmetrically inherited by the more posterior in the head while a minority comprises the endo- B8.12 cell, whose descendants are eventually dermal strand, a single column of cells that runs incorporated into the adult gonad and give rise to underneath the notochord to the posterior tip of the animal’s germ line (Takamura et al. 2002 ; the animal. The endoderm is largely undifferenti- Shirae-Kurabayashi et al. 2006 ). Thus, the B8.12 ated upon hatching, but undergoes a process of cells are the primordial germ cells (PGCs). differentiation while the larva is still swimming, Curiously, Pem1 is asymmetrically inherited by in preparation for settlement and metamorphosis B8.11, suggesting the PGCs are released from (Nakazawa et al. 2013 ). global transcriptional repression around this stage. Cell tracing experiments in Halocynthia ror- It was found that germ cells could be second- etzi revealed that the clonal boundaries of these arily induced in adults generated from larvae in cells in the larva were invariant and could be which the tail, and therefore the B8.12-derived traced to blastomeres in the pre-gastrula embryo primary germ cells, had been severed. This sug- (Hirano and Nishida 2000 ). Furthermore, differ- gests that there may be mechanisms that allow entiated adult organs could be traced to specifi c for induction of secondary germ cells in case the regions of the endoderm fate map in the larva, primary germ line is somehow lost (Takamura indicating that the “anlagen” of these organs are et al. 2002 ).

[email protected] 172 A. Stolfi and F.D. Brown

A B

Fig. 4.21 The primordial germ cells ( PGCs ) of solitary closely related colonial Botrylloides violaceus . Note that and colonial styelid ascidians. RNA-dependent helicase in both cases Vasa - positive cells are localized to a few Vasa transcripts are shown as early PGC marker. (A ) Vasa posterior cells of the early embryos. Developmental transcripts in blue for the solitary Boltenia villosa . (B ) stages are noted below each panel. Scale bar in (A ) equals Vasa transcripts marked by white arrowheads in the 50 μm (From Brown et al. ( 2007 , 2009 ))

Taken as a whole, ascidian embryos are excel- imaging of fl uorescently tagged proteins and tar- lent organisms in which to observe cellular pro- geted disruption of candidate genes (Dong et al. cesses in vivo and investigate the function of 2009 , 2011 ; Denker et al. 2013 ; Deng et al. 2013 ). genes involved in such processes. The gene regu- Finally, cell-specifi c transcriptome profi ling can latory networks of the ascidian embryo (Fig. 4.5 ) reveal those genes upregulated during morpho- have been considerably annotated, but these net- genesis, which was done for the migrating CPPs works so far have mostly concerned the connec- and revealed several candidate migration genes tions among the transcription factors and (Christiaen et al. 2008 ). Although the tissues and signaling pathways. More recently, our knowl- cell types investigated thus far are not many, the edge of these networks has started to extend to future for research on cellular morphogenesis in their interface with the realm of effector genes – ascidian embryos holds great promise. those that do not directly regulate transcription but rather interact with other proteins to carry out mechanical or enzymatic functions. The best- Early Development of Other studied effector genes are those that are respon- Tunicates sible for the unique properties of differentiated cells, also known as terminal differentiation Recently acquired evidence suggests that the genes. More challenging genes to study are those thaliaceans and larvaceans, once thought to be that are involved in transient cellular properties distinct from ascidians, appear to group within and behaviors like polarity, motility, adhesion, the paraphyletic Ascidiacea. Phylogenetic trees contractility, and other components of morpho- strongly support the placement of a monophy- genesis. There have been some notable examples letic Thaliacea squarely within the phlebobranch where the genes controlling these processes in ascidians, i.e., Enterogona (Tsagkogeorga et al. the ascidian embryo were discovered. For 2009 ). Phylogenetic support for a revised classi- instance, forward genetic screens for notochord fi cation of the larvaceans as the sister group to the morphogenesis defects in Ciona identifi ed genes stolidobranch ascidians is not as strong, but required for the convergent extension of this tis- embryological data suggest they may be highly sue (Jiang et al. 2005 ; Veeman et al. 2008 ). A derived ascidians (Stach et al 2008 ; Fujii et al. later phase of notochord morphogenesis, namely, 2008 ). Here, we discuss what is known about the formation of the notochord tube, has been developmental processes in these less- frequently studied using a combination of sophisticated studied groups.

[email protected] 4 Tunicata 173

Larvaceans: Reduction and Neoteny cately sculpted cellulosic fi ltration apparatus In the larvaceans, as the name implies, there is no (Fig. 4.22 ). This house is shaped through the pat- morphological contrast between the embryo/ terning of the head epidermis into a complex larva and the adult. Larvaceans are planktonic landscape of specialized cells that secrete spe- tunicates that retain a functional notochord and cifi c subsets of oikosins, the major protein com- paraxial musculature throughout their entire ponent of the house (Fig. 4.22 ; Spada et al. 2001 ; adult lives. These constitute a true postanal tail Thompson et al. 2001 ; Hosp et al. 2012 ). The that has been adapted to beat vigorously and gen- house is many times the size of the actual animal erate water currents through the “house,” an intri- and utilizes a series of canals and fi lters to trap

A B

C

Fig. 4.22 The larvacean house. (A ) Lateral view of an mouth. ( C ) Oikoplastic epidermis of O. dioica stained Oikopleura dioica individual and surrounding house. with Hoechst ( left ) and To-Pro3 (right ), revealing cell Anterior is to the top right . ( B ) Illustration of the animal nuclei. Different fi elds of cells have had their nuclei false and house depicted in ( A ). Animal is colored in orange. colored to highlight the different fi elds that shape the Black arrows indicate the direction of water currents that house through synthesis and secretion of various proteins, are swept into the house by beating of the animal’s tail and including oikosins (Panels ( A , B ) were adapted from through the fi lter apparatus of the house. Red arrow indi- Bouquet et al. ( 2009 ); (panel C ) was adapted from Hosp cates movement of food particles toward the animal’s et al. (2012 ))

[email protected] 174 A. Stolfi and F.D. Brown

Fig. 4.23 The ascidian and larvacean fate maps. ascidian fate map is largely based on the work of Nishida Comparison between the typical solitary ascidian and ( 1987 ) and the Oikopleura map based primarily on Stach Oikopleura dioica (larvacean) embryonic fate maps. The et al. (2008 ) and Fujii et al. (2008 ) particle food. The animal can eject from its house Strikingly, only a fraction of the Ciona noto- and does so with some regularity, some fi ve to ten chord toolkit genes are expressed in the times per day, synthesizing a new one in just Oikopleura dioica notochord (13/50), and half of 10 minutes (Alldredge 1977 ). them are completely absent from the O. dioica Occasionally imagined as representing the genome (Kugler et al. 2011 ). How larvaceans common ancestor of all tunicates, these remark- cope with a reduced number of genes to build a able creatures likely evolved from an ascidian- notochord is not yet clear. Since the larvacean tail like ancestor instead (Stach et al. 2008 ). is used to draw the water current through the ani- Secondarily evolved heterochronies superim- mal’s house, as opposed to a larval dispersal posed on the biphasic development of ascidians mechanism, it is highly elaborated during the lar- would have resulted in neoteny. This view is sup- val and adult stages, as notochord cells prolifer- ported by recent tunicate phylogenies that place ate to generate over 100 cells arranged into four larvaceans within the ascidian tree, although longitudinal columns (Søviknes and Glover these analyses are hampered by long- branch 2008 ). It is possible that this difference in gene attraction problems (Tsagkogeorga et al. 2009 ). expression underlies the adaptation of the larva- On the other hand, other analyses have placed the cean notochord for this purpose. larvaceans as the sister group to all other tuni- Another clear example of cell count reduction cates (Delsuc et al. 2006 , 2008 ). during embryogenesis is seen in the neural plate, The larvacean embryo and developmental which is comprised of just two columns of four times are even more reduced than those of soli- cells each when neurulation begins (Fujii et al. tary ascidians. The complete life cycle of the spe- 2008 ), as opposed to the eight columns and six cies Oikopleura dioica is only 5 days at 20 °C. cells of the Ciona a n d Halocynthia n e u r a l p l a t e s . Gastrulation occurs at the 32-cell stage, as Additionally, larvacean tail muscles are made of opposed to gastrulation in ascidians which hap- only a single row of ten cells on either side of the pens at the 110-cell stage (Fujii et al 2008 ). The notochord (Nishino et al. 2000 ), in contrast to the embryo is also mostly fate restricted at this stage, 18 or 21 cells per side seen in typical ascidian and this map is very similar to the ascidian fate larvae. map, in spite of the fewer cells (Nishida 2008; Given these striking parallels with the Stach et al. 2008 ). Each cell lineage develops in a slightly more complex embryos of solitary way that is very reminiscent of the homolog in ascidians, it seems reasonable to conclude that ascidian embryos, albeit clearly reduced in cell larvaceans may be descended from a sessile, numbers (Fig. 4.23 ). For instance, there are 20 solitary ascidian- like ancestor, but have special- notochord cells at the end of embryogenesis, as ized in part through reduction in cell number, opposed to 40 in most solitary ascidians (Nishino basic toolkit gene number, and acceleration of et al. 2001 ). development.

[email protected] 4 Tunicata 175

Embryonic Development of Thaliaceans that develop around the cyathozooid, which Although the thaliaceans are pelagic and free quickly degenerates afterward. swimming in their adult phase, they are very Perhaps the most specialized developmental distinct from the larvaceans and represent an mode observed in all of the tunicates occurs in alternative evolutionary trajectory from an the salps. In early embryogenesis, invading folli- ascidian-like ancestor. They resemble pelagic cle cells separate the blastomeres, which come ascidians, fi lter-feeding as they fl oat in the back together again later in development (Todaro water column. Salps use muscular contractions 1880 ; Sutton 1960 ). The function of this deliber- of the pharyngeal cavity to generate locomo- ate fragmentation of the early embryo has never tion by jet propulsion. They constantly produce been ascertained. and consume mucous nets that trap particles Doliolid embryonic development occurs from passing through the pharyngeal cavity, which internally fertilized eggs and generates a pelagic allows for simultaneous fi lter feeding and loco- chordate-like larva, which will eventually differ- motion. As a result, they can fi lter a prodigious entiate the typical barrel-shaped form of doliolid amount of water. Doliolids also move about zooids in its head, while the tail regresses to form through jet propulsion, but their feeding and the oozooid of the asexually reproducing swimming are separate functions, as they use generation. cilia to draw particles through the mucous Relatively little else is known about thaliacean feeding net, being similar in this manner to development, and much less is known about the ascidians. Pyrosome colonies move through underlying molecular mechanisms. All three the combined fl ow of each minuscule zooid, orders are capable of asexual reproduction, and which is generated by cilia as well (reviewed in salps and doliolids have complex alternation of Alldredge and Madin 1982 ). sexual and asexual generations (see below). We Pyrosomes and salps lack a caudate larval can only hope that major developmental studies stage and both have direct developing embryos in the thaliaceans will be revived quite soon, in (Julin 1912 ; Berrill 1950a , b ; Sutton 1960 ). Of time for the next round of reviews and book the three main orders in Thaliacea, only certain chapters. species of the doliolids have a tadpole-like larva (Godeaux 1955 ), which has been used as a reminder of their ascidian pedigree and to sup- LATE DEVELOPMENT port evidence for phylogenies that place them as the sister group to the salps + pyrosomes Ascidian embryonic development, as reviewed in (Tsagkogeorga et al. 2009 ). However, a more the previous section, is highly conserved, being recent and complete phylogeny that includes a nearly indistinguishable between some distantly higher number of thaliacean species supports related species. In contrast, there is a greater a salp + doliolid clade with the pyrosomes in a diversity observed in later stages of development. basal position (Govindarajan et al. 2010 ), sug- The following section will focus on some of these gesting independent losses of the tadpole stage in differences in late larval development, such as pyrosomes and salps (Fig. 4.1 ). heterochronic shifts, the distinct modes of asex- A complete review of thaliacean development ual reproduction, and blastogenesis in the colo- was last compiled by Bone (1998 ). Pyrosome nial ascidians. Therefore, while the previous embryogenesis is direct, giving rise to a peculiar section relied heavily on studies using Ciona embryo termed “cyathozooid” (Huxley 1851 ). intestinalis and Halocynthia roretzi , the follow- The cyathozooid differentiates on top of a yolk ing section will mostly focus on comparative droplet and sprouts a stolon under the droplet, work done across several species to provide a which forms four buds by constriction. These comprehensive overview of late larval, early buds will develop into the four initial blastozo- adult, and blastogenic development in the oids (called tetrazooids in this specialized case) Tunicata.

[email protected] 176 A. Stolfi and F.D. Brown

A B

C D

Fig. 4.24 Metamorphosis in ascidians. (A ) Events of tunicates shows resemblance in their organization and metamorphosis highlighting players and pathways dorsolateral iodine-containing and peroxidase activity involved (© Biodidac, 2015. All Rights Reserved). (B ) regions of the thyroid gland (in black ) (Adapted from Representation of embryonic cell lineages that contribute Fredriksson et al. ( 1988 )). ( D ) Gene expression pattern to the juvenile and adult; note that many cells undergo cell supports homology between the endostyle of cephalo- death ( black crosses ) and that mostly endoderm and chordates and ascidians with the endostyle of the lamprey mesoderm contribute to the adult (From Brown et al. (Adapted from Ogasawara et al. (1999 , 2001 ), Hiruta (2012)). ( C ) Comparison of the endostyle of different et al. (2005 ), Kluge et al. (2005 ))

Late Larval Development A few hours after hatching (or immediately after hatching, in the case of some colonial spe- Developmental timing of embryogenesis and cies), ascidian larvae acquire the ability (or larval development differs between solitary and “competency”) to respond to settlement cues colonial ascidians. Solitary free-spawning from the environment (Figs. 4.24 and 4.25 ). ascidians develop rapidly: embryogenesis Competent larvae actively seek suitable locations occurs within 8–48 h of fertilization (e.g., 15 h and substrates on which to settle. The preferred in Ciona intestinalis , 9 h i n Molgula occidenta- cues will obviously vary according to the eco- lis ), and hatched larva then swims freely for sev- logical niche of the species in question. For many eral hours or even days before settlement. In solitary species, competent larvae seek protected contrast, colonial species that brood have rela- and dark places, often settling on rocks or hard tively longer periods of embryogenesis that can substrates with indirect sunlight. Biological cues, take anywhere from days to weeks (e.g., such as naturally occurring bacteria (Roberts 5–6 days in Botryllus schlosseri or 4–6 weeks in et al. 2007 ), and physical cues such as trauma, Botrylloides violaceus ), but present shorter lar- crowding, or stress (Degnan et al. 1997 ; Davidson val periods that only last between several min- and Swalla 2002 ) have been used to artifi cially utes to a couple of hours after they have been induce metamorphosis in different ascidian released (e.g., 1–2 h in Botryllus schlosseri ). species. However, in both solitary and colonial ascidi- Billie Swalla and collaborators found that ans, the larval period is nonfeeding and serves immune genes are highly expressed at the early primarily for dispersion. onset of metamorphosis in ascidians, suggesting

[email protected] 4 Tunicata 177

A D

E

B

F

C

Fig. 4.25 Larval release and settlement (metamorphosis) Attaching larva with extended ampullae; eye spot and tail of an unidentifi ed Botryllus sp. from Santa Marta, still present. ( E ) Oozooid just before the opening of the Colombia. ( A ) Ventral view of the brooding colony hours siphons; almost regressed tail and eye spot can still be before larval release; many round larvae can be seen in observed. ( F ) Completely differentiated oozooid; Am between the zooids. ( B ) Larvae right after hatching and ampulla, As atrial siphon, En endostyle, Ht heart, Os oral release. ( C ) Larval head that shows the protruding ampul- siphon. Scale bar equals 100 μm (Courtesy of L. Restrepo lae (or papillae) in the anterior region of the head. ( D ) (Universidad de los Andes))

[email protected] 178 A. Stolfi and F.D. Brown that the activation of the innate immune system their fi nal adult forms (Fig. 4.24 ), in order for the may occur at this stage. One intriguing explana- individual to occupy a different ecological niche. tion for this is that the innate immune system is Tissue-remodeling processes during animal deployed in sensing microbiota or bacterial bio- metamorphosis are known to occur mainly fi lm as cues for settlement. Another possibility through the activation of programmed cell death may be a direct involvement of immune cells in in larval tissues and the differentiation of juvenile later metamorphic events including remodeling and adult tissues. or phagocytosis of larval tissues (Davidson and There is variation in the degree of change that Swalla 2002 ). One question that remains to be results from metamorphosis among the different answered is whether the immune system response chordate taxa: it may be absent or minimal as in observed in these cases is specifi c to ascidian mammals or birds, subtle as in fi sh that undergo metamorphosis or whether it also plays a role in slight changes in fi ns and gut, more pronounced the metamorphic events of other tunicates or as in fl atfi sh or cephalochordates that undergo chordates. rotation of parts of their bodies, or extreme as in Most solitary species’ larvae have three adhe- frogs or tunicates that can lose entire larval struc- sive papillae (palps) in the anterior end of the tures and rapidly generate adult tissues and head that function in adhesion to the substrate organs. However, this is not the same “cata- and in responding to settlement cues. In response strophic” metamorphosis of some animals such to environmental stimuli, the papilla-associated as echinoderms (cf. Chapter 1 ), ectoprocts (cf. tissues (PATs) activate the expression of EGF Vol. 2, Chapter 11 ), or others, in which an entire signaling components (e.g., Hemps, Cornichon, adult body is practically generated de novo from Meta1) and thrombospondins. Thrombospondins a few cells, with little input from embryonic pat- are secreted extracellular proteins known to bind terning. Even in tunicates, the majority of adult several ligands and are involved in many cellular structures arise from primordia that have been processes, including tissue formation or repair specifi ed, patterned, and primed for differentia- (Eri et al. 1999 ; Davidson and Swalla 2001 ; tion during the course of embryogenesis and lar- Nakayama et al. 2001 ). val development. The activation and regulation of signaling Because of the great diversity of metamorphic pathways in PATs is followed by the release of transformations observed in the chordates, these noradrenaline/adrenaline (Kimura et al. 2003 ) were long thought to have evolved independently. and the activation of transduction pathways such In the following, studies of metamorphosis are as JNK and ERK in the larval CNS (Tarallo and described for several tunicates including those on Sordino 2004 ; Chambon et al. 2007 ). Further evi- ascidian model species, focusing on how such dence supporting signaling from the PATs to the studies have contributed to our understanding of CNS comes from the disruption of sensory vesi- the origins of metamorphosis in the chordates cle retraction (a key step in metamorphosis) in and deuterostomes in general. papilla-cut larvae of Ciona intestinalis Due to the dramatic transformation of the non- (Nakayama-Ishimura et al. 2009 ). In summary, feeding larvae into the sessile feeding juveniles environmental stimuli are sensed by the PAT, and adults of ascidians, these have been used pri- which then signals to the nervous system to trig- marily as experimental study models for meta- ger the onset of metamorphosis. morphosis in the tunicates. Few studies on metamorphosis have been carried out in larva- ceans, which do not feature a strong contrast Metamorphosis between the larval and adult forms, and, to our knowledge, no work has been done on metamor- Metamorphosis involves a radical transition in phosis in Thaliacea. While some species of dolio- the life cycle of indirect-developing organisms. lid thaliaceans present tailed larvae, all species of Larval body and behavior are remodeled into salps and pyrosomes have dispensed their larval

[email protected] 4 Tunicata 179 stage entirely and instead directly develop into an Castellano et al. 2014 ). Although the same inhib- oozooid from the sexual cycle to reproduce asex- itory effect of NO on metamorphosis was ually (Bone 1998 ). observed in Boltenia villosa and Cnemidocarpa fi nmarkiensis , the opposite was found in Key Steps in Ascidian Metamorphosis Herdmania momus , in which NO appears to pro- A rapid regression of the larval tail marks the mote metamorphosis (Ueda and Degnan 2013 ). onset of metamorphosis in the ascidians. We have These results point to complex roles for NO already discussed the key events that precede immediately prior to and after the onset of meta- metamorphosis, including the activation of a sen- morphosis, which may vary from species to sitive period that allows the larva to respond to species. external stimuli (i.e., competency), the secretion of adhesives by the palps to facilitate attachment Remodeling of the Larval Body to the substrate and signaling by the larval PAT Metamorphosis proceeds with the remodeling of and CNS (Fig. 4.24A ). Upon settlement, rapid the larval body and differentiation of the juvenile regression starting at the most posterior tip of the tissues and organs for another 1 or 2 days (approx. larval tail is mediated by a few distinct mecha- 36 h in Ciona intestinalis or 24–36 h in Botryllus nisms. These include apoptosis of tunic cells, schlosseri ) (Fig. 4.25 ). Some of the major events epidermis, notochord, tail muscle, and dorsal of remodeling in the larval head concern the neural tube (Chambon et al. 2002 ), contractile regression of the papillae (extended during settle- morphogenetic events of either epidermal or ment), the release of the larval tunic, the forma- muscle cells in the tail that generate the forces for tion of epidermal extensions termed ampullae, resorption (Cloney 1982 ), or migration of cells of and a 90° rotation of the larval AP body axis so the tail (e.g., endodermal strand and PGCs) into that the oral and atrial siphons of the juvenile the head, which has now become the entire body come to lie opposite to the site of attachment to of the animal (Shirae-Kurabayashi et al. 2006 ; the substrate and sensory vesicle regression Nakazawa et al. 2013 ). (Cloney 1982 ; Nakayama-Ishimura et al. 2009 ). The relative modularity of these different Mesenchyme cells begin substantial migration steps in metamorphosis is demonstrated by the within and across the epidermis of the differenti- striking phenotypes of certain genetic mutants. In ating juvenile, giving rise to their various deriva- Ciona intestinalis , two mutants showing meta- tives including blood and tunic cells. Additional morphic defects have been studied: (1) swimming undifferentiated endoderm and mesoderm pre- juvenile ( sj ), a transposon-generated cellulose cursors develop into the organs and musculature synthase mutant strain, and (2) tail regression of the adult. failed ( trf ), a naturally isolated mutant strain. In late embryonic and early larval develop- Both mutants show that several key steps in meta- ment, apoptosis appears to occur in many mesen- morphosis can occur even in the absence of larval chymal and CNS progenitors (Tarallo and tail regression (Nakayama-Ishimura et al. 2009 ). Sordino 2004 ), but these fi ndings are in stark These mutants clearly show that the developmen- contrast to lineage tracing experiments that fi nd tal processes of metamorphosis are indeed decou- invariant contributions of specifi c embryonic pled from each other. blastomeres to specifi c juvenile tissues and In Ciona intestinalis , tail regression is pre- organs. Apoptosis is observed in mesenchymal ceded by a decrease in endogenous nitric oxide cells that generally give rise to blood and muscle (NO) signaling (Fig. 4.24A ). While NO was in the juvenile, although a role for programmed found to delay metamorphosis, in part through cell death in these progenitors is unclear inhibition of caspase-dependent apoptosis (Fig. 4.24B ). In the larval CNS, two waves of (Comes et al. 2007 ), it was also found to promote apoptosis have been reported, one pre- ERK signaling during the acquisition of compe- metamorphic anterior-posterior wave that pro- tency (Bishop et al. 2001 ; Comes et al. 2007 ; gresses from the sensory vesicle in the head along

[email protected] 180 A. Stolfi and F.D. Brown the nerve cord to the tip of the tail and a second Branchiostoma fl oridae . Therefore, there is evi- metamorphic wave from posterior to anterior that dence to support the plesiomorphic nature of thy- follows regression (Tarallo and Sordino 2004 ). roid hormone metabolism in chordates. However, In contrast to the extensive cell death observed it was found that ascidian and cephalochordate in the nerve cord, the brain, and the motor gan- TRs are mainly activated by TRIAC, a TH deriv- glion, most of the anterior sensory vesicle and the ative that is also present in mammals but at much neck region actually contribute to the adult CNS lower concentrations. Furthermore, these inverte- (Tarallo and Sordino 2004 ; Horie et al. 2011 ). brate chordate receptors cannot bind T3, the most In summary, developmental processes of active vertebrate thyroid hormone (Carosa et al. ascidian metamorphosis largely redeploy the cel- 1998 ; Paris et al. 2008 ). In contrast to the high lular processes that are universal for any develop- sequence similarity between the DNA-binding ing system, such as apoptosis, proliferation, domains of all chordate TRs, there is poor migration, and differentiation. However, more sequence conservation in the ligand-binding work will be needed to understand exactly how domains of ascidian and cephalochordate TRs. they are regulated in the metamorphosing ascid- This further suggests that TRIAC, and not canon- ian juvenile. ical THs, is the active ligand of basal chordate TRs. Surprisingly, recent evidence suggests that Thyroid Hormone Metabolism both T3 and T4 regulate the metamorphosis of in Ascidian Metamorphosis sand dollars (Saito et al. 1998 ; Heyland and The series of events that occur during metamor- Hodin 2004 ), which raises the possibility that TH phosis require regionalized responses to systemic metabolism was involved in metamorphosis even signals. Therefore, one can hypothesize that a in the last common deuterostome ancestor. common and plesiomorphic endocrine mecha- Further studies are needed in the protostomes to nism evolved to regulate metamorphosis in all test for TH involvement in metamorphosis of ear- chordates. Triiodothyronine (T3) and its prohor- lier bilaterians or even metazoans. mone, thyroxine (T4), collectively referred to as thyroid hormones (THs), regulate basal metabo- The Endostyle: Precursor to the Thyroid lism in vertebrates and serve as the trigger of Gland metamorphosis in amphibians and fi sh (Furlow The endostyles of tunicates and cephalochordates and Neff 2006 ; Gomes et al. 2014 ). In these ver- have long been recognized as homologous to the tebrates, a peak in the level of T3 in the blood is thyroid gland of vertebrates, a connection fi rst associated with the onset of metamorphosis. Ever suggested by A. Dohrn at the Stazione Zoologica since the discovery that the thyroid gland, specifi - in Naples, Italy (Dohrn 1886 ). 10 In both ascidians cally the THs produced by its follicular cells, and cephalochordates, the endostyle forms a triggers metamorphosis in frogs (Gudernatsch groove along the ventral edge of the pharynx and 1912 ; Allen 1925 ), much research has been car- contains distinct functional “zones” of special- ried out to unravel the mechanism of TH action in ized cells. The general organization of functional other vertebrates (Holzer and Laudet 2013 ). zones in the endostyles of ascidians and cephalo- However, the role of THs in the metamorphosis chordates is essentially the same, although the of invertebrates is not well understood. overall number of zones may vary (Fig. 4.24C, Studies in ascidians and cephalochordates have revealed TH synthesis activity during meta- 10 Curiously, the Stazione Zoologica has an old tradition morphosis, and T4 inhibitor treatments were of TH research. It was the same place where in 1910 found to suppress metamorphosis of competent J.F. Gudernatsch made the initial observations that the larvae (D’Agati and Cammarata 2006 ). thyroid gland acted on the induction of metamorphosis, Furthermore, thyroid hormone receptor (TR) but as he did not trust his original results because the organs he used were not fresh, he published his fi ndings a homologs have been identifi ed both in couple of years later from research done in Prague Ciona intestinalis and the cephalochordate (Gudernatsch 1912 ; Brown and Cai 2007 ).

[email protected] 4 Tunicata 181

D ): the zones in the bilateral dorsal-most crest- employed to modulate heterochronic develop- like region of the endostyle closer to the pharynx ment in ascidians (i.e., adultation). In summary, concentrate iodine and contain peroxidase activ- ascidian metamorphosis relies on information- ity essential for thyroid-like activity, whereas the carrying systems such as the nervous system, the ventral-most inner groove region contains alter- endocrine system, and particularly the immune nating zones that serve either as support elements system, in order to coordinate a response to envi- or secrete mucoproteins that coat the pharynx to ronmental cues, which involves an intricate and facilitate capture of food particles (Bone et al. highly regulated redeployment of developmental 2003 ; Hiruta et al. 2005 ; Sasaki et al. 2003 ). processes to drastically reconfi gure the individu- Physiological similarities that support homology al’s body. of the endostyle and the thyroid gland are backed by conserved gene expression patterns in the developing endostyle. Orthologs of the vertebrate Asexual Reproduction: Modes thyroid markers TPO, TTF-1, Pax2/5/8, FoxE4, of Budding (Blastogenesis) FoxQ1, and FoxA have been shown to be expressed in the endostyles of both ascidians and The considerable variation observed in the devel- cephalochordates (Fig. 4.24D ; for details, see opmental modes of budding in the tunicates Hiruta et al. 2005 ). (ascidians and thaliaceans) is indicative of the The adult solitary ascidian endostyle of Ciona independent evolution of asexual reproduction intestinalis derives from an endostyle primordium and coloniality. Several evolutionary transitions formed by anterior and ventral endodermal cells to colonialism have been identifi ed in the tuni- in the head of the larva. Within a day of metamor- cates, which are accompanied by several mor- phosis, this primordium elongates and differenti- phological, developmental, and reproductive ates (Jacobs et al. 2008 ). The rapid and anticipatory changes, such as miniaturization of individual differentiation of the endostyle observed in C. body size, primarily asexual reproduction by intestinalis and other phlebobranchs that show a budding and brooding (Davidson et al. 2004 ). relatively delayed onset of metamorphosis sug- Fully developed larvae are released from the col- gests that TH production may be crucial for the ony to rapidly metamorphose and begin the asex- completion of metamorphic differentiation. ual cycle of development, otherwise known as However, in stolidobranchs (e.g., Herdmania budding or blastogenesis. In all modes of bud- momus), the endostyle differentiates only at the ding, the epidermis of new individuals derives end of metamorphosis along with all other juve- from preexisting epidermal progenitors, suggest- nile structures, arguing against an important role ing that mechanisms of epidermal cell replenish- for the endostyle in metamorphosis. ment may act locally at the sites of budding and This apparent incongruence might be disen- may not differ all that much between solitary and tangled if other cell types, before endostyle dif- colonial species. In contrast, the endodermal ferentiation, produced THs. Seemingly resolving (pharynx, gut, etc.) and mesodermal (muscle, this issue, D’Agati and Cammarata (2006 ) heart, etc.) derivatives originate from distinct pre- reported that THs are produced by mesenchyme cursor tissues in different species. In an attempt cells of mesodermal origin, completely unrelated to identify homologous developmental mecha- to the endostyle. These cells were found scattered nisms of budding, a comparison of bud precursor throughout the head, particularly behind the papil- tissues and cells involved in this process will be lae and at the basal and posterior region of the reviewed next for several species of ascidians. head of several phlebobranch ascidian species. Although further characterization of the spe- Budding in Ascidians cifi c roles and targets of THs is necessary, a com- In colonial stolidobranchs (specifi cally in parative study of TH larvae of different species Polyandrocarpa misakiensis , Botryllus schlosseri , may reveal whether endocrine mechanisms are and Symplegma reptans ), the endoderm- derived

[email protected] 182 A. Stolfi and F.D. Brown

AB CD

E F

Fig. 4.26 Blastogenic stages in the botryllid ascidians. Representation of the origin and contribution of the dou- ( A – D ) Developmental events of budding in Botrylloides ble vesicle stage that is recapitulated during embryogene- violaceus , stage is shown below to the left , and the time in sis and blastogenesis of botryllids (Adapted from Brown days is below to the right . ( E ) Representation of botryllid et al. (2007 , 2009 ) and Tiozzo et al. (2008 )) development. Arrowheads point to primary buds. ( F ) outer epithelium surrounding the branchial sac Stoner and Weissman 1996 ). Serial transplanta- (known as the peribranchial epithelium) contrib- tions of circulatory stem cells between different utes to the formation of mesendodermal tissues colonies showed that at least two distinct popula- of the peribranchial buds. Circulatory mesenchy- tions of progenitors occur in the circulatory cells mal cells or hemoblasts also contribute to these of the colony, including germ line and at least one buds. Evaginations of the lateral peribranchial somatic stem cell lines. It was also shown that a epithelium and epidermis form an early bud that minimum of 5.000 transplanted circulatory mes- resembles a double vesicle (Fig. 4.26 ). The inner enchyme cells was suffi cient to establish a new vesicle derived from the peribranchial epithelium genotype in the chimera. From this we can esti- undergoes a series of folds that will generate mate the occurrence of one stem cell for every most organ primordia. In between the double 5.000 mesenchymal cells in the blood11 (Laird vesicle, mesenchymal progenitors or hemoblasts et al. 2005 ). In certain species of stolidobranchs are thought to contribute to the differentiation of such as Botryllus and Botrylloides spp., the dou- several tissues and organs, including the pharynx, ble vesicle stage is also observed in vascular muscles, and mantle (Sabbadin et al. 1975 ). One buds. In these buds, the outer vesicle is formed well-documented example includes the participa- from the evagination of the ectodermal vascula- tion of hemoblasts in the differentiation of mus- ture (Fig. 4.26 ), which connects the zooids of the cle cells near the oral siphon primordium, colony, whereas the inner vesicle is formed from presumably induced by the oral epidermis the accumulation and differentiation of circula- (Sugino et al. 2007 ). tory mesenchymal cells within the outer vesicle An extensive amount of research demonstrat- instead originating from an epithelium (Brown ing somatic stem cell exchange in chimeric colo- nies of Botryllus schlosseri suggests that 11 circulatory mesenchymal progenitors are respon- A stem cell ratio of 1:5.000 in the blood is a relatively high number of circulatory stem cells; for comparison, in sible for the differentiation of zooids with distinct humans 1:5.000–10.000 bone marrow cells are stem cells genotypes in the same colony (Pancer et al. 1995 ; and 1:100.000 nucleated blood cells are stem cells.

[email protected] 4 Tunicata 183 et al. 2009 ). Three possible mechanisms have vary according to the amount of initial tissue and been suggested for the origin of circulatory mes- cell types they inherit. In stolonial budding of enchymal progenitors in the stolidobranchs: (1) Clavelina lepadiformis ( Brien and Brien-Gavage dedifferentiation of the atrial epithelium and 1927 ), budding stolons sprout radially from the transdifferentiation of these cells into new epithe- base of the zooid, specifi cally from the post- lia, e.g., intestinal epithelial cells (Kawamura and thoracic region containing the epicardium. Each Fujiwara 1995 ; Kawamura et al. 2008 ), (2) self- of the stolons undergoes an enlargement of the renewal of pluri- or multipotent cells in the circu- tip, which later detach from the parental zooid to latory system that may differentiate into distinct form independently developing stolonial buds cell fates (Freeman 1964 ; Laird and Weissman (Berrill 1951 ). In contrast, pyloric buds form 2004 ; Laird et al. 2005 ; Kürn et al. 2011 ; Brown between the thoracic and abdominal regions of and Swalla 2012 ), and (3) self-renewal of pluri- the zooid where the esophagus, stomach, and or multipotent cells in stem cell niches that circu- pyloric gland are located. This corresponds to the late and differentiate into distinct cell fates site where the epicardium of the zooid prevails (Voskoboynik et al. 2008 ; Rinkevich et al. 2013 ; (reviewed in Brown and Swalla 2012 ; Nakauchi Jeffery 2014 ). 1982 ). The thoracic region of the parental zooid In aplousobranchs that are exclusively colo- develops a new abdomen, whereas the parental nial, bud development is strictly associated with abdomen develops a new thoracic region. the presence of the epicardium, an endoderm- Therefore, this budding mode forms two daugh- derived epithelium that extends from the base of ter zooids from one single parental zooid. the pharynx to the pericardium, covering the In colonial phlebobranchs (specifi cally entire abdominal or postabdominal regions of the Perophora viridis and P. japonica ), buds form at zooid. Thus, colony propagation by budding is regular intervals along tubular structures also generally activated after settlement of the larva known as stolons, which connect the zooids of and specifi cally when the oozooid has differenti- the colony. The stolons consist of epidermal ated. However, the most precocious case of bud- diverticula that project from the base of each ding in colonial ascidians has been documented zooid and spread through the substrate, thus in the Holozoidae (i.e., Distaplia and expanding the reach of the colony. Both the sto- Hypsistozoa ), in which zooids can already differ- lons of Perophora and the colonial vasculature of entiate in the larval head of the swimming larva Botryllus are ectodermally derived but differ (see below) and also generate larval buds or pro- greatly in form and function. In Perophora , sto- buds that can self-replicate or differentiate into lons are covered by a thin tunic and grow external new zooids. Other mechanisms of bud formation tubular networks in direct contact with the sub- documented in the aplousobranchs include stro- strata, whereas the vasculature of Botryllus gen- bilation (specifi cally in Aplidium spp., Eudistoma erates a network of canals within a common and spp., and Pycnoclavella spp.), stolonial budding more solid tunic of the colony. Within the lumen (specifi cally in Clavelina lepadiformis ), and of a stolon, there is a single-layered epithelium pyloric budding (specifi cally in Diplosoma liste- derived from mesenchyme or from the septum rianum and Didemnum spp.). that partitions the stolon, allowing for blood fl ow Strobilation occurs by the constriction of in opposite directions (Freeman 1964 ; Mukai abdominal or postabdominal segments (Nakauchi et al. 1983 ). 1982 , 1986 ; reviewed in Brown and Swalla Since the earliest observations of budding in 2012 ). Each segment generates an independently Perophora , Kowalevsky ( 1874 ) already noted the developing bud that detaches from the parental contribution of the septum for the formation of zooid and begins to differentiate in neighboring stolonial buds; he wrote: “… die innere Haut der areas of the tunic. As each segment develops Knospe durch eine Verdickung der Scheidewand from distinct regions along the length of the der Wurzeln …”, i.e., “… the inner layer of the abdomen or postabdomen, bud primordia may bud is formed from a thickening of the stolonial

[email protected] 184 A. Stolfi and F.D. Brown septum ….” In fact, cuboidal cells in the budding zooids by stolonial budding. Therefore, a limited zone of the septum acquire a columnar shape and input of circulatory mesenchymal cells into the form a small vesicle on the upper part of the sep- bud occurs at a very early stage before detach- tum at a close distance (approx. 100–200 μm) to ment cuts off this mesenchymal cell source, in the tips of the stolons; these are the earliest rec- contrast to constant contribution of mesenchymal ognizable structures of a stolonial bud (Deviney cells to the stolonial buds. 1934 ; Kawamura et al. 2008 ). As the vesicle grows, several circulatory mes- Budding in Thaliaceans enchymal cells are incorporated, the epidermis of In colonial, pelagic thaliaceans, stolonial bud- the stolon protrudes, and organ primordia con- ding is used to rapidly multiply when environ- tinue to differentiate (Deviney 1934 ). Among the mental conditions are optimal for growth. Salps circulatory mesenchymal cells that were observed form chains of buds that emerge by terminal con- to integrate in the developing bud, a lymphocyte- striction of internal stolons in the oozooid, which like undifferentiated cell type – similar in cyto- undergo differentiation into long chains of blas- logical characteristics to the hemoblasts in tozooids that often remain attached. The stolons botryllids – was also seen. Some cellular charac- specifi cally form from the pericardium and are teristics of these undifferentiated lymphocyte- organized into epidermis, endoderm derived from like cells have also been observed in the budding tissues of the endostyle, and mesodermal tissues zone cells of the septum, including a large nucle- surrounding the heart. As the stolon grows, con- olus and a mitochondria-rich cytoplasm, demon- striction occurs in several terminal segments strating the possible mesenchymal origin of bud forming a series of blastozooids that will develop progenitor cells (Kawamura et al. 2008 ). and mature synchronously. As chains of blasto- Furthermore, the stem cell potential of these cells zooids differentiate, older segments of blastozo- alone was demonstrated by the rescue of irradi- oids in the most distal parts of the chain release ated colonies after transplantation of isolated sperm and fertilize the eggs of the younger, more lymphocytes (Freeman 1964 ). Therefore, this proximal blastozooids. However, blastozooids experiment unequivocally demonstrates that cir- can survive either as chains or as solitary forms culatory lymphocytes in phlebobranchs are nec- when detached from the chains (Brooks 1893 ). essary and suffi cient for budding. Attempts to In contrast, pyrosomes begin to bud by stolo- rescue irradiated botryllids after hemoblast trans- nial budding during embryonic development, plants have not been successfully achieved (Laird when the developing embryo (or cyathozooid, and Weissman 2004 ), which may be suggestive see above) gives rise to an initial colony of four of distinct stem cell potentiality in these two pop- blastozooids (the tetrazooids) (Bone 1998 ). ulations of undifferentiated circulatory mesen- These blastozooids continue to generate stolons, chymal cells. The homology between these two which undergo constriction to form additional stem cell types remains to be tested. blastozooids, thus expanding the colony. There is Another mode of budding known as “terminal no solitary phase, as pyrosomes are obligate budding” has been reported in Perophora japon- colonials. ica and includes the participation of the septum Doliolids have the most complex life cycles of and the incorporation of many circulatory mesen- all thaliaceans. Fertilization and embryonic devel- chymal cells at the tip of the stolon to form a ter- opment occur internally in the gonozooid. After minal stellate bud that serves as a propagule the larvae are released, a metamorphic-like event (Mukai et al. 1983 ). The stellate buds develop results in the development of the oozooid. The three to six epidermal projections, detach from oozooid next forms a posterior-dorsal stolonial the stolon, and drift away by currents. These primordium that will continue to grow and pelagic buds will eventually settle again; the stel- undergo stolonial budding by terminal constric- late projections of the pelagic bud turn into stolo- tion. These buds then migrate independently to nial primordia that grow and begin to form new the posterior ventral edge of the doliolid and align

[email protected] 4 Tunicata 185 to form three rows of buds, which will eventually heterochrony relative to the ancestral chordate, develop into specialized blastozooids. These with larval structures quickly differentiating and include trophozooids, specialized for feeding and adult structures being slower to differentiate. responsible for nurturing the entire colony, and Even this heterochrony is not absolute. As will be phorozooids which develop into the free-living explained below, in perhaps the majority of ascid- hermaphroditic sexual gonozooids. At this stage, ian species, this boundary between larva and adult the original oozooid ceases feeding, being nur- is not clearly defi ned (due to widespread occur- tured by its trophozooids, and becomes special- rence of adultation) and in some cases does not ized for navigating and transporting the colony exist at all (i.e., in direct developing species). (Bone 1998 ; Paffenhöfer and Köster 2011 ) . Ultimately, the evolution of the tunicates from a vermiform ancestor must have involved the acquisition of a biphasic developmental mode Life History Evolution through heterochrony, with later secondarily derived heterochronies (adultation and direct On the Biphasic Nature of Ascidian development). With the exception of the germ Development cells, the intestine, and the nervous system, this Of the major chordate-specifi c traits, the noto- partitioning between adult and larval components chord, dorsal hollow nerve cord, and somites correlates with their location in either the head (represented as a single iteration of paraxial (mostly adult) or tail (mostly larval). The germ mesoderm that gives rise to the larval muscles) cells are set aside prior to gastrulation and are are only present at the larval stage. In contrast, physically contained within the tail during the other chordate characters such as the endostyle, larval phase. Immediately upon settlement, the the pharyngeal gill slits, and the branchiomeric germ cells and surrounding endoderm cells muscles are fully formed only in the juvenile. migrate into the head during tail retraction, which Thus, the claim that metamorphosis “replaces” precedes all other events that occur during meta- the chordate body plan of the larva with an morphosis. Within the nervous system, the frac- ascidian- specifi c adult plan is not accurate at a tured rudiment of the adult CNS can be found detailed anatomical level. Either stage features a interspersed with the fully differentiated compo- unique subset of chordate-specifi c anatomical nents of the larval CNS along the AP axis (Dufour structures. In this way, the chordate body plan in et al. 2006 ; Horie et al. 2011 ). This is probably ascidians is biphasic in time, but ever (partially) due to historical contingency, the result of descent present through the life of the individual. The from an ancestor with a brain-like structure ante- continuity of the chordate body plan in tunicates rior to those compartments controlling homoch- is clearest in larvaceans, for obvious reasons, and ronic feeding and locomotion functions. in those ascidian species that undergo anural In contrast, the separation of adult feeding/ development or adultation (discussed below). respiration and larval locomotor functions in From an ontological point of view, this bipha- time mirrors the strict separation of the meso- sic nature does not mean the larval and adult derm into anterior (head: branchial/pharyngeal phases are discontinuous. On the contrary, there muscles, vasculature, heart, blood cells) and is a clear linear ontological progression of all the posterior (tail: swimming muscles, notochord) major adult structures from primordia that are compartments. Whether or not this spatial specifi ed and patterned during the course of compartmentalization was already partially embryogenesis and undergo morphogenesis and completed in the last common ancestor of tuni- differentiation even as the larva is swimming. As cates and vertebrates and whether this could such, it is more akin to a “swimming embryo” have represented a precondition for additional than a true larva. temporal compartmentalization are two of the The contrast between motile, nonfeeding larva more intriguing questions surrounding the ori- and sessile, fi lter-feeding adult is likely a result of gins of tunicates.

[email protected] 186 A. Stolfi and F.D. Brown

Fig. 4.27 Adultation of ascidian larvae. A composite of of adultation or premature differentiation of the juvenile several illustrations of works by N. J. Berrill on larvae stage while the larva is still swimming (From Berrill, of several species of ascidians showing varying degrees various works)

Fig. 4.28 Evolutionary reconstruction of ascidian adulta- irrespective of species-specifi c differences are marked with tion, brooding, and budding modes. Left : head anatomy of the “brood” symbol (gray i n d i v i d u a l w i t h black s p o t s r e p r e - ascidian swimming larvae shows different degrees of adulta- senting the embryos); note that Molgula citrina and other tion (see text for details). With some variation, solitary solitary styelids can brood in the peribranchial cavity. Right : ascidian larvae show a lower degree of adult differentiation schematic representation of the distinct budding modes in the head compared to colonial forms ( shaded box ). (bold a n d capital letters ) is shown; color code represents tis- Exceptional cases include (1) two cases of anural develop- sues of endodermal (yellow ) , m e s o d e r m a l (red ) , a n d e c t o - ment (direct development) in the stolidobranchs (squared dermal origins (blue ); the epicardium (important tissue for brackets ) ; ( 2 ) Molgula citrina that presents the highest budding, see text for details) of the Enterogona is shown in degree of adultation among solitary species (solid rectan- orange ; germ layer derivatives shown for every case of bud- gle ), a labeled cross section of the head that has been ding have only been labeled in Botrylloides leachi. Minus included to show detail of the internal anatomy (Adapted (− ) signs represent absence of that character, and question from Grave 1926 ); (3) the Diazonidae that present the low- marks (? ) show cases in which some controversy occurs in est degree of larval adultation among colonial species the literature regarding the trait (Modifi ed from Millar (dashed rectangle ) . Center : species that brood their embryos (1971 ) and Brown and Swalla (2012 ))

[email protected] 4 Tunicata 187

[email protected] 188 A. Stolfi and F.D. Brown

Adultation and Heterochrony Table 4.2 Egg sizes of primitive (non-adultative) and in Development adultative species of ascidians The precocious differentiation of adult structures Primitive (non-adultative) developers during the embryonic or larval stages is known as Species Size of egg (μm) adultation (Fig. 4.27 ). This heterochrony in Ascidiella aspersa 170 development, which blurs the line between larva Boltenia echinata 170 and adult, can be observed in different ascidian Ciona intestinalis 160 species, having likely evolved independently Diazona violacea 160 multiple times. Halocynthia pyriformis 260 Molgula manhattensis 110 Among ascidians, one sees variation in the Phallusia mammillata 160 extent of adultation (Fig. 4.28 ) , w i t h s o m e s p e - Styela canopus 150 cies, like Molgula citrina , showing limited adultation, while others, like Ecteinascidia tur- Adultative developers binata , showing extensive growth and differen- Species Size of egg (μm) tiation of adult structures, supported by highly Aplidium nordmanni 380 elaborated larval structures such as a tail with Clavelina lepadiformis 260 Dendrodoa grossularia 480 increased number of muscle cells, required for Distomus variolosus 590 the larva to swim in spite of its greater mass. Hypsistozoa fasmeriana 25* This increase in size and complexity of larval Molgula citrina 200 structures (muscle, notochord, motor ganglion) Stolonica socialis 720 is termed “caudalization.” In some species of Modifi ed from Berrill ( 1930 ). Top: egg size of species that colonial ascidians that undergo adultation, lar- develop into “primitive”-type larvae that do not undergo vae may already be carrying asexually repro- adultation. This is considered the ancestral condition, due ducing buds, or blastozooids, and thus constitute to the nearly identical embryos of distantly related species a “swimming colony.” It is tempting to specu- that develop in this manner (e.g., Ciona v s . Halocynthia ). Bottom: egg sizes and species whose larvae show adulta- late that this odd confi guration of an oozooid tion. Most adultative species develop from larger and larva carrying a blastozooid bud might be an more yolk-laden eggs. Asterisk (*) indicates the exception evolutionary forerunner to the acquisition of to this, the viviparous Hypsistozoa fasmeriana , whose “catastrophic” metamorphosis. A simple heter- eggs are entirely devoid of yolk and are nurtured by the parental colony through a placenta-like structure (see text, ochrony could result in the senescence of the Fig. 4.29 , and Brewin (1956 ) for more details larva/oozooid before the initiation of metamor- phosis and in the initiation of the adult stage by Such a capacity to “anticipate” metamorphosis the blastozooid.12 This would lead to the dis- could compensate for the negative consequences continuity between embryonic patterning and of delayed metamorphosis, which might happen adult body seen in echinoderms, entoprocts, if the right environmental cues for settlement are and other taxa. not present. While this conclusion was supported In one recent study, three non-adultative phle- by the correlation between length of competency bobranch ascidian species were found to be prone period and tendency toward adultation, the phy- to slight adultation if metamorphosis was artifi - logenetic signifi cance of this fi nding awaits a cially delayed, while there was no such regulative larger sampling of species. adultation seen in the three stolidobranch ascidi- Adultation goes hand in hand with increased ans assayed (Jacobs et al. 2008 ). This suggests egg size and ovoviviparity, in which eggs, that adultation may be dynamically regulated. embryos, and hatchlings are brooded inside the parental individual or colony (Table 4.2 ) . T h i s i s 12 This is technically what occurs in pyrosomes, in which probably due to an increased threat of predation the oozooid generates a few blastozooids and immedi- on the energetically rich, yolky eggs that are ately senesces. However, because pyrosomes are direct required to sustain the development of adultative developers and the oozooids and blastozooids share a common body plan, this is not seen as a catastrophic larvae. There is also a strong association between metamorphosis. adultation and coloniality, which has also arisen

[email protected] 4 Tunicata 189

A

B

Fig. 4.29 True viviparity in Hypsistozoa fasmeriana . (A ) ana embryo and associated ectotroph, an extra-embryonic Colonies of Hypsistozoa fasmeriana (Photograph by Paul tissue. Endodermal tubes connect the brood pouch with the Caiger) (© Paul Caiger, 2015. All Rights Reserved). (B ) lumen of the gut of the embryo, presumably to transfer food Illustrations of H. fasmeriana colonial and embryonic and nutrients from the parent to the embryo. Bottom right : development. Top left : y o u n g c o l o n y . Top middle : senescing large larva with numerous buds in an advanced stage of dif- colony, preparing for shedding of larvae. Top right : anatomy ferentiation (a swimming colony) (Illustrations in ( B ) are of H. fasmeriana zooid showing embryo being brooded in reproduced with permission from Brewin (1956): h t t p : / / j c s . attached brood pouch. Bottom left : s t r u c t u r e o f H. fasmeri- biologists.org/content/s3-97/39/435.short )

[email protected] 190 A. Stolfi and F.D. Brown independently in different ascidian families individual of a colony derived from sexual repro- (Fig. 4.28 ). There are, however, some notable duction. The oozooid then activates asexual bud- exceptions, such as the adultative solitary Molgula ding cycles (i.e., blastogenesis) to generate citrina or non-adultative colonial Diazona s p p . blastozooids, individuals derived from asexual Possible explanations for the association between generations or blastogenesis (Fig. 4.26 ) . coloniality and adultation are discussed below. Although oozooids and blastozooids show cer- Some adultative species appear to have tain anatomical differences, such as different evolved true viviparity, with secondary reduction arrangement of the pharyngeal slits of the bran- in egg size since nutrients now can be passed chial sac, the general organization and body plan from parent to daughter during development. are maintained. Also, expression patterns of This appears to be the case between the closely many developmental genes that are expressed in related Botryllus a n d Botrylloides g e n e r a ( B e r r i l l embryogenesis are redeployed during the devel- 1935 ). A more extreme example of viviparity is opment of the bud. For example, Botryllus found in Hypsistozoa fasmeriana , which has schlosseri o r t h o l o g s o f Six , Eya , a n d FoxI ( t r a n - extremely small eggs (approx. 25 μm diameter), scription factors associated with the vertebrate whose growth into large “swimming colonies” or placode network) expressed in anterior and pos- compound larvae, each comprising an oozooid terior placodal regions in the metamorphosing plus several blastozooids in various stages of larva are also expressed in corresponding ante- development, depends on transfer of food through rior and posterior regions of the developing bud a connection between the colony’s brood pouch (Tiozzo et al. 2005 ; G a s p a r i n i e t a l 2013 ) . T h e and the embryo’s gut (Fig. 4.29 ; Brewin 1956 ). anterior placode region will eventually differen- tiate into cells of the oral siphon, ciliated duct, Maximum Direct Development and cerebral ganglion, whereas the posterior There are some very rare examples in which placode region will differentiate into cells of the direct development and adultation have been atrial siphon of both oozooid and blastozooid, combined in a process of “direct adultation,” also similar to their development in solitary ascidians called “maximum direct development,” in which (Mazet et al. 2005 ) and larvaceans (Bassham and early embryonic development may truly bypass Postlethwait 2005 ) . the formation of a larva (urodele or anural) prior Contrary to the reported effects of thyroid hor- to differentiation of adult structures. The maxi- mone inhibitors on Ciona intestinalis j u v e n i l e s mum direct developers potentially include during metamorphosis, the thyroid hormone inhib- Polycarpa tinctor (Millar 1962 ), Pelonaia corru- itor thiourea was found to accelerate bud develop- gata (Fig. 4.28 ; Millar 1954 ), and Molgula paci- ment and inhibit stolon growth in the colonial fi ca (Young et al. 1988 ; Bates 2002 ), although in phlebobranch Perophora orientalis. C o n v e r s e l y , all these cases a detailed stage-by-stage descrip- the thyroid derivative thyroxine inhibited bud tion of development is lacking and we are left development but induced stolon growth (Fukumoto without any certainty about how development in 1971 ). These results provide an exciting opportu- these species occurs. It is hypothesized that the nity to compare modular effects of endocrine sig- various species with anural development fi nd naling during zooid differentiation between themselves at different stages of an evolutionary metamorphosis and blastogenesis. However, the transition toward maximum direct development. expression patterns of the main thyroid pathway genes remain to be elucidated in ascidians. Embryogenesis vs. Blastogenesis Future work is needed to compare develop- Only after metamorphosis do ascidians begin to mental mechanisms of the earliest stages of feed and grow. In solitary species, the settled embryogenesis vs. blastogenesis, in order to juvenile grows and matures as a single individ- identify similarities and differences in patterning ual. In colonial species, the settled juvenile is events between bud progenitors and early blasto- called the oozooid (Fig. 4.25 ), i.e., the founding meres. In sum, the existence of dual developmental

[email protected] 4 Tunicata 191 trajectories (sexual vs. asexual) to generate the regenerative and propagative potentials (Davidson same body form in a single species provides an et al. 2004 ; Brown and Swalla 2012 ) . excellent opportunity to address questions about The current phylogeny of Deuterostomia developmental pathway co-option in evolution. supports the idea that coloniality in this group has evolved independently multiple times but Evolution of Coloniality that the ability to regenerate was likely present Within the deuterostomes, the hemichordate in the ancestral deuterostome. In the clade com- pterobranchs (Chapter 2) and many species of prised by the questionable Xenacoelomorpha + tunicates have a colonial stage in their life cycle. Ambulacraria (Fig. 4.1), coloniality only occurs Colonies are defi ned as aggregates of individuals in the pterobranchs, but regenerative potential that are “connected together, either by living has also been documented for the harrimaniids, extensions of their bodies, or by material that they ptychoderids, echinoderms, and xenacoelo- have secreted” (Barrington 1967 ) . I n t h e t u n i c a t e morphs (see Chapters 1 a n d 2 ; Vol. 1, Chapter literature, colonies with living extensions that 9 ). Among the Chordata, ascidians are the only interconnect the individuals have been referred to subphylum that show convergent evolution of as “social forms,” whereas colonies that secrete coloniality in several species that also have full material to embed the individuals have been regenerative potential (i.e., whole body regen- referred to as compound forms (Milne- Edwards eration). Regenerative potential, albeit much 1841 ; Pérez-Portela et al. 2009 ) . I n d e u t e r o - more modest (i.e., tissue, organ, and structure stomes, the colonial life stage is generally initi- regeneration), has also been reported in solitary ated after metamorphosis. Asexual reproduction ascidians, cephalochordates (Somorjai et al. of the colony occurs by clonally generating new 2012 ), some Craniata/Vertebrata, and, most individuals from preexisting juvenile or adult tis- remarkably, platyhelminths (see Vol. 2, Chapter sues. Curiously, individuals of most species that 4 for a detailed account on regeneration in the form colonies are (1) smaller in size than their latter). Therefore, we can speculate that devel- closely related solitary species clades, (2) able to opmental mechanisms of asexual reproduction brood and release larvae with a high degree of and regeneration preceded the establishment of adultation (see above), and (3) imbued with highly colonial lifestyles.

AB C

Fig. 4.30 Vertical transmission mechanisms of photo- not contain larval tunic. ( C ) Prochloron attached and symbionts in didemnid ascidian larvae. ( A ) Prochloron inserted into the rastrum, a T-shaped extension of larval cyanobacteria (green ) embedded within the larval tunic. ectoderm in the posterior-dorsal region of the larval head ( B ) Prochloron attached to hairlike projections and tunic (Modifi ed from Kott (2001 )) folds in the posterior region of the larval head that does

[email protected] 192 A. Stolfi and F.D. Brown

Conceptual arguments on life history trade- rine-like amino acids (MAAs) (Kühl et al. 2012 ) offs between individual forms and cooperative and (2) cyanobactins (toxins) to protect the entities such as colonies are starting to emerge ascidians from predators or parasites (Donia (see recent published books edited by Calcott and et al. 2011 ). In sum, didemnid ascidians provide Sterelny 2011 ; {Inserting} Bouchard and an opportunity to study the effects of symbiosis Huneman 2013 ). However, we are far from in development and the recurrence and particular understanding how selective forces act to co-opt adaptations of this biological interaction across developmental modules in the evolution of such several species. higher-level organized forms. To begin to address this question, it is necessary to fi rst identify the developmental processes associated with OPEN QUESTIONS coloniality. Tunicate/Chordate Origins Photosymbiosis in Colonial Didemnid • What was the tunicate ancestor like? Ascidians • How did it evolve from the olfactorean Last, but not least, symbiosis occurs between an ancestor? oxygenic photosynthetic cyanobacteria and sev- • Which (if existing) are the autapomorphies of eral tropical colonial ascidian species of the Olfactores (Tunicata + Vertebrata)? Didemnidae. The symbiont Prochloron spp. (Kühl et al. 2012 ) is found associated, almost Tail Loss in the Molgulids exclusively, with four different genera of didem- • How much do the multiple, independent cases nids: Diplosoma , Lissoclinum , Didemnum , and of evolutionary loss of the larval tail structures Trididemnum (Yokobori et al. 2006 ). Prochloron in the Molgulidae have in common at the is usually present in the outer surface of the tunic, molecular or cellular levels? within the common cloacal system of the colony, • Is there a common point of evolutionary con- or embedded in the tunic as either free-living vergence for the different anural species? cells or intracellularly in tunic cells. Prochloron • Did those species exhibiting direct develop- is transmitted vertically from the mother colony ment evolve from indirectly developing tail- to the daughter colonies by three distinct mecha- less species? nisms: (1) by direct attachment into a sticky pos- terior region of the larval head before the larvae Adultation of Colonial Ascidian Larvae are released, (2) by encapsulating into a special- • How are the much bigger larvae of certain ized pouch-like organ in the posterior head of the colonial ascidians formed? larvae, or (3) by the transmission of Prochloron - • How have they evolved presumably from a containing tunic cells with intracellular simple larva of the solitary kind? Prochloron from the mother colony to the devel- • Do the mechanisms involved in patterning oping larval tunics while brooding (Fig. 4.30 ; and growth of these larger embryos resemble Kott 2001 ; Yokobori et al. 2006 ; Hirose and in any way perhaps those lost in the, presum- Hirose 2007 ). ably, drastically reduced ascidian ancestor? The evolution of vertical transmission mecha- nisms of the host species suggests an obligate Evolution of Larvaceans symbiotic interaction in which the didemnid • Did Appendicularia evolve from an ascidian- ascidians provide shelter for the symbiont and like stock or did they evolve directly from the prochloron provides protection. Two protec- free-swimming, pelagic ancestors? tion mechanisms have been suggested: (1) UV • If they evolved from a sessile ascidian ances- screening mechanisms by producing mycospo- tor, what were the embryological and gene

[email protected] 4 Tunicata 193

regulatory hurdles that needed to be overcome quently lost from the less regenerative solitary in order for larvaceans to break free from the ascidians? constraints of the typical solitary ascidian embryo archetype? Gene Expression and Function • What are the mechanisms that regulate larva- • Why has the Hox complex fragmented or its cean metamorphosis, and how conserved are regulation become less cohesive in tunicates? these with the rest of the tunicates? • Has the presumed loss of segmentation in tunicate body plans relaxed the constraints on Evolution and Development of Thaliaceans Hox cluster integrity? • What are the relationships among the thali- • Do the Hox genes play a more peripheral role acean orders (pyrosomes, salps, and in gene regulatory networks governing embry- doliolids? onic development in the tunicates, relative to • How did complex life cycles evolve in this their role in other phyla? class? • What were the genes potentially co-opted by • How far do metamorphic events in the thali- the neural plate border cells in the craniate aceans resemble those of other tunicates? ancestor that might have modifi ed these into • What are the underlying molecular and genetic bona fi de neural crest cells? mechanisms of thaliacean development? • How are metamorphic processes of apoptosis, proliferation, migration, and differentiation Regeneration and Coloniality regulated in the tunicates, and to what extent • How closely do the mechanisms of blastozo- are these regulatory mechanisms redeployed oid patterning resemble those that pattern the in ascidians, larvaceans, and thaliaceans? embryo/oozooid? • How is the formation and regression of zooids Developmental System Drift modulated in colonial species? • How pervasive is developmental system drift • What are the signaling mechanisms that among the tunicates, especially those with orchestrate the synchronized differentiation of very conservative, simplifi ed embryos like blastozooids in certain species? most solitary ascidians? • When do bud progenitors fi rst reorganize • Are the genomes of tunicates rapidly evolving themselves into a new individual with a body in order to adapt to the various other environ- plan that corresponds to an embryonic stage, mental and ecological concerns faced by and which stage would that be? them, such as feeding, temperature, salinity, • Is there an analogous “phylotypic” stage for predation, reproduction, etc.? blastogenesis? • If so, do species-specifi c cis /trans compensa- • How closely do the mechanisms of blastozo- tory mechanisms evolve in order to maintain oid patterning resemble those that pattern the the same exact embryo? embryo/oozooid? • Is this invariant embryo required due to the • Given that of embryonic patterning are set up lack of regulative developmental processes by localized maternal determinants that pre- that might allow for the ontogenesis of a via- sumably would be very diffi cult to redeploy in ble sea squirt via a very different embryo? a blastogenic bud, at what point(s) in the gene • Is embryonic development more variable regulatory network(s) do the embryonic and among the colonial species (not related to the blastozooid programs converge? separate issue of adultation) because these • Was the last common ancestor of all tunicates have evolved/retained greater developmental imbued with the remarkable regenerative plasticity, as evidenced by their blastogenic potential of colonial tunicates and subse- mode of reproduction?

[email protected] 194 A. Stolfi and F.D. Brown

Prochloron Photosymbiosis Barrington EJW (1967) Invertebrate structure and function. • Why is prochloron present only in some spe- Thomas Nelson Sons Ltd., London, p 550, First Edit Bassham S, Postlethwait JH (2005) The evolutionary his- cies of didemnids? tory of placodes: a molecular genetic investigation of • How did host-symbiont interaction become the larvacean urochordate Oikopleura dioica . obligatory? Development 132:4259–4272 • How does obligate symbiosis trigger the evo- Bates WR (1995) Direct development in the ascidian Molgula retortiformis (Verrill, 1871). Biol Bull lution of specifi c mechanisms and organs for 188:16–22 vertical transmission in the host? B a t e s W R ( 2 0 0 2 ) T h e p h y l o g e n e t i c s i g n i fi cance of maxi- • Does intracellular prochloron within tunic mum direct development in the ascidian, Molgula cells represent a causative event of an older pacifi ca . Invertebr Reprod Dev 41:185–192 Bates WR, Mallett JE (1991) Anural development of the symbiotic interaction of particular lineages? ascidian Molgula pacifi ca (Huntsman). Can J Zool • Did prochloron diversify after symbiosis with 69:618–627 the host? Bateson W (1884) Memoirs: the early stages in the devel- opment of Balanoglossus (sp. incert.). Q J Microsc Sci 2:208–236 Acknowledgments Many thanks to Andreas Wanninger Bateson W (1886) Continued account of the later stages in and Ivan Dias for comments on the manuscript and to the development of Balanoglossus kowalevskii , and of Florian Razy-Krajka for thoughtful discussions. Thanks the morphology of the enteropneusta. Q J Microsc Sci to Leda Restrepo, Gabriela Agurto, Joanna Greer, and 26:511–534 Paul Caiger for contributing with some fi gures. We would Beh J, Shi W, Levine M, Davidson B, Christiaen L (2007) also like to thank all the authors who have granted us per- FoxF is essential for FGF-induced migration of heart mission to modify and adapt fi gures from their excellent progenitor cells in the ascidian Ciona intestinalis . papers. This work was supported by an “Apoio aos Novos Development 134:3297–3305 Docentes” Grant from Universidade de São Paulo, an Berrill NJ (1928) The identifi cation and validity of certain Assistant Professor Grant from Universidad de los Andes species of ascidians. J Mar Biol Assoc U K N Ser and a Prometeo Research Award from the Secretaría 15:159–175 Nacional de Ciencia, Tecnología e Innovación (Senescyt- Berrill NJ (1930) Studies in tunicate development. Part Ecuador) to F. D. B., and by a National Science Foundation I. General physiology of development of simple ascid- Postdoctoral Research Fellowship (under grant NSF- ians. Phil Trans R Soc Lond Ser B 218:37–78, 1161835) to A. S. Containing Papers of a Biological Character Berrill NJ (1931) Studies in tunicate development. Part II. Abbreviation of development in the Molgulidae. Phil Trans R Soc Lond Ser B 219:281–346, Containing References Papers of a Biological Character Berrill NJ (1932) The mosaic development of the ascidian Abbott CL, Ebert D, Tabata A, Therriault TW (2011) egg. Biol Bull 63:381–386 Twelve microsatellite markers in the invasive tunicate, Berrill NJ (1935a) Studies in tunicate development. Part Didemnum vexillum , isolated from low genome cover- IV. Asexual reproduction. Philos Trans R Soc Lond B age 454 pyrosequencing reads. Conserv Genet Resour Biol Sci 225:327–379 3:79–81 Berrill NJ (1935b) Studies in tunicate development. Part Abitua PB, Wagner E, Navarrete IA, Levine M (2012) III. Differential retardation and acceleration. Philos Identifi cation of a rudimentary neural crest in a non- Trans R Soc Lond B Biol Sci 225:255–326 vertebrate chordate. Nature 492:104–107 Berrill NJ (1936) Studies in tunicate development. Part Alldredge AL (1977) House morphology and mecha- V. The evolution and classifi cation of ascidians. Philos nisms of feeding in the Oikopleuridae (Tunicata, Trans R Soc Lond B Biol Sci 226:43–70 Appendicularia). J Zool 181:175–188 Berrill NJ (1947a) The developmental cycle of Alldredge A, Madin L (1982) Pelagic tunicates: unique Botrylloides . Q J Microsc Sci 3:393–407 herbivores in the marine plankton. Bioscience 32: Berrill NJ (1947b) The structure, development and bud- 655–663 ding of the ascidian, Eudistoma . J Morphol 81: Allen BM (1925) The effects of extirpation of the thyroid 269–281 and pituitary glands upon the limb development of Berrill NJ (1947c) The structure, tadpole and budding of anurans. J Exp Zool 42:13–30 the ascidian Pycnoclavella aurilucens Garstang. J Mar Amemiya CT, Powers TP, Prohaska SJ, Grimwood J, Biol Assoc U K 27:245–251 Schmutz J, Dickson M, Miyake T, Schoenborn MA, Berrill NJ (1948a) Budding and the reproductive cycle of Myers RM, Ruddle FH (2010) Complete HOX cluster Distaplia . Q J Microsc Sci 3:253–289 characterization of the coelacanth provides further evi- Berrill NJ (1948b) The development, morphology and dence for slow evolution of its genome. Proc Natl budding of the ascidian Diazona . J Mar Biol Assoc U Acad Sci 107:3622–3627 K 27:389–399

[email protected] 4 Tunicata 195

Berrill NJ (1948c) The gonads, larvae, and budding of the Brown FD, Tiozzo S, Roux MM, Ishizuka K, Swalla BJ, polystyelid ascidians Stolonica and Distomus . J Mar De Tomaso AW (2009) Early lineage specifi cation of Biol Assoc U K 27:633–650 long-lived germline precursors in the colonial ascidian Berrill NJ (1948d) The nature of the ascidian tadpole, Botryllus schlosseri . Development 136:3485–3494 with reference to Boltenia echinata . J Morphol Calcott B, Sterelny K (eds) (2011) The major transitions 82:269–285 in evolution revisited. The MIT Press, Cambridge, Berrill NJ (1948e) Structure, tadpole and bud formation in MA, 319 the ascidian Archidistoma . J Mar Biol Assoc U K Cameron CB, Garey JR, Swalla BJ (2000) Evolution of 27:380–388 the chordate body plan: new insights from phyloge- Berrill NJ (1950a) Budding and development in Salpa . J netic analyses of deuterostome phyla. Proc Natl Acad Morphol 87:553–606 Sci U S A 97(9):4469–4474 Berrill NJ (1950b) Budding in Pyrosoma . J Morphol Carosa E, Fanelli A, Ulisse S, Di Lauro R, Rall JE, Jannini 87:537–552 EA (1998) Ciona intestinalis nuclear receptor 1: a Berrill NJ (1951) Regeneration and budding in tunicates. member of steroid/thyroid hormone receptor family. Biol Rev 26:456–475 Proc Natl Acad Sci U S A 95(19):11152–11157 Berrill NJ (1955) The origin of vertebrates. Clarendon, Castellano I, Ercolesi E, Palumbo A (2014) Nitric oxide Oxford, p 257 affects ERK signaling through down-regulation of Bertrand V, Hudson C, Caillol D, Popovici C, Lemaire P MAP kinase phosphatase levels during larval develop- (2003) Neural tissue in ascidian embryos is induced by ment of the ascidian Ciona intestinalis . PLoS One FGF9/16/20, acting via a combination of maternal 9:e102907 GATA and Ets transcription factors. Cell Cavey MJ (1982) Myogenic events in compound ascidian 115:615–627 larvae. Am Zool 22:807–815 Bishop CD, Bates WR, Brandhorst BP (2001) Regulation Cavey MJ, Cloney RA (1976) Ultrastructure and differen- of metamorphosis in ascidians involves NO/cGMP tiation of ascidian muscle. Cell Tissue Res signaling and HSP90. J Exp Zool 289:374–384 174:289–313 Bone Q (1992) On the locomotion of ascidian tadpole Chabry L (1887). Embryologie normale et teratologique larvae. J Mar Biol Assoc U K 72:161–186 des Ascidie. In: Alcan F (ed). Paris, p 161 Bone Q (1998) The biology of pelagic tunicates. Oxford Chambon J-P, Soule J, Pomies P, Fort P, Sahuquet A, University Press, Oxford Alexandre D, Mangeat P-H, Baghdiguian S (2002) Bone Q, Carre C, Chang P (2003) Tunicate feeding fi lters. Tail regression in Ciona intestinalis (Prochordate) J Mar Biol Assoc UK 83:907–919 involves a Caspase-dependent apoptosis event associ- Bouchard F, Huneman P (2013) From groups to individu- ated with ERK activation. Development als: evolution and emerging individuality. MIT Press, 129:3105–3114 Cambridge, MA Chambon J-P, Nakayama A, Takamura K, McDougall A, Bouquet J, Spriet E, Troedsson C, Otterrå H, Chourrout Satoh N (2007) ERK-and JNK-signalling regulate D, Thompson EM (2009) Culture optimization for the gene networks that stimulate metamorphosis and emergent zooplanktonic model organism Oikopleura apoptosis in tail tissues of ascidian tadpoles. dioica. J Plankton Res 31:359–370 Development 134:1203–1219 Brewin BI (1956) The Growth and Development of a Chen JS, San Pedro M, Zeller RW (2011) miR-124 func- Viviparous Compound Ascidian, Hypsistozoa fasme- tion during Ciona intestinalis neuronal development riana. Q J Microsc Sci 97:435–454 includes extensive interaction with the notch signaling Brien P, Brien-Gavage E (1927) Contribution a l’etude de pathway. Development 138:4943–4953 la blastogenese des Tuniciers. III. Bourgeonnement de Christiaen L, Davidson B, Kawashima T, Powell W, Nolla Clavelina lepadiformis Müller. Rec Inst Zool Torley- H, Vranizan K, Levine M (2008) The transcription/ Rousseau B31–81 migration interface in heart precursors of Ciona intes- Brooks WK (1893) The genus salpa. The Johns Hopkins tinalis . Science 320:1349 Press, Baltimore, p 523 Christiaen L, Stolfi A, Davidson B, Levine M (2009) Brown DD, Cai L (2007) Amphibian metamorphosis. Dev Spatio-temporal intersection of Lhx3 and Tbx6 defi nes Biol 306(1):20–33 the cardiac fi eld through synergistic activation of Brown FD, Swalla BJ (2007) Vasa expression in a colo- Mesp . Dev Biol 328:552–560 nial ascidian, Botrylloides violaceus . Evol Dev Cloney RA (1982) Ascidian larvae and the events of meta- 9:165–177 morphosis. Am Zool 22:817–826 Brown FD, Swalla BJ (2012) Evolution and development Cole AG, Meinertzhagen IA (2004) The central nervous of budding by stem cells: ascidian coloniality as a case system of the ascidian larva: mitotic history of cells study. Dev Biol 369:151–162 forming the neural tube in late embryonic Ciona intes- Brown CD, Johnson DS, Sidow A (2007) Functional tinalis . Dev Biol 271:239–262 architecture and evolution of transcriptional elements Collins AG, Valentine JW (2001) Defi ning phyla: evolu- that drive gene coexpression. Science 317: tionary pathways to metazoan body plans. Evol Dev 1557–1560 3:432–442 B r o w n F D , P r e n d e r g a s t A , S w a l l a B J ( 2 0 0 8 ) M a n i s b u t a Comes S, Locascio A, Silvestre F, d’Ischia M, Russo GL, worm: chordate origins. Genesis 46:605–613 Tosti E, Branno M, Palumbo A (2007) Regulatory

[email protected] 196 A. Stolfi and F.D. Brown

roles of nitric oxide during larval development and Delsuc F, Brinkmann H, Chourrout D, Philippe H (2006) metamorphosis in Ciona intestinalis. Dev Biol Tunicates and not cephalochordates are the closest liv- 306:772–784 ing relatives of vertebrates. Nature 439:965–968 Conklin EG (1905a) Mosaic development in ascidian Delsuc F, Tsagkogeorga G, Lartillot N, Philippe H (2008) eggs. J Exp Zool 2:145–223 Additional molecular support for the new chordate C o n k l i n E G ( 1 9 0 5 b ) O r g a n - f o r m i n g s u b s t a n c e s i n t h e phylogeny. Genesis 46:592–604 eggs of ascidians. Biol Bull 8:205 Deng W, Nies F, Feuer A, Bočina I, Oliver D, Jiang D C o n k l i n E G ( 1 9 0 5 c ) T h e o r g a n i z a t i o n a n d c e l l - l i n e a g e o f (2013) Anion translocation through an Slc26 trans- the ascidian egg. Academy of Natural Sciences of porter mediates lumen expansion during tubulogene- Philadelphia, Philadelphia, p 175 sis. Proc Natl Acad Sci 110:14972–14977 Cooley J, Whitaker S, Sweeney S, Fraser S, Davidson B Denker E, Bočina I, Jiang D (2013) Tubulogenesis in a (2011) Cytoskeletal polarity mediates localized induc- simple cell cord requires the formation of bi-apical tion of the heart progenitor lineage. Nat Cell Biol cells through two discrete Par domains. Development 13:952–957 140:2985–2996 Corbo JC, Levine M, Zeller RW (1997) Characterization Denoeud F, Henriet S, Mungpakdee S, Aury J-M, Da of a notochord-specifi c enhancer from the Brachyury Silva C, Brinkmann H, Mikhaleva J, Olsen LC, Jubin promoter region of the ascidian, Ciona intestinalis . C, Cañestro C (2010) Plasticity of animal genome Development 124:589–602 architecture unmasked by rapid evolution of a pelagic D’Agati P, Cammarata M (2006) Comparative analysis of tunicate. Science 330:1381–1385 thyroxine distribution in ascidian larvae. Cell Tissue Deviney EM (1934) The behavior of isolated pieces of Res 323:529–535 ascidian (Perophora viridis) stolon as compared with Darras S, Nishida H (2001) The BMP/CHORDIN antago- ordinary budding. J Elisha Mitchell Sci Soc nism controls sensory pigment cell specifi cation and 49(2):185–224 differentiation in the ascidian embryo. Dev Biol D o h r n A ( 1 8 8 6 ) T h y r o i d e a b e i Petromyzon , Amphioxus 236:271–288 und den Tunicaten. Mitt Zool Sta Neapel 6:49–92 David B, Mooi R (2014) How Hox genes can shed light Dong B, Horie T, Denker E, Kusakabe T, Tsuda M, Smith on the place of echinoderms among the deuterostomes. WC, Jiang D (2009) Tube formation by complex cel- EvoDevo 5:22 lular processes in Ciona intestinalis notochord. Dev Davidson B, Levine M (2003) Evolutionary origins of the Biol 330:237–249 vertebrate heart: specifi cation of the cardiac lineage in Dong B, Deng W, Jiang D (2011) Distinct cytoskeleton Ciona intestinalis . Proc Natl Acad Sci populations and extensive crosstalk control 100:11469–11473 Ciona notochord tubulogenesis. Development 138: Davidson B, Swalla BJ (2001) Isolation of genes involved 1631–1641 in ascidian metamorphosis: epidermal growth factor Donia MS, Fricke WF, Partensky F, Cox J, Elshahawi SI, signaling and metamorphic competence. Dev Genes White JR, Phillippy AM, Schatz MC, Piel J, Haygood Evol 211:190–194 MG (2011) Complex microbiome underlying second- Davidson B, Swalla BJ (2002) A molecular analysis of ary and primary metabolism in the tunicate- Prochloron ascidian metamorphosis reveals activation of an innate symbiosis. Proc Natl Acad Sci 108:E1423–E1432 immune response. Development 129:4739–4751 Dufour HD, Chettouh Z, Deyts C, de Rosa R, Goridis C, Davidson B, Jacobs M, Swalla B (2004) The individual as Joly J-S, Brunet J-F (2006) Precraniate origin of cra- a module: metazoan evolution and coloniality. In: nial motoneurons. Proc Natl Acad Sci Schlosser G, Wagner GP (eds.) Modularity in develop- 103:8727–8732 ment and evolution. University of Chicago Press, Eri R, Arnold JM, Hinman VF, Green KM, Jones MK, Chicago, pp 443–465 Degnan BM, Lavin MF (1999) Hemps, a novel EGF- Davidson B, Shi W, Levine M (2005) Uncoupling heart like protein, plays a central role in ascidian metamor- cell specifi cation and migration in the simple chordate phosis. Development 126:5809–5818 Ciona intestinalis . Development 132:4811–4818 Ferrier DE, Holland PW (2002) Ciona intestinalis Davidson B, Shi W, Beh J, Christiaen L, Levine M (2006) ParaHox genes: evolution of Hox/ParaHox cluster FGF signaling delineates the cardiac progenitor fi eld integrity, developmental mode, and temporal colinear- in the simple chordate, Ciona intestinalis . Genes Dev ity. Mol Phylogenet Evol 24:412–417 20:2728 Field KG, Olsen GJ, Lane DJ, Giovannoni SJ, Ghiselin Degnan B, Souter D, Degnan SM, Long SC (1997) MT, Raff EC, Pace NR, Raff RA (1988) Molecular Induction of metamorphosis with potassium ions phylogeny of the animal kingdom. Science requires development of competence and an anterior 239:748–753 signalling centre in the ascidian Herdmania momus . Fischer JL (1992) The embryological oeuvre of Laurent Dev Genes Evol 206:370–376 Chabry. Dev Genes Evol 201:125–127 Dehal P, Satou Y, Campbell RK, Chapman J, Degnan B, Fredriksson G, Öfverholm T, Ericson LE (1988) Iodine De Tomaso A, Davidson B, Di Gregorio A, Gelpke M, binding and peroxidase activity in the endostyle of Goodstein DM (2002) The draft genome of Ciona Salpa Fusiformis, Thalia Democratica, Dolioletta intestinalis : insights into chordate and vertebrate ori- Gegenbauri and Doliolum Nationalis (Tunicata, gins. Science 298:2157 Thaliacea). Cell Tissue Res 253(2):403–411

[email protected] 4 Tunicata 197

Freeman G (1964) The role of blood cells in the process of Haeckel EH (1874) Anthropogenie oder asexual reproduction in the tunicate Perophora viridis . Entwickelungsgeschichte des Menschen: gemeinver- J Exp Zool 156:157–183 ständliche wissenschaftliche Vorträge über die Fujii S, Nishio T, Nishida H (2008) Cleavage pattern, gas- Grundzüge der menschlichen Keimes- und Stammes- trulation, and neurulation in the appendicularian, Geschichte. Wilhelm Engelmann, Leipzig Oikopleura dioica . Dev Genes Evol 218:69–79 Halanych KM (2004) The new view of animal phylogeny. Fukumoto M (1971) Experimental control of budding and Annu Rev Ecol Evol Syst 35(1):229–256 stolon elongation in Perophora orientalis , a compound Haupaix N, Stolfi A, Sirour C, Picco V, Levine M, ascidian. Develop Growth Differ 13:73–88 Christiaen L, Yasuo H (2013) p120RasGAP mediates Furlow JD, Neff ES (2006) A developmental switch ephrin/Eph-dependent attenuation of FGF/ERK sig- induced by thyroid hormone: xenopus laevis metamor- nals during cell fate specifi cation in ascidian embryos. phosis. Trends Endocrinol Metab 17(2):40–47 Development 140:4347–4352 Garstang W (1894) Preliminary note on a new theory of Haupaix N, Abitua PB, Sirour C, Yasuo H, Levine M, the phylogeny of the Chordata. Zool Anz 17: Hudson C (2014) Ephrin-mediated restriction of 122–125 ERK1/2 activity delimits the number of pigment cells Garstang W (1928) Memoirs: the morphology of the in the Ciona CNS. Dev Biol 394(1):170–180 Tunicata, and its bearings on the phylogeny of the Helmkampf M, Bruchhaus I, Hausdorf B (2008) Chordata. Q J Microsc Sci 2:51–187 Phylogenomic analyses of lophophorates (brachio- Gasparini F, Degasperi V, Shimeld SM, Burighel P, Manni pods, phoronids and bryozoans) confi rm the L (2013) Evolutionary conservation of the placodal Lophotrochozoa concept. Proc R Soc B Biol Sci transcriptional network during sexual and asexual 275:1927–1933 development in chordates. Dev Dyn 242:752–766 Heyland A, Hodin J (2004) Heterochronic developmental Gerhart J, Lowe C, Kirschner M (2005) Hemichordates shift caused by thyroid hormone in larval sand dollars and the origin of chordates. Curr Opin Genet Dev and its implications for phenotypic plasticity and the 15:461–467 evolution of nonfeeding development. Evolution Godeaux J (1955) Stades larvaires du Doliolum . Acad 58:524–538 Roy Belg Bull CI Sci 41:769–787 Hirano T, Nishida H (1997) Developmental fates of larval Gomes AS, Alves RN, Rønnestad I, Power DM (2014) tissues after metamorphosis in ascidian Halocynthia Orchestrating change: the thyroid hormones and roretzi. I. Origin of mesodermal tissues of the juvenile. GI-tract development in fl atfi sh metamorphosis. Gen Dev Biol 192(2):199–210 Comp Endocrinol (in press) doi: 10.1016/j. Hirano T, Nishida H (2000) Developmental fates of larval ygcen.2014.06.012 tissues after metamorphosis in the ascidian, Govindarajan AF, Bucklin A, Madin LP (2010) A molec- Halocynthia roretzi. II. Origin of endodermal tissues ular phylogeny of the Thaliacea. J Plankton Res of the juvenile. Dev Genes Evol 210(2):55–63 33:843–853 Hirose E, Hirose M (2007) Morphological process of ver- Grave C (1926) Molgula citrina (Alder and Hancock). tical transmission of photosymbionts in the colonial Activities and structure of the free-swimming larva. J ascidian Trididemnum miniatum Kott, 1977. Mar Biol Morphol 42(2):453–471 150:359–367 Grobben K (1908) Die systematische Einteilung des Hiruta J, Mazet F, Yasui K, Zhang P, Ogasawara M (2005) Tierreiches. Verh Zool Bot Ges Wien 58:l–5 Comparative expression analysis of transcription fac- Gudernatsch J (1912) Feeding experiments on tadpoles. tor genes in the endostyle of invertebrate chordates. Arch Entwicklungsmechanik Organismen 35:457–483 Dev Dyn 233:1031–1037 Gudo M, Syed T (2008) 100 years of deuterostomia Holzer G, Laudet V (2013) Thyroid hormones and post- (Grobben 1908): cladogenetic and anagenetic rela- embryonic development in amniotes. Curr Top Dev tions within the Notoneuralia domain. arXiv preprint Biol 103:397–425 arXiv:08112189 Horie T, Sakurai D, Ohtsuki H, Terakita A, Shichida Y, Hadfi eld KA, Swalla BJ, Jeffery WR (1995) Multiple ori- Usukura J, Kusakabe T, Tsuda M (2008) Pigmented gins of anural development in ascidians inferred from and nonpigmented ocelli in the brain vesicle of the rDNA sequences. J Mol Evol 40:413–427 ascidian larva. J Comp Neurol 509:88–102 Haeckel EH (1866) Generelle Morphologie der Horie T, Nakagawa M, Sasakura Y, Kusakabe TG, Tsuda Organismen allgemeine Grundzuge der organischen M (2010) Simple motor system of the ascidian larva: Formen-Wissenschaft, mechanisch begründet durch neuronal complex comprising putative cholinergic die von Charles Darwin reformirte Descendenz- and GABAergic/glycinergic neurons. Zool Sci 27: Theorie von Ernst Haeckel: allgemeine 181–190 Entwickelungsgeschichte der Organismen, vol 2, Horie T, Shinki R, Ogura Y, Kusakabe TG, Satoh N, Kritische Grundzuge der mechanischen Wissenschaft Sasakura Y (2011) Ependymal cells of chordate larvae von den entstehenden Formen der Organismen, are stem-like cells that form the adult nervous system. begründet durch die Descendenz-Theorie. Verlag von Nature 469:525–528 Georg Reimer, Berlin Hosp J, Sagane Y, Danks G, Thompson EM (2012) The Haeckel EH (1869) Ueber Arbeitstheilung in Natur und evolving proteome of a complex extracellular matrix, Menschenleben. Lüderitz, Berlin the Oikopleura house. PLoS One 7:e40172

[email protected] 198 A. Stolfi and F.D. Brown

Hotta K, Takahashi H, Erives A, Levine M, Satoh N Imai JH, Meinertzhagen IA (2007b) Neurons of the ascidian (1999) Temporal expression patterns of 39 Brachyury- larval nervous system in Ciona intestinalis : II. Peripheral downstream genes associated with notochord forma- nervous system. J Comp Neurol 501:335–352 tion in the Ciona intestinalis embryo. Dev Growth Imai K, Takada N, Satoh N, Satou Y (2000) (beta)-catenin Differ 41:657–664 mediates the specifi cation of endoderm cells in ascid- Hotta K, Mitsuhara K, Takahashi H, Inaba K, Oka K, ian embryos. Development 127:3009–3020 Gojobori T, Ikeo K (2007) A web-based interactive Imai KS, Satoh N, Satou Y (2002a) Early embryonic developmental table for the ascidian Ciona intestina- expression of FGF4 / 6 /9 gene and its role in the induc- lis , including 3D real-image embryo reconstructions: tion of mesenchyme and notochord in Ciona savignyi I. From fertilized egg to hatching larva. Dev Dyn embryos. Development 129:1729–1738 236:1790–1805 Imai KS, Satoh N, Satou Y (2002b) An essential role of a Huber JL, da Silva KB, Bates WR, Swalla BJ (2000) The FoxD gene in notochord induction in Ciona embryos. evolution of anural larvae in molgulid ascidians. Development 129:3441–3453 Semin Cell Dev Biol 11(6):419–426 Imai K, Satoh N, Satou Y (2003) A Twist-like bHLH gene Hudson C, Lemaire P (2001) Induction of anterior neural is a downstream factor of an endogenous FGF and fates in the ascidian Ciona intestinalis . Mech Dev determines mesenchymal fate in the ascidian embryos. 100:189–203 Development 130:4461–4472 Hudson C, Yasuo H (2006) A signalling relay involving Imai KS, Hino K, Yagi K, Satoh N, Satou Y (2004) Gene nodal and delta ligands acts during secondary noto- expression profi les of transcription factors and signal- chord induction in Ciona embryos. Development ing molecules in the ascidian embryo: towards a com- 133:2855–2864 prehensive understanding of gene networks. Hudson C, Yasuo H (2008) Similarity and diversity in Development 131:4047–4058 mechanisms of muscle fate induction between ascid- Imai KS, Levine M, Satoh N, Satou Y (2006) Regulatory ian species. Biol Cell 100:265–277 blueprint for a chordate embryo. Science 312:1183 Hudson C, Lotito S, Yasuo H (2007) Sequential and com- Imai KS, Stolfi A, Levine M, Satou Y (2009) Gene regula- binatorial inputs from Nodal, Delta2/Notch and FGF/ tory networks underlying the compartmentalization of MEK/ERK signalling pathways establish a grid-like the Ciona central nervous system. Development organisation of distinct cell identities in the ascidian 136:285 neural plate. Development 134:3527–3537 Izzi SA, Colantuono BJ, Sullivan K, Khare P, Meedel TH Hudson C, Ba M, Rouvière C, Yasuo H (2011) Divergent (2013) Functional studies of the Ciona intestinalis mechanisms specify chordate motoneurons: evidence myogenic regulatory factor reveal conserved features from ascidians. Development 138:1643–1652 of chordate myogenesis. Dev Biol 376:213–223 Hudson C, Kawai N, Negishi T, Yasuo H (2013) β-catenin- Jacobs MW, Degnan BM, Bishop JD, Strathmann RR driven binary fate specifi cation segregates germ layers (2008) Early activation of adult organ differentiation in ascidian embryos. Curr Biol 23:491–495 during delay of metamorphosis in solitary ascidians, Huxley TH (1851) Observations upon the anatomy and and consequences for juvenile growth. Invertebr Biol physiology of Salpa and Pyrosoma . Philos Trans R 127:217–236 Soc Lond 141:567–593 Jeffery WR (2014) Closing the wounds: one hundred and Hyman L (1959) The invertebrates: smaller coelomate twenty fi ve years of regenerative biology in the ascid- groups, Chaetognatha, Hemichordata, Pogonophora, ian Ciona intestinalis. Genesis 00:1–18 Phoronida, Ectoprocta, Brachiopoda, Sipunculida, the Jeffery WR, Swalla BJ (1992) Evolution of alternate coelomate Bilateria, vol V. McGraw-Hill Book modes of development in ascidians. Bioessays Company, New York 14:219–226 Ikuta T (2011) Evolution of invertebrate deuterostomes Jeffery WR, Swalla BJ, Ewing N, Kusakabe T (1999) and Hox/ParaHox genes. Genomics Proteomics Evolution of the ascidian anural larva: evidence from Bioinforma 9:77–96 embryos and molecules. Mol Biol Evol 16:646–654 Ikuta T, Saiga H (2007) Dynamic change in the expression Jeffery WR, Strickler AG, Yamamoto Y (2004) Migratory of developmental genes in the ascidian central nervous neural crest-like cells form body pigmentation in a system: revisit to the tripartite model and the origin of urochordate embryo. Nature 431:696–699 the midbrain–hindbrain boundary region. Dev Biol Jeffery WR, Chiba T, Krajka FR, Deyts C, Satoh N, Joly 312:631–643 J-S (2008) Trunk lateral cells are neural crest-like cells Ikuta T, Yoshida N, Satoh N, Saiga H (2004) Ciona intes- in the ascidian Ciona intestinalis : insights into the tinalis Hox gene cluster: its dispersed structure and ancestry and evolution of the neural crest. Dev Biol residual colinear expression in development. Proc Natl 324:152–160 Acad Sci U S A 101:15118–15123 Jiang D, Smith WC (2007) Ascidian notochord morpho- Ikuta T, Satoh N, Saiga H (2010) Limited functions of genesis. Dev Dyn 236:1748–1757 Hox genes in the larval development of the ascidian Jiang D, Munro EM, Smith WC (2005) Ascidian prickle Ciona intestinalis . Development 137:1505–1513 regulates both mediolateral and anterior-posterior cell Imai JH, Meinertzhagen IA (2007a) Neurons of the ascidian polarity of notochord cells. Curr Biol 15:79–85 larval nervous system in Ciona intestinalis : I. Central Johnson DS, Davidson B, Brown CD, Smith WC, Sidow nervous system. J Comp Neurol 501:316–334 A (2004) Noncoding regulatory sequences of Ciona

[email protected] 4 Tunicata 199

exhibit strong correspondence between evolutionary ecology of the chlorophyll b-containing symbiotic constraint and functional importance. Genome Res Cyanobacterium Prochloron in the didemnid ascidian 14:2448–2456 Lissoclinum patella. Front Microbiol 3:402 Johnson DS, Zhou Q, Yagi K, Satoh N, Wong W, Sidow A Kumano G, Nishida H (2009) Patterning of an ascidian (2005) De novo discovery of a tissue-specifi c gene embryo along the anterior–posterior axis through spa- regulatory module in a chordate. Genome Res tial regulation of competence and induction ability by 15:1315–1324 maternally localized PEM. Dev Biol 331:78–88 Julin C (1912) Recherches sur le développement embry- Kumano G, Yamaguchi S, Nishida H (2006) Overlapping onnaire de Pyrosoma giganteum . Zool Jahrb Suppl 15: expression of FoxA and Zic confers responsiveness to 775–863 FGF signaling to specify notochord in ascidian Katikala L, Aihara H, Passamaneck YJ, Gazdoiu S, José- embryos. Dev Biol 300:770–784 Edwards DS, Kugler JE, Oda-Ishii I, Imai JH, Nibu Y, Kumano G, Takatori N, Negishi T, Takada T, Nishida H Di Gregorio A (2013) Functional brachyury binding (2011) A maternal factor unique to ascidians silences sites establish a temporal read-out of gene expression the germline via binding to P-TEFb and RNAP II reg- in the Ciona notochord. PLoS Biol 11:e1001697 ulation. Curr Biol 21:1308–1313 Kawamura K, Fujiwara S (1995) Establishment of cell Kürn U, Rendulic S, Tiozzo S, Lauzon RJ (2011) Asexual lines from multipotent epithelial sheet in the budding propagation and regeneration in colonial ascidians. tunicate, Polyandrocarpa misakiensis. Cell Struct Biol Bull 221:43–61 Funct 20:97–106 Kusakabe T, Swalla BJ, Satoh N, Jeffery WR (1996) Kawamura K, Sugino Y, Sunanaga T, Fujiwara S (2008) Mechanism of an evolutionary change in muscle cell Multipotent epithelial cells in the process of regenera- differentiation in ascidians with different modes of tion and asexual reproduction in colonial tunicates. development. Dev Biol 174:379–392 Develop Growth Differ 50:1–11 Kusakabe T, Yoshida R, Kawakami I, Kusakabe R, Kim GJ, Nishida H (2001) Role of the FGF and MEK Mochizuki Y, Yamada L, Shin-i T, Kohara Y, Satoh N, signaling pathway in the ascidian embryo. Develop Tsuda M (2002) Gene expression profi les in tadpole Growth Differ 43:521–533 larvae of Ciona intestinalis . Dev Biol 242:188–203 Kimura Y, Yoshida M, Morisawa M (2003) Interaction Laird DJ, Weissman IL (2004) Telomerase maintained in between noradrenaline or adrenaline and the β1 - self-renewing tissues during serial regeneration of the adrenergic receptor in the nervous system triggers urochordate Botryllus schlosseri . Dev Biol early metamorphosis of larvae in the ascidian, Ciona 273:185–194 savignyi . Dev Biol 258:129–140 Laird DJ, De Tomaso AW, Weissman IL (2005) Stem cells Kluge B, Renault N, Rohr KB (2005) Anatomical and are units of natural selection in a colonial ascidian. molecular reinvestigation of lamprey endostyle devel- Cell 123:1351–1360 opment provides new insight into thyroid gland evolu- Lamarck JB (1816) Histoire Naturelle Des Animaux Sans tion. Dev Genes Evol 215(1):32–40 Vertèbres. Verdière, Paris Kott P (2001) The Australian Ascidiacea part 4, Lankester ER (1877) Memoirs: notes on the embryology Aplousobranchia (3), Didemnidae. Mem Queensland and classifi cation of the animal kingdom: comprising a Mus 47:1–408 revision of speculations relative to the origin and signifi - Kourakis MJ, Smith WC (2007) A conserved role for FGF cance of the germ-layers. Q J Microsc Sci 2:399–454 signaling in chordate otic/atrial placode formation. Lemaire P (2009) Unfolding a chordate developmental Dev Biol 312:245–257 program, one cell at a time: invariant cell lineages, Kowalevsky A (1874) Ueber die knospung der Ascidien. short-range inductions and evolutionary plasticity in Arch Mikrosk Anat 10:441–470 ascidians. Dev Biol 332:48–60 Kowalewski AO (1866) Entwicklungsgeschichte der ein- Lemaire P (2011) Evolutionary crossroads in develop- fachen Ascidien. In: mémoires de L’Académie mental biology: the tunicates. Development Impériale des Sciences de St.-Pétersbourg, vol. 10. St. 138(11):2143–2152 Petersburg: commissionäre der Kaiserlichen Marino R, Melillo D, Di Fillippo M, Yamada A, Pinto M, Akademie der Wissenschaften De Santis R, Brown ER, Matassi G (2007) Ammonium Kozloff EN (1990) Invertebrates. Saunders College Pub, channel expression is essential for brain development Philadelphia, p 866 and function in the larva of Ciona intestinalis. J Comp Krieg M, Arboleda-Estudillo Y, Puech P-H, Käfer J, Neurol 503:135–147 Graner F, Müller D, Heisenberg C-P (2008) Tensile Matthysse AG, Deschet K, Williams M, Marry M, White forces govern germ-layer organization in zebrafi sh. AR, Smith WC (2004) A functional cellulose synthase Nat Cell Biol 10:429–436 from ascidian epidermis. Proc Natl Acad Sci U S A Kugler JE, Kerner P, Bouquet J-M, Jiang D, Di Gregorio 101:986–991 A (2011) Evolutionary changes in the notochord Mazet F, Hutt JA, Milloz J, Millard J, Graham A, Shimeld genetic toolkit: a comparative analysis of notochord SM (2005) Molecular evidence from Ciona intestina- genes in the ascidian Ciona and the larvacean lis for the evolutionary origin of vertebrate sensory Oikopleura . BMC Evol Biol 11:21 placodes. Dev Biol 282:494–508 Kühl M, Behrendt L, Trampe E, Qvortrup K, Schreiber U, Meedel TH, Farmer SC, Lee JJ (1997) The single MyoD Borisov SM, Larkum AWD (2012) Microenvironmental family gene of Ciona intestinalis encodes two

[email protected] 200 A. Stolfi and F.D. Brown

differentially expressed proteins: implications for the cleavage-plane orientation and unequal cell divisions evolution of chordate muscle gene regulation. in ascidians. Curr Biol 17:1014–1025 Development 124:1711–1721 Nicol D, Meinertzhagen I (1988a) Development of the Meedel TH, Lee JJ, Whittaker J (2002) Muscle develop- central nervous system of the larva of the ascidian, ment and lineage-specifi c expression of CiMDF , the Ciona intestinalis L: I. The early lineages of the neural MyoD family gene of Ciona intestinalis . Dev Biol plate. Dev Biol 130:721–736 241:238–246 Nicol D, Meinertzhagen I (1988b) Development of the Meedel TH, Chang P, Yasuo H (2007) Muscle develop- central nervous system of the larva of the ascidian, ment in Ciona intestinalis requires the b-HLH myo- Ciona intestinalis L: II. Neural plate morphogenesis genic regulatory factor gene Ci -MRF . Dev Biol and cell lineages during neurulation. Dev Biol 302:333–344 130:737–766 Millar R (1954) The breeding and development of the Nielsen C (1995) Animal evolution: interrelationships of ascidian Pelonaia corrugata Forbes and Goodsir. J the living phyla. Oxford University Press, Oxford Mar Biol Assoc U K 33:681–687 Nielsen C (2002) The phylogenetic position of entoprocta, Millar R (1962) The breeding and development of the ectoprocta, phoronida, and brachiopoda. Integr Comp ascidian Polycarpa tinctor . Q J Microsc Sci Biol 42:685–691 3:399–403 Nishida H (1987) Cell lineage analysis in ascidian Millar RH (1971) The biology of ascidians. Adv Mar Biol embryos by intracellular injection of a tracer enzyme 9:1–100. doi: 1 0 . 1 0 1 6 / S 0 0 6 5 - 2 8 8 1 ( 0 8 ) 6 0 3 4 1 - 7 , Elsevier III. Up to the tissue restricted stage. Dev Biol Milne-Edwards H (1841) Observations sur les ascidies 121:526–541 composées des côtes de la Manche. Chez Fortin- Nishida H (1992) Regionality of egg cytoplasm that pro- Masson et Cie, Paris motes muscle differentiation in embryo of the ascid- Milne-Edwards H (1843) Élémens de Zoologie, Ou, ian, Halocynthia roretzi . Development 116:521–529 Leçons Sur L’anatomie, La Physiologie, La Nishida H (1993) Localized regions of egg cytoplasm that Classifi cation et Les Moeurs Des Animaux. Fortin promote expression of endoderm-specifi c alkaline Masson, Paris phosphatase in embryos of the ascidian Halocynthia Minokawa T, Yagi K, Makabe KW, Nishida H (2001) roretzi . Development 118:1–7 Binary specifi cation of nerve cord and notochord cell Nishida H (1994a) Localization of determinants for for- fates in ascidian embryos. Development mation of the anterior-posterior axis in eggs of the 128:2007–2017 ascidian Halocynthia roretzi . Development M o c h i z u k i Y, S a t o u Y, S a t o h N ( 2 0 0 3 ) L a r g e ‐scale charac- 120:3093–3104 terization of genes specifi c to the larval nervous system Nishida H (1994b) Localization of egg cytoplasm that in the ascidian Ciona intestinalis . G e n e s i s 3 6 : 6 2 – 7 1 promotes differentiation to epidermis in embryos of Mukai H, Koyama H, Watanabe H (1983) Studies on the the ascidian Halocynthia roretzi . Development reproduction of three species of Perophora 120:235–243 (Ascidiacea). Biol Bull 164:251–266 Nishida H (1996) Vegetal egg cytoplasm promotes gastru- Nakashima K, Yamada L, Satou Y, Azuma J, Satoh N lation and is responsible for specifi cation of vegetal (2004) The evolutionary origin of animal cellulose blastomeres in embryos of the ascidian Halocynthia synthase. Dev Genes Evol 214:81–88 roretzi . Development 122:1271–1279 Nakauchi M (1982) Asexual development of ascidians: its Nishida N (2008) Development of the appendicularian biological signifi cance, diversity, and morphogenesis. Oikopleura dioica: culture, genome, and cell lineages. Am Zool 22:753–763 Dev Growth Differ 50:S239–S256 Nakauchi M (1986) Oozooid development and budding in Nishida H, Sawada K (2001) macho - 1 encodes a localized the polyclinid ascidian, Parascidia fl emingii mRNA in ascidian eggs that specifi es muscle fate dur- (Urochordata). J Zool 208:255–267 ing embryogenesis. Nature 409:724–729 Nakayama A, Satou Y, Satoh N (2001) Isolation and char- Nishikawa T (1991) The ascidians of the Japan sea. acterization of genes that are expressed during Ciona II. Publ Seto Mar Biol Lab 35:25–170 intestinalis metamorphosis. Dev Genes Evol Nishino A, Satou Y, Morisawa M, Satoh N (2000) Muscle 211:184–189 actin genes and muscle cells in the appendicularian, Nakayama-Ishimura A, Chambon J-p, Horie T, Satoh N, Oikopleura longicauda : phylogenetic relationships Sasakura Y (2009) Delineating metamorphic path- among muscle tissues in the urochordates. J Exp Zool ways in the ascidian Ciona intestinalis . Dev Biol 288:135–150 326:357–367 Nishino A, Satou Y, Morisawa M, Satoh N (2001) Nakazawa K, Yamazawa T, Moriyama Y, Ogura Y, Kawai Brachyury ( T ) gene expression and notochord devel- N, Sasakura Y, Saiga H (2013) Formation of the diges- opment in Oikopleura longicauda (Appendicularia, tive tract in Ciona intestinalis includes two distinct Urochordata). Dev Genes Evol 211:219–231 morphogenic processes between its anterior and poste- Nishino A, Okamura Y, Piscopo S, Brown ER (2010) A rior parts. Dev Dyn 242:1172–1183 glycine receptor is involved in the organization of Negishi T, Takada T, Kawai N, Nishida H (2007) swimming movements in an invertebrate chordate. Localized PEM mRNA and protein are involved in BMC Neurosci 11:6

[email protected] 4 Tunicata 201

Nishitsuji K, Horie T, Ichinose A, Sasakura Y, Yasuo H, Prodon F, Sardet C, Nishida H (2008) Cortical and cyto- Kusakabe TG (2012) Cell lineage and cis-regulation plasmic fl ows driven by actin microfi laments polarize for a unique GABAergic/glycinergic neuron type in the cortical ER-mRNA domain along the a–v axis in the larval nerve cord of the ascidian Ciona intestinalis . ascidian oocytes. Dev Biol 313:682–699 Develop Growth Differ 54:177–186 Raff RA, Love AC (2004) Kowalevsky, comparative evo- Noda T, Satoh N (2008) A comprehensive survey of cad- lutionary embryology, and the intellectual lineage of herin superfamily gene expression patterns in Ciona evo‐devo. J Exp Zool B Mol Dev Evol 302:19–34 intestinalis . Gene Expr Patterns 8:349–356 Razy-Krajka F, Lam K, Wang W, Stolfi A, Joly M, Oda-Ishii I, Bertrand V, Matsuo I, Lemaire P, Saiga H Bonneau R, Christiaen L (2014) Collier/OLF/EBF- (2005) Making very similar embryos with divergent dependent transcriptional dynamics control pharyn- genomes: conservation of regulatory mechanisms of geal muscle specifi cation from primed Otx between the ascidians Halocynthia roretzi and cardiopharyngeal progenitors. Dev Cell 29:263–276 Ciona intestinalis . Development 132:1663–1674 Rinkevich Y, Voskoboynik A, Rosner A, Rabinowitz C, Paz Ogasawara M, Di Lauro R, Satoh N (1999) Ascidian G, Oren M, Douek J, Alfassi G, Moiseeva E, Ishizuka KJ homologs of mammalian thyroid peroxidase genes are (2013) Repeated, long-term cycling of putative stem cells expressed in the thyroid-equivalent region of the endo- between niches in a basal chordate. Dev Cell 24:76–88 style. J Experimental Zool 285(2):158–169 Roberts B, Davidson B, MacMaster G, Lockhart V, Ma E, Ogasawara M, Shigetani Y, Suzuki S, Kuratani S, Satoh N Wallace SS, Swalla BJ (2007) A complement response (2001) Expression of thyroid transcription factor-1 may activate metamorphosis in the ascidian Boltenia (TTF-1) gene in the ventral forebrain and endostyle of villosa . Dev Genes Evol 217:449–458 the agnathan vertebrate, Lampetra japonica. Genesis Roegiers F, Djediat C, Dumollard R, Rouvière C, Sardet C 30(2):51–58 (1999) Phases of cytoplasmic and cortical reorganiza- Ogura Y, Sakaue-Sawano A, Nakagawa M, Satoh N, tions of the ascidian zygote between fertilization and Miyawaki A, Sasakura Y (2011) Coordination of mitosis fi rst division. Development 126:3101–3117 and morphogenesis: role of a prolonged G2 phase during Romer AS (1967) Major steps in vertebrate evolution. chordate neurulation. Development 138:577–587 Science (New York, NY) 158(3809):1629–1637 Okuda Y, Ogura E, Kondoh H, Kamachi Y (2010) B1 Rothbächer U, Bertrand V, Lamy C, Lemaire P (2007) A SOX coordinate cell specifi cation with patterning and combinatorial code of maternal GATA, Ets and morphogenesis in the early zebrafi sh embryo. PLoS β-catenin-TCF transcription factors specifi es and pat- Genet 6:e1000936 terns the early ascidian ectoderm. Development Paffenhöfer G-A, Köster M (2011) From one to many: on 134:4023–4032 the life cycle of Dolioletta gegenbauri Uljanin Roure A, Lemaire P, Darras S (2014) An Otx/Nodal regu- (Tunicata, Thaliacea). J Plankton Res 33:1139–1145 latory signature for posterior neural development in Pancer Z, Gershon H, Rinkevich B (1995) Coexistence ascidians. PLoS Genet 10:e1004548 and possible parasitism of somatic and germ cell lines Sabbadin A, Zaniolo G, Majone F (1975) Determination in chimeras of the colonial urochordate Botryllus of polarity and bilateral asymmetry in palleal and vas- schlosseri . Biol Bull 189:106–112 cular buds of the ascidian Botryllus schlosseri . Dev Paris M, Brunet F, Markov GV, Schubert M, Laudet V Biol 46:79–87 (2008) The amphioxus genome enlightens the evolu- Saito M, Seki M, Amemiya S, Yamasu K, Suyemitsu T, tion of the thyroid hormone signaling pathway. Dev Ishihara K (1998) Induction of metamorphosis in the Genes Evol 218:667–680 sand dollar Peronella japonica by thyroid hormones. Pasini A, Amiel A, Rothbächer U, Roure A, Lemaire P, Develop Growth Differ 40:307–312 Darras S (2006) Formation of the ascidian epidermal Sander K, Fischer J-L (1992) How to dart ascidian blasto- sensory neurons: insights into the origin of the chor- meres: the embryological micro-tools of Laurent date peripheral nervous system. PLoS Biol 4:e225 Chabry (1855–1893). Dev Genes Evol 201:191–193 Pasini A, Manenti R, Rothbächer U, Lemaire P (2012) Sardet C, Speksnijder J, Inoue S, Jaffe L (1989) Antagonizing retinoic acid and FGF/MAPK pathways Fertilization and ooplasmic movements in the ascidian control posterior body patterning in the invertebrate egg. Development 105:237–249 chordate Ciona intestinalis . PLoS One 7:e46193 Sardet C, Nishida H, Prodon F, Sawada K (2003) Maternal Pérez-Portela R, Bishop J, Davis AR, Turon X (2009) mRNAs of PEM and macho 1, the ascidian muscle Phylogeny of the families Pyuridae and Styelidae determinant, associate and move with a rough endo- (Stolidobranchiata, Ascidiacea) inferred from mito- plasmic reticulum network in the egg cortex. chondrial and nuclear DNA sequences. Mol Development 130:5839–5849 Phylogenet Evol 50:560–570 Sardet C, Paix A, Prodon F, Dru P, Chenevert J (2007) Perrier E (1898) Note sur la Classifi cation des Tuniciers. From oocyte to 16‐cell stage: cytoplasmic and cortical CR Acad Sci Paris 124:1758–1762 reorganizations that pattern the ascidian embryo. Dev Picco V, Hudson C, Yasuo H (2007) Ephrin-Eph signal- Dyn 236:1716–1731 ling drives the asymmetric division of notochord/neu- Sasaki A, Miyamoto Y, Satou Y, Satoh N, Ogasawara M ral precursors in Ciona embryos. Development (2003) Novel endostyle-specifi c genes in the ascidian 134:1491–1497 Ciona intestinalis . Zool Sci 20:1025–1030

[email protected] 202 A. Stolfi and F.D. Brown

Sasakura Y, Nakashima K, Awazu S, Matsuoka T, Somorjai IM, Somorjai RL, Garcia-Fernàndez J, Escrivà Nakayama A, Azuma J, Satoh N (2005) Transposon- H (2012) Vertebrate-like regeneration in the inverte- mediated insertional mutagenesis revealed the func- brate chordate amphioxus. Proc Natl Acad Sci tions of animal cellulose synthase in the ascidian Ciona 109:517–522 intestinalis. Proc Natl Acad Sci U S A 102:15134 Sorrentino M, Manni L, Lane N, Burighel P (2000) Sasakura Y, Kanda M, Ikeda T, Horie T, Kawai N, Ogura Evolution of cerebral vesicles and their sensory organs Y, Yoshida R, Hozumi A, Satoh N, Fujiwara S (2012) in an ascidian larva. Acta Zool 81:243–258 Retinoic acid-driven Hox1 is required in the epidermis Søviknes AM, Glover JC (2008) Continued growth and for forming the otic/atrial placodes during ascidian cell proliferation into adulthood in the notochord of metamorphosis. Development 139:2156–2160 the appendicularian Oikopleura dioica. Biol Bull Satoh N (2013) Developmental genomics of ascidians. 214:17–28 Wiley-Blackwell, Hoboken, p 216 Spada F, Steen H, Troedsson C, Kallesøe T, Spriet E, Satou Y (1999) posterior end mark 3 ( pem -3 ), an ascidian Mann M, Thompson EM (2001) Molecular pattern- maternally expressed gene with localized mRNA ing of the oikoplastic epithelium of the larvacean encodes a protein with Caenorhabditis elegans M E X - tunicate Oikopleura dioica . J Biol Chem 3- like KH domains. Dev Biol 212:337–350 276:20624–20632 Satou Y, Satoh N (1997) posterior end mark 2 ( pem - 2 ), Spagnuolo A, Ristoratore F, Di Gregorio A, Aniello F, pem - 4 , pem - 5 , and pem -6 : maternal genes with local- Branno M, Di Lauro R (2003) Unusual number and ized mRNA in the ascidian embryo. Dev Biol genomic organization of Hox genes in the tunicate 192:467–481 Ciona intestinalis . Gene 309:71–79 Satou Y, Takatori N, Yamada L, Mochizuki Y, Squarzoni P, Parveen F, Zanetti L, Ristoratore F, Hamaguchi M, Ishikawa H, Chiba S, Imai K, Kano Spagnuolo A (2011) FGF/MAPK/Ets signaling ren- S, Murakami SD (2001) Gene expression profi les in ders pigment cell precursors competent to respond to Ciona intestinalis tailbud embryos. Development Wnt signal by directly controlling Ci - Tcf transcrip- 128:2893–2904 tion. Development 138:1421–1432 Satou Y, Takatori N, Fujiwara S, Nishikata T, Saiga H, Stach T, Winter J, Bouquet J-M, Chourrout D, Schnabel R Kusakabe T, Shin-i T, Kohara Y, Satoh N (2002) Ciona (2008) Embryology of a planktonic tunicate reveals intestinalis cDNA projects: expressed sequence tag traces of sessility. Proc Natl Acad Sci 105:7229–7234 analyses and gene expression profi les during embryo- Stefaniak L, Zhang H, Gittenberger A, Smith K, Holsinger genesis. Gene 287:83–96 K, Lin S, Whitlatch RB (2012) Determining the native Satou Y, Kawashima T, Kohara Y, Satoh N (2003) Large region of the putatively invasive ascidian Didemnum scale EST analyses in Ciona intestinalis . Dev Genes vexillum Kott, 2002. J Exp Mar Biol Ecol 422:64–71 Evol 213:314–318 Stolfi A, Christiaen L (2012) Genetic and genomic tool- Satou Y, Imai KS, Satoh N (2004) The ascidian Mesp box of the chordate Ciona intestinalis . Genetics gene specifi es heart precursor cells. Development 192:55–66 131:2533–2541 Stolfi A, Levine M (2011) Neuronal subtype specifi cation Seo HC, Kube M, Edvardsen RB, Jensen MF, Beck A, in the spinal cord of a protovertebrate. Development Spriet E, Gorsky G, Thompson EM, Lehrach H, 138:995–1004 Reinhardt R (2001) Miniature genome in the marine Stolfi A, Gainous TB, Young JJ, Mori A, Levine M, chordate Oikopleura dioica . Science 294:2506 Christiaen L (2010) Early chordate origins of the ver- Seo H, Edvardsen RB, Maeland AD, Bjordal M, Jensen tebrate second heart fi eld. Science 329:565 MF, Hansen A, Flaat M, Weissenbach J, Lehrach H, Stolfi A, Wagner E, Taliaferro JM, Chou S, Levine M Wincker P, Reinhardt R, Chourrout D (2004) Hox (2011) Neural tube patterning by Ephrin, FGF and cluster disintegration with persistent anteroposterior Notch signaling relays. Development 138:5429–5439 order of expression in Oikopleura dioica. Nature Stolfi A, Lowe EK, Racioppi C, Ristoratore F, Brown CT, 431:67–71 Swalla BJ, Christiaen L (2014) Divergent mechanisms Sherrard K, Robin F, Lemaire P, Munro E (2010) regulate conserved cardiopharyngeal development and Sequential activation of apical and basolateral contrac- gene expression in distantly related ascidians. eLife. tility drives ascidian endoderm invagination. Curr Biol doi: 10.7554/eLife.03728 20:1499–1510 Stoner DS, Weissman IL (1996) Somatic and germ cell Shi W, Levine M (2008) Ephrin signaling establishes parasitism in a colonial ascidian: possible role for a asymmetric cell fates in an endomesoderm lineage of highly polymorphic allorecognition system. Proc Natl the Ciona embryo. Development 135:931–940 Acad Sci 93:15254–15259 Shirae-Kurabayashi M, Nishikata T, Takamura K, Tanaka Sugino YM, Matsumura M, Kawamura K (2007) Body KJ, Nakamoto C, Nakamura A (2006) Dynamic redis- muscle-cell differentiation from coelomic stem cells tribution of vasa homolog and exclusion of somatic in colonial tunicates. Zool Sci 24:542–546 cell determinants during germ cell specifi cation in Sutton MF (1960) The sexual development of Salpa fusi- Ciona intestinalis . Development 133:2683–2693 formis (Cuvier) Part I. J Embryol Exp Morpholog Small K, Brudno M, Hill M, Sidow A (2007) A haplome 8:268–290 alignment and reference sequence of the highly poly- Swalla B (2006) Building divergent body plans with simi- morphic Ciona savignyi genome. Genome Biol 8:R41 lar genetic pathways. Heredity 97:235–243

[email protected] 4 Tunicata 203

Swalla BJ, Jeffery WR (1990) Interspecifi c hybridization Todaro F (1880) Sui primi fenomeni dello sviluppo delle between an anural and urodele ascidian: differential Salpe. Atti R Accad Lincei 3 Trans 4(3):86–89 expression of urodele features suggests multiple Tokuoka M, Satoh N, Satou Y (2005) A bHLH transcrip- mechanisms control anural development. Dev Biol tion factor gene, Twist - like1 , is essential for the forma- 142:319–334 tion of mesodermal tissues of Ciona juveniles. Dev Swalla BJ, Jeffery WR (1992) Vestigial brain melanocyte Biol 288:387–396 development during embryogenesis of an anural ascid- Tokuoka M, Kumano G, Nishida H (2007) FGF9/16/20 ian. Develop Growth Differ 34:17–25 and Wnt-5α signals are involved in specifi cation of Swalla BJ, Cameron CB, Corley LS, Garey JR (2000) secondary muscle fate in embryos of the ascidian, Urochordates are monophyletic within the deutero- Halocynthia roretzi . Dev Genes Evol 217:515–527 stomes. Syst Biol 49:52–64 Tolkin T, Christiaen L (2012) Development and evolution Tagawa K, Jeffery WR, Satoh N (1997) The recently- of the ascidian cardiogenic mesoderm. Curr Top Dev described ascidian species Molgula tectiformis is a Biol 100:107 direct developer. Zool Sci 14:297–303 Treen N, Yoshida K, Sakuma T, Sasaki H, Kawai N, Takahashi H, Mitani Y, Satoh G, Satoh N (1999) Yamamoto T, Sasakura Y (2014) Tissue-specifi c and Evolutionary alterations of the minimal promoter for ubiquitous gene knockouts by TALEN electroporation notochord-specifi c Brachyury expression in ascidian provide new approaches to investigating gene function embryos. Development 126:3725–3734 in Ciona . Development 141:481–487 Takamura K, Fujimura M, Yamaguchi Y (2002) True JR, Haag ES (2001) Developmental system drift and Primordial germ cells originate from the endodermal fl exibility in evolutionary trajectories. Evol Dev strand cells in the ascidian Ciona intestinalis . Dev 3:109–119 Genes Evol 212:11–18 T s a g k o g e o r g a G , T u r o n X , H o p c r o f t R R , T i l a k M - K , T a k a t o r i N , K u m a n o G , S a i g a H , N i s h i d a H ( 2 0 1 0 ) Feldstein T, Shenkar N, Loya Y, Huchon D, Douzery Segregation of germ layer fates by nuclear migration- EJ, Delsuc F (2009) An updated 18S rRNA phylogeny dependent localization of Not mRNA. Dev Cell 19: of tunicates based on mixture and secondary structure 589–598 models. BMC Evol Biol 9:187 Tang WJ, Chen JS, Zeller RW (2013) Transcriptional Tsagkogeorga G, Cahais V, Galtier N (2012) The population regulation of the peripheral nervous system in Ciona genomics of a fast evolver: high levels of diversity, func- intestinalis. Dev Biol 378:183–193 tional constraint, and molecular adaptation in the tuni- Tarallo R, Sordino P (2004) Time course of programmed cate Ciona intestinalis . Genome Biol Evol 4:852–861 cell death in Ciona intestinalis in relation to mitotic Tsuda M, Sakurai D, Goda M (2003) Direct evidence for activity and MAPK signaling. Dev Dyn 230:251–262 the role of pigment cells in the brain of ascidian larvae Tassy O, Dauga D, Daian F, Sobral D, Robin F, Khoueiry by laser ablation. J Exp Biol 206:1409–1417 P, Salgado D, Fox V, Caillol D, Schiappa R (2010) The Tzahor E, Evans SM (2011) Pharyngeal mesoderm develop- ANISEED database: digital representation, formaliza- ment during embryogenesis: implications for both heart tion, and elucidation of a chordate developmental pro- and head myogenesis. Cardiovasc Res 91:196–202 gram. Genome Res 20:1459–1468 Ueda N, Degnan SM (2013) Nitric oxide acts as a positive Tatián M, Lagger C, Demarchi M, Mattoni C (2011) regulator to induce metamorphosis of the ascidian Molecular phylogeny endorses the relationship Herdmania momus . PLoS One 8:e72797 between carnivorous and fi lter‐feeding tunicates Veeman MT, Nakatani Y, Hendrickson C, Ericson V, Lin (Tunicata, Ascidiacea). Zool Scr 40:603–612 C, Smith WC (2008) Chongmague reveals an essential Telford MJ, Copley RR (2011) Improving animal phylog- role for laminin-mediated boundary formation in chor- enies with genomic data. Trends Genet 27:186–195 date convergence and extension movements. Terakubo HQ, Nakajima Y, Sasakura Y, Horie T, Konno Development 135:33–41 A, Takahashi H, Inaba K, Hotta K, Oka K (2010) Veeman MT, Newman-Smith E, El-Nachef D, Smith WC Network structure of projections extending from (2010) The ascidian mouth opening is derived from peripheral neurons in the tunic of ascidian larva. Dev the anterior neuropore: reassessing the mouth/neural Dyn 239:2278–2287 tube relationship in chordate evolution. Dev Biol Thompson EM, Kallesøe T, Spada F (2001) Diverse genes 344:138–149 expressed in distinct regions of the trunk epithelium Vinson JP, Jaffe DB, O'Neill K, Karlsson EK, Stange- defi ne a monolayer cellular template for construction Thomann N, Anderson S, Mesirov JP, Satoh N, Satou of the oikopleurid house. Dev Biol 238:260–273 Y, Nusbaum C (2005) Assembly of polymorphic Tiozzo S, Christiaen L, Deyts C, Manni L, Joly JS, genomes: algorithms and application to Ciona savi- Burighel P (2005) Embryonic versus blastogenetic gnyi . Genome Res 15:1127–1135 development in the compound ascidian Botryllus Voskoboynik A, Soen Y, Rinkevich Y, Rosner A, Ueno H, schlosseri : insights from Pitx expression patterns. Dev Reshef R, Ishizuka KJ, Palmeri KJ, Moiseeva E, Dyn 232:468–478 Rinkevich B (2008) Identifi cation of the endostyle as a Tiozzo S, Brown FD, De Tomaso AW (2008) Regeneration stem cell niche in a colonial chordate. Cell Stem Cell and stem cells in ascidians. In: Bosch TCG (ed) Stem 3:456–464 cells from hydra to man. Springer, Dordrecht, Voskoboynik A, Neff NF, Sahoo D, Newman aM, pp 95–112 Pushkarev D, Koh W, Quake SR (2013) The genome

[email protected] 204 A. Stolfi and F.D. Brown

sequence of the colonial chordate, Botryllus schlos- Willey A (1894) Amphioxus and the ancestry of the verte- seri. eLife 2:e00569–e00569 brates. Macmillan, New York, p 316 Wada H (1998) Evolutionary history of free-swimming Winchell CJ, Sullivan J, Cameron CB, Swalla BJ, Mallatt and sessile lifestyles in urochordates as deduced from J (2002) Evaluating hypotheses of deuterostome phy- 18S rDNA molecular phylogeny. Mol Biol Evol logeny and chordate evolution with new LSU and SSU 15:1189–1194 ribosomal DNA data. Mol Biol Evol 19:762–776 Wada H (2001) Origin and evolution of the neural crest: a Yagi K, Satou Y, Satoh N (2004) A zinc fi nger transcrip- hypothetical reconstruction of its evolutionary history. tion factor, ZicL , is a direct activator of Brachyury in Develop Growth Differ 43:509–520 the notochord specifi cation of Ciona intestinalis . W a d a S , S a i g a H ( 2 0 0 2 ) HrzicN , a new Zic family gene of Development 131:1279–1288 ascidians, plays essential roles in the neural tube and Yagi K, Takatori N, Satou Y, Satoh N (2005) Ci - Tbx6b notochord development. Development 129:5597–5608 and Ci -Tbx6c are key mediators of the maternal effect Wada H, Holland PWH, Sato S, Yamamoto H, Satoh N gene C i - macho1 in muscle cell differentiation in (1997) Neural tube is partially dorsalized by overex- Ciona intestinalis embryos. Dev Biol 282: pression of HrPax-37 : the ascidian homologue of Pax- 535–549 3 and Pax-7 . Dev Biol 187:240–252 Yasuo H, Hudson C (2007) FGF8 /17 /18 functions Wada H, Saiga H, Satoh N, Holland P (1998) Tripartite together with FGF9 / 16 / 20 during formation of the organization of the ancestral chordate brain and the notochord in Ciona embryos. Dev Biol 302:92–103 antiquity of placodes: insights from ascidian Pax- 2 / 5 / 8 , Yasuo H, Satoh N (1998) Conservation of the develop- Hox a n d Otx genes. Development 125:1113–1122 mental role of Brachyury in notochord formation in a Wagner E, Levine M (2012) FGF signaling establishes the urochordate, the ascidian Halocynthia roretzi . Dev anterior border of the Ciona neural tube. Development Biol 200:158–170 139:2351–2359 Yokobori S, Kurabayashi A, Neilan BA, Maruyama T, Wagner E, Stolfi A, Choi YG, Levine M (2014) Islet is a Hirose E (2006) Multiple origins of the ascidian- key determinant of ascidian palp morphogenesis. Prochloron symbiosis: molecular phylogeny of photo- Development 141:3084–3092 symbiotic and non-symbiotic colonial ascidians Wang W, Razy-Krajka F, Siu E, Ketcham A, Christiaen L inferred from 18S rDNA sequences. Mol Phylogenet (2013) NK4 antagonizes Tbx1/10 to promote cardiac Evol 40:8–19 versus pharyngeal muscle fate in the ascidian second Yokoyama T, Hotta K, Oka K (2014) Comprehensive heart fi eld. PLoS Biol 11:e1001725 morphological analysis of individual peripheral neu- W h i t t a k e r J ( 1 9 7 3 ) S e g r e g a t i o n d u r i n g a s c i d i a n e m b r y o - ron dendritic arbors in ascidian larvae using the photo- genesis of egg cytoplasmic information for tissue- convertible protein kaede. Dev Dyn 243:1362–1373 specifi c enzyme development. Proc Natl Acad Sci Young CM, Gowan RF, Dalby J, Pennachetti CA, 70:2096–2100 Gagliardi D (1988) Distributional consequences of Whittaker J (1977) Segregation during cleavage of a fac- adhesive eggs and anural development in the ascidian tor determining endodermal alkaline phosphatase Molgula pacifi ca (Huntsman, 1912). Biol Bull development in ascidian embryos. J Exp Zool 174:39–46 202:139–153 Zeng L, Swalla BJ (2005) Molecular phylogeny of the pro- Whittaker J (1982) Muscle lineage cytoplasm can change tochordates: chordate evolution. Can J Zool 83:24–33 the developmental expression in epidermal lineage Zeng L, Jacobs MW, Swalla BJ (2006) Coloniality has cells of ascidian embryos. Dev Biol 93:463–470 evolved once in stolidobranch ascidians. Integr Comp Wicht H, Lacalli TC (2005) The nervous system of amphi- Biol 46:255–268 oxus: structure, development, and evolutionary signifi - Zimmer R, Larwood G (1973) Living and fossil bryozoa. cance. Can J Zool 83:122–150 Academic, Lawrence

[email protected]