Evolution of Allorecognition in Botryllid Ascidians Inferred from a Molecular Phylogeny Author(s): C. Sarah Cohen, Yasunori Saito, Irving L. Weissman Source: Evolution, Vol. 52, No. 3 (Jun., 1998), pp. 746-756 Published by: Society for the Study of Evolution Stable URL: http://www.jstor.org/stable/2411269 . Accessed: 28/02/2011 02:19

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at . http://www.jstor.org/action/showPublisher?publisherCode=ssevol. .

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Society for the Study of Evolution is collaborating with JSTOR to digitize, preserve and extend access to Evolution.

http://www.jstor.org Evolution, 52(3), 1998, pp. 746-756

EVOLUTION OF ALLORECOGNITION IN BOTRYLLID ASCIDIANS INFERRED FROM A MOLECULAR PHYLOGENY

C. SARAH COHEN,1,3 YASUNORI SAITO,2 AND IRVING L. WEISSMAN1 IDepartmentof DevelopmentalBiology and HopkinsMarine Station,Stanford University, Stanford, California 94305 2ShimodaMarine Research Station,University of Tsukuba,5-10-1 Shimoda,Shizuoka 415, Japan

Abstract.-Despitethe functionaland phyleticubiquity of highlypolymorphic genetic recognition systems, the evo- lutionand maintenanceof theseremarkable loci remainan empiricaland theoreticalpuzzle. Manyclonal invertebrates use polymorphicgenetic recognition systems to discriminatekin fromunrelated individuals during behavioral inter- actionsthat mediate competition for space. Space competitionmay have been a selectiveforce promoting the evolution of highlypolymorphic recognition systems, or preexistingpolymorphic loci may have been cooptedfor the purpose of mediatingspace competition.Ascidian species in the familyBotryllidae have an allorecognitionsystem in which fusionor rejectionbetween neighboring colonies is controlledby allele-sharingat a single,highly polymorphic locus. The behavioralsequence involvedin allorecognitionvaries in a species-specificfashion with some species requiring extensiveintercolony tissue integrationprior to the allorecognitionresponse, while otherspecies contactopposing colonies at onlya fewpoints on theouter surface before resolving space conflicts.Due to an apparentspecies-specific continuumof behavioralvariation in the degree of intercolonytissue integrationrequired for allorecognition,this systemlends itself to a phylogeneticanalysis of the evolution of an allorecognitionsystem. We constructeda molecular phylogenyof the botryllidsbased on 18S rDNA sequence and mappedallorecognition behavioral variation onto the phylogeny.Our phylogenyshows the basal allorecognitioncondition for the groupis the mostinternal form of the recognitionreaction. More derivedspecies show progressivelymore external allorecognition responses, and in some cases loss of some featuresof internalfunction. We suggestthat external allorecognition appears to be a secondary functionof a polymorphicdiscriminatory system that was alreadyin place due to otherselective pressures such as gamete,pathogen, or developmentalcell lineage recognition. Key words.-Allorecognition,ascidian, behavioral evolution, histocompatibility, space competition.

Received November18, 1997. AcceptedMarch 6, 1998.

Highly polymorphic loci form the genetic basis of rec- 1990; Lang and Chornesky1990). The broad phyleticdis- ognition systemsin diverse taxa fromfungal and angiosperm tributionof tissue recognitionsystems in invertebrateshas mating compatibilityloci to invertebrateand vertebratehis- led to speculationthat these early metazoan historecognition tocompatibilityloci (e.g., reviews in: de Nettancourt 1977; systemsmay containcomponents that are the evolutionary Buss and Green 1985; Grosberg 1988a; papers in Grosberg antecedentsof vertebratetissue recognitionsystems (e.g., et al. 1988; Potts and Wakeland 1990; Cook and Crozier Burnet1971; Scofieldet al. 1982b; Buss and Green 1985; 1995; Kahmann and Bolker 1996; Nauta and Hoekstra 1996). Weissman1988; Du Pasquier1989; Humphreysand Reinherz The evolution and maintainance of highly polymorphic loci 1994). This hypothesisis intriguinggiven the absence of in natural populations has been attributedto a variety of naturalselective scenarios for vertebratehistorecognition evolutionary forces, but theirexistence remains an empirical systems.Since vertebratesdo not,in general,exchange tissue and theoretical mystery. Potential evolutionary scenarios betweenindividuals (although maternal/fetal tissue interac- have invoked a variety of selective and nonselective forces tionsare a notableexception), the selectiveforces that pro- operating eitherwithin the organism or externallyto mediate motedthe evolution of the vertebrate historecognition system intra-or interspecificchallenges (e.g., Buss and Green 1985; are enigmatic.Since manyinvertebrates do possess histo- Klein 1987; Grosberg 1988a; papers in Grosberg et al. 1988; recognitionloci thatare of obviousselective importance dur- Harvell 1990; Potts and Wakeland 1990; Lee and Vacquier ing space competition,an understandingof theevolution of 1992; Brown and Eklund 1994; De Boer 1995; Grosberg et invertebrateallorecognition systems should shed light on al. 1996; Parham and Ohta 1996; Apanius et al. 1997). Ex- mechanismspromoting the development of highlypolymor- isting evolutionary forces may have promoted the de novo phic loci in a broad rangeof taxa (e.g., Burnet1971; Buss appearance of highly polymorphic recognition loci to func- and Green 1985; Grosberg1988a; Weissman1988; Du Pas- tion in theircurrent context, or those evolutionarypressures quier 1989; Brownand Eklund1994; Davidson 1994; Hum- may be responsible for the cooption of preexistingmolecular phreysand Reinherz1994). systems (e.g., discussion in: Burnet 1971; Buss 1982; Buss Single-speciesstudies have been used frequentlyto address et al. 1984; Harp et al. 1988; Weissman 1988; Haig 1996). the questionof evolutionaryforces responsible for allore- Many clonal invertebratetaxa (e.g., sponges, cnidarians, cognitionsystems in clonal invertebrates.Empirical studies bryozoans, and ascidians) possess tissue recognitionsystems have attemptedto measurethe costs and benefitsof fusion, that mediate naturally occuring competition between geno- rejection,and aggressionwithin a singlespecies (e.g., Buss types forlimited space (as reviewed in Grosberg 1988a; Buss 1982; Rinkevichand Weissman1992; Ayre and Grosberg 1995, 1996; Chadwick-Furmanand Weissman, in press). The- 3Presentaddress: Zoology Department,Shoals Marine Labora- oreticalapproaches have modeledcosts and benefitsto look tory,University of New Hampshire,Durham, New Hampshire forevolutionary stable strategies that would lead to theevo- 03824; E-mail: [email protected]. lutionor maintenanceof highlypolymorphic loci (e.g., Cro- 746 ? 1998 The Society for the Study of Evolution. All rightsreserved. ALLORECOGNITION IN BOTRYLLID ASCIDIANS 747 zier 1988; Neigel 1988; Grosbergand Quinn 1989). In con- solid substratawhere they encounter other conspecifics (e.g., trast,we take a multispeciesphylogenetic approach to the Grave 1933; Yamaguchi1975; Grosberg1988b; Bermanet evolutionarypuzzle of highlypolymorphic loci and recog- al. 1992). Interactionsbetween contacting colonies occuras nitionsystems by looking specificallyat the evolutionof a set of species-specificbehaviors involving the outerbody allorecognitionbehavior in a clade of ascidians.Construction covering(tunic) and terminalblood-filled bulbs (ampullae) of phylogenies,followed by independentcharacter-mapping, at the peripheryof each colony.The allorecognitioninter- has provedan extraordinarilypowerful means of gainingin- actioninvolves both cellular (blood and tunic)and humoral sightinto evolutionaryprocesses in many areas including (plasma) componentsassociated with the tunic and ampullae behavior(e.g., Brooksand McClennan1991; Maddisonand of each colony.During a rejectionreaction between two col- Maddison 1992; Martins1996). onies,classic cytotoxicresponses are visibleunder the light Specificallyfor the problem of allorecognitionloci, a phy- microscope(e.g., Scofieldand Nagashima1983). logeneticapproach has been consideredon a broad,whole Naturallycontacting conspecific colonies show species- metazoanscale in generalverbal terms (e.g., Buss and Green specificbehavioral variation in thedegree of intercolonytis- 1985; Harvell 1990; Sima and Vetvicka1990). Some authors sue integrationrequired for a recognitionresponse to occur have pointedout the need for a more carefulphylogenetic (Taneda et al. 1985; Hirose et al. 1988; Saito et al. 1994; analysis of the distributionof allorecognitionsystems in Fig. 1; Table 1). This behavioralvariation produces an ef- metazoans(e.g., Harvell 1990). However,since the genetic fectivedifference in the locationof the allorecognitionre- basis (and ecological details) of most of these invertebrate sponse betweenspecies, with variationin this clade from allorecognitionsystems are knownonly in themost cursory limitedexternal intercolony contact to extensiveinternal tis- fashion,inferences of selectivescenarios based on phylo- sue integration.At one end of thespectrum, recognition fol- geneticinference at thistaxonomic scale are currentlylim- lows minimalouter tissue contact between the two colonies ited.Even withinthe phylum Cnidaria, a multitudeof effector as theirgrowing edges contact and push up againsteach other. structureshave been described,and thereis no reason to There may be limitedblood cell leakage fromthe terminal suppose thesestructures are homologousacross thephylum ampullarvessels and perhapspartial fusion of outertunics (Buss et al. 1984; Buss 1990). at selectedpoints of contact.At the otherextreme in allo- Here,we use a narrower-scalephylogenetic approach (in- recognitionbehavioral sequences are otherspecies that com- trafamily-levelphylogeny), to inferthe ancestralstate and pletelyfuse tunics along thecontacting border and thensub- directionalevolution of allorecognitionbehavior in the as- sequentlyfuse blood vessels and actuallyexchange blood cidian familyBotryllidae. This groupof colonialprotochor- cells withinthe fused vessels beforethe recognition reaction dates are an ideal clade fora comparativeapproach to the occurs.Between these two extremesare a numberof species evolutionof a tissue allorecognitionsystem. Species of co- showingintermediate levels of intercolonyfusion and blood lonial ascidiansin the botryllidfamily use a simplegeneti- exchange beforerecognition occurs. All botryllidspecies cally controlledallorecognition system to mediatecompeti- testedto dateshow the allorecognition response in encounters tion forspace betweenneighboring colonies (Oka and Wa- withconspecifics (reviewed in Rinkevich1992; Hiroseet al. tanabe 1957, 1960; Sabbadin 1982; Scofieldet al. 1982a,b). 1994; Saito et al. 1994). In botryllids,the single locus controllingthe fusion/rejection A phylogenyof thebotryllids based on morphologicalsys- determinationis highlypolymorphic in naturalpopulations tematicsis notpossible at thistime due to uncertaintyabout (Milkman 1967; Mukai and Watanabe1975a,b; Scofieldet the reliabilityof characters(Monniot and Monniot 1987; al. 1982a,b; Grosbergand Quinn 1986; Rinkevichand Saito Monniot1988; Boyd et al. 1990). Here, we presenta phy- 1992; Yund and Feldgarden1992; Rinkevichet al. 1994, logenyof the botryllidascidians using a nuclearribosomal 1995). Dependingon whetherat least one allele at thislocus DNA locus independentof the allorecognitionlocus itself. is sharedbetween these competing conspecific colonies, in- This allows us to independentlyevaluate the evolution of the dividualsmay fuse tissue and becomea potentiallychimeric group,and subsequentlymap the allorecognitionbehavior individual,or they may reject each other'stissue and establish traitonto the phylogeny. We use thisapproach to infer,based a new boundarybetween them. The phylogeneticcharacter- on thebasal characterstate and thepattern of charactertrans- mappingapproach taken here is based on the factthat bo- formationin subsequentlymore derived species, whichse- tryllidsshow species-specificdifferences in allorecognition lectiveforces were primaryin theevolution of allorecogni- behavior.In the botryllidfamily, the apparentsimilarity of tion behaviorin the botryllidsand which selectiveforces responses,effectors, and genetics of the locus, combined with werelikely secondary. If internalevolutionary pressures such the close phylogeneticrelationship of the species involved, as defenseagainst internal pathogens, somatic or germcell makethis a good groupto use forinferences of theselective parasitism,gametic incompatibility systems, or developmen- forcesinvolved in the evolutionand maintenanceof the al- tal lineage sortingled to the evolutionof this locus, then lorecognitionlocus. species with the most internalallorecognition behaviors Botryllidcolonies, like all colonial ascidians,consist of should be basal. Conversely,if externalpressures such as individualfeeding and reproductiveunits called zooids that energeticefficiency in space competitionwere primaryse- are encased in a commonouter covering, the tunic.Zooids lective forcesfor this locus, thena species withthe most in a singlecolony are connectedto each otherand to sausage- limitedtype of tissueinteraction is expectedto be basal in shapedterminal blood-filled bulbs at thetunic periphery by the botryllidphylogeny. a commonblood vascular system.Botryllid colonies fre- We chose thenuclear small ribosomal subunit (18S rDNA) quentlygrow in crowdedconditions in a sheetlikeform on locus for sequence informationbecause it had previously 748 C. SARAH COHEN ET AL.

TABLE 1. Botryllidspecies and locationsused in this studyand locationof thespecies-specific allorecognition response (Taneda et al. 1985; Hirose et al. 1988; Saito et al. 1994; Y. Saito, unpub. obs.). Notes on ingroup(botryllid) samples includedin thisphy- logeny:species designationsattached to some samplesin thisstudy are notwell resolvedin thecurrent literature. In particularlydifficult cases, the samples are labeled in parentheseshere by collecting location.Further molecular work is in progressto resolvethe tax- onomicidentity of thesepopulations.

Stage of allorecognition Geographic Species response location B. scalaris ampullarfusion Japan B. primigenus ampullarpenetration Japan Botryllusscalaris B. sexiens ampullarfusion? Japan B. fuscus partialfusion of tunic Japan (fusiononly in cut assays) B. simodensis partialfusion of tunic Japan B. leachi unknown Italy B. diegense? unknown U.S., NE: Woods (WHD) Hole, MA B. diegense? unknown U.S., SW: Mission (SD) Bay, CA B. violaceous partialfusion of tunic U.S., W: Monterey, (fusiononly in cut CA A A~ A assays) lije ~AA AA A ~~~~2AAQ 7 Monniot et al. 1991) as the closest outgroupto the botryllids. We also include more distantoutgroup species: fromthe Mo- Botryllusprimigenus gulidae (another family in the same order, ) and from a differentorder (Phlebobranchia).

MATERIALS AND METHODS Specimens A)-A2yg The botryllid species sampled for this phylogeny (Table 1) include all species known to exhibit the most extreme differences in allorecognition behavior. Also included are most species described as having intermediate sorts of al- lorecognitionbehavior. In addition, in the full analysis, some specimens are included frombotryllid populations where life- historyand reproductivetraits have been described, or where accidental introductionshave raised questions about the iden- tityof newly invading species. Botrylloidessimodensis At the most extreme end of the behavioral spectrum of botryllid allorecognition, ampullar fusion (the complete FIG. 1. Location of the allorecognitionresponse in threespecies membrane fusion of the vascular systems of opposing col- of botryllids.Each of thethree illustrations show two colonies,on onies and subsequent exchange of blood cells between the theright and leftsides of thedrawing, and theinteraction between two colonies), has been described in them.The tunicborder is delineatedby a thinblack line vertically Botryllusscalaris (Saito drawndown the centerof the illustration.A dashedline indicates and Watanabe 1982) and is also thoughtto be the allorecog- areas of potentialtunic fusion between the two colonies. The blood nition behavior of B. sexiens (Y. Saito, unpubl. obs.). Fol- vascularsystem of each colony,including the protruding ampullar lowing extensive tunic and vascular fusion and blood cell ends, are showncontaining round and triangularblack blood cells exchange, elements in the blood (cellular and plasma) interact withinthe vessels. Blood cells outsidethe vessels indicateleakage into thetunic. between the two colonies. In a rejection reaction, these in- teractingcells turndark and lyse. Parts of the shared vascular system and tunic area become necrotic and the two colonies shown to be appropriate for constructionof molecular phy- both shrinkaway fromthe zone of interaction.Fused vascular logenies at the familylevel in ascidians, including the family elements may be pinched offfrom the parent colony and left under consideration here (Hadfield et al. 1995). To to degenerate in a necrotic zone. Both colonies erect a thick root the botryllid phylogeny, we use solitary styelids, the fibrousbarrier in this zone that is easily visible with a light solitary group basal to the colonial botryllids (Berrill 1950; microscope. ALLORECOGNITION IN BOTRYLLID ASCIDIANS 749

A less extensiveinteraction between opposing colonies DNA was extractedand PCR-amplifiedgenerally follow- involvingextensive outer tunic fusion and penetrationof one ing the methodsand using the primersof Hadfieldet al. colony'sampullae into the opposing colony's tunic (but, with- (1995) withthe following modifications. Most materialwas out fusionbetween opposing colonies' ampullae)is foundin preservedin 70% or 95% ethanol.Ethanol-preserved samples B. primigenus(Oka 1970; Taneda et al. 1985). Here blood were blottedon Kimwipesto remove as much ethanolas and plasma interactionsoccur withinthe tunic,but outside possible and thenfinely minced with a razor blade before the vascularsystems, as blood cells and plasma are leaked extraction.Some extractionseither included CTAB in the into the shared tunic area. Again, as in the vessel-fusing initialextraction buffer or in a second roundof purification speciesdescribed above, if the colonies reject, a darknecrotic afterthe initial extraction.Some otherextractions used a reactionis observedaround the interacting vessels and cells. Chelex 100 resin(BioRad) protocol(adapted from Walsh et Vessels are thenpinched off and abandonedin a necroticarea al. 1991). thatis demarcatedby a thickfibrous barrier between the two Agar bands were purifiedusing a handpackedglasswool colonies. spin columnor a QlAquick gel extractioncolumn (Qiagen In Botrylloidessimodensis, B. violaceous, and B. fuscus, the Inc, Chatsworth,CA) and precipitatedone or two timeswith outersurface of the 's body is the site of allogenic 100% ethanoland 3M NaAc, pH 5.2. PCR productswere tissuefusion and rejection reactions (Hirose et al. 1988, 1990, directlysequenced or cloned withthe TA CloningKit (In- 1994). Fusion of the outercolony covering(tunic) occurs vitrogen,San Diego, CA). Most sequencingwas carriedout onlyat limitedpoints of contactin thesespecies. In thislast withan AppliedBiosystems, Inc., automatedsequencer. Sev- groupof species,blood cells and plasma leak at the points eral sequenceswere also sequencedmanually using the Pro- of contactbetween the two colonies and cytotoxicreactions mega Cycle Sequencingkit and DNA labelledwith 33P. For- occurwithin the tunic between blood cells and plasmaof the wardand reversesequencing or sequencingof a secondsam- opposingcolonies. In thefinal phase of therejection reaction ple was carriedout foreach species. in these species, a new bordercomposed of a thin,fibrous materialis formedbetween the opposing colonies (Hirose et PhylogeneticAnalyses al. 1990). Otherbotryllid samples included in the fullphylogenetic Outgroup Choice.-Trees were rooted with outgroup spe- analysisinclude Botrylloides leachi fromVenice Lagoon, It- cies of differentgenetic distances (based on commonlyac- aly, and specimenscollected from Woods Hole, Massachu- cepted phylogeniesof ascidians; Berrill 1936; Kott 1969; setts,and San Diego, California,that may both represent the Monniotand Monniot1973; Wada et al. 1992). The fullcom- species B. diegensisor B. leachi. Althoughdetailed descrip- plementof outgroups,used in the parsimonyanalysis, in- tionsof thenatural allorecognition behavior of thesespecies cludes otherspecies fromthe familyStyelidae outside the or populationsare not generallyagreed upon or have not botryllidclade (sometimesrecognized as a subfamily,Bo- been published,these samples were included in thephylog- tryllinae;Berrill 1950), anotherspecies fromthe orderSto- enybecause theyrepresent historically and ecologicallyim- lidobranchiatain thefamily Molgulidae, and a species from portantpopulations, and because theirtaxonomic status rel- anotherorder, Phlebobranchiata. Outgroup sequences were ative to otherbotryllids is in question.Current work is in obtainedfrom GenBank (accession numbers are listed in Had- progressto resolvetheir systematic identities and allorecog- fieldet al. 1995). nitionbehavior (C. S. Cohen,University of New Hampshire). Differentmethods of phylogeneticinference were used to In some cases, wheretaxonomic details separatingsym- comparethe resultsobtained with full and partialdatasets patricspecies are particularlydifficult to resolve,field-col- usingmethods that make different assumptions about the data lected specimenswere transferredto glass slides and raised (Felsenstein1988; Hillis et al. 1993; Swofford1996). Phy- in a controlledenvironment to confirmdiagnostic reproduc- logeneticanalyses were carriedout withprograms from the tive and life-historytraits. This culturingwork was carried PHYLIP softwarepackage (vers. 3.56c; Felsenstein1993) out forall of the samplesobtained from Japan. for distanceand maximum-likelihoodanalyses and PAUP vers. 3.1.1; Swofford1990) forthe parsimony analysis. Ge- DNA Sequences neticdistances between sequences were estimated using Ki- mura'stwo-parameter model in theDNADIST program.Phy- Specimenswere used eitherfresh or afterfreezing or eth- logeneticinference according to theneighbor-joining method anol preservation.Matching voucher specimens were kept in was carriedout using the NEIGHBOR program.Statistical ethanoland formalinwhere possible for morphological com- confidenceestimates expressed as bootstrapvalues wereob- parisons. tainedusing SEQBOOT, and consensustrees were generated Molecularstudies with ascidians, and botryllidsin partic- withthe program CONSENSE. Treeswere drawn with PAUP ular,have been difficultdue to the presenceof secondary (parsimony)or withthe hypercard program Tree Draw Deck compoundsassociated withthe DNA (Kumar et al. 1988; (Gilbert1990), whichmakes use of theprograms DrawTree Panceret al. 1995). Experimentationwith a wide varietyof and DrawGramfrom PHYLIP. extractiontechniques permitted this comparative study of bo- Using parsimonyand distance,we experimentedwith dif- tryllidspecies withvarying body chemistry.For thisstudy, ferentcombinations of ingroupand outgrouptaxa to see we developed an extractionprotocol for ethanol-preservedwhich nodes mightbe subject to change (Lecointreet al. species to facilitateshipping of samples fromdistant loca- 1993; Swofford1996). For the distancetests, we restricted tions. theoutgroups to twoto threespecies following the suggestion 750 C. SARAH COHEN ET AL.

TABLE 2. Pairwise distance matrix of all 15 species used in phylogenetic analyses. Numbers below the diagonal represent absolute distances and numbers above the diagonal are mean distances (corrected for missing data). Species abbreviations as in Figure 2.

Ascid 0.127 0.120 0.113 0.117 0.104 0.107 0.104 0.117 0.117 0.121 0.134 0.150 0.101 0.142 fus 39 0.026 0.062 0.042 0.062 0.062 0.062 0.023 0.029 0.026 0.040 0.139 0.059 0.056 leach 37 8 0.039 0.032 0.042 0.042 0.045 0.003 0.010 0.007 0.020 0.125 0.036 0.036 scal 35 19 12 0.052 0.045 0.052 0.042 0.039 0.039 0.042 0.056 0.141 0.039 0.058 prim 36 13 10 16 0.058 0.045 0.055 0.029 0.036 0.033 0.046 0.131 0.049 0.045 Splic 32 19 13 14 18 0.023 0.013 0.042 0.042 0.046 0.059 0.134 0.029 0.061 Cfin 33 19 13 16 14 7 0.016 0.042 0.049 0.046 0.059 0.128 0.026 0.055 Pcorr 32 19 14 13 17 4 5 0.046 0.045 0.049 0.062 0.134 0.023 0.058 WHD 36 7 1 12 9 13 13 14 0.007 0.003 0.016 0.129 0.039 0.036 simo 36 9 3 12 11 13 15 14 2 0.003 0.023 0.132 0.039 0.042 SD 37 8 2 13 10 14 14 15 1 1 0.02 0.129 0.043 0.039 viol 41 12 6 17 14 18 18 19 5 7 6 0.143 0.056 0.049 Mblei 46 42 38 43 40 41 39 41 39 40 39 43 0.125 0.131 Dgro 31 18 11 12 15 9 8 7 12 12 13 17 38 0.059 sex 44 17 11 18 14 19 17 18 11 13 12 15 40 18

of Swofford(1996) and alternatelypicking combinations of PhylogeneticAnalyses ingroupand outgroup species from the full 15-species dataset. Because of the close relationshipsbetween these species For the maximum-likelihoodanalysis, using DNAML in (as shownin the distancematrix, Table 2), the assumption PHYLIP, we used as manyspecies as computationallypos- of similarityof evolutionaryrates between taxa, an assump- sible includingimportant members of theingroup exhibiting tion thatmay be crucialto robustphylogenetic analysis, is therange of allorecognitionbehavior. Outgroup members are strongin thisstudy (e.g., Swofford1996). This assumption limitedto two solitaryStyelidae and one solitaryAscidia in is apparentlyless criticalto maximum-likelihoodmethods theorder Phlebobranchiata. Bootstrap values forthe distance (Hillis et al. 1993), and ourresults with maximum likelihood and maximum-likelihoodanalyses were generatedwith the are in good agreementwith analyses using parsimonyand PHYLIP programsSEQBOOT, DNAML with inputorder distancemethods. In the parsimonyanalysis, of 314 total jumbledthree times, and CONSENSE. sites 64 were variable (20.0%) and 31 were informative (9.9%). RESULTS The botryllidingroup is supportedin all analyses.Botryllid Sequences fromthe most variablecentral portion of the monophylyis not questionedin any ascidian systematiclit- nuclearsmall subunitribosomal locus were trimmedto 314 erature,and thereis no reasonto supposelack of monophyly bases. This portionof 18S rDNA was chosen for analysis here.All analysesplaced colonial botryllidsas derivedwith based on thesurprisingly high levels of variability previously respectto thesolitary styelids. These analysesdo notprovide foundin molgulidascidians (Hadfield et al. 1995). Hadfield informationon theparticular relationship of specificsolitary et al. (1995) suggestthat relative to othertaxa, the small styelidspecies to the botryllids,and thereare manymore subunitribosomal DNA in ascidianseither has a fasterrate species of solitaryand colonial styelids(including colonial of mutation(as also foundin Drosophila;Friedrich and Tautz species outsidethe botryllid clade) thatwould likely need to 1995), or the divergenceof these lineages is more ancient be addedto thephylogeny to gainthis information. Placement thanpreviously suspected. of B. scalaris as thebasal botryllidis particularlywell sup- Sequences were initially aligned using the program portedby themolecular results. Bootstrap support measures GeneWorks(Intelligenetics). Alignments were adjustedby for the placementof B. scalaris were 74% using the full eye. Sequences obtained for this study are deposited in parsimonyanalysis, 88% in a neighbor-joiningdistance anal- GenBank under accession numbersAF008422-AF008429. ysis, and 92% in a maximum-likelihoodanalysis. Accession numbersof previouslysequenced species are Concerningthe branching order of thebotryllids, we con- foundin Hadfieldet al. (1995). Estimationof thetransition/ sistently found the same topologywith regard to the place- transversionratios for the datasetshowed an approximately mentof B. scalaris as basal, followedby otherspecies in the 2:1 ratio and approximatelyequal frequenciesof all bases Botryllus (B. primigenus and B. sexiens), and finally, (rangingfrom 0.20 to 0.33). Botrylloidesspecies as most derived. Botrylloidesappears as a monophyleticcluster in 84% of the maximum-parsimony Species Identifications bootstraps.In the full parsimonyanalysis, Botryllus primi- genusand B. sexiensappear as a sistergroup to theBotryl- The identityof thelocation-specified samples in thisstudy loides species, and both groupsare derivedwith respect to remainsin questionat present.Neither of the specimensof B. scalaris. uncertaintaxonomic identity (WHD and SD) had 18S se- quencesthat showed complete matching with any other sam- DISCUSSION ple thus far. Botrylloides diegense? (SD) and B. diegense? Tree Topology and AllorecognitionCharacter Mapping (WHD), whichcould potentiallyboth have been B. leachior B. diegense,instead appear to be differentfrom each other Phylogeneticanalyses using parsimony, distance, and max- and fromItalian B. leachi as well. imum-likelihoodalgorithms all agreed on the placementof ALLORECOGNITION IN BOTRYLLID ASCIDIANS 751

Ascid Phlebobranchiata fus leach

8 4 WHD Botrylloides simo 5 6 7 4 MSD Botryllidae viol

6 0 prim

sex Botryllus

Stol idobranchiata scal Splic 8 5 - Cfin Styelidae

Pcor 2% sequence divergence Dgro Mblei Molgulidae FIG. 2. Full 15-speciesphylogenetic tree producedwith parsimony analysis. Phylogenetic analyses using parsimony,distance, and maximum-likelihoodalgorithms all agreedon theplacement of all botryllidspecies wherebootstrap support is indicatedat thenodes in Figures2 and 3. Outgroupspecies to thebotryllids include the following solitary ascidians: (1) orderPhlebobranchiata, family Ascidiidae: Ascid, Ascidia ceratodes;(2) orderStolidobranchiata, family Molgulidae: Mblei, Molgula bleizi; (3) orderStolidobranchiata, family Styelidae: Dgro, Dendrodoa grossularia;Pcor, Pelonaia corrugata;Cfin, Cnemidocarpa finmarkiensis; Splic, Styelaplicata. Ingroup botryllidspecies are: scal, Botryllusscalaris; sex,Botryllus sexiens; prim, Botryllus primigenus; fus, Botrylloidesfuscus; leach, Botrylloides leachi; simo,Botrylloides simodensis; viol, Botrylloidesviolaceous. The treeshown is a strictconsensus of the six shortesttrees found using the branchand bound methodwith the furthestaddition sequence optionin PAUP, vers. 3.1.1. These six treesall had an equal lengthof 122 steps and a consistencyindex (CI) of 0.820 (HI = 0.180). Numbersat nodes representpercent of 100 bootstrapsthat supportthat node. all botryllidspecies wherebootstrap support is indicatedat allorecognition behaviors (B. sexiens and B. primigenus) to the nodes in Figures2 and 3. Analysesusing three types of theBotrylloides species withthe mostexternal forms of al- phylogeneticalgorithms (parsimony [Fig. 2], distance,and lorecognitionbehaviors (B. simodensis, B. fuscus, and B. vi- maximumlikelihood [Fig. 3]), whichmake differentkinds olaceous). The allorecognitionmechanisms of some species of assumptionsabout the data and theprocess of molecular are unknowneither from lack of studyor ambiguitycon- evolutionarychange, are in agreementon thebasal stateand cerningspecies identifications(e.g., B. diegensisand B. lea- patternof evolutionwithin the botryllidclade. Further,the chi). Botryllus sexiens is thought to behave similarly to B. topologyof thetree regarding the basal botryllidspecies and scalaris (Y. Saito, unpubl.obs.). the subsequentbotryllid branching arrangement were robust Interestingly,two of the most derivedspecies, B. fuscus to changesin the compositionand numberof ingroupand and B. violaceoushave outertunic recognition abilities, but outgroupspecies. In particular,there is a highlevel of boot- appearto have lostthat ability internally (Hirose et al. 1994). strappedconfidence using all analysesfor the placementof This abilityis assessed using cut colony assays wherethe B. scalaris as the basal botryllidin thisphylogeny. Phylo- outerampullar regions of two colonies are cut away and the geneticreconstructions using DNA sequence data fromthe internalareas are juxtaposedon a slide. Most botryllidspe- 18S nuclearrDNA locus stronglysupport the hypothesis of cies testedthus far show similar rejection responses in fusion internalallorecognition as thebasal statein thebotryllids. assays using whole colonies (outerampullar contacts) and There is also a high level of supportfor the subsequent cut colonies. However,B. fuscus and B. violaceous, species branchingorder from species withinternal or intermediate thatdo notnaturally undergo intercolony internal tissue con- 752 C. SARAH COHEN ET AL.

34 simo

80 leach

55 viol

92 fus Botrylloides 80 prim

scal Botryllus Botryllids

89 Pcor

Splic Styelids Stolidobranchia Ascid Phlebobranchia FIG. 3. Maximum-likelihoodphylogeny of a subsetof the original15 species concentratingon the botryllids.Adjacent cartoons of interactingtunic and ampullarareas forthree species show the locationof the allorecognitionresponse. tact,appear to accept all fusionsusing cut assays,including responsefound in cut colonyassays of thesespecies (as de- fusionwith individuals that they reject in whole colonyas- scribedabove). Experimentationwith a varietyof different says. Interestingly,B. simodensis,another of the species species in an effortto isolate blood componentsof therec- showingthe most external form of allorecognitionbehavior, ognitionsystem has revealed some commoncomponents, does show rejectionin cut assays (Hirose et al. 1990). whichare discussedbelow. Whilelittle is knownabout physiological mechanisms reg- ulatingallorecognition and allowingthis behavioral variation Solitary and Colonial Ascidian AllorecognitionSystems in intercolonyintegration to occurbetween species, it seems likelythat all speciesshare components of a commoneffector Identificationof internalallorecognition as the ancestral system(Taneda et al. 1985; Hiroseet al. 1990). Scofieldand stateis compatiblewith observations using allogeneic and Nagashima(1983; Scofield1987) suggestthat the continuum syngeneicblood mixturesin solitaryascidians. As previously in allorecognitionbehavior represents different species-spe- mentioned,solitary ascidians are ancestralto colonialascid- cific thresholdsfor activationof effectorresponses for re- ians accordingto modernphylogenetic reconstructions of as- jection. Hirose et al. (1994) proposethat differential distri- cidian evolutionusing eithermorphological (Berrill 1936, butionof allorecognitioneffector components in variousspe- 1955; Kott1974; Monniotet al. 1991) ormolecular characters cies, and specifically,a concentrationof theeffectors at the (Wada et al. 1992). Severalspecies of solitaryascidians show tunicwhere allorecognition normally occurs in B. fuscusand cytotoxicreactions in allogeneicblood mixtures(e.g., Hal- B. violaceous,and an absence of those effectorsat interior ocynthiaroretzi: Fuke 1980; Sawada and Ohtake 1994; Styela locationsin the colony,may explainthe lack of a rejection plicata: Raftosand Hutchinson1995). Cytotoxicityhas been ALLORECOGNITION IN BOTRYLLID ASCIDIANS 753 observedin allogeneicbotryllid blood mixturesas well (Ta- TABLE 3. Botryllidlife historycharacters (Berrill 1936, 1947, neda and Watanabe1982; Saito and Watanabe1984; Ballarin 1950, 1955; Kott 1974; Saito et al. 1981a,b; Saito and Watanabe 1982, 1985; Mukai et al. 1987; Monniotet al. 1991). Comparison et al. 1995). Both solitaryand colonial ascidians show a of the two generashows an increasein larval size at hatching,a varietyof cellular and humoral allorecognition response char- decrease in the numberof eggs per zooid, and an increasein de- acteristicsincluding lectin-binding; response to vertebrate velopmenttime between Botryllus and Botrylloidesspecies. antigens;antimicrobial activity, and phagocytic,cytotoxic, Larval # of and encapsulationbehaviors (Ballarin et al. 1993, 1994; Mil- size eggs/ Embry. dvt. Type of lar and Ratcliffe1994; Parrinelloet al. 1996). For botryllids Species (mm) zooid time (d) brooding specifically,allorecognition responses at theblood level are Botryllus observedas a darkeningeither within the cells involvedor scalaris 1.5 2-4 4-5 no pouch associatedwith blood cells thathave leaked out of theiram- primigenus 1.5-1.7 2-4 4-5 pouch pullar vessels and contactedcells or plasma fromthe op- sexiens 1.5-1.8 4 4-5 no pouch posing colony.The darkeninghas been shownto be asso- Botrylloides ciated witha phenoloxidasereaction in at least one species fuscus 2.6-2.7 usu. 1 12-14 pouch (Ballarinet al. 1995). simodensis 1.7-2.0 1 or 2 4-5 pouch It remainsto be establishedthat the blood componentsin leachi 2 2 "a few days" pouch solitaryascidians thatare involvedin internalrecognition violaceous 3 1 or 2 >30 pouch functionsare homologousto blood componentsinvolved in the allorecognitionreaction of colonial ascidians (Satoh brood structuresand most hold their free in the 1994). Laboratory-graftrejection experiments in solitaryas- embryos peribranchialchamber (Berrill 1950). cidiansindicate that for some species thehistocompatibility and Watanabe a series matchingsystem apparently requires complete matching of Saito (1985) propose transformation of correlatedreproductive characters with increasingspe- all alleles at morethan one locus, while forothers a single cializationand care for from sharedallele, as in thebotryllid recognition system, appears parental embryosdirectionally Botryllusspecies to variousBotrylloides species (and other to be sufficientfor graft acceptance (Raftos and Briscoe 1990; authorshave noted these Berrill Rinkevich1992; Davidson 1994). Since solitaryascidians trends,e.g., 1947, 1950; Sabbadin et al. Characters in Table associ- rarely,if ever (but see Schmidt1982), undergonatural fusion 1992). (listed 3) ated with the transformationseries include generaltrends of intactindividuals, the existenceof homologouscellular fromBotryllus species to Botrylloidesspecies of fewereggs and blood-basedrecognition systems in solitaryand colonial perzooid, increasingdevelopment time per embryo, increas- ascidians arguesfor selective forces other than space com- inglarval size at hatching,and development of more enclosed petitionas the evolutionaryimpetus for an allorecognition brood structures.Similar correlated trends in some of these systemsubsequently coopted for botryllidallorecognition. traitshave been for a numberof As researchon blood-basedascidian effector systems and the developmental proposed ascidian families geneticsof historecognitionresponses advances, it will be (Berrill1935; Kott 1969). While two namesfor are stillin interestingto look forhomologies between solitary styelids generic botryllids general a of and other and colonial botryllids.The existenceof clear species iden- use, reanalysis reproductive morphological charactersthat are the basis forthe separationhas led to a tifications(based on moleculardata) and a phylogeneticcon- proposedsynonomization of the genera(Monniot and Mon- textfor interspecific variation should contribute substantially niot 1987; Monniot1988). This molecularphylogeny does to mechanisticstudies on the physiologyand geneticbasis not contradictthe genericdistinction since the Botrylloides forthe allorecognitionresponse. andBotryllus species each clustertogether, however this clus- teringin themolecular phylogeny does notnecessarily justify Concordance of the Molecular Phylogenywith Other separategeneric status either. Characters A detailedmorphological character analysis is not avail- PhylogeneticHypotheses on Allorecognition able at this time for the botryllidsdue to ambiguitiesin Applicationof a comparativephylogenetic approach to the systematiccharacters in this group(Monniot and Monniot evolutionof invertebrateallorecognition systems has been 1987; Monniot1988). However,this phylogeny does agree consideredat much broaderscales (e.g., across phyla,Du withsome previoussuggestions on the ancestraltype of al- Pasquier 1989; Harvell 1990; Davidson 1994; and morespe- lorecognitionbehavior based on morphologicaland behav- cificallywithin the phylum Cnidaria; Buss et al. 1984; Buss ioral observations(Taneda et al. 1985; Hirose et al. 1988, 1990). However,because analysisat thesehigher levels most 1994; Saito et al. 1994). Historically,morphological analysis likely deals with nonhomologousallorecognition effector of this grouphas separatedthe botryllidsinto two genera, structures(Buss 1990) and thegenetics of allorecognitionfor Botryllusand Botrylloides.The placement of Botrylloidesspe- nearlyall ofthese systems is currentlyunknown, the question cies as derivedwith respect to Botryllusspecies is supported of selectiveforces promoting the evolutionof the polymor- by developmentaland reproductivecharacters (Berrill 1947, phic recognitionloci functionallyrelated to thesestructures 1950; Mukai 1977; Saito and Watanabe1985; Mukai et al. becomes morecomplicated. The advantageof usingthe bo- 1987; Table 3). For example,all describedBotrylloides spe- tryllidclade for a phylogeneticanalysis of allorecognition cies to datehave separatespecialized pouches where embryos lies in the factthat there is substantialbehavioral variation are brooded,however few Botryllus species have specialized based on a singleeffector system. 754 C. SARAH COHEN ET AL.

Phylogeneticreconstruction of theevolution of allorecog- . 1950. The Tunicatawith an accountof theBritish species. nitionbehavior using molecularsystematics in this study Royal Societyof London,London. . 1955. The originof the vertebrates.Oxford Univ. Press, showsthat allorecognition at thewhole colonylevel forpur- London. poses of space competitionis unlikelyto have been theorig- BOYD, H., I. WEISSMAN, AND Y. SAITO. 1990. Morphologicaland inal evolutionaryforce selecting for a highlypolymorphic geneticverification that Monterey Botryllus and Woods Hole recognitionsystem in botryllidascidians, and perhaps,other Botryllusare the same species. Biol. Bull. 178:239-250. BROOKS, D., AND D. McLENNAN. 1991. Phylogeny,ecology, and colonial protochordatesas well. In styelidascidians, it ap- behavior.Univ. of Chicago Press,Chicago. pears morelikely that internal discriminatory functions se- BROWN, J., AND A. EKLUND. 1994. Kin recognitionand themajor lectedfor blood-mediated recognition mechanisms in solitary histocompatibilitycomplex: an integrativereview. Am. Nat. 143: ascidians,and thesesystems were subsequentlycoopted for 435-461. externalspace competitivepurposes in colonial botryllids. BURNET, F 1971. "Self-recognition"in colonialmarine forms and floweringplants in relationto theevolution of immunity.Nature Whetherthis pattern of internalto externalallorecognition 232:230-235. functionin botryllidsis reflectedin the evolutionof allo- Buss, L. 1982. Somaticcell parasitismand the evolution of somatic recognitionsystems in othercolonial invertebratesremains tissuecompatibility. Proc. Nat. Acad. Sci. 79:5337-5341. to be seen. . 1990. Competitionwithin and betweenencrusting clonal invertebrates.Trends Ecol. Evol. 5:352-356. Buss, L., AND D. R. GREEN. 1985. Histoincompatibilityin verte- ACKNOWLEDGMENTS brates:the relict hypothesis. Dev. Comp. Immunol.9:191-201. We thankW. Buss, L., C. McFADDEN, AND D. KEENE. 1984. Biology of hy- Jeffrey(U.C. Davis and Bodega MarineLab) dractiniidhydroids. 2. Histocompatibilityeffector system/com- and his laboratory,including C. Olsen and K. Hadfield;and petitivemechanism mediated by nematocystdischarge. Biol. M. Smith(Simon FraserUniversity) and his laboratory,in- Bull. 167:139-158. cludingK. Beckenbach.For assistancein obtainingspeci- CHADWICK-FURMAN, N., AND I. WEISSMAN. 1998. Fitnesscosts of mens,we thankJ. Carlton,A. N. Cohen,R. Dukas, D. Har- allorgeneiccontacts in field-growncolonies of the protochordate Botryllus schlosseri. J. Exp. Zool. In press. vell, K. Ischizuka,I. Mackenzie,D. McHugh,C. Mills, K. COOK, J.,AND R. CROZIER. 1995. Sex determinationand population Palmeri,B. Rinkevich,R. da Rocha,A. Sabbadin,A. Sewell, biologyin the hymenoptera.Trends Ecol. Evol. 10:281-286. M. Sewell, and K. Wasson. For discussionof problemsof CROZIER, R. 1988. Kin recognitionusing innatelabels: a central botryllididentity, we thankS. Bartl,J. Carlton, R. Grosberg, role forpiggy-backing? Pp. 143-156 in R. Grosberg,D. Hed- gecock,and K. Nelson, eds. Invertebratehistorecognition. Ple- and T. Newberry.For assistancewith phylogenetic inference numPress, New York. programs,we thankA. C. Cohen and J. Mountain.We also DAVIDSON, E. 1994. Stepwiseevolution of major functionalsys- thankS. Craig for commentson an earlierversion of the temsin vertebrates,including the immune system. Pp. 133-142 manuscript.Financial support for this study was providedby in J. Hoffman,C. Janeway,and S. Natori,eds. Phylogenetic a StanfordUniversity Medical School Dean's postdoctoral perspectivesin immunity:the insect host defense. R. G. Landes, Austin,TX. fellowshipto CSC, Grant-in-Aidfor ScientificResearch DE BOER, R. 1995. The evolutionof polymorphiccompatibility #08459005from the Ministry of Education, Science and Cul- molecules.Mol. Biol. Evol. 12:494-502. tureof Japanto YS, and by a grantfrom Systemix/Sandoz DE NETTANCOURT, D. 1977. Incompatiblityin angiosperms. to IW. Springer-Verlag,New York. Du PASQUIER, L. 1989. Evolutionof theimmune system. Pp. 139- 165 inW. Paul, ed. Fundamentalimmunology. Raven Press, New LITERATURE CITED York. APANIUS, V., D. PENN, P. SLEV, L. RUFF, AND W. POTTS. 1997. The FELSENSTEIN, J. 1988. Phylogeniesfrom molecular sequences: in- natureof selectionon the major histocompatibilitycomplex. ferenceand reliability.Annu. Rev. Genetics22:521-565. Crit.Rev. Immunol.17:179-224. 1993. PHYLIP-phylogeny inferencepackage. Vers. 3.5c. AYRE, D., AND R. GROSBERG. 1995. Aggression,habituation, and Univ. of Washington,Seattle, WA. clonal coexistencein the sea anemoneAnthopleura elegantis- FRIEDRICH, M., AND D. TAUTZ. 1995. RibosomalDNA phylogeny sima. Am. Nat. 146:427-453. of themajor extant arthropod classes and theevolution of myr- . 1996. Effectsof social organizationon interclonaldom- iapods. Nature376:165-167. inance relationshipsin the sea anemoneAnthopleura elegantis- FUKE, M. 1980. Contactreaction between xenogeneic or allogeneic sima. Anim.Behav. 51:1233-1245. coelomic cells of solitaryascidians. Biol. Bull. 158:304-315. BALLARIN, L., F CIMA, AND A. SABBADIN. 1993. Histoenzymatic GILBERT, D. 1990. Tree draw deck, Biology Department,Indiana stainingand characterizationof the colonial ascidianBotryllus Univ.,Bloomington, IN. schlosserihemocytes. Bull. Zool. 60:19-24. GRAVE, B. 1933. Rate of growth,age at sexual maturity,and du- 1994. Phagocytosisin the colonial ascidian Botryllus rationof life of certainsessile organisms,at Woods Hole, MA. schlosseri.Dev. Comp. Immunol.18:467-481. Biol. Bull. 65:375-386. . 1995. Morula cells and histocompatibilityin the colonial GROSBERG, R. 1988a. The evolutionof allorecognitionspecificity ascidian Botryllus schlosseri. Zool. Sci. 12:757-764. in clonal invertebrates.Q. Rev. Biol. 63:377-412. BERMAN, J.,L. HARRIS, W. LAMBERT, M. BUTTRICK, AND M. DUF- . 1988b. Life-historyvariation within a populationof the RESNE. 1992. Recentinvasions of theGulf of Maine: threecon- colonial ascidian Botryllusschlosseri. I. The geneticand envi- trastingecological histories.Conserv. Biol. 6:435-441. ronmentalcontrol of seasonal variation.Evolution 42:900-920. BERRILL, N. 1935. Studies on tunicatedevelopment. III. Differ- GROSBERG, R., AND J. F QUINN. 1986. The geneticcontrol and entialretardation and acceleration.Philo. Trans.R. Soc. Lond. consequencesof kinrecognition by thelarvae of a colonialma- B Biol. Sci. 225:255-326. rineinvertebrate. Nature 322:456-459. . 1936. Studiesin tunicatedevelopment. V. The evolution . 1989. The evolutionof selectiveaggression conditioned and classificationof ascidians.Philo. Tran. R. Soc. Lond. B Biol. on allorecognitionspecificity. Evolution 43:504-515. Sci. 226:43-70. GROSBERG, R., D. HEDGECOCK, AND K. NELSON. 1988. Invertebrate . 1947. The developmentalcycle of Botrylloides.Q. J. Mi- historecognition.Plenum Press, New York. croscop.Sci. 88:393-407. GROSBERG, R., D. LEVITAN, AND B. CAMERON. 1996. Evolutionary ALLORECOGNITION IN BOTRYLLID ASCIDIANS 755

geneticsof allorecognitionin the colonial hydroidHydractinia compoundascidian Botrylloides violaceous. Proc. Jpn. Acad. Sci. symbiolongicarpus. Evolution 50:2221-2240. 51:48-50. HADFIELD, K., B. SWALLA, AND W. JEFFERY. 1995. Multipleorigins MUKAI, H., Y. SAITO, AND H. WATANABE. 1987. Viviparousde- of anural developmentin ascidians inferredfrom rDNA se- velopmentin Botrylloides(compound ascidians). J. Morphol. quences. J. Mol. Evol. 40:413-427. 193:263-276. HAIG, D. 1996. Gestationaldrive and the green-beardedplacenta. NAUTA, N., AND R. HOEKSTRA. 1996. Vegetativeincompatibility Proc. Nat. Acad. Sci. 93:6547-6551. in ascomycetes:highly polymorphic but selectivelyneutral? J. HARP, J., C. TSUCHIDA,I. WEISSMAN,AND V. SCOFIELD. 1988. Au- Theor.Biol. 183:67-76. toreactiveblood cells and programmedcell deathin growthand NEIGEL, J. 1988. Recognitionof self or nonself?Theoretical im- developmentof protochordates.J. Exp. Zool. 247:257-262. plicationsand and empiricaltest. Pp. 127-142 in R. Grosberg, HARVELL, C. D. 1990. The evolutionof inducibledefense. Para- D. Hedgecock, and K. Nelson. Invertebratehistorecognition. sitology100:S53-S61. PlenumPress, New York. HILLIS, D., M. ALLARD, AND M. MIYAMOTO. 1993. Analysis of OKA, H. 1970. Colonyspecificity in compoundascidians. Pp. 196- DNA sequencedata: phylogeneticinference. Methods Enzymol. 206 inH. Yukawa,ed. Profilesof Japanese science and scientists. 224:456-487. Kodansha,Tokyo. HIROSE, E., Y. SAITO, AND H. WATANABE. 1988. A new typeof OKA, H., AND H. WATANABE. 1957. Colonyspecificity in compound manifestationof colony specificityin the compoundascidian, ascidiansas testedby fusionexperiments. Proc. Jpn.Acad. 33: Botrylloides violaceus Oka. Biol. Bull. 175:240-245. 657-659. . 1990. Allogeneicrejection induced by cut surfacecontact . 1960. Problemsof colonyspecificity in compoundascid- in the compound ascidian, Botrylloides simodensis. Invertebr. ians. Bull. Mar. Biol. Stn. Asamushi10:153-155. Reprod.Dev. 17:159-164. PANCER, Z., H. GERSHON, AND B. RINKEVICH. 1995. Isolation of . 1994. Surgicalfusion between incompatible colonies of highmolecular weight DNA fromcolonies of botryllid ascidians the compoundascidian, Botrylloidesfuscus. Dev. Comp. Im- (). Tech. Fish Immunol.4:23-25. munol. 18:287-294. PARHAM,P., AND T. OHTA. 1996. Populationbiology of antigen HUMPHREYS, T., AND E. REINHERZ. 1994. Invertebrateimmune rec- presentationby MHC class I molecules.Science 272:67-74. ognition,natural immunity, and theevolution of positiveselec- PARRINELLO,N., M. CAMMARATA, AND V. ARIZZA. 1996. Univa- tion.Immunol. Today 15:316-320. cuolar refractilehemocytes from the tunicateCiona intestinalis KAHMANN, T., AND M. BOLKER. 1996. Self/nonselfrecognition in are cytotoxicfor mammalian erythrocytes in vitro.Biol. Bull. fungi:old mysteriesand simplesolutions. Cell 85:145-148. 190:418-425. KLEIN, J. 1987. Originof the major histocompatibilitycomplex POTTS, W., AND E. WAKELAND. 1990. Evolutionof diversityat the polymorphism:the trans-species hypothesis. Hum. Immunol. 19: major histocompatibilitycomplex. TrendsEcol. Evol. 5:181- 155-162. 187. KOTT, P. 1969. AntarcticAscidiacea. Antarct. Res. Ser.Washington RAFTOS, D., AND D. BRISCOE. 1990. Geneticbasis of histocom- 13:1-239. patibilityin thesolitary urochordate Styela plicata. J.Hered. 81: . 1974. The evolutionand distributionof Australian tropical 96-100. Ascidiacea. Pp. 405-423 in Proceedingsof the second inter- RAFTOS, D., AND A. HUTCHINSON. 1995. Cytotoxicityreactions in nationalcoral reefsymposium 1. The GreatBarrier Reef Com- the solitary Styela plicata. Dev. Comp. Immunol. 19: mittee,Brisbane, Australia. 463-471. KUMAR, S., B. DEGNAN, AND M. LAVIN. 1988. Cloningof a major RINKEVICH,B. 1992. Aspectsof the incompatibilitynature in bo- repeatDNA sequencefrom Pyura stolonifera. DNA 7:433-440. tryllid ascidians. Anim. Biol. 1:17-28. LANG, J., AND E. CHORNESKY. 1990. Competitionbetween scler- RINKEVICH,B., AND Y. SAITO. 1992. Self-nonself recognition in actinianreef corals-a reviewof mechanismsand effects.Pp. the colonial protochordateBotryllus schlosseri fromMutsu Bay, 209-252 in Z. Dubinsky,ed. Coral reefs.Elsevier, Amsterdam. Jpn.Zool. Sci. 9:983-988. LECOINTRE, G., H. PHILIPPE, H. LE, AND H. GUYADER. 1993. Species RINKEVICH,B., AND I. WEISSMAN. 1992. Chimerasvs genetically samplinghas a major impacton phylogeneticinference. Mol. homogeneousindividuals: potential fitness costs and benefits. Phylog.Evol. 2:205-224. Oikos 63:119-124. LEE, Y.-H., AND V. D. VACQUIER. 1992. The divergenceof species- RINKEVICH,B., T LILKER-LEVAV,AND M. GOREN. 1994. Allore- specificabalone spermlysins is promotedby positiveDarwinian cognition and xenorecognition responses in Botrylloides (Asci- selection.Biol. Bull. 182:97-104. diacea) subpopulations from the Mediterranean coast of Israel. MADDISON, W., AND D. MADDISON. 1992. MacClade: analysisof J. Exp. Zool. 270:302-313. phylogenyand characterevolution. Sinauer, Sunderland, MA. RINKEVICH,B., R. PORAT, AND M. GOREN. 1995. Allorecognition MARTINS, E. 1996. Phylogeniesand the comparativemethod in elements on urochordate histocompatibility locus indicate un- animalbehavior. Oxford Univ. Press,New York. precedented extensive polymorphism.Proc. R. S. Lond. B Biol. MILKMAN, R. 1967. Geneticand developmental studies on Botryllus Sci. 259:319-324. schlosseri.Biol. Bull. 132:229-243. SABBADIN,A. 1982. Formal genetics of ascidians. Am. Zool. 22: MILLAR, D., AND N. RATCLIFFE. 1994. Invertebrates.Pp. 29-68 in 765-773. R. Turner,ed. Immunology:a comparativeapproach. Wiley, New SABBADIN,A., P. BURIGHEL,P., AND G. ZANIOLO. 1992. Some as- York. pects of reproductionin ascidians. Pp. 103-115 in R. Dallai, ed. MONNIOT, C. 1988. Ascidiesde Nouvelle-Caledonie.IV. Styelidae Sex origin and evolution 6. Mucchi, Modena, Italy. (suite). Bull. Mus. Nat. Hist. Nat. Paris 10:163-196. SAITO, Y., AND H. WATANABE. 1982. Colony specificity in the MONNIOT, C., AND F MONNIOT. 1973. Clefmondialedes genres compound ascidian, Botryllus scalaris. Proc. Jpn. Acad. 58B: d'ascidies. Arch.Zool. Exp. Gen. 113:311-367. 105-108. . 1987. Les ascidiesde polynesefrancaise. Mem. Mus. Nat. . 1984. Partialbiochemical characterization of humoral fac- Hist. Nat. Paris 136:1-154. torsinvolved in the nonfusionreaction of a botryllidascidian, MONNIOT, C., F. MONNIOT, AND P. LABOUTE. 1991. Coral reefas- Botrylloides simodensis. Zool. Sci. 1:229-235. cidians of New Caledonia. ORSTOM, Paris. . 1985. Studieson Japanesecompound styelid ascidians IV. MUKAI, H. 1977. Comparativestudies on the structureof repro- Three new species of the genus Botrylloidesfrom the vicinity ductiveorgans of fourbotryllid ascidians. J. Morphol. 152:363- of Shimoda.Publ. Seto Mar. Biol. Lab. 30:227-240. 380. SAITO, Y., H. MUKAI, AND H. WATANABE. 1981a. Studiesof Jap- MUKAI, H., AND H. WATANABE. 1975a. Distribution of fusion in- anese compoundstyelid ascidians I. Two new species of Bo- compatibility types in natural populations of the compound as- tryllusfrom the vicinity of Shimoda.Publ. Seto Mar. Biol. Lab. cidian Botryllusprimigenus. Proc. Jpn. Acad. Sci. 51:44-47. 26:347-355. . 1975b. Fusibility of colonies in natural populations of the . 1981b. Studieson Japanesecompound styelid ascidians. 756 C. SARAH COHEN ET AL.

II. A new species of the genus Botrylloides and redescription of C. Moritz,and B. Mable, eds., Molecularsystematics. Sinauer, B. violaceus Oka. Publ. Seto Mar. Biol. Lab. 26:357-368. Sunderland,MA. SAITO, Y., E. HIROSE, AND H. WATANABE. 1994. Allorecognition TANEDA, K., AND H. WATANABE. 1982. Studies on colony speci- in compound ascidians. Int. J. Dev. Biol. 38:237-247. ficityin the compoundascidian, Botryllus primigenus Oka. II. SATOH, N. 1994. Developmental biology of ascidians. Cambridge In vivo bioassay foranalyzing the mechanismof "nonfusion" Univ. Press, Cambridge, U.K. reaction.Develop. Comp. Immunol.6:243-252. SAWADA, T., AND S.-I. OHTAKE. 1994. Mixed-incubation of allo- TANEDA, K., Y. SAITO, AND H. WATANABE. 1985. Self or non-self geneic hemocytes in tunicate Halocynthia roretzi. Zool. Sci. 11: discriminationin ascidians.Zool. Sci. 2:433-442. 817-820. WADA, K., W. MAKABE, M. NAKAUCHI, AND N. SATOH. 1992. Phy- SCOFIELD, V. 1987. Control of fusibilityin colonial . Pp. logeneticrelationships between solitary and colonial ascidians, 29-48 in A. Greenberg, ed. Invertebratemodels: cell receptors as inferredfrom the sequence of the centralregion of theirre- and cell communication. S. Karger AG, Basel, Switzerland. spective18S rDNAs. Biol. Bull. 183:448-455. SCOFIELD, V., AND L. NAGASHIMA. 1983. Morphology and genetics WALSH, P, D. METZGER, AND R. HIGUCHI. 1991. Chelex 100 as a of rejection reactions between oozooids from the tunicate Bo- mediumfor simple extractionof DNA for PCR-based typing tryllusschlosseri. Biol. Bull. 165:733-744. fromforensic material. BioTechniques 10:506-513. SCOFIELD, V., J. SCHLUMPBERGER, AND I. WEISSMAN. 1982a. Colony WEISSMAN, I. 1988. Was the MHC made forthe immunesystem, specificity in the colonial tunicate Botryllus and the origins of or did immunitytake advantage of an ancientpolymorphic gene vertebrateimmunity. Am. Zool. 22:783-794. familyencoding cell surfaceinteraction molecules? A specu- SCOFIELD, V., J. SCHLUMPBERGER, J. WEST, AND I. WEISSMAN. lative essay. Intern.Rev. Immunol.3:393-413. 1982b. Protochordate allorecognition is controlled by a MHC- YAMAGUCHI, M. 1975. Growthand reproductivecycles of thema- like gene system. Nature 295:499. rine fouling ascidians Ciona intestinalis,Styela plicata, Botryl- loides violaceus, and Leptoclinum mitsukuriiat Aburatsubo-Mo- SIMA, P., AND V. VETVICKA. 1990. Evolution of immune reactions. roiso Inlet (CentralJapan). Mar. Biol. 29:253-259. CRC Press, Boston, MA. YUND, P., AND M. FELDGARDEN. 1992. Rapid proliferationof his- SWOFFORD, D. 1990. PAUP: phylogenetic analysis using parsi- torecognitionalleles in populationsof a colonialascidian. J. Exp. mony. Vers. 3.1.1. Illinois Natural History Survey, Champaign, Zool. 263:442-452. IL. . 1996. Phylogenetic inference. Pp. 514-544 in D. Hillis, CorrespondingEditor: T. Markow