Self-Nonself Recognition in the Colonial Protochordate Botryllus schlosseri from Mutsu Bay, Japan(Immunology)

6 Pages·1992·2.3 MB·English

by  Baruch Rinkevich

ZOOLOGICAL SCIENCE 9: 983-988 (1992) © \992 Zoological SocietyofJapan
Self-Nonself Recognition in the Colonial Protochordate
Botryllus schlosseri from Mutsu Bay, Japan1
Baruch Rinkevich2 and Yasunori Saito3
^National Institute ofOceanography, Israel Oceanographic & Limnological
Research, Tel-Shikmona, P.O.B. 8030, Haifa 31080, Israel and
^Shimoda Marine Research Center, University of Tsukuba
5-10-1 Shimoda-shi, Shizuoka 415, Japan
—
ABSTRACT WildBotryllusschlossericollectedfroma5x5mareainMutsuBay(Aomori Prefecture,
Japan) were tested foralloresponses in intrapopulation colonyallorecognition assays (CAAs). Results
indicate that rejection patterns are similar to those recorded previously in the populations from
Monterey and Santa Barbara. California, from the Mediterranean coast of Israel, and from the
Venetianlagoon, Italy. Theonlydifferencewasthemarkedaccumulationofbright-yellowbloodcellsin
the tips of interacting ampullae. Pairwise CAAs which were performed on all combinations (n=91)
from 14 colonies resulted in 12.1% fusions, which gives a populational estimation of32 alleles on the
fusibilitylocus. Afusibilitychartforthese 14genotypesrevealed 18-19different, notequallyfrequent,
allorecognitionalleles,ofwhich 12wereassignedtoonlyonegenotype,each;5occurredtwice,andeach
oneof2 alleleswas present in three differentgenotypes. It isconcluded that the sampled areawastoo
small to represent the probable higher number offusibility alleles residing in this population.
This species has been studied for intrapopulation
INTRODUCTION
allorecognition responses in 4 remote localities:
It is now almost 90 years that studies on the the population from the Venetian lagoon, Italy [6,
colonial tunicate Botryllus schlosseri (Pallas) have 9], from the Mediterranean coast of Israel [10],
shown its capacity for colony specificity [1], a from Woods Hole, Massachusetts, Atlantic Ocean
histocompatibility systemwhich resemblesthe ma- [2, 11, 12], and from the Monterey and Santa
jor histocompatibility complex (MHC) of the Barbara areas, California, Pacific Ocean [11-15].
vertebrates in many aspects [2]. The genetic basis Results indicated that allogenic interactions be-
for the colony specificity resides in a single, highly tween ampullae of noncompatible Botryllus con-
polymorphic haplotype (called the fusibility/histo- specifics exhibited not only the species-specific
compatibility locus, Fu/HC [3]), which possesses characteritics of rejection processes, but also
multiple codominantly expressed alleles. Interact- population-specific rejection types which were
ing colonies which do not share any allele on this constantly expressed even in interpopulation en-
locus rejecteach other, while colonies which share counters[11. 12]. Whentwoincompatiblecolonies
in common at least one allele on the Fu/HC locus of the Woods Hole population came into tunic-
may undergo a natural transplantation (fusion) by tuniccontact, a limited fusion ofthe cortical layers
forming vascular anastemoses between their of both colonies prevailed, which resulted in a
peripheral ampullae [2, 4-6]. continuous tunic matrix between both partners.
B. schlosseri is a cosmopolitan inhabitant of This permitted a reciprocal ampullae penetration
shallow water, hard bottom communities [7, 8]. which was followed by ampullae amputation and/
or haemorrhages formation, the development of
Accepted July 1, 1992 dark-brown necrotic areas, points of rejection
Received April 16, 1992
(PORs). In the other 3 studied B. schlosseri
Contribution number 541 from the Shimoda Marine
Research Center. populations, on the otherhand, the outer layers of
To whom all correspondence should be addressed. the tunics did not fuse during allogeneic encoun-
984 B. RlNKEVICH AND Y. SaITO
ters. As a result, the cortical layers were clearly reciprocally positioned in tip-tip orientation where
observed as demarcating lines between colonies. the cortical layers of the allogeneic partners were
PORs, therefore, were developed without recip- clearlyobserved asdemarcatinglinesalongcontact
rocal ampullae penetration. areas. When compatible pairs were assayed, the
The above studies [12, 13] have suggested that tunic matrices eventually fused in limited or in
an extended comparison of intrapopulation broader areas. Ampullae of both partners, or of
allogeneic interactions would be of great benefit only one partner in a pair, penetrated into the
for a better understanding of self/nonself histo- tunic matrixofthe othercolony through the fusion
compatibility alloresponses of this cosmopolitan areas and were positioned in tip-ampulla base
species. Here we study allorecognition of a B. (proximal part) orientations. This resulted, within
schlosseri population from Mutsu Bay, Japan, and less than 24hr, in allogeneic anastomosis of blood
try to evaluate the polymorphism pattern of the vessels, the formation ofchimeras. Chimeraswere
Fu/HC locus by calculating frequencies of fusion followed upforaperiodofone month. Duringthis
within a population sampled from a small area. period, chimerism in several assays already ended
in the resorption ofall zooids ofone partnereach.
When nonfusible partners come into direct con-
MATERIALS AND METHODS
tact, tunics did not fuse together, so ampullae
Wild B. schlosseri colonies were collected engaged each other reciprocally but indirectly,
(November 1991) in Mutsu Bay, Aomori (Aomori through both cortical layers (Fig. la-i). During
Prefecture), where they were grown in shallow this process, the tips of interacting ampullae, and
water on floating Pecten culturing cages. All sometimes also the tips of all peripheral ampullae
colonies were collected from a5x5 m area. Large (even those of the other sides of the colony), of
healthy colonies were removed from substrates by one or both partners in a pair, became very
razorblades, tiedwiththin cotton fibersonto2.5X distinctive in bright fluorescent yellow color. This
7.5 cm glass slides, and shipped to Shimoda resulted from aggregations of bright yellow blood
Marine Research Center, where they were kept cells which accumulated in the ampullar tips (Fig.
vertically in slots within a wooden culture box, la. b). This phenomenon was also observed in
submerged in Nabeta Bay, Shimoda. For the interacting colonies of the Monterey population
colony allorecognition assays (CAAs), we used [11, 12] and the Mediterranean colonies [10]. By
small groups of zooids at the growing edges of employing histological examinations, we (Terta-
colonies, carefully isolated from colonies. These kover and Rinkevich, unpublished) characterized
subclones were attached in pairs on 5x7.5 cm the accumulated cells as morula cells. However,
glass slides, as described previously [11-15]. Dur- the Mediterranean and Monterey populations did
ing the experiments, the colonies were kept in not exhibit the deeper intensity in color and the
17-liter standing seawater tanks, aerated by air- high frequency of cases as it is recorded in the
stones and maintained at 17-18°C by aquarium Japanese population. In most of the yellow col-
heaters. CAAs were observed daily and cleaned ored tips, the intenseyellowcolorgraduallyabated
during the observations by soft small brush. Food and disappeared within the next 24-72 hr. while in
was supplied daily (artificial diet; Liquifry Marine, some of the tips the color became dark-brown,
England). blood cells infiltrated out ofthe ampullar tips, and
the formation ofPORs through haemorrhageswas
documented (Fig. la-i). In a few cases, ampullae
RESULTS
were amputated from the peripheral blood vessels
and gradually disintegrated. Only a few (1-8) out
Intrapopulation alloresponses of many (up to tens) of interacting ampullae
Contactsbetweenextended ampullaeofencoun- produced cytotoxic lesions (Fig. la. e). Following
ter colonies were established within 24 hr after the acute phase of allogeneic response, where all
formingthe CAA. Tipsofmarginal ampullae were PORs were developed within a short period of a
Allorecognition in Japanese Botryllus 985
Fig. 1. Allogeneic interactions between B. schlosseri colonies: genotype combination 3 (left) vs 2 (a-d), and
genotype combination 10 (left) vs3 (e-i). a-48hrafterCAA. Six small reciprocal PORs along the contact area
(arrowheads) outof12vs 17 interactingampullae. Tipsofampullae (genotype 3) still possessed aggregationsof
brightyellowcellsinthecontactareaandinotherperipheralampullae(confinedbetweentwowhitearrowheads).
b-Twodaysthereafter. NomorePORswereadded. Ampullaewereclearedofyellowcellaggregations,andonly
9 vs 7 were still interacting. Genotype 2 started retreat growing [14]. c-Two days later. Only a few colony 2
ampullaewerestillincontactwiththoseofcolony3. Colony2zooidsretreated 1-2mm. d-7days later. Zooids
of colony 2 retreated more than 3mm from the initial location, leaving an unvascularized tunic matrix with
rudimentsofbloodvessels/ampullaeandmassesofdegeneratedinfiltratingbloodcells(smallblack arrowheads).
e-48hrafterdoingCAA. 17vs 14interactingampullae, respectively. 7PORs(arrowheads)wereallproducedby
colony 10, but some were not yeat well developed. In genotype 10, the upper peripheral ampullae still had the
yellowcolor. f-Twodaysthereafter. AllformerPORswerefullydevelopedandnomorewereadded. Ampullae
were cleared of yellow color, g-h-4 days later. Some of the large necrotic lesions were diffused and dispersed
(arrowheads), forming a black line along the contact area between the two genotypes. Interacting ampullae of
both colonies (predominantly of genotype 10) retreated, leaving behind degenerated ampullae and masses of
dying cells (small black arrowheads), i-5 days thereafter. Ampullae of both genotypes reciprocally retreated,
leaving degenerating areas. Scale ofbars: a-g, i=1 mm. h=0.25 mm.
few days (usually within 48 hr), other ampullae enhanced by the "retreat growth phenomenon"
continued to interact for longer periods, but with- [14], where fewer buds than zooids per generation
outtheformationofany more PORs (Fig. Id, g-i). were developed in the contact area, resulting in a
It was also evident in many cases that during POR directional colony growth form, away from in-
development and thereafter, ampullae of one or teracting zones (Fig. lb-d).
both colonies in a pair withdrew from the contact
areas-, leavingabare tunicwhichgraduallydeterio- Pairwise allorecognition assays
rated (Fig. Id, h, i). This process was sometimes We examined 14Japanese B. schlossericolonies
986 B. RlNKEVICH AND Y. SaITO
in a pairwise allorecognition panel of all 91 com- pattern as in the populations from the Venetian
binations. Fusions were recorded in 11 (12.1%) lagoon, fromthe Mediterranean coastofIsraeland
allogeneic assays (Fig. 2a, striped squares) as well from the Monterey-Santa Barbara areas [6, 9, 10-
as in all the controls, the isogeneic combinations 15]. The only difference which we could detect
(Fig. 2a, striped triangles). An Fu/HC chart for which characterized the Japanese population from
the studied 14 genotypes (Fig. 2b) revealed 18 to all the above is the marked accumulation ofyellow
19 different fusibility alleles. Genotypes 11 and 2 colored cell aggregations in the tips of interacting
are clearly distinguished from each other since ampullae. Therefore, out of the 5 studied B.
they possess different color morphs. However, schlosseri populations up to date, only the Woods
according to the predictions ofthe fusibility model Hole population differs significantly from all the
(legend to Fig. 2b), they share in common either others when comparing between the effector
one or both Fu/HC alleles. Twelve out of the mechanisms which are expressed during intra- or
maximum 19 Fu/HC alleles (63.2%) are assigned interpopulation interactions [11, 12, 16]. Since B.
to only one colony each, while 5 alleles (F, I, J, P, schlosseri is found in many additional localities
R) (26.3%) occurred twice, and 2 alleles (M, N) around the world [7, 8, 11], further study is
(10.5%) occurred three times. required to find whether the Woods Hole popula-
tion possesses a unique variation of alloresponses
within this taxon.
1 2 3 4 5 6 9 10 11 12 21 22 23 25 Different populations of B. schlosseri are prob-
^ &< ablycharacterizedbya highly polymorphicpattern
A \N^\^^^ \\\>^ ,N ^ osfamaplllionregactfirvoimty.a IvnertyhesmMaulltsuareBaayofpo5pulbayti5onm,
^ revealed only 12.1% of fusions. In the Woods
\ I Hole population, two different studies have found
B ^ 6.2% offusions out of 1,262contiguous borders of
Con2lo31on AAAlBBleCCliDDcEcoFmbGinHatUioKn Co2l13ony ALLlleMMMMliNNcOcOomPbiQnaRtiSon^N| v-^ i cc[1oo8ll]oo.nniieeSssimc[io1l7la]lreaclntoeddw4rp.ae2nr%dceofnmutlsayigoeanslnoounugmtbaeofr20s50mo0ftpfraauinsrsiseocnotsf
2 EF 25 ^
11 (E)FG 9 NNOP were obtained from the Israeli B. schlosseri
22 HI 12 PR population (Porat and Rinkevich. unpublished).
4 IJ 5
6 JK 10 Ft S These studies and others [19] reflect the existence
Fig. 2. A pairwise allograft panel between 14Japanese of approximately 100 histocompatibility alleles in
B. schlosseri colonies (A) and their Fu/HC allelic
combination chart (B). A-Each rejecting CAA set each studied B. schlosseripopulation. In Botryllus
ismarkedbyablanksquare,fusiblepairbyastriped primigemis, pairwise fusibility tests revealed a
square. All isografts resulted in fusions (striped similar low percentage of fusions [5, 20, 21]. In
triangles). B-The chart depicting predictions for a Botrylloides fuscus, on the other hand, panels of
single Mendelian locus model of partial genetic pairwise combinations between colonies which
smiavtechhientgerfoozryghoissittoycomopfattihbeiliFtuy/[H2,C4-l6o]c,usaninexwcillu-d were collected from 3 localities, 1-3 km apart,
genotypes [2, 12], and a codominant expression by resulted in 64.0-74.0% offusions. This indicatesa
Fu/HCalleles[2,4-6]. Colony 11 ismarkedalsoby much lower polymorphism on the Fu/HC locus of
allele E in brackets to note that it can share with this species [22].
colony 2 one or both alleles on the Fu/HC ha- The 14 genotypes from Mutsu Bay which were
plotype.
collected randomly from a very small site possess
18 to 19 alleles on their fusibility locus, ofwhich at
least 2 are commonly shared with the Monterey,
DISCUSSION
CA, population (colonics 5. 9. 11 fused with
Intrapopulation allorejection processes in B. Monterey colonies [16]). Based on the results of
.schlosseri from Mutsu Bay, Japan, arc of like 12.195 fusions, we may estimate [23] the numbers
Allorecognition in Japanese Botryllus 987
of alleles on the Fu/HC haplotype in the Mutsu 4 Oka, H. and Watanabe, H. (1957) Colony specific-
Bay population as approximately 32. However, it ity incompoundascidiansastestedbyfusionexperi-
is clearthat thisestimation is minimal and that our ments. Proc. Japan Acad., 33: 657-659.
5 Oka, H. and Watanabe, H. (I960) Problems of
sample does not represent the probable higher
colonyspecificity incompoundascidians. Bull. Mar.
number of alleles residing in the Fu/HC locus of Biol. Stn. Asamushi, 10: 153-155.
this population, since colonies were collected from 6 Sabbadin, A. (1962) La basi genetiche della capaci-
a very limited (5x5 m) area. Moreover, calcula- ta di fusione fra colonie in Botryllus schlosseri
tions are basedon the prediction that all the alleles (Ascidiacea). Rend. Accad. Lincei, 32: 1031-1035.
on the fusibility locus are equally frequent [23], 7 Berrill, N. J. (1950) The Tunicata. Ray Society.
London.
which isnot thecase ofthe presentstudy (Fig. 2b).
8 Tokioka, T. (1953) Ascidians of Sagami Bay. Iwa-
This skews the figure for the number of histocom- nami-Shoten, Tokyo.
patibility alleles and also marks the difficulty of 9 Sabbadin, A. and Astorri, C. (1988) Chimeras and
using CAAs as the only implementation for asses- histocompatibility in the colonial ascidian Botryllus
sing polymorphism on the Fu/HC haplotype. schlosseri. Dev. Comp. Immun., 12: 737-747.
Some of these colonies responded to intrapopula- 10 Lilker-Levav, T. (1992) Allogenic responses in
tion allogeneic encounters by the "retreat growth Botryllusschlosseri and Botrylloides leachifrom the
Mediterranean coast of Israel. M. Sc. Dissertation,
phenomenon" [14]. In assays resulting in fusion, Tel-Aviv Univ. (with English summary)
one of the partners in the chimera was usually 11 Boyd, H. C, Weissman. I. L. and Saito, Y. (1990)
resorbed [3, 9, 13], a result which was recorded Morphologic and genetic verification that Monterey
before in other B. schlosseri populations as well. Botryllus and Woods Hole Botryllus are the same
These outcomes suggest that botryllid ascidians species. Biol. Bull., 178: 239-250.
have highly complex systems of effector mechan- 12 Rinkevich, B. and Weissman, I. L. (1991) Interpo-
pulational allogeneic reactions in the colonial pro-
isms, allcontrolled by avarietyofhistocompatibil- tochordate Botryllus schlosseri. Int. Immun., 3:
ity genes. 1265-1272.
13 Rinkevich, B. andWeissman, I. L. (1992) Chimeras
vs genetically homogeneous individuals: potential
ACKNOWLEDGEMENTS
fitness costs and benefits. Oikos, 63: 119-124.
Thanks are due to Dr. Numakunai for helping in 14 Rinkevich, B. and Weissman, I. L. (1988) Retreat
colonycollectionandtoT. IshiiandA. Dasaiforgeneral growth in the ascidian Botryllus schlosseri: a con-
assistance. The staff of the Shimoda Marine Research sequence of nonself recognition. In "Invertebrate
Center is acknowledged for their hospitality and assist- Historecognition". Ed. by R. K. Grosberg, D.
ance. Thisstudywassupported byagrantfromS. Price, Hedgecock and K. Nelson. Plenum, New York, pp.
North American FriendsofIOLR, byaJSPS Fellowship 93-109.
for Research in Japan (B.R.), and bya Grant-in-Aid for 15 Rinkevich, B. andWeissman, I. L. (1992) Incidents
Scientific Research from the Ministry of Education, ofrejectionandindifferenceinFu/HCincompatible
Science and Culture of Japan, nos. 03455008 and protochordate colonies. J. Exp. Zool., 263: 105—
01304007 (Y.S.). 111.
16 Rinkevich, B., Shapira, M., Weissman, I. L. and
Saito, Y. (1992) Allogenicresponses betweenthree
REFERENCES remote populations of the cosmopolitan ascidian
Botryllusschlosseri. Zool. Sci., 9: 989-994.
1 Bancroft, F. W. (1903) Variation and fusion of 17 Karakashian, S. and Milkman, R. (1967) Colony
colonies in compound ascidians. Proc. Calif. Acad.
fusion compatibility types in Botryllus schlosseri.
Sci.,3: 137-186.
Biol. Bull., 133: 473.
2 Scofield, V. L., Schlumpberger, J. M., West, L. A. 18 Grosberg, R. K. and Quinn, J. F. (1986) The
and Weissman, I. L. (1982) Protochordate genetic control and consequencesofkin recognition
allorecognition is controlled by an MHC-like gene by the larvae of a colonial marine invertebrate.
system. Nature, 295: 499-502.
Nature, 322: 456-459.
3 Weissman, I. L., Saito, Y. and Rinkevich, B. (1990) 19 Scofield, V. L. and Nagashima. L. S. (1983) Mor-
Allorecognition histocompatibilityMiHnCa protochor- phology and genetics ofrejection reactions between
datespecies: Isthe relationshipto semanticor oozooids from the tunicate Botryllus schlosseri.
structural? Immun. Rev., 113: 227-241. Biol. Bull., 165: 733-744.
988 B. RlNKEVICH AND Y. Saito
20 Tanaka, K. and Watanabe, H. (1973) Allogeneic 22 Hirose, E. (1990) Colony specificity and possible
inhibition in a compound ascidian, Botryllus pri- allorecognition sitesinBotrylloides(Ascidiae, Com-
migenus Oka. I. Processes and features of "nonfu- positae). Ph. D. Dissertation, Univ. Tsukuba.
sion" reaction. Cell. Immun., 7: 410-426. Japan.
21 Mukai, H. andWatanabe, H. (1975) Distributionof 23 Grosberg, R. K. (1988) The evolution of
fusion incompatibility types in natural populations allorecognition specificity in colonial invertebrates,
of the compound ascidian, Botryllus primigenus. Quart. Rev. Biol., 63: 377-412.
Proc. Japan Acad., 51: 44-47.