Immunogenetics (2009) 61:463–481 DOI 10.1007/s00251-009-0373-z ORIGINAL PAPER Origin and evolution of the vertebrate leukocyte receptors: the lesson from tunicates Ivana Zucchetti & Rosaria De Santis & Simona Grusea & Pierre Pontarotti & Louis Du Pasquier Received: 10 February 2009 / Accepted: 3 April 2009 / Published online: 30 April 2009 # Springer-Verlag 2009 Abstract Two selected receptor genes of the immunoglobulin superfamily (IgSF), one CTX/JAM family member, and one poliovirus receptor-like nectin that have features of adhesion molecules can be expressed by Ciona hemocytes, the effectors of immunity. They can also be expressed in the nervous system (CTX/JAM) and in the ovary (nectin). The genes encoding these receptors are located among one set of genes, spread over Ciona chromosomes 4 and 10, and containing other IgSF members homologous to those encoded by genes present in a tetrad of human (1, 3+X, 11, 21+19q) or bird chromosomes (1, 4, 24, 31) that include the leukocyte receptor complex. It is proposed that this tetrad is due to the two rounds of duplication that affected a single prevertebrate ancestral region containing a primordial leukocyte receptor complex involved in immunity and other developmental regulatory functions. Electronic supplementary material The online version of this article (doi:10.1007/s00251-009-0373-z) contains supplementary material, which is available to authorized users. I. Zucchetti (*) : R. De Santis (*) Laboratory of Animal Physiology and Evolution, Stazione Zoologica Anton Dohrn, Villa Comunale, 80121 Naples, Italy e-mail: ivana.zucchetti@szn.it e-mail: rosaria.desantis@szn.it S. Grusea : P. Pontarotti LATP UMR 6632 CNRS Evolution biologique et Modélisation, Université de Provence, case 19, 3 place Victor Hugo, 13331 Marseille Cedex 03, France L. Du Pasquier Institute of Zoology and Evolutionary Biology, University of Basel, Vesalgasse 1 CH 4051 Basel, Switzerland Keywords Linkage . Duplication . Evolution . Tunicates . Leukocyte receptors . Hemocytes Introduction In Metazoa, the immunoglobulin superfamily (IgSF) domains participate in the edification of many cell surface molecules and receptors involved in cell adhesion and immunity. Ig domains consist in beta barrels of about 100 amino acids and are classified into different types according to structural criteria linked to the number, length, and position of the strands and loops (variable V and constant C domains). In gnathostome vertebrates, the large number of cell surface IgSF receptors involved in immunity-related functions can be subdivided into two major categories: 1. The antigen-specific receptors of lymphocytes, the ligand-interacting domains of which are generated somatically (V domains of BCR and TCR) or encoded in the germline (MHC-specific NK cell receptors in some species). 2. Other receptors (the ligands of which often also belong to the IgSF, e.g., CD200/CD200R, CD96/CD155, etc.) involved in the differentiation, the regulation of the immunocytes, and their interaction with dendritic cells, stroma, endothelia, etc.; these receptors participate, in one way or the other, to the generation of an immune response (innate or adaptive). This category encompasses a great variety of molecules, since about 34% of leukocyte membrane polypeptides contain IgSF domains (Barclay et al. 1997). Reconstituting the history that led these genes to their involvement in one or the other sector of immunity and to their diversification is not easy. Which came first? How and 464 in which order were members of the same family recruited in the immune system? And were they expressed in other systems? For addressing such issues, a comparative approach might be useful. Several IgSF molecules related to the above have been identified in nonvertebrate chordates such as the cephalochordate amphioxus (Cannon et al. 2002), in colonial tunicates (De Tomaso et al. 2005), and in solitary ascidians (Azumi et al. 2003). We chose to study further the Tunicates, a group of chordates relatively close to the vertebrates (Delsuc et al. 2006). In this group, the solitary ascidian Ciona intestinalis has been the object of a genome project, and in a first survey oriented towards finding relatives of the BCR and TCR (Du Pasquier et al. 2004), three types of molecules of the IgSF have been identified based on the presence of a distal IgSF V domain. These three types belong to nectin (poliovirus receptor, PVR and homologs, CD155), CTX/JAM, and CD166 families, several members of which are expressed on immunocytes of different species of vertebrates (Barclay et al. 1997). Members of these families can be involved also in differentiation of other tissues and systems in particular in the nervous system (Boulanger et al. 2001). Are some of those used to regulate the immune systems of organisms such as Ciona that do not have TCR or BCR? A first step toward answering this question is to investigate the expression pattern of those genes to see whether they are expressed in immunocytes. Therefore, in the first part of this paper, the expression of those Ciona members in hemocytes, presumptive hematopoietic tissues, and other tissues will be presented. We selected two genes: the CTX/JAM member (Cioin 221378), because it is homologous to several genes in vertebrates, some of them expressed on lymphocytes surfaces (CTX and CHT1), and the ciNec2 (Cioin 288444), because the Ciona nectins are relatively close homologs to the nectin-like moiety of the Fu/HC of Botryllus, a locus involved in allorecognition. The architecture of these IgSF receptors can be summarized as (V or I)n-(C)n-TM-Cy. Their genes are frequently duplicated in vertebrates and some homologs are present in agnathans as well, even though not used for the antigen-specific receptor function (Pancer et al. 2004b). In the gnathostome species, many of these genes are not distributed at random in the genome. They are located on a limited number of chromosome segments that is not difficult to gather into tetrads thought to be due to the two rounds of genome duplications and reminding what was described for the MHC (Kasahara 1998). In human several homologs of PVR, CTX, CD166, and -like genes are located on chromosomes 1q, 3q, 11q, 21q (Du Pasquier et al. 2004). Other structurally related IgSF members form sometimes large multigene families located on specific regions of chromosomes, such as the leukocyte receptor Immunogenetics (2009) 61:463–481 complex (LRC) on 19q13 in human and some related regions, for instance, on human chromosome 1q (Fayngerts et al. 2007; Volz et al. 2001). Therefore, in the second part of the paper, we investigated with the partial physical map of Ciona (http://hoya. zool.kyoto-u.ac.jp/chromosomeall.html and http://genome. jgi-psf.org/Cioin2/Cioin2.home.html) the location of the above-mentioned genes and possibly the presence of other ones in an organism that does not have a tetraploid genome. In other words, this will allow investigating the possible conservation of an important genetic region involved in immune system regulation. Materials and methods Embryo and tissue preparation C. intestinalis adults were taken from the Gulf of Naples. Eggs and sperm were collected surgically from the gonoducts and used for in vitro fertilization. Embryos were raised in Millipore-filtered sea water (SW) at 18–20°C and either fixed for whole-mount in situ hybridization or collected for RNA extraction (Zucchetti et al. 2008). Blood from both lipopolysaccharide (LPS; Escherichia coli, serotype 055:B5; Sigma–Aldrich, St. Louis, MO)treated (LT) and non-LPS-treated control animals (NLT) was collected from the basal sinus with a syringe in the presence of Ca2+- and Mg2+-free artificial SW (0.47 M NaCl, 10 mM KCl, 1 mM Na2SO4, 2.5 mM NaHCO3, pH7.5) containing 1 mM EDTA to prevent cell clotting. Samples were used immediately for in situ hybridization or processed for RNA extraction as described (Zucchetti et al. 2008). Tissues from adult animals were surgically dissected and processed either for in situ hybridization or for total RNA extraction. The LPS treatment of C. intestinalis individuals was carried out as previously described (Zucchetti et al. 2008). Local inflammatory reaction was induced by LPS injection into the tunic of adult individuals between the outer layer and the epidermis in the area between the two siphons, as described (Pinto et al. 2003). RNA preparation, RT-PCR amplification, and preparation of RNA probes Total RNA was extracted with the SV Total RNA Isolation System (Promega, Madison, WI). Oligo(dT) single-strand cDNA was synthesized from ∼2,5µg total RNA with the SuperScript™ First-Strand Synthesis System for RT-PCR (Invitrogen, Carlsbad, CA). PCR amplification screening was carried out using two primers (for ciNec2: sense primer 5′ CCAATTACCCGTAGACCAGTCATATCA3′ and antisense Immunogenetics (2009) 61:463–481 primer 5′AGCAACAGCGACAACTCCTCCAATCAC3′; for ciCTX/JAM: sense primer 5′TGTGTTGCTGTAATG GAGTGCGAGTGA3′ and antisense primer 5′TCCAAA CACAATTCCGACTATCATC3′) designed on the basis of the cDNA sequence (gene models ciCTX/JAM: ci0100131719; ciNEC2: ci0100134019 C. intestinalis genome version 1.0) available in JGI database. A single band of the expected size (ciNec2, 1039 bp; ciCTX/JAM, 720 bp) was detected in all the samples examined. These products were cloned into the pCRII-TOPO®-TA-cloning vector (Invitrogen) and verified by sequencing. No quantitative PCR assay was performed. For the in situ hybridization experiments, the cloned cDNAs were used as a template for in vitro transcription of antisense and sense RNA probes, carried out with a digoxigenin (DIG) RNA Labeling Kit (Roche Diagnostic, Mannheim, Germany) according to the supplier’s instructions. In situ hybridization on paraffin sections and circulating blood cells In situ hybridization was performed as previously described (Simeone et al. 1995). For circulating blood cells, a modified protocol has been followed (Zucchetti et al. 2008). Overnight hybridization was carried out with 100 ng/slide of DIGlabeled riboprobes at 60°C. Control experiments were run in parallel using the corresponding sense RNA probes. For microscopy, sections were dehydrated and mounted in a nonaqueous medium (Eukitt Mounting Medium, O. Kindler GmbH, Freiburg, Germany). The blood cells were mounted in PBS/glycerol 1:1. Slides were observed under a light microscope (Zeiss, Göttingen, Germany) in brightfield or by Nomarsky optics. Whole-mount in situ hybridization on larval stages Whole-mount in situ hybridization was performed with 0.1 ng/μl DIG-labeled antisense or sense transcripts at 55°C (Zucchetti et al. 2008). The embryos were observed under a light microscope (Zeiss) by Nomarsky optics. Sequence analysis The Ciona sequences accessible via the JGI C. intestinalis genome versions 1.0 and 2.0 were used for gene model identification and chromosomal position. In addition SMART (http://smart.embl-heidelberg.de/) and PHYRE servers (http://www.sbg.bio.ic.ac.uk/∼3dpssm/) were used for further improvement or correction of the models. Chromosomal position was also checked with the site http://hoya.zool.kyoto-u.ac.jp/chromosomeall.html. Sequences were screened with PROSITE and manually for signalization motifs. BLAST was done against the various 465 genomic and EST database on the NCBI site (http://www. ncbi.nlm.nih.gov/). The human and chicken gene positioning was done with the help of OMIM (http://www.ncbi.nlm.nih. gov/sites/entrez? db = OMIM), ENSEMBL (http://www. ensembl.org/index.html) and UCSC Blat servers (http:// genome.ucsc.edu/). Drosophila genes were investigated on the FlyBase server (http://flybase.bio.indiana.edu/). Phylogenetic trees were obtained with the neighbor joining (EXPASY, http://www.ebi.ac.uk/To) or maximum likelihood PhyML algorithms (http://atgc.lirmm.fr/phyml/) following Clustal alignments. To track the origin of the primordial IgSF receptors, searching for classical synteny group appears inadequate due to long-term sequence drift and extensive recombination. However, tracking long-distance syntenies of welldefined related molecules may be interesting to define loose, ancient gene groups duplicated into large paralog sets. This is made possible by the fact that intrachromosomal recombinations are much more frequent than interchromosomal recombinations (Richard et al. 2003). This is the approach used in the present study that does not focus on finding the orthologs but rather the co-orthologs of all genes concerned, their gene structure, and their linkage (often in disorder) on large chromosomal segments. The genes presented in the Fig. 4 and Fig. S1 in the chicken have been chosen in the following way: after the discovery of IgSF members in Ciona and Branchiostoma with CTX-JAM, nectin features, or CD166 features (Du Pasquier et al. 2004), it was noticed that the warm-blooded vertebrate homologs of these genes segregated on a tetrad of chromosomes 1, 3, 11, 21. So adjacent IgSF genes were scrutinized to see whether they would also segregate on the same chromosomal segments and build a bigger complex. In parallel, other IgSF members with related architectures and signaling motifs of the tyrosine-based immunomodulatory inhibitory and activation motifs (ITIMs and ITAMs) were looked for in Ciona. Domain sequence comparison, domain architectures, and number of domains per molecule, as well as intron–exon organization, were used to group the genes and to extend the homology to members present on other gene segments (see below 19q and Xq). Phylogenies that confirm a putative generation of the tetrad by gene duplication were obtained for CTX/JAM families and nectin but were not systematically constructed for all members (see Fig. 9). As stated above, the purpose of the paper was to see if there was a conservation of broad linkages for gene families in human and birds as compared to Ciona rather than to follow individual members throughout all organisms. Orthologs could have been eliminated, which could make the phylogenetic studies more difficult. Non-IgSF genes were identified as anchors in the regions determined as putative region of interest (yellow, orange in the figures) and used to confirm the paralogous nature of the chromosomal segment. 466 Not all genes belonging to the families of interest, although most of them (see Table S1) were located in the tetrad and statistics were necessary to evaluate the meaning of the suspected linkages. Immunogenetics (2009) 61:463–481 orthologs of the gene i from the reference region will all have a weight of f1 . i We consider the following counting measure. For every interval I from [0,1], the weight of the interval I is the sum of the weights of the orthologs belonging to this interval, i.e., Statistical evaluation The syntenic relationships of the IgSF members and cosegregating markers on human chromosome 1, 3, 11, 21, and 19q were first observed “manually” (Daeron et al. 2008; Du Pasquier et al. 2004) and confirmed by Clever Agent System for Synteny Inheritance and Other Phenomena in Evolution in a previously published paper (Hansen et al. 2008). More specifically, the solidity of the conservation of linkage between the C. intestinalis genetic regions described in this paper and above-human chromosomal regions was estimated with a specific statistical test to assess the significance of the predicted conserved gene clusters. For assessing the significance of a given conserved genomic region, we used a compound Poisson approximation for computing its p-value under the null hypothesis of random gene order. We will explain here the main ideas of our approach (for more details, see: Grusea, S., Compound Poisson approximation and testing for conserved genomic regions. submitted). Our test is adapted to the reference region approach, when one starts with a fixed genomic region in the genome of a certain species A and searches for orthologous regions in the genome of another species B. We take into account the existence in the genome B of multigene families, i.e., the fact that for a given gene in the reference region, we may have several orthologs in the genome B. Suppose that there are m genes in the reference region that have at least one ortholog in the genome B. For each of those genes i=1,...,m, we denote i the number of orthologs it has in the genome B. We let n:=1 +...+m denote the total number of genes in B which are orthologs of genes in the reference region. We also need the total size N of the genome B. We will use a pure significance test, with the null hypothesis being the hypothesis H0: random gene order in the genome B. We use a simplified model in which a genome is seen as an ordered sequence of genes, without separation into chromosomes. Based on the fact that we are in the case m << N, we make a further approximation and consider the genome B as the continuous interval [0, 1], in which the “new” positions of the orthologs are obtained by renormalization, i.e., by dividing their real positions in the genome by N. For taking into account the existence in B of multiple orthologs for the genes in the reference region, we weigh the orthologs in inverse proportion to the size of the multigene family to which they belong. For example, the mm ð I Þ :¼ n X k¼1 wk 1fUðkÞ 2I g ; where for every k, wk denotes the weight of the kth orthologous gene, in position U(k). For an event A, 1A denotes its indicator function. In the simple case of no multigene families (all the i’s equal to one) μm(I) is simply the number of orthologs lying in the interval I. Suppose that we are interested in evaluating the significance m P ni of an observed cluster having the weight h ¼ ; where for f i¼1 i every i=1,…,m, ni is the number of orthologs of the gene i from the reference region which belong to the cluster. Suppose also that the cluster is of (renormalized) length r in [0,1]. The p-value of this cluster is the probability, under the null hypothesis H0, of finding somewhere in the genome B a cluster of weight greater than h and of length smaller than r. We will call such a cluster of type (h/r). We denote by U(1), …, U(n) the ordered positions in B of the n orthologs, i.e., the order statistics of n independent and identically random variables uniformly distributed on [0,1]. Let Wm denote the random variable representing the number of (possibly overlapping) clusters of type (h/r) in the genome B. If we denote   
Ak ¼ mm UðkÞ ; Uðk Þ þ r  h the event of having in B a cluster of type (h/r) starting with the kth ortholog, in position U(k), then Wm ¼ n X dheþ1 1Ak : k¼1 The p-value of the cluster equals P ð W m  1Þ ¼ P [ k ! Ak : Note that the events Ak are very strongly dependent locally and almost independent at large distance. Because of the strong local dependence, the events Ak will tend to arrive in clumps. This is the reason why we calculate this probability using a compound Poisson approximation. Immunogenetics (2009) 61:463–481 467 We approximate the distribution of Wm by a compound Poisson distribution of parameters li ¼ ð n l1 ¼ ð n dhe þ 1Þpqi 1 ð1 qÞ2 ; for i ¼ 2; . . . ; dhe dhP e 1 ili ; dhe þ 1Þp 1; i¼2 where p ¼ PðA1 Þ; q ¼ PðA2 jA1 Þ: The parameter λi can be interpreted as the mean number of clumps formed by i consecutive events. For computing the probabilities π and q we sum over all possible orderings in B of the n orthologs and we use classic results on the distribution of order statistics for uniform random variables. We then approximate the p-value of the cluster by the probability that a compound Poisson random variable with the above parameters takes a value greater or equal to 1, i.e. P ð W m  1Þ  1 exp ( dX he 1 i¼1 ) li : In addition, there were several Ciona IgSF members that could not yet be mapped and that had to be ignored given the framework of this study. In addition, several gene models looked not structurally possible (respective position of extra and intracellular domains, position of hydrophobic segments, etc.). The assembly of the genome version 2.0 is certainly useful but not yet without many mistakes. We are aware that our choice might be biased and may have to be revised 1 day. Results and discussion Ciona nectins and CTX/JAM genes can be expressed by hemocytes Several genes of the vertebrate IgSF can be expressed by many different tissues especially the nervous system (Boulanger et al. 2001; Du Pasquier 2005). Hence, it does not follow automatically that one or the other Ciona homologs function as hemocyte receptors. In human, some IgSF receptors can be restricted to immunocytes (e.g., CRTAM, CD96), but their best homologs in insects are detected on cells of the nervous system (Amalgam, Beat). Expression can also be shared by the two different systems, and this varies with the species (CD166 in vertebrates, DSCAM in arthropods; Du Pasquier 2005). For understanding the history of the immune system and of the receptors encoded by the LRC and paralogous regions in human, as well as to imagine a scenario for the recruitment of the antigen-specific IgSF receptor, it is interesting to see what the situation in Ciona is. The selected CTX/JAM family member (Cioin 221378) and the ciNec2 (Cioin 288444), as well as the related nectin1 (Cioin 29266), are presented in Figs. 1 and 2 with their putative ITIM- and ITAM-like and their predicted domain architecture, as briefly described earlier (Du Pasquier 2004; Du Pasquier et al. 2004). No attempt has been made at this stage to model the putative ligand binding regions. RT-PCR and in situ hybridization analyses have been performed in different developmental stages of the embryo, in organs of the adult animal, such as ovary, testis and neural ganglion, and in tissues related to the immune system. In particular, our attention has been focused on circulating blood cells, primary effectors of the immune response, stomach, considered a hematopoietic tissue (Ermak 1976), and the tunic, the outpost for the interaction with parasites and pathogens. The well-known responsiveness of Ciona hemocytes to LPS (Pinto et al. 2003; Zucchetti et al. 2008), an inducer of the inflammatory reaction, led us to use this compound in the course of the experiments in order to compare the expression of the genes in normal and treated animals. PCR amplification screening has been carried out on oligo(dT) first-strand cDNAs synthesized from total RNAs using primers designed on the basis of the cDNA sequences available in JGI database. The ciNec2 and ciCTX/JAM transcripts of the expected size are present in all the samples examined (data not shown). The products have been cloned, sequenced, and found to correspond to the gene model sequences. These products have been used as templates to prepare both antisense and sense riboprobes for the in situ hybridization experiments. These reveal an interesting expression pattern in some of the tissue examined, as described below. Expression of ciNec2 ciNec2 has nectin characteristics (Takai et al. 2003). Mammalian nectins are Ca2+-independent Ig-like cell–cell adhesion molecules, which comprise a family of four members, nectin-1, nectin-2, nectin-3, and nectin-4 (Takai et al. 2003; Takai and Nakanishi 2003). Several nectin-like molecules are not specifically expressed by hematopoietic tissue cells, but some can be expressed on leukocyte surfaces. All the bona fide nectin family members, except nectin-1β, nectin-1γ, nectin-3γ, and nectin-4, have a conserved motif of four amino acid residues (Glu/Ala-XTyr-Val) at their carboxy-termini and this binds the PDZ domain of afadin, an actin-filament (F-actin)-binding protein. The nectin-like molecules of Ciona do not have 468 Immunogenetics (2009) 61:463–481 Fig. 1 Position of Ciona IgSF genes (and non IgSF linked genes in yellow) on Ciona chromosomes 4, 10q and 10p. Cioin: Ciona identification number JGI version 2.0 the PDZ-binding motif. In that sense, they resemble more the CD155 of mammals. In situ hybridization experiments performed on stomach sections of Ciona show that, while no mRNA expression is present in stomach of NLT animals, a strong expression of ciNec2 mRNA has been found in the majority of the granular amebocytes of LT animals (Fig. 3). This seems to indicate a participation of ciNec2-expressing cells in a response whether inflammatory or immune. No significant signal was recorded in the tunic or in the circulating blood cells, suggesting differential regulation of expression among the hemocytes present in different body compartments. The expression of a nectin by cells of a presumptive Ciona hematopoietic tissue (Ermak 1976) and its possible role in controlling migration towards an inflammation site is consistent with several roles attributed to this family of molecules (Reymond et al. 2004). An example of the involvement of nectin in hematopoiesis is provided by the expression of poliovirus receptor-related 2 (nectin/PRR2) in a restricted cell lineage of the human bone marrow where it could regulate the hematopoietic or endothelial cell functions (Lopez et al. 1998). As to the migration control, it has been demonstrated that CD155 is expressed on B cells and monocyte-derived cells, the latter a type with which Ciona amebocytes could share properties (e.g., motility and phagocytosis; Lange et al. 2001). Expression of ciCTX/JAM The Ciona molecule resembles one receptor with characteristics of both JAM and CTX receptors. As suggested earlier (Du Pasquier 2004), this set of rather similar genes exists always as pair in the four paralogs of the human tetrad (Fig. 4), but Ciona seems to have a preduplicated version or it has lost one of the members. The alternative forms, not noticed earlier, with or without a tyrosine-based phosphorylation motif, suggest a role in signaling (Fig. 2). Immunogenetics (2009) 61:463–481 Fig. 1 (continued) 469 470 Immunogenetics (2009) 61:463–481 Fig. 1 (continued) The vertebrate molecules of the CTX family have acquired different specializations as cell adhesion molecules in various tissues and some function as thymocyte surface receptors in chicken and frogs but not in human (Chrétien et al. 1996). ciCTX/JAM is expressed in blood cells present in different body compartments of the adult. In circulating blood cells, ciCTX/JAM gene is expressed in the granular amebocytes following treatment of the animals with LPS (Fig. 5). Moreover, a close relationship with local inflammatory reaction has been observed. In fact, ciCTX/ JAM mRNA expression follows LPS treatment of the tunic (Fig. 6) where it is also detected in granular amebocytes. As for ciNec2, the failure in the detection of mRNA in situ hybridization on blood of control animals could be due to resolution constrains, as confirmed by the successful isolation of ciCTX/JAM mRNA fragment by PCR analysis on total RNAs isolated from blood of NLT animals. In addition, as already indicated by the expression of other immunity-related genes (Marino et al. 2002; Melillo et al. 2006; Zucchetti et al. 2008), these experiments point to the granular amebocytes, the defense effectors that infiltrate the tunic as the carriers of the molecules that could have a key role in local inflammation. IgSF members’ expression in nonhematopoietic tissues ciNec2 in the ovary It is known that the ovarian oocytes at early stages of maturation are covered by undifferentiated accessory cells. As maturation proceeds, these cells differentiate into test and follicle cells (Cotelli et al. 1981). In situ hybridization experiments conducted on ovary sections show a strong labeling of the follicle cells of young previtellogenic oocytes and a rather faint signal in the cytoplasm of the oocytes at the same maturation stage (Fig. 7). Interestingly, although this issue is still controversial, it has been suggested that the oocytes’ accessory cells originate from amoeboid cells of mesenchymal origin, often interpreted as hemoblasts (Mancuso 1965). In mammals, nectins are known to be involved in differentiation of male gonads (Lui et al. 2006); this is not the case of Ciona in which no hybridization signal has been found in the testis. ciCTX/JAM in nervous system In larval stages, when no differentiation of hemocytes has occurred yet, the expression of the gene is limited to structures related to the nervous system. The expression is restricted to a quite small area of the posterior part of the sensory vesicle of the young larvae, corresponding to the body of some coronet cells (Fig. 8b). At the same stage, the mRNA is also expressed in the papillary region containing the papillary neurons (Fig.8a). In summary, IgSF genes of two selected families of receptors can be expressed in hemocytes of Ciona that are known to function as immunocytes, but they are not exclusively expressed by those cells. For technical reasons, a systematic quantitative estimation of message upregulation has not been performed during these studies, but the patterns of expression and the structure of the receptors Immunogenetics (2009) 61:463–481 471 Fig. 2 Ciona IgSF molecules of the chr. 4-10 linkage group, including the ones with tyrosine based activation and inhibitory motifs. In each identity box the closest homolog (global architecture, not single domain) found in the human genome is indicated. Motifs are considered canonical and labeled in red if they correspond to the sequence V,L,I,S,T,XYXXV,L,I. Other motifs underscored in orange are ITIM-like and fit with the description of ITIM-like motifs described by Sweeney et al. (Sweeney et al. 2005) or found in LRC molecule of different species. For ITAM the consensus is (E/D)-X-X-Y-X-X-(L/I)X6-8-Y-X-X-(L/I) and the fact that YxxP can be part of a functional ITAM has been published (Lee et al. 1998). Predicted SMART structures are represented in Fig. 1. Ig-fold characteristic residues are in blue. Canonical IgSF cysteines and canonical ITIMs are in red. Non canonical ITIMs or ITAMs are in orange. Transmembrane segments are in pink. Ig: Immunoglobulin superfamily domain. TM: transmembrane segment. Cy: cytoplasmic segment. Cioin: Ciona identification number JGI version 2.0 (external IgSF domains and cytoplamic signaling segments) suggest, but do not prove, some role in immunity. They do not show variability and resemble more regulatory receptors than antigen receptors. A further comparison with vertebrates to see whether these genes might belong to some conserved linkage groups with members indeed involved in immunoregulation could help reinforcing the hypothesis. 472 Immunogenetics (2009) 61:463–481 Fig. 3 Expression of ciNec2 mRNA in the stomach of LT and NLT animals. While no cells are labeled in the sections from NLT animals (D), some amebocytes are strongly labeled following infection (A, B). C Section hybridized with sense riboprobe. Arrow, amebocytes; asterisk, unlabelled amebocytes; URG, univacuolar refractile granulocyte; Scale bars, 10µm The genes encoding Ciona CTX/JAM, nectins, and CD166-like are located in two linkage groups on chromosomes 4 and 10 that contain newly characterized genes encoding proteins structurally similar to some human LRC and LRC-linked receptors of chromosome 19q13 The whole genome of C. intestinalis consists in 160–180 Mb (i.e., roughly the length of human chromosome 6) with ca. 15,000 genes distributed on 14 pairs of chromosomes. Thanks to the still partial physical map (available on JGI C. intestinalis genome version 2.0), it has been possible to locate the previously described V(n)-C(n)-TM-Cy IgSF genes coding for molecules with nectin (PVR-like, CD155like), CTX/JAM, and CD166 architectures (Azumi et al. 2003; Du Pasquier et al. 2004) on chr. 4 and 10 (Fig. 1). Those chromosomes were then inspected for the possible presence of other IgSF members and for the presence of conserved non-Ig markers. Several IgSF members were characterized from the gene prediction models of the JGI site, some were confirmed, other had to be revisited and their supposedly correct architecture is represented in the Figs. 1 (position and overall architecture) and 2 (sequence). Several genes encoding putative receptors with a multiIg domain (C2-I set)n-TM-Cy architecture (n varying from 1 to 6) were discovered on Ciona chr. 4 and 10 in relatively close linkage with the above mentioned V-set containing genes (all genes within 3.5–4 Mb). Some did not have a transmembrane segment, but it cannot yet be determined whether this corresponds to reality or to incomplete sequencing data. Some of the receptors with a cytoplasmic tail showed canonical ITIMs or ITIM-related motifs (221319, 292449, AK174180, 221378; Fig. 2). ESTs for the CTX/JAM gene 221378 (Fig. 2) revealed an alternative splice variant encoding an ITIM-related motif. This reminds of modifications encountered in other IgSF receptors in vertebrates and invertebrates (Brites et al. 2008; Daeron et al. 2008). Such a form would resemble a receptor used in immunity. The original CTX receptor, described in Xenopus, and expressed by thymocytes, also has an ITIM motif, whereas its chicken homolog CHT1 has an ITIMlike motif. The human ortholog CTX-HUMX (Du Pasquier 2000; now called VSIg1), not expressed in leukocytes, does not have an ITIM motif (Scanlan et al. 2006). Another receptor on chromosome 10q, ciNectin1 (Du Pasquier 2004) contains a putative ITAM (Cioin 292667, Fig. 2) or several endocytosis motifs. Altogether, the products encoded by these genes have features of molecules belonging to conserved families found in many different species of vertebrates, such as the teleost fish LITRs (Stafford et al. 2006), NITRs (Yoder et al. 2001), Xenopus XILR, XFLs (Guselnikov et al. 2008; Ohta et al. 2006), chicken CHIRs (Dennis et al. 2000), human leukocyte immunoglobulin-like receptors (LILR), killer immunoglobulin-like receptors (KIR; Vivier and Colonna 2006) and murine PIR-A and PIR-B (Takai 2005). Yet, like in other species, comparisons among the genes yielded no clear orthologies as if independent evolution had taken place in each phylum or class starting Immunogenetics (2009) 61:463–481 473 Fig. 4 Graphic representations of the paralogous relationships between Ciona 10-4 region and the human chromosomes. A Nectin (red lines) and CTX/ JAM (black lines) families between Ciona and human chromosomes; B Nectin (red lines) and CTX/JAM (black lines) families in the tetrad of human chromosomes. Non-IgSF members (orange lines and boxes); C The 19q13 relationship with human chromosome (black lines) and the relationship of this segment with Ciona or the nonIgSF members (orange lines and boxes); D The other IgSF paralogs between Ciona and the tetrad of human chromosomes (black lines). Pale gray lines from pattern A have been superimposed and permit to evaluate all IgSF members from a few now disappeared members, by a birth and death process of evolution (Nei et al. 1997). However, because of the structural properties of the Ciona chr. 4–10 receptors and because of some conservation of synteny observed from fish to mammals, many of these vertebrate IgSF subfamilies have been related to the LRC (Kelley et al. 2005). The human LRC is located on chromosome 19q13.4 in a region spanning approximately 1.0 Mb. It comprises sets of genes encoding natural killer receptors, the KIRs, as well as other related IgSF members encoding regulatory receptors, including the LILRs and the leukocyte-associated immunoglobulin-like receptors involved in ITIM/ITAMmediated inhibitory and activating regulation (Daeron et al. 2008). In other species, this complex of regulatory 474 Immunogenetics (2009) 61:463–481 Fig. 4 (continued) receptors is located on mouse chromosome 7 (Martin et al. 2002), rat chromosome 1 (Berg et al. 1999), and chicken chromosome 31 (Viertlboeck et al. 2005). We, therefore, looked for the conservation of homologs of LRC-linked IgSF member genes in Ciona chr. 4 and 10. We found several homologies at the level of some domains or at the level of the whole molecule for some of them. For instance CD22, CD166, Kirre, CD155, all linked to the LRC, are the closest homologs of the Ciona 292449, 288392, 222185, [292667(CiNec1) and 288444 (CiNec2)] genes, respectively, as determined by BLAST and domain structure prediction (SMART and PHYRE). We also looked for Ciona homologs of 19q13 non-IgSF genes to test the hypothesis of the conservation of a larger Immunogenetics (2009) 61:463–481 475 gnathostomes, it generated originally four paralogous segments now distributed on five or perhaps six human chromosomes (Fig. 4) due to secondary translocations. Hypothesis: the human LRC is a segment of the tetrad 1q, 3q, 11q, 21q, originally located on chromosome 21 but translocated on human 19q13 region during evolution Fig. 5 In situ hybridization carried out on circulating blood cells from LT animals by using antisense ciCTX/JAM riboprobe showing the expression of the gene in granular amebocytes. Unstained granular amebocytes are also found. Insert, blood cells hybridized with sense riboprobe. Arrow, granular amoebocyte. Scale bar, 10µm genetic region. We found on chromosome 10 of Ciona the homologs (Cioin identification numbers in brackets) of the following human 19q13 markers (labeled in yellow in Fig. 1): CBL-c (29273), DNA ligase (260784), BCL3 (GenBank AB210323). In addition, we identified on chromosome 4 of Ciona the homologs of the branched chain keto acid dehydrogenase E1, alpha polypeptide (BCKDHA) located on human 19q13 (Figs. 1, 4),1 and of the ethylmalonic encephalopathy protein 1 (ETHE1; 288593). The presence of these BCKDHA and ETHE1-like genes on chromosome 4 (without a paralog on 10) suggests that the two groups of genes on chromosome 4 and 10 are not Ciona paralogous segments due to block duplication in the tunicates. In summary, even though several IgSF-duplicated members could suggest paralogy of the region (two nectins for instance are present), the different non-IgSF genes serving as 19q13 markers are not paralogous of each other. In fact, the sum of the Ciona non-Ig markers on chr. 10 and 4 corresponds to the series encountered on the single human 19q13 (Fig. 4). The presence of many IgSF genes is then more likely to be due to Ciona-specific tandem duplications. Therefore, the chromosome 4 and 10 regions should probably be considered as one region broken in two segments. The physical distance between the genes in Ciona is compatible with a rather short fragment that could be comparable or even shorter than the 19q13 region. This putative single region may have been located originally on one ancestral chromosome that was duplicated in vertebrates following the two whole genome duplications (Kasahara 2007). During evolution towards 1 The homologies mentioned do not concern in each case the full molecule, but a characteristic domain. The above-mentioned human chromosome 19q13 region is not an isolated region. In fact, many of the genes found in the LRC or in its vicinity have paralogs on one to three out of the four human chromosomal regions: 1q, 3q, 11q, 21q (Table S2a–e). These regions that form a tetrad of paralogs similar to the tetrad of MHC paralogs (Kasahara 1999) have been noticed earlier on the basis of the positions of genes belonging to the nectin, CTX, and JAM families as well as many other IgSF molecules especially nectins (Du Pasquier et al. 2004). This point is reinforced and illustrated (Table S2a–e) with more genes for the nectin and CTX/JAM gene families. The paralogous nature of the nectins and the CTX and JAM members has been investigated by phylogenies (Fig. 9). Of all the chromosomes forming the human tetrad, chromosome 21 stands out as the one with the smallest number of homologs or paralogs. It does not contain any nectin for instance. It is tempting to consider the 19q region with its two nectins, CD155 and PVRL 2, as a fragment of this tetrad that would have been translocated. Such translocations have been observed or suggested several times when investigating tetrad of chromosomes in the context of whole genome duplications (Kasahara 1999). A statistical evaluation of this hypothesis found the homology of the Ciona linkage group 4+10 to the 19q+ 21q human region highly significant. We have calculated the p-value of this region using a compound Poisson approximation, as previously described in the section of “Material and methods.” In this case, we have m=14 genes in the Ciona reference region 4q+10q which have at least one ortholog in the human genome (Table S1). The sizes i of the orthologous multigene families in the human genome are the following: 1,1,1,1,1,2,2,2,3,3,4,4,7,16. The total number of genes in the human genome which are orthologous to genes from the Ciona reference region is n=48. The 19q+21q human region is of weight h=5.6399, as it contains three genes of weight 1, two genes of weight 1/2, two genes of weight 1/3, two genes of weight 1/4, two genes of weight 1/7, and three genes of weight 1/16. As said before, the weight of a gene is the inverse of the size of the multigene family to which it belongs. The (renormalized) length of the region is r=1123/36396, as it contains 1123 genes and the total size of the human genome is 36396. With our test we find a p-value of 1.91×10-6, so this region is highly significant. 476 Immunogenetics (2009) 61:463–481 Fig. 6 In situ hybridization of section of tunic of LT animals (A, B, C) using ciCTX/JAM antisense riboprobe indicates that the gene is expressed in amebocytes (arrows) following inflammation. (D) Control run with sense riboprobe. Asterisk, unstained amoebocyte; Scale bar, 50µm If we consider only the region 19q, which has the weight h=4.9524 and the length r=803/36396, we obtain a p-value of 7.32×10-6, so this region is also highly significant. The region 11q has the weight h=4.744 and the length r= 998/36396. We obtain a p-value of 5.77×10-5, hence this region is very significant too. The regions 1q and 3q do not appear significant with our test. A fourth JAM member was expected but it is missing, unless the X-linked pair of CTXHUMX and ZF39 (a JAM homolog) (Du Pasquier 1999) represents the missing fragment that could have been tanslocated from chromosome 3 to X. In addition there exists another nectin family gene on chromosome 3 (IgSF 4d, Table S2 b), but on the p arm, together with homologs of Roundabout. In Gallus (Table S2 f–i) these genes are part of the conserved homologous linkage group. Their position in man could be therefore the result of a pericentric inversion as it has been proposed for MHC paralogous elements on chromosome 1p and 1q (Kasahara 1999). The duplications yielding to the tetraploid genomes of vertebrates are supposed to have occurred during the evolution of chordates. Ciona could therefore have retained a primitive genome organization with only one set of the above-mentioned IgSF-linked genes instead of four. If the hypothesis is correct one expects that homologs of Ciona genes located on chr. 4 and 10 (that we consider in this paper as a single set) would be found not only in the LRC region but also on at least some, if not all, the members of the tetrad resulting from the duplication. This is indeed the case for Ciona CTX/JAM and nectin genes but also for Kirre-like (Kirrel) gene, a nephrin family member with homologs (Fig. 4) on human 19q13 (Kirrel-2), 1q21-q25 (Kirrel) and 11q24 (NEPHRIN-like 2). This is also the case for Cioin 2988576 IgSF domain, homologous to neurotrimin, with homologs on human chr. 11q23 (OBCAM), 3q13 (LSAMP) and 19q13 (OBCAM-L), all members of the IgLON family (Gil et al. 1998). For non-IgSF genes CiCbl has homologs Cbl-b located on 19q13 and Cbl-c located on 3q. Furthermore Ci-BCL3 has a homolog on 19q13 and a BCL3-like on chr. 3q. The best example of a conserved set of paralogs remains a sorting nexin 26 homolog (Cioin 251138 on chr. 10) of which four coorthologs can be found on all human paralogs: 1q, 3q, 11q, and 19q (Fig. 4). The solidity of the tetrad 1, 3, 11, 21 hypothesis is reinforced by its conservation in fish (Hansen et al. 2008) and the chicken (Table S2 f–i) where to the favor of translocations the homologous genes are found in different locations on the chromosomes of the tetrad without synteny conservation. The fourth chicken paralog is missing, but it is likely that the chicken chr. 1 contains in fact two paralogs, since its synteny map on ENSEMBL shows the presence of large segments of human chr. 3q and 21q. Because several tandem duplications occurred in addition to the polyploidization in vertebrates and also perhaps because of some losses during evolution of Ciona, more IgSF members than the few encountered in Ciona can be found in the above-mentioned tetrad. Among those IgSF members, many play a role in the control of lymphocyte Immunogenetics (2009) 61:463–481 477 function or in some nervous system function in the different phyla (NCAM, OBCAM, LSAMP, DSCAM) and some have in fact retained a duality of expression (e.g. CD166 in mammals, DSCAM in arthropods) (Du Pasquier 2005). In several cases ligands and receptors are present within this tetrad (CD96/CD155, VSIg9/CD155) (Seth et al. 2007). Altogether the Gallus and Homo tetrad sensu lato, i.e. including the LRC, originated from Ciona, and built up a fundamental primordial complex involved, in its modern form with many duplicates, in many leukocyte regulatory functions: control of killing, control of proliferation, and migration etc. via, among other means, activation and inhibition using ITAMs and ITIMs. The molecules encoded by this complex might have actually provided the appropriate context for the development of the somaticallygenerated antigen-specific receptors: either the variable lymphocyte receptor (VLR) of agnathans (Cannon et al. 2005) or the BCR and TCR of gnathostomes. This genetic region may also have indeed been involved in immunity in another context as its apparent homolog in Botryllus controls the histocompatibility reactions of this species (De Tomaso et al. 2005). Fig. 7 Expression of ciNec2 in C. intestinalis ovary. A, B) In situ hybridization of section of ovary using antisense riboprobe indicates that the gene is expressed in the follicle cells of young previtellogenic oocytes (arrow) and in the cytoplasm of the same oocyte maturation stages. C) Control run with sense riboprobe showing an unspecific labeling of the oocytes nuclei. Scale bar, 25µm activation in vertebrates (CD200, CD80, CD86, CD96, CRTAM, CD16, CD3 complex, CD47, CD166) (Barclay et al. 1997) or in phagocytosis even in invertebrates (Watson et al. 2005). Some could be committed to leukocytes Fig. 8 Expression pattern of ciCTX/JAM in whole mount in situ hybridization of young larva. A) Lateral view; B) dorsal view. The gene is expressed in the palps (arrow) and in the posterior part of the sensory vesicle corresponding to the body of coronet cells (arrowhead). C) Control run with sense riboprobe. Scale bar, 100µm (A, C); 50µm (B) 478 Immunogenetics (2009) 61:463–481 Fig. 9 Phylogenies (ML=Maximum likelihood) of CTX, JAM and nectin members. The trees support the 2 duplications model to generate those IgSF members. The presence of a single CTX/JAM member in Ciona suggests that a pre-2 rounds of duplication occurred. This separated the two families JAM and CTX, which afterwards underwent the 2 rounds of duplication, as suggested by the tree Is there homology of the Ciona LRC precursor region with the Botryllus FuHC-Fester complex? In the distal region of Ciona chromosome 10p several genes interspersed with the IgSF members (and that are not so far the object of a gene annotation in the JGI version 2.0) were found to belong to the highly diverse family of vCRL1, recently described (Kürn et al. 2007). These duplicated genes are highly polymorphic and code for a transmembrane protein containing several recognizable short consensus repeat domains (SCR/CCP). The protein is structurally similar to vertebrate complement receptors. vCRL1 shows an unprecedented high degree of amino acid variations among Ciona individuals and is expressed in follicle cells as well as in hemocytes. It has been proposed that in the absence of MHC, Ciona uses variable components of the complement system as individuality markers (Kürn et al. 2007). This suggestion takes perhaps even more meaning when one realizes that in fact the core region of vCRL1, with its SCR/CCP, shows recognizable homology with Fester of Botryllus (Kürn et al. 2007). The gene encoding Fester, one of the putative receptors involved in allorecognition, is physically linked to the Fu/HC locus (De Tomaso et al. 2005; Nyholm et al. 2006). The Fu/HC product contains, embedded in the gene, a nectin–like element with some homology to Ciona nectins (best with nectin 2 on chromosome 4) and to human CD155 (encoded in the 19q13 region) and chicken IgSF 4 encoded in the same set of paralogs (see Fig. S1). In other words the chromosome 10+4 regions of Ciona linked to the history of the 19q13 linkage group in human and its paralogs, could be the homolog of the FuHC locus+Fester of Botryllus. Conclusions Many IgSF genes encoding molecules important for leukocyte biology are located in some specific regions of the genome in vertebrates (human or chicken investigated here) forming, like the four MHC paralogous regions, a tetrad that we propose also includes the LRC region. This tetrad likely originated by whole genome duplication from a single ancestral region of which a related region remains in the modern tunicates (chr. 4 and chr. 10 of Ciona). Perhaps they formed originally a logical adaptive cluster essential for hemocyte-immunocyte functions that were preserved throughout evolution. By finding that some of these genes can be expressed in hemocytes of a tunicate such as Ciona and establishing their genetic linkage on chromosomes, this paper suggests that a fundamental vertebrate gene complex, encoding regulatory receptors of the IgSF (including the LRC), finds its origin in an ancient conserved linkage group present in the ancestor of tunicates and vertebrates, already involved in cellular interactions among which hemocyte regulation. Our data demonstrate that hemocytes, in particular those cells participating to an LPS-induced response, can express at least two receptors encoded in this region: one nectin and the CTX/JAM family member. Similarly, members of the homologous region Fu/HC in Botryllus, a colonial form of tunicate, are expressed on Botryllus hemocytes (De Tomaso et al. 2005). In vertebrates the homologs of the ciCTX/ JAM, CTX in Xenopus and CHT1 in the chicken, are expressed on thymocytes whereas the homolog of the ciNectins, CD155, can be expressed by monocytes (Maier et al. 2007), i.e. on cells of the hematopoietic lineage. Furthermore, these genes are also expressed in other tissues, such as gonads and nervous system (Guzman et al. 2006; Preissner and Bronson 2007). This is not surprising from members of the IgSF, whose high structural diversity and many possibilities of interactions, either with other IgSF molecules or with other proteins, make them the elective material to modulate highly specific mechanisms of cell interaction. Our results point to a recurrent motif of a parallel expression between the cells of immune system and structures of the nervous system, thus adding further evidence to the current speculations about gene co-option between these two systems in Metazoa (Boulanger et al. Immunogenetics (2009) 61:463–481 2001; Du Pasquier 2005; Zucchetti et al. 2008), which is characteristic of several members of the tetrad presented here. In addition, the expression of ciNec2 in the follicle cells reminds that of nectin by Sertoli cells in vertebrate gonads (Lui et al. 2006). When compared to another tunicate, the Ciona genetic region defined in this paper appears to encompass the homolog of the Fu/HC-Fester region of Botryllus, with some big differences when considering receptor architecture. Although the nectin moiety of Fu/HC is homologous to the Ciona nectin, it does not exist as a single receptor in Botryllus, but it is embedded in a longer polypeptide, making the role of the Ig domain of the resulting molecule difficult to interpret. The colonial (Botryllus) versus solitary (Ciona) mode of life, with or without possibility of contamination with germ cells (Stoner and Weissman 1996), might have placed the region under different selection pressures in the two organisms. It is also suggestive that the expression of ciNec2 in the ovary overlaps with the localization of the ci-vCLR1 mRNA (accession numbers: DQ792834, DQ792835) (Kürn et al. 2007), in both the blood cells and the oocyte follicle cells, particularly in view of the above mentioned hypothesis that these two genes may belong to the same original linkage group. Interestingly, this is also reminiscent of what has been described for the two Botryllus linked genes Fu/HC and Fester (De Tomaso et al. 2005; Nyholm et al. 2006): these, putative homologues of ciNec2 and ci-vCLR1 respectively, show an overlapping expression pattern in ampullar epithelia and blood cells. Noticeably, this expression site is shared by all the genes mentioned above in both Botryllus and Ciona. Furthermore, another leading theme is their involvement in self-nonself recognition systems. In fact, it has been demonstrated that Fu/Hc and Fester are directly involved in somatic self-nonself recognition (De Tomaso et al. 2005; Nyholm et al. 2006) and that ci-vCRL1 might be involved in gamete self-nonself recognition (Kürn et al. 2007). Although no conclusion can be drawn about its function, it is intriguing that, during oogenesis, also ciNec2 is expressed in the follicle cells, oocytes accessory cells demonstrated to be directly involved in the onset of self sterility (De Santis and Pinto 1991; Marino et al. 1999; Marino et al. 1998). Hence, these observations prompt to speculate about a possible shared evolutionary history of some molecular bases of the somatic and gamete recognition. It will now be interesting to monitor this complex and its polymorphism in other species of tunicates such as Halocynthia or Oikopleura to see if linkage is visible there, despite the big differences in genome size and organization. One should of course also investigate a separate group of Protochordates, the cephalochordate amphioxus, now that its genome is accessible. 479 When compared to the Ciona IgSF genes, the members of the agnathan IgSF receptors, members with ITIM, the socalled V-preB (Cannon et al. 2005), TCR prototypes (Pancer et al. 2004b), and the nectins, are in fact all part of the same family as the one described in Ciona (data not shown). The TCR and V-preB nomenclature can be in fact misleading. We suggest that all these genes form in fact a fundamental and ancient complex that provided elements essential to many aspects of the regulation of lymphocyte biology (control of proliferation, effector molecule production, migration etc.) and that has nothing to do with the selection of the actual antigen-specific receptor. According to our hypothesis, antigen-specific receptors were recruited later and independently in agnathans and gnathostomes, LRR in the VLR for one (Pancer et al. 2004a), IgSF in the TCR and BCR for the other. Which set of genes contained the IgSF V member that ultimately acquired somatic rearrangement in gnathostomes is not elucidated. The APAR genes of hagfish (Suzuki et al. 2005) with their V domains that are the most similar to V domain of antigenspecific receptor, could represent an example of the intermediary form. It would be interesting to locate them in the genome of the hagfish to see whether they correspond to an independent lineage or whether they might be linked to our hypothetical complex presented in this paper. So far no homolog of APAR has been identified in Ciona. In order to probe for the early history of this genetic region, the existence of a genome project in the cnidarian Nematostella, where some three IgSF domain nectin-like are detectable (e.g. XP_001626417 with ITIM, XM_001626370 without ITIM, Du Pasquier unpublished), leaves the door open to many possibilities. In fact, in the same phylum the analysis of the recently identified allorecognition region in Hydractinia (Cadavid et al. 2004) will perhaps tell us about conservation or variation on the theme of the immunological involvement of these IgSF receptors in evolution, to the light of what has been found in colonial tunicates, such as Botryllus. Finally relationships between the members of this tetrad and that of the MHC with which it shares some related IgSF members (Du Pasquier 2002), remain to be explored further. Acknowledgment We thank Olivier Chabrol for discussions and Pierre Boudinot and Maria Rosaria Pinto for their critical reading of the manuscript. We would like to acknowledge the Gene Expression Service at the Stazione Zoologica Anton Dohrn for technical support. References Azumi K, De Santis R, De Tomaso A, Rigoutsos I, Yoshizaki F, Pinto MR et al (2003) Genomic analysis of immunity in a Urochordate 480 and the emergence of the vertebrate immune system: "waiting for Godot". Immunogenetics 55:570–581. doi:10.1007/s00251-0030606-5 Barclay AN, Brown MH, Law SKA, McKnight AJ, Tomlinson MG, van der Merwe PA (1997) The leucocyte antigen. Academic Press, San Diego Berg SF, Fossum S, Dissen E (1999) NILR-1, a novel immunoglobulinlike receptor expressed by neutrophilic granulocytes, is encoded by a leukocyte receptor gene complex on rat chromosome 1. Eur J Immunol 29:2000–2006. doi:10.1002/(SICI)1521-4141(199906) 29:06&lt;2000::AID-IMMU2000>3.0.CO;2-5 Boulanger LM, Huh GS, Shatz CJ (2001) Neuronal plasticity and cellular immunity: shared molecular mechanisms. Curr Opin Neurobiol 11:568–578. doi:10.1016/S0959-4388(00) 00251-8 Brites D, McTaggart S, Morris K, Anderson J, Thomas K, Colson I et al (2008) The Dscam homologue of the crustacean Daphnia is diversified by alternative splicing like in insects. Mol Biol Evol 25:1429–1439. doi:10.1093/molbev/msn087 Cadavid LF, Powell AE, Nicotra ML, Moreno M, Buss LW (2004) An invertebrate histocompatibility complex. Genetics 167:357–365. doi:10.1534/genetics.167.1.357 Cannon JP, Haire RN, Litman GW (2002) Identification of diversified genes that contain immunoglobulin-like variable regions in a protochordate. Nat Immunol 3:1200–1207. doi:10.1038/ni849 Cannon JP, Haire RN, Pancer Z, Mueller MG, Skapura D, Cooper MD et al (2005) Variable domains and a VpreB-like molecule are present in a jawless vertebrate. Immunogenetics 56:924–929. doi:10.1007/s00251-004-0766-y Chretien I, Robert J, Marcuz A, Garcia-Sanz JA, Courtet M, Du Pasquier L (1996) CTX, a novel molecule specifically expressed on the surface of cortical thymocytes in Xenopus. Eur J Immunol 26:780–791 Cotelli F, Andronico F, De Santis R, Monroy A, Rosati F (1981) Differentiation of the vitelline coat in the ascidian Ciona intestinalis: An ultrastructural study. Dev Biol 190:252–258 Daeron M, Jaeger S, Du Pasquier L, Vivier E (2008) Immunoreceptor tyrosine-based inhibition motifs: a quest in the past and future. Immunol Rev 224:11–43. doi:10.1111/j.1600-065X.2008.00666.x Delsuc F, Brinkmann H, Chourrout D, Philippe H (2006) Tunicates and not cephalochordates are the closest living relatives of vertebrates. Nature 439:965–968. doi:10.1038/nature04336 Dennis G Jr, Kubagawa H, Cooper MD (2000) Paired Ig-like receptor homologs in birds and mammals share a common ancestor with mammalian Fc receptors. Proc Natl Acad Sci USA 97:13245– 13250. doi:10.1073/pnas.230442897 De Santis R, Pinto MR (1991) Gamete self-discrimination in ascidians: a role for the follicle cells. Mol Reprod Dev 29:47– 50. doi:10.1002/mrd.1080290108 De Tomaso AW, Nyholm SV, Palmeri KJ, Ishizuka KJ, Ludington WB, Mitchel K et al (2005) Isolation and characterization of a protochordate histocompatibility locus. Nature 438:454–459. doi:10.1038/nature04150 Du Pasquier L (1999) Relationship among the genes encoding MHC molecules and the specific antigen receptor. In: Kasahara M (ed) Major Histocompatibility Complex Evolution, Structure, and Function. Spring, Tokyo, pp 53–65 Du Pasquier L (2000) The phylogenetic origin of antigen-specific receptors. Curr Top Microbiol Immunol 248:160–185 Du Pasquier L (2002) Several MHC-linked Ig superfamily genes have features of ancestral antigen-specific receptor genes. Curr Top Microbiol Immunol 266:57–71 Du Pasquier L (2004) Innate immunity in early chordates and the appearance of adaptive immunity. C R Biol 327:591–601. doi:10.1016/j.crvi.2004.04.004 Du Pasquier L (2005) Immunology. Insects diversify one molecule to serve two systems. Science 309:1826–1827. doi:10.1126/science.1118828 Immunogenetics (2009) 61:463–481 Du Pasquier L, Zucchetti I, De Santis R (2004) Immunoglobulin superfamily receptors in protochordates: before RAG time. Immunol Rev 198:233–248. doi:10.1111/j.0105-2896.2004.00122.x Ermak TH (1976) The hematogenic tissues of tunicates. In: Wright RK, Cooper EL (eds) Phylogeny of thymus and bone marrow-bursa cells. Elsevier/North-Holland Biomedical Press, Amsterdam, pp 45–56 Fayngerts SA, Najakshin AM, Taranin AV (2007) Species-specific evolution of the FcR family in endothermic vertebrates. Immunogenetics 59:493–506. doi:10.1007/s00251-007-0208-8 Gil OD, Zanazzi G, Struyk AF, Salzer JL (1998) Neurotrimin mediates bifunctional effects on neurite outgrowth via homophilic and heterophilic interactions. J Neurosci 18:9312–9325 Guselnikov SV, Ramanayake T, Erilova AY, Mechetina LV, Najakshin AM, Robert J et al (2008) The Xenopus FcR family demonstrates continually high diversification of paired receptors in vertebrate evolution. BMC Evol Biol 8:148. doi:10.1186/1471-2148-8-148 Guzman G, Oh S, Shukla D, Engelhard HH, Valyi-Nagy T (2006) Expression of entry receptor nectin-1 of herpes simplex virus 1 and/or herpes simplex virus 2 in normal and neoplastic human nervous system tissues. Acta Virol 50:59–66 Hansen JD, DuPasquier L, Lefranc MP, Lopez V, Benmansour A, Boudinot P (2008) The B7 family of immunoregulatory receptors: A comparative and evolutionary perspective. Mol Immunol (in press). Kasahara M (1998) What do the paralogous regions in the genome tell us about the origin of the adaptive immune system? Immunol Rev 166:159–175. doi:10.1111/j.1600-065X.1998.tb01261.x Kasahara M (1999) The chromosomal duplication model of the major histocompatibility complex. Immunol Rev 167:17–32. doi:10.1111/j.1600-065X.1999.tb01379.x Kasahara M (2007) The 2R hypothesis: an update. Curr Opin Immunol 19:547–552. doi:10.1016/j.coi.2007.07.009 Kelley J, Walter L, Trowsdale J (2005) Comparative genomics of natural killer cell receptor gene clusters. PLoS Genet 1:129–139. doi:10.1371/journal.pgen.0010027 Kürn U, Sommer F, Hemmrich G, Bosch TC, Khalturin K (2007) Allorecognition in urochordates: identification of a highly variable complement receptor-like protein expressed in follicle cells of Ciona. Dev Comp Immunol 31:360–371. doi:10.1016/j. dci.2006.06.008 Lange R, Peng X, Wimmer E, Lipp M, Bernhardt G (2001) The poliovirus receptor CD155 mediates cell-to-matrix contacts by specifically binding to vitronectin. Virology 285:218–227. doi:10.1006/viro.2001.0943 Lee H, Guo J, Li M, Choi JK, DeMaria M, Rosenzweig M et al (1998) Identification of an immunoreceptor tyrosine-based activation motif of K1 transforming protein of Kaposi’s sarcoma-associated herpesvirus. Mol Cell Biol 18:5219–5228 Lopez M, Aoubala M, Jordier F, Isnardon D, Gomez S, Dubreuil P (1998) The human poliovirus receptor related 2 protein is a new hematopoietic/endothelial homophilic adhesion molecule. Blood 92:4602–4611 Lui WY, Sze KL, Lee WM (2006) Nectin-2 expression in testicular cells is controlled via the functional cooperation between transcription factors of the Sp1, CREB, and AP-1 families. J Cell Physiol 207:144–157. doi:10.1002/jcp. 20545 Maier MK, Seth S, Czeloth N, Qiu Q, Ravens I, Kremmer E et al (2007) The adhesion receptor CD155 determines the magnitude of humoral immune responses against orally ingested antigens. Eur J Immunol 37:2214–2225. doi:10.1002/eji.200737072 Mancuso V (1965) An electron microscope study of the test cells and follicle cells of Ciona intestinalis during oogenesis. Acta Embryol Morphol Exp 8:239–266 Marino R, Pinto MR, Cotelli F, Lamia CL, De Santis R (1998) The hsp70 protein is involved in the acquisition of gamete self-sterility in the ascidian Ciona intestinalis. Development 125:899–907 Immunogenetics (2009) 61:463–481 Marino R, De Santis R, Giuliano P, Pinto MR (1999) Follicle cell proteasome activity and acid extract from the egg vitelline coat prompt the onset of self-sterility in Ciona intestinalis oocytes. Proc Natl Acad Sci USA 96:9633–9636. doi:10.1073/ pnas.96.17.9633 Marino R, Kimura Y, De Santis R, Lambris JD, Pinto MR (2002) Complement in urochordates: cloning and characterization of two C3-like genes in the ascidian Ciona intestinalis. Immunogenetics 53:1055–1064. doi:10.1007/s00251-001-0421-9 Martin AM, Kulski JK, Witt C, Pontarotti P, Christiansen FT (2002) Leukocyte Ig-like receptor complex (LRC) in mice and men. Trends Immunol 23:81–88. doi:10.1016/S1471-4906(01) 02155-X Melillo D, Sfyroera G, De Santis R, Graziano R, Marino R, Lambris JD et al (2006) First identification of a chemotactic receptor in an invertebrate species: structural and functional characterization of Ciona intestinalis C3a receptor. J Immunol 177:4132–4140 Nei M, Gu X, Sitnikova T (1997) Evolution by the birth-and-death process in multigene families of the vertebrate immune system. Proc Natl Acad Sci USA 94:7799–7806. doi:10.1073/pnas.94.15.7799 Nyholm SV, Passegue E, Ludington WB, Voskoboynik A, Mitchel K, Weissman IL et al (2006) fester, A candidate allorecognition receptor from a primitive chordate. Immunity 25:163–173. doi:10.1016/j. immuni.2006.04.011 Ohta Y, Goetz W, Hossain MZ, Nonaka M, Flajnik MF (2006) Ancestral organization of the MHC revealed in the amphibian Xenopus. J Immunol 176:3674–3685 Pancer Z, Amemiya CT, Ehrhardt GR, Ceitlin J, Gartland GL, Cooper MD (2004a) Somatic diversification of variable lymphocyte receptors in the agnathan sea lamprey. Nature 430:174–180. doi:10.1038/nature02740 Pancer Z, Mayer WE, Klein J, Cooper MD (2004b) Prototypic T cell receptor and CD4-like coreceptor are expressed by lymphocytes in the agnathan sea lamprey. Proc Natl Acad Sci USA 101:13273–13278. doi:10.1073/pnas.0405529101 Pinto MR, Chinnici CM, Kimura Y, Melillo D, Marino R, Spruce LA et al (2003) CiC3–1a-mediated chemotaxis in the deuterostome invertebrate Ciona intestinalis (Urochordata). J Immunol 171:5521–5528 Preissner KT, Bronson RA (2007) The role of multifunctional adhesion molecules in spermatogenesis and sperm function: Lessons from hemostasis and defense? Semin Thromb Hemost 33:100–110. doi:10.1055/s-2006-958468 Reymond N, Imbert AM, Devilard E, Fabre S, Chabannon C, Xerri L et al (2004) DNAM-1 and PVR regulate monocyte migration through endothelial junctions. J Exp Med 199:1331–1341. doi:10.1084/jem.20032206 Richard F, Lombard M, Dutrillaux B (2003) Reconstruction of the ancestral karyotype of eutherian mammals. Chromosome Res 11:605–618. doi:10.1023/A:1024957002755 Scanlan MJ, Ritter G, Yin BW, Williams C Jr, Cohen LS, Coplan KA et al (2006) Glycoprotein A34, a novel target for antibody-based cancer immunotherapy. Cancer Immun 6:2 Seth S, Maier MK, Qiu Q, Ravens I, Kremmer E, Forster R et al (2007) The murine pan T cell marker CD96 is an adhesion receptor for CD155 and nectin-1. Biochem Biophys Res Commun 364:959–965. doi:10.1016/j.bbrc.2007.10.102 481 Simeone A, Avantaggiato V, Moroni MC, Mavilio F, Arra C, Cotelli F et al (1995) Retinoic acid induces stage-specific antero-posterior transformation of rostral central nervous system. Mech Dev 51:83–98. doi:10.1016/0925-4773(95) 96241-M Stafford JL, Bengten E, Du Pasquier L, McIntosh RD, Quiniou SM, Clem LW et al (2006) A novel family of diversified immunoregulatory receptors in teleosts is homologous to both mammalian Fc receptors and molecules encoded within the leukocyte receptor complex. Immunogenetics 58:758–773. doi:10.1007/ s00251-006-0134-1 Stoner DS, Weissman IL (1996) Somatic and germ cell parasitism in a colonial ascidian: possible role for a highly polymorphic allorecognition system. Proc Natl Acad Sci USA 93:15254– 15259. doi:10.1073/pnas.93.26.15254 Suzuki T, Shin IT, Fujiyama A, Kohara Y, Kasahara M (2005) Hagfish leukocytes express a paired receptor family with a variable domain resembling those of antigen receptors. J Immunol 174:2885–2891 Sweeney MC, Wavreille AS, Park J, Butchar JP, Tridandapani S, Pei D (2005) Decoding protein-protein interactions through combinatorial chemistry: sequence specificity of SHP-1, SHP-2, and SHIP SH2 domains. Biochemistry 44:14932–14947. doi:10.1021/bi051408h Takai T (2005) Paired immunoglobulin-like receptors and their MHC class I recognition. Immunology 115:433–440. doi:10.1111/ j.1365-2567.2005.02177.x Takai Y, Nakanishi H (2003) Nectin and afadin: novel organizers of intercellular junctions. J Cell Sci 116:17–27. doi:10.1242/ jcs.00167 Takai Y, Irie K, Shimizu K, Sakisaka T, Ikeda W (2003) Nectins and nectinlike molecules: roles in cell adhesion, migration, and polarization. Cancer Sci 94:655–667. doi:10.1111/j.1349-7006.2003.tb01499.x Viertlboeck BC, Habermann FA, Schmitt R, Groenen MA, Du Pasquier L, Gobel TW (2005) The chicken leukocyte receptor complex: a highly diverse multigene family encoding at least six structurally distinct receptor types. J Immunol 175:385–393 Vivier E, Colonna M (eds) (2006) Immunobiology of Natural Killer Cell receptors. Curr Top Microbiol Immunol, vol 298, Springer, Heidelberg. Volz A, Wende H, Laun K, Ziegler A (2001) Genesis of the ILT/LIR/MIR clusters within the human leukocyte receptor complex. Immunol Rev 181:39–51. doi:10.1034/j.1600-065X.2001.1810103.x Watson FL, Puttmann-Holgado R, Thomas F, Lamar DL, Hughes M, Kondo M et al (2005) Extensive diversity of Ig-superfamily proteins in the immune system of insects. Science 309:1874– 1878. doi:10.1126/science.1116887 Yoder JA, Mueller MG, Wei S, Corliss BC, Prather DM, Willis T et al (2001) Immune-type receptor genes in zebrafish share genetic and functional properties with genes encoded by the mammalian leukocyte receptor cluster. Proc Natl Acad Sci USA 98:6771– 6776. doi:10.1073/pnas.121101598 Zucchetti I, Marino R, Pinto MR, Lambris JD, Du Pasquier L, De Santis R (2008) ciCD94–1, an ascidian multipurpose C-type lectin-like receptor expressed in Ciona intestinalis hemocytes and larval neural structures. Differentiation 76:267–282. doi:10.1111/ j.1432-0436.2007.00214.x