Resembling Those of Antigen Receptors Receptor Family with A

Hagfish Leukocytes Express a Paired Receptor Family with a Variable Domain Resembling Those of Antigen Receptors

This information is current as Takashi Suzuki, Tadasu Shin-I, Asao Fujiyama, Yuji Kohara of September 29, 2021. and Masanori Kasahara J Immunol 2005; 174:2885-2891; ; doi: 10.4049/jimmunol.174.5.2885 http://www.jimmunol.org/content/174/5/2885 Downloaded from

References This article cites 37 articles, 11 of which you can access for free at: http://www.jimmunol.org/content/174/5/2885.full#ref-list-1 http://www.jimmunol.org/ Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

by guest on September 29, 2021 *average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2005 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Hagﬁsh Leukocytes Express a Paired Receptor Family with a Variable Domain Resembling Those of Antigen Receptors1,2

Takashi Suzuki,* Tadasu Shin-I,§ Asao Fujiyama,†¶ʈ Yuji Kohara,‡§ and Masanori Kasahara3*

Jawed vertebrates are equipped with TCR and BCR with the capacity to rearrange their V domains. By contrast, jawless vertebrates, represented by hagfish and lampreys, apparently lack such receptors. We describe in this study a family of hagfish genes carrying a single V-type domain resembling those of TCR/BCR. This multigene family, which we call agnathan paired receptors resembling Ag receptors (APAR), is expressed in leukocytes and predicted to encode a group of membrane glycoproteins with organizations characteristic of paired Ig-like receptors, consisting of activating and inhibitory forms. APAR has a J region in its V-type domain, and its V and J regions are encoded in a single exon. Thus, APAR is a member of the emerging families of diversified, innate immune-type receptors with TCR/BCR-like V-type domains and has many of the features expected for a

primordial TCR/BCR-like receptor. The extracellular domain of APAR may be descended from a V-type domain postulated to Downloaded from have acquired recombination signal sequences in a jawed vertebrate lineage. The Journal of Immunology, 2005, 174: 2885–2891.

wo types of Ag receptor, TCR and BCR, are central ele- call agnathan paired receptors resembling Ag receptors (APAR).4 ments of the adaptive immune system. Through recom- The overall domain organization of APAR resembles those of T bining V and J or V, D, and J gene fragments, they gen- novel immune-type receptors (NITR) (14) and signal-regulatory erate enormous variability in their Ag-binding regions (1, 2). proteins (SIRP) (15) in that it is a member of paired Ig-like re- http://www.jimmunol.org/ Although the strategies for generating BCR diversity differ in an- ceptor families carrying TCR/BCR-like, V-type domains. The imals in different taxa, all jawed vertebrates, including the most close structural similarity of the V-type domain of APAR to those phylogenetically primitive extant representatives, such as sharks, of TCR/BCR (and, to a lesser extent, those of NITR) reinforces an are equipped with both TCR and BCR (3–6). By sharp contrast, all evolutionary link between innate and adaptive immune receptors attempts to identify TCR, BCR, or MHC have been unsuccessful (16, 17) and raises the possibility that the V-type domain of APAR in jawless vertebrates and invertebrates (7–10), suggesting that is descended from a hypothetical V-type domain that, after inser- these molecules exist only in jawed vertebrates. Indeed, recent tion of recombination signal sequences (RSS), gave rise to the V work in the sea lamprey indicates that this jawless vertebrate regions of TCR/BCR in a jawed vertebrate lineage. mounts anticipatory immune responses by using a novel receptor by guest on September 29, 2021 structurally unrelated to TCR or BCR (11). Materials and Methods As the most advanced animals with no clear evidence for the Animals presence of TCR or BCR, the jawless fish represented by hagfish and lampreys offer unique opportunities for investigating the origin Adult hagfishes, E. burgeri, were captured in the Pacific Ocean off the coast of Misaki, Kanagawa Prefecture, Japan. of Ag receptors (12). As an initial step in understanding their immune systems, we recently conducted large-scale expressed se- Screening of the cDNA library quence tag analysis of leukocytes isolated from the inshore hag- A cDNA clone predicted to encode a TCR/BCR-like V-type domain was fish, Eptatretus burgeri (13). Although this survey yielded no identified in the course of expressed sequence tag analysis of hagfish leu- evidence for the existence of TCR, BCR, or MHC, we found a kocytes. Part of this clone (nt positions 86–426; GenBank accession no. cDNA clone predicted to encode a V-type Ig domain resembling AB177615) was used as a probe to screen the leukocyte cDNA library 5 those of TCR/BCR. In this study we show that this cDNA clone prepared from two individuals of adult hagfish (13). A total of 5 ϫ 10 encodes a member of the paired Ig-like receptor family, which we clones were screened with a digoxigenin (DIG)-labeled probe as previously described (18) following the instructions of the manufacturer (Roche). Hy- bridization was performed at 38°C in a DIG Easy Hyb solution (Roche). Final washing conditions were 1ϫ SSC/0.1% SDS at 50°C.

*Department of Biosystems Science, School of Advanced Sciences, and Departments Southern blot analysis of †Informatics and ‡Genetics, School of Life Sciences, Graduate University for Ad- vanced Studies (Sokendai), Hayama, Japan; §Center for Genetic Resource Informa- Genomic DNAs digested by restriction enzymes were size-fractionated on tion and ¶Division of Theoretical Genetics, National Institute of Genetics, Mishima, a 0.8% agarose gel, blotted to a nylon membrane, and hybridized with Japan; and ʈInformation Research Division, National Institute of Informatics, Tokyo, DIG-labeled probes according to the instructions of the manufacturer Japan (Roche) as previously described (19). Hybridization was performed at Received for publication August 17, 2004. Accepted for publication December 3, 2004. 3 Address correspondence and reprint requests to Dr. Masanori Kasahara at his current ad- The costs of publication of this article were defrayed in part by the payment of page dress: Department of Pathology, Hokkaido University Graduate School of Medicine, North-15 charges. This article must therefore be hereby marked advertisement in accordance West-7, Sapporo 060-8638, Japan. E-mail address: [email protected] with 18 U.S.C. Section 1734 solely to indicate this fact. 4 Abbreviations used in this paper: APAR, agnathan paired receptor resembling Ag 1 This work was supported by a Grant-in-Aid for Scientiﬁc Research on Priority Areas receptor; BAC, bacterial artiﬁcial chromosome; CP, connecting peptide; CYT, cyto- (C) Genome Science from the Ministry of Education, Culture, Sports, Science, and plasmic tail; DIG, digoxigenin; NAR, nurse shark novel Ag receptor; NICIR, novel Technology of Japan. ITAM-containing Ig superfamily receptor; NITR, novel immune-type receptor; RSS, 2 Sequence data reported in this paper have been submitted to the DDBJ/EMBL/GenBank recombination signal sequence; SIRP, signal regulatory protein; TM, transmembrane; databases under accession numbers AB177610–AB177620. UTR, untranslated region; VCBP, V region-containing chitin-binding protein.

42°C in a DIG Easy Hyb solution. Final washing conditions were 2ϫ made up of six distinct sequences (B1–B6) and predicted type I SSC/0.1% SDS at 45°C. membrane proteins with a single extracellular V-type domain, fol- Real-time quantitative PCR lowed by the CP, the TM region without any charged residues, and the CYT containing an ITIM. Thus, the APAR gene family has Expression patterns of APAR were analyzed by real-time RT-PCR using a features characteristic of paired Ig-like receptors (14, 22, 23); GeneAmp 5700 sequence detection system (Applied Biosystems). Briefly, total cellular RNAs isolated from leukocytes and various organs of adult APAR-A molecules are likely to associate with an adaptor mole- hagfish were converted to cDNA with the Superscript II kit (Invitrogen cule with an ITAM and function as activating receptors. In con- Life Technologies). PCR was set up in triplicate on 2 ␮l (20 ng) of cDNA trast, APAR-B molecules are likely to function as inhibitory using a SYBR Green PCR kit (Applied Biosystems). Primer sequences receptors. were 5Ј-TCATCAGGAAATGGAAGATTGA-3Ј and 5Ј-AATACACAGT TGGGTTGCTGGA-3Ј for the activating form of APAR,5Ј-ACAGTCTT Signal peptides of 27 and 25 aa residues were predicted for GGTGTTTGCGAGA-3Ј and 5Ј-ATCTTCCAGTGTGCCATTTCC-3Ј for APAR-A and APAR-B, respectively. Thus, mature APAR-A pro- the inhibitory form of APAR, and 5Ј-AACCCATGGAAAATGGAAGG-3Ј teins would contain 238–264 aa, with a calculated molecular mass Ј Ј and 5 -GTCACGCTCCTGGAAAACTG-3 for GAPDH (a reference of 26,616–29,839; mature APAR-B proteins are predicted to have housekeeping gene). Cycling conditions were one cycle at 50°C for 2 min, one cycle at 95°C for 10 min, followed by 40 cycles of denaturation at 258–264 aa, with a calculated molecular mass of 28,315–29,078. 95°C for 15 s and annealing/extension at 55°C for 1 min. A dissociation The extracellular V-type domains of APAR-A and APAR-B con- curve was generated at the end of each PCR to verify the amplification of tain three or four and three to five potential N-liked glycosylation a single product. Standard curves were generated for each gene from serial sites, respectively. The V-type domains of APAR-A and APAR-B dilutions of known amounts of cloned cDNA fragments. The cycle thresh- are of the same size and show 86–94% amino acid and 94–96%

old value was calculated with the software provided by the GeneAmp 5700 Downloaded from sequence detection system with an automatic baseline setting and a fluo- nucleotide sequence identity; by contrast, all other regions of rescence threshold of 0.85. Cycle threshold values for positive samples APAR-A and APAR-B show no significant sequence similarity to ranged from 16 to 35 cycles compared with no template control values of each other at either the nucleotide or the amino acid level. Fur- Ͼ APAR 39 cycles. Expression levels of mRNA were normalized to thermore, the length of the 3Ј-untranslated region (UTR) differs GAPDH mRNA levels. Leukocytes were enriched by Percoll gradient centrifugation (13). When markedly between APAR-A and APAR-B; the activating form had examined by light microscopy, leukocyte-enriched fractions mainly con- the 3Ј-UTR ranging from 210 to 295 bp excluding the poly(A) tail, tained lymphocyte-like cells and spindle cells. Other cell populations that whereas the length of the 3Ј-UTR in the inhibitory form was 1,773 http://www.jimmunol.org/ Ͻ occupied 10% in total were granulocytes, monocytes, and RBC. Spindle bp in B1 and B2, 1,149 bp in B3 and B4, and 1,776 bp in B5 and cells are leukocytes thought to be akin to lymphocyte-like cells (20). B6. In B1/B2 and B5/B6, canonical polyadenylation signals Analysis of gene structures (AATAAA) were found 16 bp upstream from the beginning of the We constructed a hagfish bacterial artificial chromosome (BAC) library poly(A) tail. By contrast, B3 and B4 lacked canonical polyadenyl- with an average insert size of 100 kb, estimated to represent the 3ϫ ge- ation signals. Thus, B3 and B4 appear to use a noncanonical se- nome coverage (T. Suzuki, T. Ota, A. Fujiyama, and M. Kasahara, unpub- quence, TATAAA, located 16 bp upstream from the beginning of lished observations). Three-dimensional pools of this library were screened the poly(A) tail as a polyadenylation signal. Although B1, B2, B5, by PCR using two sets of primers specific for the inhibitory form of APAR:

5Ј-ACAATCACCACCAATGTAAATGGAC-3Ј and 5Ј-CAGGTTCT and B6 also contain this sequence at corresponding positions, anal- by guest on September 29, 2021 CATGTTCTTGTGTTTCA-3Ј, and 5Ј-GAAGTGCTGCGAGGAAAAAT ysis of the cDNA clones provided no evidence that it was used as G-3Ј and 5Ј-CGACTGTGGTGGAATGAGAGTTT-3Ј. We determined the polyadenylation signals. organization of the APAR gene by amplifying two adjacent exons by PCR using BAC DNA as a template and then sequencing exon-intron boundaries. DNA sequencing and sequence analysis Hagfish genome contains multiple copies of APARA and APARB genes Plasmid DNA was sequenced using the CEQ 2000XL DNA Analysis Sys- tem (Beckman Coulter). Amino acid sequences were aligned with the The CP of A5 contains an insertion of 18 aa, which is an incom- Clustal W program (21). The alignment was then adjusted by eye to max- plete copy of the adjacent sequence (Fig. 1). A2 has an 8-bp de- imize sequence similarity. Signal peptides were predicted using the Sig- letion in the region where the stop codon is located in the remain- nalIP version 2.0.b2 server (͗www.cbs.dtu.dk/services/SignalP-2.0/͘). Transmembrane (TM) regions were predicted using the SOSUI program ing APAR-A clones, resulting in the addition of 26 aa at its C (͗http://sosui.proteome.bio.tuat.ac.jp/sosuiframe0.html͘). terminus. Otherwise, the five APAR-A sequences are highly ho- mologous to one another, including the 3Ј-UTR. Hence, it is dif- Results ficult to infer which sequences are allelic and which are derived APAR cDNAs encode paired Ig-like receptors from different loci. Nevertheless, identification of five distinct se- A cDNA clone predicted to encode an Ag receptor-like V region quences from the library prepared from two individuals indicates was identified in the course of transcriptome analysis of hagfish that hagfish have at least two genes of the activating type. In com- leukocytes. We screened the hagfish leukocyte cDNA library un- parison with the ARAR-A clones, the APAR-B clones have higher der low stringency conditions using part of this clone as a probe. sequence diversity. On the basis of sequence similarity and the A total of 26 positive clones were isolated from 5 ϫ 105 plaques; length of the 3Ј-UTR, three pairs of sequences, presumably de- all of them were sequenced partially from both 5Ј and 3Ј ends. This rived from three distinct loci, appear to exist; B1 and B2 are pre- analysis revealed the existence of 11 distinct cDNA sequences. We sumably allelic, as are B5 and B6. B3 and B4 have identical nu- chose a representative cDNA clone for each sequence and deter- cleotide sequences where the comparison can be made, except for mined complete sequences (GenBank accession no. AB177610– the insertion of 18 bp in the CYT. Because this insertion occurs AB177620). On the basis of sequence similarity, these cDNA precisely at one of the exon-intron boundaries (see below), these clones could be classified into two major groups (Fig. 1). One two sequences may represent a splice variant of the same allele, group, designated APAR-A, contained five distinct sequences with the insertion in B4 resulting from the alternative usage of a (A1–A5) and predicted type I membrane proteins with a single splice donor site. extracellular V-type domain, followed by the connecting peptide To estimate the copy number of APAR genes, we performed (CP), the TM region with a charged arginine residue, and the cy- Southern blot analysis using region-specific probes (Fig. 2). The toplasmic tail (CYT). Another group, designated APAR-B, was V-type domain-specific probe, which should hybridize with both The Journal of Immunology 2887 Downloaded from http://www.jimmunol.org/ by guest on September 29, 2021

FIGURE 1. Deduced amino acid sequences of APAR-A and APAR-B molecules. V-type domain sequences of A and B types are aligned to highlight their sequence similarity. All other regions are aligned separately for A and B types. F and underlines indicate conserved cysteine residues and potential indicate identity with the top sequence and ء N-linked glycosylation sites, respectively. Œ in APAR-B indicate exon-intron boundaries. The signs Ϫ and absence of residues, respectively. ࡗ at the end of the V-type domain indicates J-like elements. The ITIM, deﬁned as (I/V/L/S)-X-Y-X-X-(L/V), is boxed. The arginine residue in the TM region, which is presumed to be involved in the interaction with an adaptor molecule with an ITAM motif, is indicated by ϩ. Brackets in the CP region indicate duplicated sequence blocks in APAR-A5. The predicted location of ␤ strands (A–G) in the V-type domain is indicated above the sequence.

APARA and APARB, detected multiple hybridizing bands in all The V-type domain of APAR contains a J region and resembles restriction enzyme digests. The number of bands ranged from four those of TCR/BCR in EcoRI digests to six in BamHI digests. The probe specific for the BLASTP searches of the GenBank database using the deduced Ј activating form (designed in the CYT/3 -UTR) detected two to amino acid sequence of A1 as a query identified a rainbow trout four bands; the four HindIII bands were of almost equal intensity, TCR ␤-chain (accession no. CAD57331) as the most similar pro- suggesting the existence of up to four activating genes. In contrast, tein (E value ϭ 0.001). This was followed by the nurse shark novel the probe specific for the inhibitory form (designed in the CYT/ Ag receptor (NAR; accession no. AAC83716; E value ϭ 0.016), 3Ј-UTR) detected only one or two bands; however, the single TCR ␤-chains of several species, and the Ig L chain type III of bands in EcoRI, PvuII, and PstI digests, and the upper band in Xenopus laevis (AAL40102; E value ϭ 0.14). A similar search EcoRV digests were intense and appeared to represent multiple using the deduced amino acid sequence of B1 as a query identified overlapping bands. Taken together, we estimate that the hagfish a mouse TCR ␣-chain (accession no. AAK49779) as the most genome contains at least five and presumably more than six APAR similar protein (E value ϭ 0.008). Inspection of the deduced genes. The CYT/3Ј-UTR probes yielded identical hybridization APAR sequences indicates that their extracellular Ig-like domain patterns in three hagfishes, whereas the hybridization pattern with has features characteristic of the V-type (Fig. 3). First, APAR has the V-type domain-specific probe showed a variation in one indi- six of the seven canonical residues shared by V- and I-type Ig vidual (data not shown). domains (Cys38, Trp54, Leu101, Asp110, Tyr114, and Cys116) (24). 2888 AGNATHAN PAIRED RECEPTORS RESEMBLING Ag RECEPTORS

tine). Interestingly, APAR-B transcripts were more abundant than APAR-A transcripts in the intestine.

Exon-intron organization of the APAR gene We chose one of the BAC clones identiﬁed by screening with APAR-B-speciﬁc primers for characterization. This BAC clone contained a single APAR gene of the inhibitory type whose sequence was most closely related to that of the B5/6 cDNA in exons 1–3 and that of the B1/2 cDNA in exons 4–7. We deduced precise exon-intron boundaries of this gene by comparing the genomic and cDNA sequences, which revealed the presence of seven exons located within a distance of ϳ30 kb (Fig. 5). Like many other V-type Ig superfamily members, exon 1 encoded the 5Ј-UTR and the signal peptide. The C-terminal four residues of the signal peptide and FIGURE 2. Southern blot analysis of the APAR gene family. Genomic the entire V-type domain, including the J region, were encoded by DNA was digested with the indicated enzymes, size-separated, and blotted. exon 2; there was no interrupting intron between the V and J re- Hybridization was conducted (from left to right) using a V-type domain-

gions. Exon 3, which started immediately after the J region, en- Downloaded from specific probe (nt positions 169–388 of APARB1; accession no. AB177615), an activating type-specific probe (nt positions 774-1094 of coded the CP. Exon 4 encoded the rest of the CP, the TM, and part APARA1; accession no. AB177610), and an inhibitory type-specific probe of the CYT. Exons 5 and 6 encoded the CYT. The 3Ј-UTR and the (nt positions 813-1320 of APARB1; accession no. AB177615). remaining part of the CYT were encoded by exon 7. Interestingly, the ITIM was located in a junction between exons 5 and 6, with the C-terminal consensus residue (Val) encoded by exon 6 and the rest

by exon 5. The nucleotide sequences surrounding the exon-intron http://www.jimmunol.org/ Second, unlike the I-type domain, in which the distance between boundaries all conformed to the GT/AG rule. The locations of the two conserved cysteine residues is normally Ͻ70 residues, the two boundaries are indicated in Fig. 1 (Œ). cysteine residues of APAR are 78 residues apart, thus indicating the presence of the CЈЈ strand unique to the V-type domain. The V regions of TCR and BCR share several highly conserved residues Discussion crucial for their structural integrity (25). Some of such residues, Accumulated evidence indicates that RAGs essential for V(D)J re- 21 55 115 most notably Gln , Tyr , and Phe , are present in APAR combination were originally derived from a retrotransposon of (Fig. 3). prokaryotic origin, and that the RAG transposon-mediated inser- Another indication that the V-type domain of APAR is closely tion of an RSS enabled the emergence of a common ancestor of by guest on September 29, 2021 related to those of TCR/BCR V regions is the fact that it has a TCR/BCR, with characteristic split organizations of V, (D), and J sequence stretch almost identical with the consensus J region se- segments (7, 28–30). Although this widely accepted scenario pre- quence (F/W-G-X-G-T-X-L-X-V). Like type I NAR (26), APAR dicts that the gene that became a target of RSS insertion had a has cysteine at the first position of the J region. In NAR, this V-type domain resembling those of TCR/BCR, its identity is not cysteine has been suggested to form a disulfide bridge with a non- known (31, 32). A recent analysis of the draft genome sequence of canonical cysteine in the second framework region to enhance the the ascidian, Ciona intestinalis (33), revealed that this protochor- stability of its V region (27). Interestingly, the V-type domain of date has at least three types of gene with receptor-like architecture APAR also contains a noncanonical cysteine residue (Cys58)ata (12, 34). However, none of them had a V-type domain that qual- region predicted to correspond to the second framework region. ified as a likely immediate ancestor of TCR/BCR. Likewise, the Although the V-type domain is almost invariant in APAR-A, it V-type domains of V region-containing chitin-binding proteins shows sequence polymorphism in APAR-B (Fig. 1). It is notable (VCBP) identified in amphioxus, Branchiostoma floridae (35), that most, if not all, of these polymorphic residues are located in those regions predicted to correspond to the second and third showed little homology to those of TCR/BCR; thus, the ancestor of complementarity-determining regions in TCR/BCR (Figs. 1 VCBP is again an unlikely candidate for RSS insertion. As one of and 3). the most advanced animals that apparently lack RAG, we reasoned that the hagfish might offer clues to the nature of the target gene. Tissue distribution of APAR transcripts In the present study we identified a hagfish gene family named To determine the tissue distribution of APAR mRNA, we con- APAR, carrying a V-type domain that resembled those of TCR/ ducted quantitative real-time PCR analysis using two sets of BCR (Figs. 1 and 3). The V-type domain of APAR has an unsplit primers: one set specific for the activating form and another set gene organization, with its V and J regions encoded in a single specific for the inhibitory form (Fig. 4). Both forms were ex- exon (Fig. 5). Therefore, this domain structurally resembles a hy- pressed most abundantly in leukocytes. The expression level of pothetical primordial V-type domain into which RSS was inserted the activating form in leukocytes was ϳ5-fold higher than that and hence is likely to share a common ancestry with V regions of of the inhibitory form. The higher expression of the activating TCR/BCR. over the inhibitory form is consistent with the observation that A key question that remains unanswered is how close the V-type of the 26 clones isolated by library screening, seven were of the domain of APAR and variable regions of TCR/BCR are related. APAR-B type and 19 were of the APAR-A type (data not Unlike TCR or BCR, APAR has no constant regions. Hence, it is shown). Other tissues that expressed the APAR genes were possible that the jawless fish have a gene or gene family with V-C those thought to be involved in hemopoiesis (kidney, intestine, architecture that resembles more closely TCR/BCR than does and liver) or defenses in the body surface (skin, gill, and intes- APAR. Indeed, we have previously described a hagfish molecule The Journal of Immunology 2889 Downloaded from http://www.jimmunol.org/ by guest on September 29, 2021

FIGURE 3. Deduced amino acid sequences of the APAR V-type domain and those of the V-type domains of representative Ag receptors or Ag receptor-like molecules. Red circles indicate seven canonical residues shared by V- and I-type Ig domains (24). Arginine likely to form a salt bridge with Asp110 is located at position 80 (indicated by blue circles). Residues in yellow are those frequently found in V domains of Ig and TCR (25), of which those residues that are frequently found in TCRA and TCRB are indicated in green (25). Complementarity-determining regions in Ig and TCR are highlighted in pink. A–G, Predicted locations of ␤ strands. Arrowheads indicate the location of the consensus J region sequence (F/W-G-X-G-T-X-L-X-V). For APARA1 and APARA2, residues are numbered as in Fig. 1. Residues numbers referred to in the text are those of APAR molecules. Abbreviations for species names are: Eb, E. burgeri; Hs, Homo sapiens; Dr, Danio rerio (zebrafish); Gc, Ginglymostoma cirratum (nurse shark); Sn, Sphoeroides nephelus (pufferfish); and Bf, Branchiostoma floridae (amphioxus). Database accession numbers are: Hs_TCRA, AAA60635.1; Hs_TCRB, CAA25134.1; Hs_T- CRD, CAA33331.1; Hs_TCRG, AAA61110.1; Hs_Ig-heavy, AAO38726.1; Hs_IgL-␬, BAC01752.1; Hs_IgL-␭, BAC01836.1; Dr_Ig-␫, AAG31729.1; Gc_IgNAR, AAB48208.1; Gc_IgNARC, AAB08972.1; Sn_NITR, AAC23609.1; Hs_SIRP〈1, NP_542970.1; Hs_SIRPB1, NP_006056.1; Hs_CD8B, NP_742097.1; Hs_CD79B, NP_000617.1; Hs_CD7, NP_006128.1; Hs_DORA, NP_005840.1; Hs_CRTAM, NP_062550.1; Hs_CTLA4, NP_005205.2; Hs_NKp30, NP_667341.1; Hs_PVR, NP_006496.2; Hs_nectin, NP_002846.2; Hs_CTH, NP_055127.2; Bf_VCBP1V1, AAN62848.1; Bf_VCBP1V2, AAN62848.1; Eb_NICIR, BAD52303.1; and Pm_TCR-like, AAU09668.1

named novel ITAM-containing Ig superfamily receptor (NICIR) BCR, such as Val/Leu/Ile19, Gln21, Phe55, and Gly131. Thus, the with V-C2 architecture (13). Furthermore, while this manuscript V-type domain of APAR appears more closely related to a hypo- was under review, Pancer et al. (36) described a sea lamprey mol- thetical, primordial domain that became a target of RSS insertion. ecule with V-C2 architecture. The V-type domains of these mol- To date, large-scale expressed sequence tag analysis of leukocytes ecules, which are the most closely related to each other among has yielded no evidence for the existence of a C1-type domain in known V-type domains, show almost identical levels of amino acid the jawless ﬁsh (10, 13, 36). Thus, it is possible that the V-C1-type similarity to those of TCR/BCR as do the V-type domains of organization observed in TCR/BCR emerged ﬁrst at the level of a APAR. However, they lack residues frequently found in TCR/ common ancestor of jawed vertebrates. 2890 AGNATHAN PAIRED RECEPTORS RESEMBLING Ag RECEPTORS

FIGURE 5. Exon-intron organization of the APARB gene. f, Exons. Exon size (in base pairs) is shown below each box. SP, signal peptide.

Acknowledgments We thank Dr. Martin F. Flajnik for critical reading, Dr. Tatsuya Ota and Jun Kasamatsu for helpful discussions, Dr. Masaaki Morisawa for supply- ing us with the hagﬁsh, and Kaori Kuno and Dr. Taeko Nagata for technical assistance.

Disclosures The authors have no ﬁnancial conﬂict of interest.

FIGURE 4. Real-time RT-PCR analysis of APAR expression in adult

References Downloaded from hagfish tissues and leukocytes. Errors are given as the SDs of the means. 1. Tonegawa, S. 1983. Somatic generation of antibody diversity. Nature 302:575. 2. Davis, M. M., and P. J. Bjorkman. 1988. T-cell antigen receptor genes and T-cell recognition. Nature 334:395. 3. Litman, G. W., M. K. Anderson, and J. P. Rast. 1999. Evolution of antigen binding receptors. Annu. Rev. Immunol. 17:109. APAR is a member of the emerging families of diversified, in- 4. Flajnik, M. F. 2002. Comparative analyses of immunoglobulin genes: surprises and portents. Nat. Rev. Immunol. 2:688. nate immune-type, paired Ig-like receptors carrying TCR/BCR- 5. Marchalonis, J. J., S. F. Schluter, R. M. Bernstein, S. Shen, and http://www.jimmunol.org/ like V-type domains. In jawed vertebrates, two such receptor fam- A. B. Edmundson. 1998. Phylogenetic emergence and molecular evolution of the ilies have been identified (17): a family of bony fish molecules immunoglobulin family. Adv. Immunol. 70:417. 6. Laird, D. J., A. W. De Tomaso, M. D. Cooper, and I. L. Weissman. 2000. 50 known as NITR, which is proposed to be related to the receptors million years of chordate evolution: seeking the origins of adaptive immunity. encoded in the mammalian leukocyte receptor complex (14, 37), Proc. Natl. Acad. Sci. USA 97:6924. 7. Thompson, C. B. 1995. New insights into V(D)J recombination and its role in the and a family of mammalian proteins known as SIRP (15). Struc- evolution of the immune system. Immunity 3:531. tural features of APAR suggest that it is akin to NITR and SIRP. 8. Klein, J., A. Sato, and W. E. Mayer. 2000. Jaws and AIS. In Major Histocom- However, BLASTP searches of the GenBank database using the patibility Complex: Evolution, Structure, and Function. M. Kasahara, ed. Spring- er-Verlag, Tokyo, Berlin, Heidelberg, New York, p. 3. full-length APAR sequences identified neither NITR nor SIRP as 9. Flajnik, M. F., and M. Kasahara. 2001. Comparative genomics of the MHC:

sequences producing significant alignments with APAR. Further- glimpses into the evolution of the adaptive immune system. Immunity 15:351. by guest on September 29, 2021 10. Mayer, W. E., T. Uinuk-Ool, H. Tichy, L. A. Gartland, J. Klein, and more, detailed domain organizations of NITR and SIRP differ from M. D. Cooper. 2002. Isolation and characterization of lymphocyte-like cells from that of APAR; a typical member of the NITR family has an ex- a lamprey. Proc. Natl. Acad. Sci. USA 99:14350. tracellular region consisting of a single V-type domain and a V- 11. Pancer, Z., C. T. Amemiya, G. R. A. Ehrhardt, J. Ceitlin, G. L. Gartland, and M. D. Cooper. 2004. Somatic diversification of variable lymphocyte receptors in like C2-type domain, and a typical SIRP family member has an the agnathan sea lamprey. Nature 430:174. extracellular region composed of one V-type domain and two C1- 12. Kasahara, M., T. Suzuki, and L. Du Pasquier. 2004. On the origins of the adaptive immune system: novel insights from protochordates and cold-blooded verte- type domains. However, like APAR, some members of the NITR brates. Trends Immunol. 25:105. family have a single V-type domain (38), and the draft human 13. Suzuki, T., T. Shin-I, Y. Kohara, and M. Kasahara. 2004. Transcriptome analysis genome sequence predicts the existence of additional SIRP mem- of hagfish leukocytes: a framework for understanding the immune system of jawless fishes. Dev. Comp. Immunol. 28:993. bers with different domain organizations (17). Furthermore, a 14. Litman, G. W., N. A. Hawke, and J. A. Yoder. 2001. Novel immune-type re- BLASTP search using the TM and CYT regions of APAR-B receptor genes. Immunol. Rev. 181:250. ϭ 15. Cant, C. A., and A. Ullrich. 2001. Signal regulation by family conspiracy. Cell. vealed weak sequence similarity (E value 0.76) to the corre- Mol. Life Sci. 58:117. sponding regions of the rainbow trout NITR4 (accession no. 16. Strong, S. J., M. G. Mueller, R. T. Litman, N. A. Hawke, R. N. Haire, AAL83817.1). Although this may be partly accounted for by the A. L. Miracle, J. P. Rast, C. T. Amemiya, and G. W. Litman. 1999. A novel multigene family encodes diversified variable regions. Proc. Natl. Acad. Sci. USA sharing of ITIM, it is also possible that APAR is more closely 96:15080. related to members of the paired Ig-like receptor family, such as 17. van den Berg, T. K., J. A. Yoder, and G. W. Litman. 2004. On the origins of NITR, than to an ancestor of TCR/BCR. adaptive immunity: innate immune receptors join the tale. Trends Immunol. 25:11. Despite the many unresolved issues discussed above, it is clear 18. Takahashi, T., M. Yawata, T. Raudsepp, T. L. Lear, B. P. Chowdhary, that the V-type domain of APAR has a J region and is more closely D. F. Antczak, and M. Kasahara. 2004. Natural killer cell receptors in the horse: evidence for the existence of multiple transcribed LY49 genes. Eur.J. Immunol. related to those of TCR/BCR than any other V-type domain iden- 34:773. tified to date outside jawed vertebrates. In addition, APAR is a 19. Kasahara, M., Y. Watanabe, M. Sumasu, and T. Nagata. 2002. A family of MHC multigene family encoding diversified receptors presumably capa- class I-like genes located in the vicinity of the mouse leukocyte receptor complex. Proc. Natl. Acad. Sci. USA 99:13687. ble of transmitting a signal across the cell membrane. Hence, 20. Fa¨nge, R. 1998. Hagfish blood cells and their formation. In The Biology of Hag- APAR is equipped with most, if not all, the cardinal features ex- fishes. J. M. Jørgensen, J. P. Lomholt, R. E. Weber, and H. Malte, eds. Chapman & Hall, London, p. 287. pected for a primordial TCR/BCR-like receptor (32, 39). In this 21. Thompson, J. D., D. G. Higgins, and T. J. Gibson. 1994. CLUSTAL W: improv- regard, identification of this gene family represents an important ing the sensitivity of progressive multiple sequence alignment through sequence step toward elucidation of the origin of Ag receptors. We are likely weighting, positions-specific gap penalties and weight matrix choice. Nucleic Acids Res. 22:4673. to have a better understanding of the relationship of APAR to other 22. Kubagawa, H., P. D. Burrows, and M. D. Cooper. 1997. A novel pair of immu- TCR/BCR-like V-type domain-containing molecules, such as Ag noglobulin-like receptors expressed by B cells and myeloid cells. Proc. Natl. Acad. Sci. USA 94:5261. receptors, NITR, SIRP, NKp30, and NKp44, when the draft ge- 23. Arase, H., and L. L. Lanier. 2004. Specific recognition of virus-infected cells by nome sequence of the hagfish or lamprey becomes available. paired NK receptors. Rev. Med. Virol. 14:83. The Journal of Immunology 2891

24. Harpaz, Y., and C. Chothia. 1994. Many of the immunoglobulin superfamily 33. Dehal, P., Y. Satou, R. K. Campbell, J. Chapman, B. Degnan, A. De Tomaso, domains in cell adhesion molecules and surface receptors belong to a new struc- B. Davidson, A. Di Gregorio, M. Gelpke, D. M. Goodstein, et al. 2002. The draft tural set which is close to that containing variable domains. J. Mol. Biol. 238:528. genome of Ciona intestinalis: insights into chordate and vertebrate origins. Sci- 25. Chothia, C., D. R. Boswell, and A. M. Lesk. 1988. The outline structure of the ence 298:2157. ␣␤ T-cell receptor. EMBO J. 7:3745. 34. Azumi, K., R. De Santis, A. De Tomaso, I. Rigoutsos, F. Yoshizaki, M. R. Pinto, 26. Greenberg, A. S., D. Avila, M. Hughes, A. Hughes, E. C. McKinney, and R. Marino, K. Shida, M. Ikeda, M. Ikeda, et al. 2003. Genomic analysis of M. F. Flajnik. 1995. A new antigen receptor gene family that undergoes rear- immunity in a urochordate and the emergence of the vertebrate immune system: rangement and extensive somatic diversification in sharks. Nature 374:168. waiting for Godot. Immunogenetics 55:570. 27. Roux, K. H., A. S. Greenberg, L. Greene, L. Strelets, D. Avila, E. C. McKinney, and M. F. Flajnik. 1998. Structural analysis of the nurse shark (new) antigen 35. Cannon, J. P., R. N. Haire, and G. W. Litman. 2002. Identification of diversified receptor (NAR): molecular convergence of NAR and unusual mammalian im- genes that contain immunoglobulin-like variable regions in a protochordate. Nat. munoglobulins. Proc. Natl. Acad. Sci. USA 95:11804. Immunol. 3:1200. 28. Bernstein, R. M., S. F. Schluter, H. Bernstein, and J. J. Marchalonis. 1996. Pri- 36. Pancer, Z., W. E. Mayer, J. Klein, and M. D. Cooper. 2004. Prototypic T cell mordial emergence of the recombination activating gene 1 (RAG1): sequence of receptor and CD4-like coreceptor are expressed by lymphocytes in the agnathan the complete shark gene indicates homology to microbial integrases. Proc. Natl. sea lamprey. Proc. Natl. Acad. Sci. USA 101:13273. Acad. Sci. USA 93:9454. 37. Yoder, J. A., M. G. Mueller, S. Wei, B. C. Corliss, D. M. Prather, T. Willis, 29. Agrawal, A., Q. M. Eastman, and D. G. Schatz. 1998. Transposition mediated by R. T. Litman, J. Y. Djeu, and G. W. Litman. 2001. Immune-type receptor genes RAG1 and RAG2 and its implications for the evolution of the immune system. in zebrafish share genetic and functional properties with genes encoded by the Nature 394:744. mammalian leukocyte receptor cluster. Proc. Natl. Acad. Sci. USA 98:6771. 30. Hiom, K., M. Melek, and M. Gellert. 1998. DNA transposition by the RAG1 and RAG2 proteins: a possible source of oncogenic translocations. Cell 94:463. 38. Hawke, N. A., J. A. Yoder, R. N. Haire, M. G. Mueller, R. T. Litman, 31. Du Pasquier, L. 2000. The phylogenetic origin of antigen-specific receptors. A. L. Miracle, T. Stuge, L. Shen, N. Miller, and G. W. Litman. 2001. Extraor- Curr. Top. Microbiol. Immunol. 248:160. dinary variation in a diversified family of immune-type receptor genes. Proc. 32. Cannon, J. P., R. N. Haire, J. P. Rast, and G. W. Litman. 2004. The phylogenetic Natl. Acad. Sci. USA 98:13832. origins of the antigen-binding receptors and somatic diversification mechanisms. 39. Kaufman, J. 2002. The origins of the adaptive immune system: whatever next? Immunol. Rev. 200:12. Nat. Immunol. 3:1124. Downloaded from http://www.jimmunol.org/ by guest on September 29, 2021