Description of an elasmobranch TCR coreceptor: CD8α from Rhinobatos productus

Developmental and Comparative Immunology 35 (2011) 452–460

Contents lists available at ScienceDirect

Developmental and Comparative Immunology journal homepage: www.elsevier.com/locate/dci

Description of an elasmobranch TCR coreceptor: CD8␣ from Rhinobatos productus夽 John D. Hansen a,∗ , Thomas J. Farrugia b , James Woodson a , Kerry J. Laing c a b c

U.S. Geological Survey, Western Fisheries Research Center, Seattle, WA 98115, United States Department of Biological Sciences, California State University, Long Beach, CA 90840, United States Vaccine and Infectious Diseases Division, Fred Hutchinson Cancer Research Center, Seattle, WA 98109, United States

a r t i c l e

i n f o

Article history: Received 22 September 2010 Received in revised form 18 November 2010 Accepted 18 November 2010 Available online 24 November 2010 Keywords: Teleost CD8 Epigonal IgNAR Elasmobranch

a b s t r a c t Cell-mediated immunity plays an essential role for the control and eradication of intracellular pathogens. To learn more about the evolutionary origins of the first signal (Signal 1) for T-cell activation, we cloned CD8␣ from an elasmobranch, Rhinobatos productus. Similar to full-length CD8␣ cDNAs from other vertebrates, Rhpr-CD8␣ (1800 bp) encodes a 219 amino acid open reading frame composed of a signal peptide, an extracellular IgSF V domain and a stalk/hinge region followed by a well-conserved transmembrane domain and cytoplasmic tail. Overall, the mature Rhpr-CD8␣ protein (201 aa) displays ∼30% amino acid identity with mammalian CD8␣ including absolute conservation of cysteine residues involved in the IgSf V domain fold and dimerization of CD8␣␣ and CD8␣␤. One prominent feature is the absence of the LCK association motif (CXC) that is needed for achieving signal 1 in tetrapods. Both elasmobranch and teleost CD8␣ protein sequences possess a similar but distinctly different motif (CXH) in the cytoplasmic tail. The overall genomic structure of CD8␣ has been conserved during the course of vertebrate evolution both for the number of exons and phase of splicing. Finally, quantitative RTPCR demonstrated that elasmobranch CD8␣ is expressed in lymphoid-rich tissues similar to CD8 in other vertebrates. The results from this study indicate the existence of CD8 prior to the emergence of the gnathostomes (>450 MYA) while providing evidence that the canonical LCK association motif in mammals is likely a derived characteristic of tetrapod CD8␣, suggesting potential differences for T-cell education and activation in the various gnathostomes. Published by Elsevier Ltd.

1. Introduction Cytotoxic and helper T lymphocytes are essential for cellmediated immunity in vertebrates based upon their recognition of endogenously and exogenously derived antigens presented by MHC class I and II molecules respectively. This recognition process occurs due to the interaction of the TCR with MHC, which is stabilized by the TCR co-receptors CD8 and CD4. Together with the TCR ␣␤ or ␥␦ chains and the associated CD3 and ␨-chain signaling components, the T-cell co-receptors form vital components of the overall TCR complex, and define specificity of function of T-cells (Smith-Garvin et al., 2009). CD4+ lymphocytes respond to exogenous peptides presented by MHC class II molecules to produce cytokines that facilitate an appropriate immune response against a

夽 Sequences have been deposited in GenBank under accession numbers: EF559247 and EF559248; GU205320–321 and GU205319. ∗ Corresponding author at: USGS, Western Fisheries Research Center, 6505 NE, 65th Street, Seattle, WA 98115, United States. Fax: +1 206 526 6654. E-mail addresses: [email protected], [email protected] (J.D. Hansen). 0145-305X/$ – see front matter. Published by Elsevier Ltd. doi:10.1016/j.dci.2010.11.014

given pathogen. CD8+ lymphocytes become activated upon detection of endogenous peptides that are presented by MHC class I molecules and are essential for the elimination of virus infected cells and tumors. T-cell activation is dependant on two signals: signal one is derived from the interaction between TCR and MHC, whereas signal two is transmitted via a co-stimulatory molecule (CD28) following engagement with its ligands (CD80 and CD86) on antigen-presenting cells. Signal one, which is antigen dependent, is a critical step requiring a signal through the co-receptor that is mediated by lymphocyte specific kinase (LCK). Once associated with the co-receptors, LCK phosphorylates immunoreceptor tyrosine-based activation motifs (ITAMs) on CD3 and ␨-chains to initiate downstream signaling events that culminate with the production of IL-2 for T-cell proliferation and further effector functions of the T-cell (Smith-Garvin et al., 2009). Based upon the expression of key genes – including recombination-activating genes (RAG), LCK, Ikaros, TCR and CD3 chains – the thymus is the site of T-cell development and education for all gnathostomes (Hansen and Zapata, 1998). At this site, T-cells commit to either the CD4+ or CD8+ lineage. Several models have been put forth to explain T-cell lineage choice (Germain, 2002;

J.D. Hansen et al. / Developmental and Comparative Immunology 35 (2011) 452–460

He and Kappes, 2006; von Boehmer, 2000) including the stochastic, instructive and the more recent kinetic signaling model. The stochastic model describes the random predisposition of double positive (DP) thymocytes to follow either the CD4 or CD8 lineages through chance downregulation of one of the two co-receptors. If the wrong choice is made, downstream signaling events at later stages of T-cell development lead to the death of thymocytes with mismatched co-receptors. In the instructive model, differential intracellular signaling events are induced when the pre-TCR complex and CD8 co-engage MHC class I or the pre-TCR and CD4 co-engage MHC class II; the unique signals result in commitment to a particular lineage. In this model, the quality of the interaction between the TCR and MHC-peptide ligands, and the subsequent recruitment and activation of LCK, is essential to the signaling events leading to lineage choice. Considering that CD4 binds more efficiently with LCK than CD8␣ (Campbell et al., 1995; Wiest et al., 1993), it is not surprising that signals from LCK and the downstream extracellular signal-regulated kinase (ERK) are higher in those cells destined for the CD4 lineage and that low signal levels lead to commitment to the CD8 lineage. In the kinetic signaling model, opposing intensities of signals result from in a programmed transient reduction in CD8 expression during thymocyte development (He and Kappes, 2006). In this scenario, signals propagated in CD4hi /CD8lo T-cells vary in duration depending on the quality of interaction with class II ligand, such that those with a long lasting signal are programmed to permanently downregulate CD8 transcription, while those with a signal of short duration will reinitiate CD8 transcription and terminate transcription of CD4 (Singer, 2002). Thus differences between class I and II restricted TCRs are programmed at the CD4hi /CD8lo stage during T cell development and are likely transcriptionally regulated (Sarafova et al., 2005). The current view is that both strength (instructive) and duration (kinetic) of signal shape the routes for T-cell lineage choice. CD8 is a membrane-bound glycoprotein found on the surface of cytotoxic T-cells and some NK cell subsets as CD8␣␣ homodimers or CD8␣␤ heterodimers. NK cells and intraepithelial lymphocytes (IELs–TCR␥␦+ ) mainly express CD8␣␣ whereas the majority of TCR␣␤ lymphocytes express the CD8␣␤ heterodimer. Both subunits, CD8␣ or CD8␤, are structurally similar with each being composed of a leader peptide, an immunoglobulin superfamily (IgSf) V-domain, a heavily glycosylated stalk domain, followed by a trans-membrane region and cytoplasmic tail (Parnes, 1989). Biochemical properties within the extracellular IgSf V domain of CD8 allow it to associate with the ␣2 and ␣3 domains of MHC class IA molecules. Following interaction of the TCR with non-self peptides presented by MHC class IA, CD8 associates with LCK via a LCK binding motif (CXC) in its cytoplasmic tail. This latter interaction results in the phosphorylation of CD3, in mammals, and subsequent activation of the cytotoxic T-cells via downstream signaling events. A similar motif is found in the cytoplasmic tail of CD4 that is associated with LCK interaction. However, differing from CD8␣, the intracellular tail of mammalian CD8␤ does not possess a CXC motif and, hence, does not associate with LCK. Therefore, the CD8␣ chain of CD8 not only participates in MHC class I association but is indispensable for the activation of cytotoxic T-cells. The gnathostomes are represented by three major extant vertebrate lineages including cartilaginous fish (elasmobranches and holocephalians), ray-finned fish (teleosts) and sarcopterygii (lobe finned fish and tetrapods), the latter of which are thought to have eventually evolved to form the modern tetrapod lineages. We are interested in the origins and evolutionary forces that have shaped cell-mediated immunity in the jawed vertebrates. Cell-mediated cytotoxicity has been reported in several teleost models including crucian carp, rainbow trout and catfish (Fischer et al., 2003; Nakanishi et al., 2002; Shen et al., 2002; Toda et al., 2009; Utke et al., 2007). In addition, CD8 and MHC class IA genes have been

453

described from various teleosts (Buonocore et al., 2006; Hansen and Strassburger, 2000; Hansen et al., 1999; Moore et al., 2005, 2009; Suetake et al., 2007). On the same note, MHC class II genes and more recently CD4 and CD4-related genes have been described in teleost fish supporting the notion that cellular immunity is controlled by distinct cytotoxic (CD8) and helper (CD4) T-cell subsets, similar to that found for tetrapods (Dijkstra et al., 2006; Laing et al., 2006; Moore et al., 2009; Stet et al., 2003; Suetake et al., 2004; Sun et al., 2007). At the base of all gnathostomes, elasmobranchs represent a unique phylogenetic position for studies examining the evolution of vertebrate immunity. The major primary and secondary lymphoid tissues and many molecular components of adaptive immunity, including genes for MHC class IA, TCR (␣, ␤, ␥ and ␦), immunoglobulin heavy and light chains, and transcriptional regulators of lymphocyte development (e.g., Ikaros family, GATA-3, and Runx3) have been described in several elasmobranch species (Anderson et al., 2004; Bernstein et al., 1996; Criscitiello and Flajnik, 2007; Dooley and Flajnik, 2006; Fleurant et al., 2004; Miracle et al., 2001; Ohta et al., 2000; Rumfelt et al., 2004). In this report; we present the cloning and initial characterization of an elasmobranch CD8␣ homologue from the shovelnose guitarfish, Rhibatos productus. Our results indicate conservation of structural composition, genomic organization and expression patterns for CD8␣ during vertebrate evolution. Importantly the elasmobranch CD8␣ gene contains the conserved CHX motif found in all teleost sequences indicating that the “canonical” LCK association motif (CXC) found in tetrapods is a feature that was derived after the divergence of the tetrapod and fish lineages. 2. Materials and methods 2.1. cDNA cloning of Rhinobatos CD8 A short degenerate primer (DegenR-IgSF V) corresponding to the “F/Y-I/V-WAPL” sequence within the transmembrane region of tetrapod and some teleost CD8A genes was used to amplify the majority of CD8A from a directional (ZAP Express, Stratagene) guitarfish splenic cDNA library (kindly provided by Dr. Martin Flajnik, University of Maryland). Dilutions of the cDNA library were first heated to 95 ◦ C for 10 min in 1× PCR buffer and then used as templates for PCR under the following conditions: 3 cycles of 94 ◦ C for 15 s, 48 ◦ C for 30 s and 72 ◦ C for 30 s, followed by 32 cycles of 94 ◦ C for 15 s, 50 ◦ C (plus 0.15 ◦ C increase per cycle) for 30 s and 72 ◦ C for 30 s, and final incubation at 72 ◦ C for 10 min. PCR amplicons were cloned into pTOPO2.1 (Invitrogen, San Diego, CA) according to the manufacturer’s specifications. Twenty–five random transformants were selected for sequencing. Two primers, Rhpr-CD8Vfwd and Rhpr-CD8VRev, were designed against one clone, which demonstrated high similarity to mammalian CD8␣, to amplify a 240 bp probe to facilitate isolation of full-length Rhpr-CD8␣. The 240 bp probe was randomly labeled (Ready-to-go beads, Amersham) with [32P]dCTP (Amersham) and approximately 6 × 105 PFUs from the splenic library were plated and screened under high stringency (0.2× SSC/0.2% SDS at 65 ◦ C final wash) as previously described (Hansen et al., 1999). Upon secondary screening, putative positive plaques were confirmed by PCR using the Rhpr-CD8VFwd/Rev primer set. Three full-length clones were completely sequenced (ABI 3130) using BigDye v3.1 chemistry. 2.2. Genomic analysis of Rhinobatos CD8A The genomic structure of the guitarfish CD8A gene was obtained by long-range PCR amplification (Elongase, Invitrogen Corporation) followed by partial sequencing at predicted exon/intron

454

J.D. Hansen et al. / Developmental and Comparative Immunology 35 (2011) 452–460

Table 1 Primers and probes. Name of primer

Sequence 5 –3

Information

DegenR-IgSF V Rhpr-CD8Vfwd Rhpr-CD8Vrev Rp-CD8-F Rp-CD8-R Rp-CD8-Stop Rp-IgNAR-F Rp-IgNAR-R Uni-ARP-F Uni-ARP-R Rp-qCD8-F Rp-qCD8-R Rp-qIgNAR-F Rp-qIgNAR-R Rp-qARP-F Rp-qARP-R

TWCRTHTGGGCNCCNYT (W = A/T, R = A/G, H = A/C/T & Y = C/T) GATGAAAATTATCGGATTCT GTTTTCACGGTCAGAGAAAC CGATGAAAATTATCGGATTC GGTCTTCGAATATCTTCAGGTATC GAATTAACTTGTTAATGGTAG GCAGCAACATTCACAAAGAC AGATTGTGAAGGTGGGTGC GATGCCCAGGGAAGACAGG CCTGGAAGAAGGAGGTCTTCTC CCCTTGCACTCCGACCAA GACGAAGATCAGTGCGAGAAACA CGGACAGGAAAGACAAAAGCAA TAGAATCCTGTTATCAAGCATGTGAGTGT CCGTATATCACTGCGTGGTAAGG ATATGTGGCAGGAGCTTTTCCA

Reverse degenerate primer for amplification of CD8 along with T3 PCR probe for library screening PCR probe for library screening Amplification of CD8 Amplification of “nearly” full-length CD8 Amplification of full-length CD8 Amplification of a partial IgNAR cDNA clone Amplification of a partial IgNAR cDNA clone Universal ARP forward primer; partial clone Universal ARP reverse primer: partial clone Rp-CD8 forward primer for QRTPCR Rp-CD8 reverse primer for QRTPCR Rp-IgNAR forward primer for QRTPCR Rp-IgNAR reverse primer for QRTPCR Rp-ARP forward primer for QRTPCR Rp-ARP reverse primer for QRTPCR

QPCR probes Rp-CD8A Rp-IgNAR Rp-ARP

AAATGTCCTGCCACTTATTCGTCTGGGCTC CAAAGCTCCGACTCTTGCCCCCG TCCGTGGCCACCTGGAGAACAACC

QPCR probe for CD8 QPCR probe for IgNAR QPCR probe for ARP

boundaries. Briefly, 250 ng of Rhinobatos productus genomic DNA (kind gift from Dr. Martin Flajnik) was used as a template in conjunction with a forward (Rhpr-F and reverse primer set (RhprSTOP) as shown in Table 1). PCR conditions were as follows: 94 ◦ C for 1.5 min followed by 32 cycles of 94 ◦ C for 30 s, 58 ◦ C for 30 s and 68 ◦ C for 10 min. A single product was amplified and cloned into pBlunt (Invitrogen). Based upon the genomic structure of salmonid CD8␣ (Hansen and Strassburger, 2000; Moore et al., 2005; Suetake et al., 2007) and BLAT analysis of zebrafish CD8␣, primers corresponding to predicted exon/intron boundaries were used for sequencing the Rhpr-CD8␣ genomic clone as shown in Table 1. Methods for Southern blotting have been previously described (Hansen et al., 1999). Briefly, genomic DNA (10 ␮g) from a single individual was digested with indicated restriction enzymes, transferred to Nytran (Amersham) using 0.4N NaOH and hybridized with the RhprV-domain probe (same as for cDNA library screen) and washed under high stringency (0.2× SSC/0.2% SDS at 65 ◦ C). Radio-labeled blots were exposed to X-ray film for 72 h. 2.3. Quantitative RT-PCR for Rhpr-CD8A, IgNAR and ARP Quantitative RTPCR (qRTPCR) was used to define the tissuespecific expression of CD8 and IgNAR from a naïve, shovelnose guitarfish. An adult, male shovelnose guitarfish (R. productus) was caught via a baited hook at Seal Beach, CA (coordinates 33.7392◦ N 118.1100◦ W). The shovelnose guitarfish was then transferred to a flow-through seawater tank at the California State University (Long Beach) Biology Department and euthanized with buffered MS222. Indicated tissues (stomach, spleen, intestine-outer lining of the spiral valve, spiral valve, gill, liver, Leydig’s organ and Epigonal tissue) were harvested and placed into RNAlater (Qiagen) and stored at 80 ◦ C until processing. Total RNA was later processed from tissues using the RNAeasy kit (Qiagen) (Laing et al., 2006). Spleen and intestinal cDNA were used to amplify CD8A, IgNAR and acidic ribosomal protein P0 (ARP) with the following primer sets (Table 1): Rp–CD8–F/R, Rp–IgNAR–F/R and Uni-ARP–F/R to design specific QRTPCR primer and probe sets. Primers for CD8 (Rp–CD8–F/R) were based upon the CD8 sequence from guitarfish isolated during this investigation; (deposited in GenBank, accession number: EF559247). IgNAR primers (Rp–IgNAR–F/R) were designed against a guitarfish EST (AY524298) encoding IgNAR (Rumfelt et al., 2004) and ARP primers were derived from a multiple sequence alignment of teleost, Xenopus, chicken and murine ARP nucleic acid sequences. PCR conditions were as follows: 94 ◦ C

for 20 s followed by 30 cycles of 94 ◦ C for 30 s, 58 ◦ C for 30 s and 68 ◦ C for 1.5 min using proofreading (Platimum Pfx, Invitrogen) DNA polymerase. Amplified products were cloned into pBlunt (Invitrogen) and sequenced as described in Section 2.1. The following cloned amplicon sequences were deposited into GenBank: GU205321 (Rhpr-CD8A, 651 bp), GU20520 (Rhpr-IgNAR, 626 bp) and GU205319 (Rhpr-ARP, 419 bp). 2.4. Computational analysis Methods for multiple sequence alignments and the construction of phylogenetic trees have been previously described (Laing et al., 2006). Comparisons of nucleotide and amino acid sequences with sequences in Genbank and Swissprot databases were performed using BLAST-based (McGinnis and Madden, 2004) searches at NCBI (www.ncbi.nlm.nih.gov/BLAST). In addition, CD8 and CD7 sequences were identified from genome and EST indices using tBLASTn at NCBI and ENSEMBL (www.ensembl.org/index.html). Syntenic relationships were used to establish the Xenopus sequences and gene structures from available genomes were deduced using BLAT (http://genome.ucsc.edu/cgi-bin/hgBlat) and ENSEMBL. Signal peptides, IgSf domains and transmembrane regions were predicted using the Simple Modular Architecture Research Tool at http://smart.embl-heidelberg.de/. N- and Olinked glycosylation sites were assessed at the CBS main server (www.cbs.dtu.dk/services/). 3. Results and discussion A limiting factor for tracing the origins of cell-mediated immunity is the lack of information from elasmobranch species as they form the base for all gnathostomes in vertebrate phylogeny. We are interested in the origins of cellular immunity and, in particular, signaling events that are required for activating naïve T-cell subsets and controlling lineage commitment of immature lymphocytes to distinct T-cell subsets. Intrigued by the presence of a distorted LCK binding motif in all known teleost CD8A genes, we cloned CD8A from a representative elsamobranch to determine if this specific feature was unique to teleost fish. 3.1. Sequence features of R. productus CD8˛ A relatively short (17 bp), reverse degenerate primer located within the transmembrane region of CD8␣ was used for anchored

J.D. Hansen et al. / Developmental and Comparative Immunology 35 (2011) 452–460

455

Fig. 1. Multiple sequence alignment of CD8␣ amino acids from representative gnathostomes. The five distinct regions of CD8␣ (leader, IgSf V, stalk, transmembrane and cytoplasmic domains) are indicated above the appropriate regions with a raised, ended bar. The 9 ␤-strands (A–G) that comprise the human IgSf V domain are underlined with the letters above and the three complementary determining regions (CDRs) that are thought to interact with the MHC class I A3 domain are also listed. Absolutely conserved residues are shaded where cysteine residues involved in the Ig fold are shaded/boxed and the two cysteines responsible for the dimerization potential of CD8A are shaded and indicated by arrows. Cysteine residues within the transmembrane domain are boxed. Circled residues in the IgSf domain (A strand) indicate the beginning of the mature CD8 polypeptides as predicted by SMART analysis and circled residues within the stalk region represent O-linked glycosylation sites as predicted by the NetOGlyc 3.1 server. Finally, the LCK association motif in the cytoplasmic domain is boxed and labeled.

PCR using dilutions of a guitarfish unidirectional, splenic cDNA library as the template. Over 20 different amplicons in the appropriate size range (500–600 bp) were sequenced and analyzed by BlastX analysis. Two clones differing only in the length of the 5 UTR demonstrated significant amino acid identity (averaging 28%) to the V-domains of CD8␣ and TCR-␣/␥ from a variety of vertebrates. The nucleotide sequence corresponding to the V-domain was amplified and used as a homologous probe to screen the splenic library, yielding a full-length clone. Rhpr-CD8A (EF559247) codes for a single open reading frame of 219 amino acids followed by 1110 bp of 3 UTR that includes a polyadenylation site (bp 1749: AATAAA) and polyA-tail. After removing the predicted 18 amino acid leader sequence, BlastP analysis indicated that the mature peptide (201 aa) was ∼28–35% identical to teleost, avian and mammalian CD8␣ over a span of >180 amino acids. In fact, the first 30 BLASTP matches were for CD8␣ (E values: 4e−8 –3e−4 ) followed by TCRA/B, IgL and CD8B. An amino acid alignment (Fig. 1) was generated to display the relative conservation of CD8␣ across 5 different vertebrate classes and to support classification of the guitarfish sequence as CD8␣. All five species possess recognizable hydrophobic leader sequences followed by the IgSf V domain characterized by a 9 ␤strand structure in the order A-B-C-C -C -D-E-F-G. The hallmark of a true IgSf domain is the presence of 2 cysteine residues in ␤strands B and F that, via disulphide bonding, that is responsible for the stabilizing Ig-domain fold. These 2 cysteines are conserved in the CD8 orthologs of all species examined. An additional cysteine residue in the C strand, which forms a unique intradomain bond in mammalian CD8␣ (Kirszbaum et al., 1989) is not found in the other vertebrate taxa. The V domain is critical for the function of CD8, using the A/B ␤-strands and the complementary determining regions (CDRs), to associate with MHC class I-peptide complexes. Outside of the critical residues that define an IgSf domain, there is little overall conservation within the CD8␣ V domain throughout vertebrates. However, all of the A and B strands do possess acidic and basic residues – Arg4 and Lys21 in the A/B strands of human CD8␣ associate with a charged stretch of amino acids in the ␣3 region of MHC class I. CDR2, in all species, is lysine rich and therefore may also be involved in associating with the MHC class I solvent exposed loop throughout vertebrates. Finally, human CD8␣ associates with the ␣3 domain using Ser34 found on the border of CDR2/strand C and Tyr51 located at the start of

CDR2. With the exception of teleost CD8␣ where it is replaced with serine, the tyrosine of CDR2 is conserved, suggesting that this residue may be important for class I association in lower vertebrates (Buonocore et al., 2006). The serine residue of CDR1 is not conserved in ectotherms. In the hinge/stalk region there are two additional evolutionary conserved cysteine residues. Cys193 is absolutely critical for, and Cys178 plays a lesser role in, CD8␣␣ and CD8␣␤ dimerization (through interchain disulphide bonding) and cell surface expression, in humans, (Hennecke and Cosson, 1993; Kirszbaum et al., 1989; Leahy et al., 1992). Based upon the conservation of these two residues, it is apparent that the dimerization potential of CD8 has existed for over 450 million years of evolution. The proline/threonine-rich stalk/hinge region of CD8 also contains multiple O-linked glycosylation sites. Sialation of the O-linked sugar groups is proposed to create a more rigid structure in the hinge region (Jentoft, 1990) that enables CD8 to extend and associate with the MHC class I ␣2 and ␣3 regions (Leahy et al., 1992). Comparison of the hinge/stalk region between tetrapods, teleosts and the guitarfish sequence indicates that the guitarfish stalk region is similar in composition and size to that of CD8␣ vs. CD8␤ (Supplementary Fig. 1). The overall stalk length and composition of CD8␣ and ␤ is important for bending CD8␣␣ and CD8␣␤ dimers to ensure proper selection of CD8 subsets and responses of peripheral T-cells (Rettig et al., 2009). Neural network analysis (NetOGlyc3.1, Julenius et al., 2005) predicted multiple O-linked sites within all of the vertebrate CD8␣ sequences shown in Fig. 1 indicating that O-linked glycosylation of this region is likely conserved, constituting an essential structural feature for the function of CD8␣. The highest level of amino acid identity shared between CD8␣ sequences was found in the transmembrane domain (42% human vs. guitarfish) and to a lesser degree within the cytoplasmic region (33% human vs. guitarfish). The CD8␣ transmembrane region includes the presence of a relatively conserved cysteine residue that is not found in any CD8␤ sequence except in mice (Suetake et al., 2007). Interestingly, Hennecke and Cosson (Hennecke and Cosson, 1993) previously demonstrated that the transmembrane region of CD8␣ but not CD8␤ plays a minor but significant role in the formation of CD8␣␣ and CD8␣␤ homo- and heterodimers in addition to the major roles of the two conserved cysteines in the stalk/hinge region of CD8␣.

456

J.D. Hansen et al. / Developmental and Comparative Immunology 35 (2011) 452–460

The cytoplasmic region contains a striking feature, a canonical LCK association motif (CXC), which is responsible for the association with LCK in human T-cells; this interaction results in the phosphorylation of the TCR complex and initiates T-cell activation upon recognition of non-self. Mutational analysis (Turner et al., 1990) demonstrated that, in mammals, the two cysteines in this motif are important for the interaction with a cysteine-based motif (CXXC) found in the N-terminal portion of LCK. More precisely, the mutation analysis implied that the first cysteine in the CXC motif plays a larger role for the association with LCK. The CXXC motif in LCK is absolutely conserved from teleost fish through mammals

(Brenner et al., 2002; Laing et al., 2007; Langenau et al., 2004). However, as shown in Supplementary Fig. 2, the guitarfish CD8␣ sequence does not contain the CXC motif; it instead contains a CXH motif that is conserved within both CD8␣ and ␤ in teleost fish (Hansen and Strassburger, 2000; Moore et al., 2005; Suetake et al., 2007). The length (30 amino acids) and composition of the guitarfish CD8 cytoplasmic region is consistent with that of other CD8␣ (25–34 amino acids) cytoplasmic domains as compared to that found for CD8␤ (13–17 amino acids). Recently, Hayashi and colleagues (Hayashi et al., 2010) answered a very important question for the evolution of T-cell activation by demonstrating

Fig. 2. Phylogenetic tree depicting the relationships of CD7 and CD8 among various vertebrates. NJ tree based upon the V-domain (exon 2). A NJ tree based upon the entire mature ORF for the sequences produced identical branching orders. Node values represent the percentage of 1000 bootstrap replicants. CD7 and 8 accession numbers are as follows: CD7 human, X06180; mouse, D10329; chicken, CF257794 (Chr. 18 next to MSFB); feline, BAD23973. Both human and chicken CD7 are flanked by MSFB. Potential xenopus CD7: CX854356 (linked to MSFB). CD8A: carp, AB186398; zebrafish, AB186400 (Chr. 21 6.04MB); rainbow trout, AF178053; brown trout, AY701523; flounder, AB082958; Dog, NM 001002935; Xenopus, EL675625 (scaf 280); sea bass, AJ846849; human, M12828; mouse, P01731; chicken, NM 205235; duck, AF378373; bovine, 31783; feline, NM 001009843. CD8B: chicken, NM 205247; mouse, NM 009858; human, Y00805; axolotl, AF242416; rainbow trout, AY563420; feline, NM 001009867; Xenopus, EB645787 and Scaf 280; Zebrafish CD8B: EB981219 plus Genscan analysis of Chr. 21 at region 6.06MB. T: tetrapod, F: fish and arrow: guitarfish.

J.D. Hansen et al. / Developmental and Comparative Immunology 35 (2011) 452–460

457

Fig. 3. Genomic structure of CD8␣ in gnathostomes. The exon/intron organization of CD8 from 5 classes of vertebrates. The Rhpr-CD8␣ genomic region was amplified by long-range PCR and sequenced to determine splice sites and phase of introns. The genomic structure for trout CD8␣ has been previously reported (Hansen and Strassburger, 2000) and the structures for the other CD8␣ genes were deduced using BLAT and ENSEMBL analysis. Amino acids encoded for each exon are shown below the exon (e.g., Hosa exon encodes 16AA) and intron phase is indicated between exons. Two exons encode the cytoplasmic domain (C1 and C2). Hosa: Homo sapien, Mumu: Mus musculus, Xetr: Xenopus tropicalis, Onmy-Oncorhynchus mykiss and Rhpr: Rhinobatos productus.

that the CXH-containing cytoplasmic domain of trout CD8␣ can indeed interact with LCK – thus further implying the critical role of the first cysteine for LCK association. The overall conservation of the CD8␣ and CD8␤ cytoplasmic domains from bony fish coupled with the guitarfish sequence lends for speculation that the CXH motif represents the ancestral LCK interaction motif. During the divergence and evolution of tetrapods, there were likely early mutation and selection events that led to the “canonical” CXC motif – possibly for fine tuning TCR:MHC signaling in tetrapods. To further test guitarfish CD8 as a CD8␣ ortholog, we conducted a phylogenetic analysis using all available CD8␣ andCD8␤ sequences along with the structurally similar CD7 gene. As depicted in Fig. 2, the tetrapod and teleost CD8␣ andCD8␤ sequences cluster into their own distinct clades with the guitarfish CD8 sequence preferentially grouping with the overall CD8␣ clade – strongly supported by bootstrap analysis – thus further supporting the notion that the Rhinobatos sequence is likely a CD8␣ ortholog. 3.2. The Rhinobatos CD8˛ gene One key feature of vertebrate CD8 is its gene organization. As further support for confirming its orthology to vertebrate CD8␣, we utilized a long-range PCR approach to amplify the guitarfish CD8␣ gene from genomic DNA. The structure of the shovelnose guitarfish CD8␣ gene is conserved relative to human and teleostei CD8␣ in both exon organization and intron phase (Fig. 3). The various CD8␣ exons encode distinct regions for CD8 and all are of approximately the same size across the vertebrate taxa. Exon 1 codes for the 5 UTR and hydrophobic signal sequence followed by exon 2, which encodes the IgSf V-like domain. Exon 3 codes for the membrane-proximal hinge/stalk region, exon 4 the transmembrane domain followed by exons, 5 and 6 which together, encode the cytoplasmic domain. Exon 5 encodes for the first half of the cytoplasmic domain that includes the LCK association site. Identical for all investigated vertebrate CD8 genes, the exons for the shovelnose guitarfish CD8␣ gene are split by phase 1 introns for the first four splice sites and by a phase 2 intron for the last splice site.

Fig. 4. Rhpr-CD8␣ is a single copy gene as determined by Southern blot analysis. R. productus gDNA from a single individual was digested with BamH1, HindIII, EcoRV or EcoRI, blotted to nylon and hybridized under stringent conditions using a labeled probe corresponding to Rhpr-CD8␣ exon 2. Molecular weight designations are in kilobase pairs.

458

J.D. Hansen et al. / Developmental and Comparative Immunology 35 (2011) 452–460

Fig. 5. Relative mRNA expression of R. productus CD8␣ and IgNAR in adult tissues. Expression levels for CD8␣ and IgNAR were normalized relative to the expression of acidic ribosomal protein (ARP). Values shown represent CD8 (light shading, black outline) and IgNAR (black shading, white outline) fold change compared to the Leydig’s gland (set to 1). Tissues are as follows: Sto, stomach; Int, outer wall of spiral valve; Epi, epigonal tissue; Gill, Ley, Leydig’s gland; Liv, liver and spleen.

A probe derived from the Rhpr-CD8␣ V domain (exon 2) was used to detect CD8␣ gene(s) in genomic DNA digested with either BamHI, HindIII, EcoRI or EcoRV during Southern blotting (Fig. 4). One band was found for all digests except EcoRI – a cut site for the latter enzyme exists in exon 2 of guitarfish CD8␣ – suggesting that CD8␣ exists as a single copy gene in elasmobranchs, as in most other vertebrates. An exception is found in avian species; multiple partial exons for the CD8␣ V-domain are present in the chicken CD8␣ locus that generate a variety of polymorphic CD8␣ alleles in a process analogous to gene conversion (Liaw et al., 2007). This “repertoire” for chicken CD8␣ is suggested to be under co-evolutionary pressure to accommodate the polymorphic nature of the chicken MHC class I ␣3 region, which interacts with CD8.

3.3. Tissue-specific expression of T and B cell markers in R. productus We then assessed the relative expression of CD8␣ (a cytotoxic T-cell marker) to that of IgNAR (a B-cell marker) in an adult, naive shovelnose guitarfish. To do so, a recently caught R. productus specimen was used to generate a panel of tissue cDNAs for expression analysis. Prior to qRTPCR, CD8␣ and IgNAR were amplified from this individual to allow the design of specific primer and probe sets. Surprisingly, the CD8␣ and IgNAR nucleotide sequences from the R. productus specimen caught near Longbeach, CA were different from the CD8 sequence obtained during this study (Section 3.1) and a previously reported full-length IgNAR sequence (82% identity to AY524298, Rumfelt et al., 2004), both derived from the same Rhinobatos cDNA library. This raised the question as to whether shovelnose guitarfish encode multiple CD8␣ genes or whether the two sequences (EF559247, GU205321: 83% nucleotide identity) are distant alleles. Primer sets specific for the two guitarfish CD8␣ sequences could only amplify CD8 from the guitarfish spleen library or from splenic cDNA from the recently caught shovelnose guitarfish but not both (data not shown): consistent with CD8␣ being a single copy gene as suggested by Southern blot analysis. There are >9 species within Rhinobatos, thus the tissue provided to make the Rhinobatos cDNA library (Rumfelt et al., 2004) was most likely a closely related congener within Rhinobatos. As shown in Fig. 5, Rhpr-CD8␣ is mainly expressed in the spleen and the mucosal/gut-associated immune tissues (gill, intestine – outer wall of spiral valve, and the spiral valve itself) and to a lesser

extent in the liver. This mirrors expression of elasmobranch TCR␣␤ mRNA, which is limited to the thymus, spleen, spiral valve, intestine and gills (Criscitiello et al., 2010; Miracle et al., 2001), and upholds the hypothesis that elasmobranch CD8␣ is T cell associated. We were unable to locate an identifiable thymus in this specimen, an adult guitarfish, likely owing to thymic involution. In contrast, IgNAR (one of the three elasmobranch IgH isotypes: IgM, IgX/W and IgNar) expression (Fig. 5) was primarily localized to the epigonal tissue and the spleen followed by the liver and gills. This analysis included the Leydig’s gland, a posterior portion of the mesonephros involved in semen secretion, not to be confused with the Leydig’s organ. The epigonal and Leydig organs (the latter of which is not found in all elasmobranchs) are gonad and esophageal associated tissues, which are unique to elasmobranchs and – by virtue of Ig heavy chain, RAG, TdT and B-cell specific transcription factor expression within them – are considered major sites of Bcell development for both skates and sharks (Anderson et al., 2004; Miracle et al., 2001; Rumfelt et al., 2002). We were not identify a Leydig’s organ in our biopsy of R. productus. Thus, the high mRNA expression of IgNAR in the epigonal tissue strongly suggests that this tissue is largely responsible for B-cell development in adult Rhinobatos differing to the situation in rays, where the Leydig’s tissue has been suggested play a more prominent role for B-cell development in adults (Miracle et al., 2001). Therefore the expression of Rhpr-CD8␣ and IgNAR are consistent with results from other cartilaginous fish for identifying the sites of T and B cell lymphocyte populations.

4. Concluding remarks One main question that arises after this study is whether the gene we have described as Rhpr-CD8␣ truly represents CD8␣ or -␤. We contend that our analyses, which included amino acid identity profiles, molecular phylogeny, primary structural features, genomic architecture, presence of a cysteine in the transmembrane and length differences of the stalk and cytoplasmic regions between CD8␣ and ␤, supports our argument that this sequence is a likely ortholog to mammalian CD8␣. There is considerable interest in the participation of CD8 for TCR co-receptor functionality and, moreover, in the overall evolution of this gene family as they play major roles in T-cell lineage development and functionality. Both CD8 genes are tightly linked

J.D. Hansen et al. / Developmental and Comparative Immunology 35 (2011) 452–460

in all vertebrate classes examined including mammals (30–50 kbp), birds (∼200 kbp), amphibians (Xenopus tropicalis-32 kbp) and bony fish (3–20 kbp) strongly supporting the hypothesis that the CD8 gene family arose from a simple cis-duplication event prior to the divergence of the osteichthyes. CD8␣ is more representative of the ancestral form of this two-member gene family as CD8␣␣ homodimers are expressed by NK and ␥␦ T-cells, which are considered to be the most primitive forms of lymphocytes based upon their innate roles in immunity. CD8␣ possesses feature(s), not only for association with MHC class I, but also for association with the membrane-associated kinase LCK for eliciting an activation signal in T-cells; cytotoxic T-cells can associate with classical MHC class I molecules, in the absence of CD8␤ but not CD8␣ in mammals, and CD8␣ but not CD8␤, associates with both LCK and LAT by means of its cytoplasmic tail (Cheroutre and Lambolez, 2008). However, CD8␤, but not CD8␣, appears to play a major role in directing TCR and CD3 molecules to the lipid raft for enhanced T-cell signaling (Cheroutre and Lambolez, 2008). Increasing evidence suggests CD8␣␣ homodimers are negative regulators (co-repressors) of cytotoxic T-lipid rafts. These CD8␣␣ homodimers, in mice, have a strong preference to interact with thymus leukemia antigen (TL), a non-classical MHC class IB molecule (Liu et al., 2003). The CD8␣␣ positive lymphocytes display an “always active” phenotype, thought to be controlled by repression from CD8␣␣. Sharks possess a variety of non-classical MHC class I genes in addition to classical MHC class I genes, supporting potential for conserved interactions of CD8␣ with both types of class I molecules (Hashimoto et al., 1999; Ohta et al., 2000; Wang et al., 2003). In mammals, non-disulphide bonding occurs between LCK and the tails of both CD4 and CD8␣ and is stabilized by a Zn2+ ion co-ordinated between the four cysteines (Kim et al., 2003). The relevant cysteine pairs within LCK and CD4-like molecules are conserved in lower vertebrates, presumably reflecting conservation of this interaction (Buonocore et al., 2008; Dijkstra et al., 2006; Laing et al., 2006, 2007; Suetake et al., 2004). However, in the CD8␣ proteins of both cartilaginous and bony fish, the CXC motif is replaced with a conserved CXH motif (Fig. 2). Studies have shown the metal cation binding capacity of histidine residues; this amino acid, by virtue of its imidazole side chains, is a receptor for Cu2+ , Ni2+ and Zn2+ ions (Jakab et al., 2008). In addition CXH motifs have been shown to co-operate with CXXC motifs as Zn2+ binding sites, providing the ion with its preferred tetrahedral interaction (Li et al., 2006; Simonson and Calimet, 2002). Zinc dependent binding between trout LCK and CD8␣ does occur with similar affinities as that for human LCK and CD8␣ (Hayashi et al., 2010), and importantly, this interaction was impaired by zinc chelators demonstrating that the molecular mechanism of the CD8:LCK interaction has been conserved during evolution. Based upon these studies, we would speculate that elasmobranch CD8␣ would also associate with LCK to guide T-cell education and activation. Thus it appears that CD8␤ likely evolved to enhance the expression and function of CD8 on TCR␣␤ T-cells as part of the evolution of the adaptive immune response as CD8␣␣ homoodimers are expressed mainly by TCR␥␦ T-cells. Again, gene duplication followed by divergence and selection represents a logical scenario for the generation of the CD8␣ and −␤ genes and their functional diversification. Finally, the analysis presented here coupled with prior work on teleost and mammalian CD8 evolution suggests that the CXC LCK association motif found in tetrapods is a derived feature.

Acknowledgements We would like to thank Drs. Hans Dijkstra and Carl Luer for their comments on the manuscript. The use of trade, firm, or corporation names in this publication is for the information and convenience of

459

the reader. Such use does not constitute an official endorsement or approval by the U.S. Department of Interior or the U.S. Geological Survey of any product or service to the exclusion of others that may be suitable. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.dci.2010.11.014. References Anderson, M.K., Pant, R., Miracle, A.L., Sun, X., Luer, C.A., Walsh, C.J., Telfer, J.C., Litman, G.W., Rothenberg, E.V., 2004. Evolutionary origins of lymphocytes: ensembles of T cell and B cell transcriptional regulators in a cartilaginous fish. J. Immunol. 172, 5851–5860. Bernstein, R.M., Schluter, S.F., Bernstein, H., Marchalonis, J.J., 1996. Primordial emergence of the recombination activating gene 1 (RAG1): sequence of the complete shark gene indicates homology to microbial integrases. Proc. Natl. Acad. Sci. U.S.A. 93, 9454–9459. Brenner, S., Venkatesh, B., Yap, W.H., Chou, C.F., Tay, A., Ponniah, S., Wang, Y., Tan, Y.H., 2002. Conserved regulation of the lymphocyte-specific expression of LCK in the Fugu and mammals. Proc. Natl. Acad. Sci. U.S.A. 99, 2936–2941. Buonocore, F., Randelli, E., Bird, S., Secombes, C.J., Costantini, S., Facchiano, A., Mazzini, M., Scapigliati, G., 2006. The CD8alpha from sea bass (Dicentrarchus labrax L.): cloning, expression and 3D modelling. Fish Shellfish Immunol. 20, 637–646. Buonocore, F., Randelli, E., Casani, D., Guerra, L., Picchietti, S., Costantini, S., Facchiano, A.M., Zou, J., Secombes, C.J., Scapigliati, G., 2008. A CD4 homologue in sea bass (Dicentrarchus labrax): molecular characterization and structural analysis. Mol. Immunol. 45, 3168–3177. Campbell, K.S., Buder, A., Deuschle, U., 1995. Interactions between the amino-terminal domain of p56LCK and cytoplasmic domains of CD4 and CD8 alpha in yeast. Eur. J. Immunol. 25, 2408–2412. Cheroutre, H., Lambolez, F., 2008. Doubting the TCR coreceptor function of CD8alphaalpha. Immunity 28, 149–159. Criscitiello, M.F., Flajnik, M.F., 2007. Four primordial immunoglobulin light chain isotypes, including lambda and kappa, identified in the most primitive living jawed vertebrates. Eur. J. Immunol. 37, 2683–2694. Criscitiello, M.F., Ohta, Y., Saltis, M., McKinney, E.C., Flajnik, M.F., 2010. Evolutionarily conserved TCR binding sites, identification of T cells in primary lymphoid tissues, and surprising trans-rearrangements in nurse shark. J. Immunol. 184, 6950–6960. Dijkstra, J.M., Somamoto, T., Moore, L., Hordvik, I., Ototake, M., Fischer, U., 2006. Identification and characterization of a second CD4-like gene in teleost fish. Mol. Immunol. 43, 410–419. Dooley, H., Flajnik, M.F., 2006. Antibody repertoire development in cartilaginous fish. Dev. Comp. Immunol. 30, 43–56. Fischer, U., Utke, K., Ototake, M., Dijkstra, J.M., Kollner, B., 2003. Adaptive cell-mediated cytotoxicity against allogeneic targets by CD8-positive lymphocytes of rainbow trout (Oncorhynchus mykiss). Dev. Comp. Immunol. 27, 323–337. Fleurant, M., Changchien, L., Chen, C.T., Flajnik, M.F., Hsu, E., 2004. Shark Ig light chain junctions are as diverse as in heavy chains. J. Immunol. 173, 5574–5582. Germain, R.N., 2002. T-cell development and the CD4-CD8 lineage decision. Nat. Rev. Immunol. 2, 309–322. Hansen, J.D., Strassburger, P., 2000. Description of an ectothermic TCR coreceptor CD8 alpha, in rainbow trout. J. Immunol. 164, 3132–3139. Hansen, J.D., Strassburger, P., Thorgaard, G.H., Young, W.P., Du Pasquier, L., 1999. Expression, linkage, and polymorphism of MHC-related genes in rainbow trout, Oncorhynchus mykiss. J. Immunol. 163, 774–786. Hansen, J.D., Zapata, A.G., 1998. Lymphocyte development in fish and amphibians. Immunol. Rev. 166, 199–220. Hashimoto, K., Okamura, K., Yamaguchi, H., Ototake, M., Nakanishi, T., Kurosawa, Y., 1999. Conservation and diversification of MHC class I and its related molecules in vertebrates. Immunol. Rev. 167, 81–100. Hayashi, N., Takeuchi, M., Nakanishi, T., Hashimoto, K., Dijkstra, J.M., 2010. Zinc-dependent binding between peptides derived from rainbow trout CD8alpha and LCK. Fish Shellfish Immunol. 28, 72–76. He, X., Kappes, D.J., 2006. CD4/CD8 lineage commitment: light at the end of the tunnel? Curr. Opin. Immunol. 18, 135–142. Hennecke, S., Cosson, P., 1993. Role of transmembrane domains in assembly and intracellular transport of the CD8 molecule. J. Biol. Chem. 268, 26607–26612. Jakab, I.N., Lorincz, O., Jancso, A., Gajda, T., Gyurcsik, B., 2008. Approaching the minimal metal ion binding peptide for structural and functional metalloenzyme mimicking. Dalton Trans. 1, 6987–6995. Jentoft, N., 1990. Why are proteins O-glycosylated? Trends Biochem. Sci. 15, 291–294. Julenius, K., Molgaard, A., Gupta, R., Brunak, S., 2005. Prediction, conservation analysis, and structural characterization of mammalian mucin-type O-glycosylation sites. Glycobiology 15, 153–164. Kim, P.W., Sun, Z.Y., Blacklow, S.C., Wagner, G., Eck, M.J., 2003. A zinc clasp structure tethers LCK to T cell coreceptors CD4 and CD8. Science 301, 1725–1728.

460

J.D. Hansen et al. / Developmental and Comparative Immunology 35 (2011) 452–460

Kirszbaum, L., Sharpe, J.A., Goss, N., Lahnstein, J., Walker, I.D., 1989. The alpha-chain of murine CD8 lacks an invariant Ig-like disulfide bond but contains a unique intrachain loop instead. J. Immunol. 142, 3931–3936. Laing, K.J., Dutton, S., Hansen, J.D., 2007. Molecular and biochemical analysis of rainbow trout LCK suggests a conserved mechanism for T-cell signaling in gnathostomes. Mol. Immunol. 44, 2737–2748. Laing, K.J., Zou, J.J., Purcell, M.K., Phillips, R., Secombes, C.J., Hansen, J.D., 2006. Evolution of the CD4 family: teleost fish possess two divergent forms of CD4 in addition to lymphocyte activation gene-3. J. Immunol. 177, 3939–3951. Langenau, D.M., Ferrando, A.A., Traver, D., Kutok, J.L., Hezel, J.P., Kanki, J.P., Zon, L.I., Look, A.T., Trede, N.S., 2004. In vivo tracking of T cell development, ablation, and engraftment in transgenic zebrafish. Proc. Natl. Acad. Sci. U.S.A. 101, 7369–7374. Leahy, D.J., Axel, R., Hendrickson, W.A., 1992. Crystal structure of a soluble form of the human T cell coreceptor CD8 at 2.6—a resolution. Cell 68, 1145–1162. Li, Z., Cao, R., Wang, M., Myers, M.P., Zhang, Y., Xu, R.M., 2006. Structure of a Bmi-1-Ring1B polycomb group ubiquitin ligase complex. J. Biol. Chem. 281, 20643–20649. Liaw, H.J., Chen, W.R., Huang, Y.C., Tsai, C.W., Chang, K.C., Kuo, C.L., 2007. Genomic organization of the chicken CD8 locus reveals a novel family of immunoreceptor genes. J. Immunol. 178, 3023–3030. Liu, Y., Xiong, Y., Naidenko, O.V., Liu, J.H., Zhang, R., Joachimiak, A., Kronenberg, M., Cheroutre, H., Reinherz, E.L., Wang, J.H., 2003. The crystal structure of a TL/CD8alphaalpha complex at 2.1 A resolution: implications for modulation of T cell activation and memory. Immunity 18, 205–215. McGinnis, S., Madden, T.L., 2004. BLAST: at the core of a powerful and diverse set of sequence analysis tools. Nucl. Acids Res. 32, W20–25. Miracle, A.L., Anderson, M.K., Litman, R.T., Walsh, C.J., Luer, C.A., Rothenberg, E.V., Litman, G.W., 2001. Complex expression patterns of lymphocyte-specific genes during the development of cartilaginous fish implicate unique lymphoid tissues in generating an immune repertoire. Int. Immunol. 13, 567–580. Moore, L.J., Dijkstra, J.M., Koppang, E.O., Hordvik, I., 2009. CD4 homologues in Atlantic salmon. Fish Shellfish Immunol. 26, 10–18. Moore, L.J., Somamoto, T., Lie, K.K., Dijkstra, J.M., Hordvik, I., 2005. Characterisation of salmon and trout CD8alpha and CD8beta. Mol. Immunol. 42, 1225–1234. Nakanishi, T., Fischer, U., Dijkstra, J.M., Hasegawa, S., Somamoto, T., Okamoto, N., Ototake, M., 2002. Cytotoxic T cell function in fish. Dev. Comp. Immunol. 26, 131–139. Ohta, Y., Okamura, K., McKinney, E.C., Bartl, S., Hashimoto, K., Flajnik, M.F., 2000. Primitive synteny of vertebrate major histocompatibility complex class I and class II genes. Proc. Natl. Acad. Sci. U.S.A. 97, 4712–4717. Parnes, J.R., 1989. Molecular biology and function of CD4 and CD8. Adv. Immunol. 44, 265–311. Rettig, L., McNeill, L., Sarner, N., Guillaume, P., Luescher, I., Tolaini, M., Kioussis, D., Zamoyska, R., 2009. An essential role for the stalk region of CD8B in the coreceptor function of CD8. J. Immunol. 182, 121–129. Rumfelt, L.L., Diaz, M., Lohr, R.L., Mochon, E., Flajnik, M.F., 2004. Unprecedented multiplicity of Ig transmembrane and secretory mRNA forms in the cartilaginous fish. J. Immunol. 173, 1129–1139.

Rumfelt, L.L., McKinney, E.C., Taylor, E., Flajnik, M.F., 2002. The development of primary and secondary lymphoid tissues in the nurse shark Ginglymostoma cirratum: B-cell zones precede dendritic cell immigration and T-cell zone formation during ontogeny of the spleen. Scand. J. Immunol. 56, 130–148. Sarafova, S.D., Erman, B., Yu, Q., Van Laethem, F., Guinter, T., Sharrow, S.O., Feigenbaum, L., Wildt, K.F., Ellmeier, W., Singer, A., 2005. Modulation of coreceptor transcription during positive selection dictates lineage fate independently of TCR/coreceptor specificity. Immunity 23, 75–87. Shen, L., Stuge, T.B., Zhou, H., Khayat, M., Barker, K.S., Quiniou, S.M., Wilson, M., Bengten, E., Chinchar, V.G., Clem, L.W., Miller, N.W., 2002. Channel catfish cytotoxic cells: a mini-review. Dev. Comp. Immunol. 26, 141–149. Simonson, T., Calimet, N., 2002. Cys(x)His(y)–Zn2+ interactions: thiol vs. thiolate coordination. Proteins 49, 37–48. Singer, A., 2002. New perspectives on a developmental dilemma: the kinetic signaling model and the importance of signal duration for the CD4/CD8 lineage decision. Curr. Opin. Immunol. 14, 207–215. Smith-Garvin, J.E., Koretzky, G.A., Jordan, M.S., 2009. T cell activation. Annu. Rev. Immunol. 27, 591–619. Stet, R.J., Kruiswijk, C.P., Dixon, B., 2003. Major histocompatibility lineages and immune gene function in teleost fishes: the road not taken. Crit. Rev. Immunol. 23, 441–471. Suetake, H., Araki, K., Akatsu, K., Somamoto, T., Dijkstra, J.M., Yoshiura, Y., Kikuchi, K., Suzuki, Y., 2007. Genomic organization and expression of CD8alpha and CD8beta genes in fugu Takifugu rubripes. Fish Shellfish Immunol. 23, 1107–1118. Suetake, H., Araki, K., Suzuki, Y., 2004. Cloning, expression, and characterization of fugu CD4, the first ectothermic animal CD4. Immunogenetics 56, 368–374. Sun, X.F., Shang, N., Hu, W., Wang, Y.P., Guo, Q.L., 2007. Molecular cloning and characterization of carp (Cyprinus carpio L.) CD8beta and CD4-like genes. Fish Shellfish Immunol. 23, 1242–1255. Toda, H., Shibasaki, Y., Koike, T., Ohtani, M., Takizawa, F., Ototake, M., Moritomo, T., Nakanishi, T., 2009. Alloantigen-specific killing is mediated by CD8-positive T cells in fish. Dev. Comp. Immunol. 33, 646–652. Turner, J.M., Brodsky, M.H., Irving, B.A., Levin, S.D., Perlmutter, R.M., Littman, D.R., 1990. Interaction of the unique N-terminal region of tyrosine kinase p56LCK with cytoplasmic domains of CD4 and CD8 is mediated by cysteine motifs. Cell 60, 755–765. Utke, K., Bergmann, S., Lorenzen, N., Kollner, B., Ototake, M., Fischer, U., 2007. Cellmediated cytotoxicity in rainbow trout, Oncorhynchus mykiss, infected with viral haemorrhagic septicaemia virus. Fish Shellfish Immunol. 22, 182–196. von Boehmer, H., 2000. T-cell lineage fate: instructed by receptor signals? Curr. Biol. 7, R642–645. Wang, C., Perera, T.V., Ford, H.L., Dascher, C.C., 2003. Characterization of a divergent non-classical MHC class I gene in sharks. Immunogenetics 55, 57–61. Wiest, D.L., Yuan, L., Jefferson, J., Benveniste, P., Tsokos, M., Klausner, R.D., Glimcher, L.H., Samelson, L.E., Singer, A., 1993. Regulation of T cell receptor expression in immature CD4+ CD8+ thymocytes by p56LCK tyrosine kinase: basis for differential signaling by CD4 and CD8 in immature thymocytes expressing both coreceptor molecules. J. Exp. Med. 178, 1701–1712.