|Home | About | Journals | Submit | Contact Us | Français|
Transmembrane bound receptors comprised of extracellular immunoglobulin (Ig) or lectin domains play integral roles in a large number of immune functions including inhibitory and activating responses. The function of many of the activating receptors requires a physical interaction with an adaptor protein possessing a cytoplasmic regulatory motif. The partnering of an activating receptor with an adaptor protein relies on complementary charged residues in the two transmembrane domains. The mammalian natural killer (NK) and Fc receptors (FcR) represent two of many receptor families, which possess activating receptors that partner with adaptor proteins for signaling. Zebrafish represent a powerful experimental model for understanding developmental regulation at early stages of embryogenesis and for efficiently generating transgenic animals. In an effort to understand developmental aspects of immune receptor function, we have accessed the partially annotated zebrafish genome to identify six different adaptor molecules: Dap10, Dap12, Cd3ζ, Cd3ζ-like, FcRγ and FcRγ-like that are homologous to those effecting immune function in mammals. Their genomic organizations have been characterized, cDNA transcripts have been recovered, phylogenetic relationships have been defined and their cell lineage-specific expression patterns have been established.
The cells of the mammalian immune system rely on an intricate network of signaling pathways in order to differentiate between “self” and “non-self”. The activation or inhibition of these signaling pathways relies on specific membrane receptors on immune cells engaging specific ligands. In general, these receptors can be classified as inhibitory or activating based on the functional outcome of ligand recognition. For example, when an activating natural killer (NK) cell receptor binds its ligand, the NK cell is activated to kill the target cell; in contrast, when an inhibitory NK receptor binds its ligand, NK cell-mediated killing is repressed. Similarly, engagement of T-cell antigen receptor (TCR) or Fc receptor (FcεRI) with the appropriate ligand (e.g. peptide-MHC complex or IgE, respectively) leads to a direct cell-mediated immune response.
Despite differences in receptor structures, the cytoplasmic signaling utilized by NK receptors, TCR and Fc receptors, are well conserved (Billadeau and Leibson 2002; Cerwenka and Lanier 2000; Yokoyama and Kim 2006). Inhibitory NK and Fc receptors typically possess one or more cytoplasmic immunoreceptor tyrosine-based inhibition motifs (ITIMs). In contrast, activating NK, Fc and other receptors including TCR partner with adaptor proteins, which ultimately transduce signals to the cell nucleus. These receptors physically associate with an adaptor protein via oppositely charged residues within their transmembrane domains, e.g. a positive charge in the transmembrane domain of the activating receptor and a negative charge in the transmembrane domain of the adaptor protein. The majority of activating NK receptors, including most KIRs and Ly49, utilize the adaptor protein, DAP12; the activating NK receptor NKG2D, another lectin-type receptor, utilizes the adaptor DAP10 (Hyka-Nouspikel and Phillips 2006; Takaki et al 2006). TCR and FcR along with other immune related activating receptors, including: CD16, NKp30, NKp46 and NKR-P1C and KIR2DL4, utilize FcRγ or CD3ζ (Cerwenka and Lanier 2000; Tassi et al 2006). The DAP12, FcRγ and CD3ζ adaptors utilize cytoplasmic immunoreceptor tyrosine-based activation motifs (ITAMs: YxxLX6-12YxxL/I) for signaling (Pitcher and van Oers 2003), whereas DAP10 uses a YxxM motif similar to CD28 (Wu et al 1999).
The zebrafish is becoming a more broadly recognized model for infection and immunity and is particularly well suited for examining gene function during embryogenesis (Deiters and Yoder 2006; Lieschke and Currie 2007; Phelps and Neely 2005; Traver et al 2003a; Trede et al 2004; Van Der Sar et al 2004; Yoder et al 2002). As part of an ongoing effort to characterize immune receptors in zebrafish and to understand the signaling pathways mediated by their immune cells, we have identified and characterized the adaptor molecules: Dap12, Dap10, CD3ζ, CD3ζ-like, FcRγ, and FcRγ-like.
A catfish (Ictalurus punctatus) DAP10 cDNA sequence (GenBank: AAZ16504) was used as the query for a tBLASTn search of the NCBI zebrafish sequence database (Wheeler et al 2007). Zebrafish bacterial artificial chromosome (BAC) clone DKEY-29H14 (GenBank: BX571853) from chromosome 16 encodes a sequence similar to catfish DAP10 (E value = 1e-10). A sequential RACE strategy was used to clone the full-length open reading frame (ORF) of dap10 (hcst) from zebrafish kidney/spleen RNA (GeneRacer, Invitrogen). Initially, 3′ RACE was completed with nested forward primers (designed from the DKEY-29H14 sequence) ZFDAP10-F1 and ZFDAP10-F2 which identified the 3′ untranslated region of the dap10 mRNA. Subsequently, nested, reverse primers ZFDAP10-3′UTR-R1 and ZFDAP10-3′UTR-R2 were designed within the 3′ UTR of dap10 and 5′ RACE was completed to generate a cDNA encoding the ORF of dap10. Primer sequences are listed in Table 1. PCR products were cloned into pGEM-T easy (Promega) or pCR4-TOPO (Invitrogen) and sequenced.
tBLASTn analyses of the zebrafish nucleotide (nr), genome and EST databases, using human DAP12 as a query, failed to identify a similar sequence (no sequence identified with an E value <0.05). As DAP10 and DAP12 are adjacent genes in mammals and pufferfish (Guselnikov et al 2003b), the zebrafish BAC DKEY-29H14, which encodes dap10, was examined for dap12 sequence. A BLASTx analysis using the nucleotide sequence of DKEY-29H14 as a query identified a short sequence (108 nucleotides) within this BAC with extremely low similarity (E value = 1.5) to the transmembrane domain of mouse TCRδ (GenBank: AAH62807) which, like the transmembrane domain of DAP12, includes a negatively charged residue. We predicted that this novel transmembrane domain encoded by BAC DKEY-29H14 would represent an exon of zebrafish dap12 (tyrobp) based on the following observations: 1) the transmembrane domains encoded by this novel zebrafish sequence and mammalian DAP12 include a negatively charged residue; 2) this candidate zebrafish transmembrane domain is adjacent to dap10; and 3) no other sequences within DKEY-29H14 shares similarity with DAP12. Subsequently, a full-length open reading frame cDNA of this sequence was cloned (as follows) and confirmed to encode dap12. Initially, 3′ RACE was completed from zebrafish kidney/spleen RNA with nested forward primers (designed from the DKEY-29H14 novel transmembrane sequence) ZFDAP12-F1 and ZFDAP12-F2 which identified the 3′ untranslated region of the dap12 mRNA. Subsequently, nested, reverse primers ZFDAP12-R3 and ZFDAP12-R4 were designed within the ORF of dap12 and 5′ RACE was completed to generate a cDNA encoding the 5′ UTR of dap12. Finally, nested, forward primers ZFDAP12-5′UTR-F3 and ZFDAP12-5′UTR-F4 were designed within the 5′ UTR of dap12 and a cDNA encoding the ORF of dap12 was derived by 3′ RACE. Primer sequences are listed in Table 1. PCR products were cloned into pGEM-T easy (Promega) or pCR4-TOPO (Invitrogen) and sequenced.
A zebrafish kidney EST (GenBank BG799671), which encodes FcRγ (fcer1g), was identified from a tBLASTn search using a catfish FcRγ sequence (GenBank AF538721) as a query. In order to confirm this sequence, a 5′ RACE strategy was performed as described above using spleen RNA and the nested reverse primers, ZFFcRg-3′UTR-R1 and ZFFcRg-3′UTR-R2, which are complementary to the 3′ UTR of FcRγ. Primer sequences are listed in Table 1. PCR products were cloned into pGEM-T easy (Promega) or pCR4-TOPO (Invitrogen) and sequenced.
A zebrafish FcRγ-like (fcer1gl) EST sequence was identified (GenBank BC124700) from a tBLASTn search using catfish FcRγ-like sequence (GenBank AAN38001: Shen et al 2003) as a query. As this zebrafish EST encoded the entire open reading frame (ORF) of FcRγ-like, forward and reverse primers, ZFFcRg-B-F, and ZFFcRg-B-R, were designed to amplify the entire ORF from zebrafish spleen cDNA. Primer sequences are listed in Table 1. The FcRγ-like ORF was amplified from zebrafish spleen cDNA using 40 cycles and cloned into pCRII-TOPO (Invitrogen) and sequenced.
A zebrafish cd3ζ (cd247) EST sequence was identified (GenBank EE715400) from a tBLASTn search using a trout CD3ζ sequence (GenBank CA344798; Rexroad, III et al 2003) as a query. As this zebrafish EST encoded the entire open reading frame (ORF) of cd3ζ, forward and reverse primers, ZFNEWCd3z-F, and ZFNEWCd3z-R, were designed to amplify the entire ORF from zebrafish spleen cDNA. Primer sequences are listed in Table 1. The cd3ζ ORF was amplified from zebrafish lymphocyte cDNA using 40 cycles and annealing at 70° C and subsequently cloned into pCRII-TOPO (Invitrogen) and sequenced.
A predicted zebrafish cd3ζ-like (cd247l) sequence was identified (GenBank XM_688285) from a tBLASTn search using the catfish CD3ζ-like sequence (GenBank AAZ16505) as a query. The same RACE strategy as described above for dap10 and dap12 was used to clone the full-length open reading frame of cd3ζ-like from zebrafish spleen RNA. Initially, 3′ RACE was completed with nested forward primers (complementary to a conserved cytoplasmic immunoreceptor tyrosine-based activation motif [ITAM]) ZFCD3z-F1 and ZFCD3z-F2 which confirmed the 3′ untranslated region of the cd3ζ-like mRNA. Subsequently, nested, reverse primers ZFCD3z-3′UTR-R1 and ZFCD3z-3′UTR-R2 were designed within the 3′ UTR of cd3ζ-like and 5′ RACE was completed to generate a cDNA encoding the ORF of cd3ζ-like. Primer sequences are listed in Table 1. PCR products were cloned into pGEM-T easy (Promega) or pCR4-TOPO (Invitrogen) and sequenced.
Predicted leader and transmembrane domains of protein sequences were identified with SMART software (http://smart.embl-heidelberg.de/; Letunic et al 2004) and proteins were aligned by ClustalW (http://www.ebi.ac.uk/clustalw/#; Chenna et al 2003). Neighbor-joining trees (Saitou and Nei 1987) were constructed from pairwise Poisson correction distances with 2000 bootstrap replications by MEGA2.1 software (Kumar et al 2001). Sequences used for this alignment and analyses are listed in Table 2.
Zebrafish (AB strain) embryos were collected by natural mating and maintained at 28° C as described (Westerfield 2000). Tissues from adult zebrafish were dissected and myeloid and lymphoid cell lineages were purified from anterior kidney by cell sorting using forward and side scatter as described (Traver et al 2003b) and cDNAs were generated as described (Panagos et al 2006). One μl of cDNA was subjected to thermal cycling with gene-specific primers (see RT-PCR primers in Table 1) and Titanium Taq DNA polymerase (BD Biosciences) in a 20 μl PCR reaction and 10 μl were analyzed by agarose gel electrophoresis. The number of PCR cycles used for detecting dap10, dap12, FcRγ, FcRγ-like, Cd3ζ, Cd3ζ-like, and actin transcripts were 30 (annealing at 65° C), 40 (annealing at 65° C), 40 (annealing at 65° C), 30 (annealing at 70° C), 40 (annealing at 70° C), 40 (annealing at 65° C), and 25 (annealing at 65° C), respectively. Myeloperoxidase (mpx) transcripts were detected using 25 PCR cycles (annealing at 70° C) and TCRα transcripts were detected using 40 PCR cycles (annealing at 60° C). Forward and reverse primers span at least 1 intron and are listed in Table 1.
As a first step in characterizing immune signaling patterns in zebrafish, we data-mined the zebrafish genome and EST databases to identify sequences corresponding to six candidate adaptor proteins. Full-length cDNAs encoding these proteins termed, Dap12 (tyrobp), Dap10 (hcst), FcRγ (fcer1g), FcRγ-like (fcer1gl), Cd3ζ (cd247) and Cd3ζ-like (cd247l), were cloned by RACE or RT-PCR and sequenced. The predicted protein sequences for these zebrafish adaptors are shown aligned to orthologous proteins in Figure 1. All six zebrafish adaptors encode a peptide leader sequence, a transmembrane domain with a negatively charged amino acid (Asp) and a cytoplasmic tail with at least one activating motif.
Zebrafish Dap12 is 33% identical (46% similar) to human DAP12 and possesses a single cytoplasmic ITAM (Figure 1A). The zebrafish Dap10 sequence identified here differs from a previously described Dap10 EST (GenBank BQ261988) by a single residue, is 26% identical to human DAP10 (40% similar) and possesses the cytoplasmic tyrosine motif, YMNV, which is similar to the YINM motif encoded by mammalian DAP10 (Figure 1B and Guselnikov et al 2003b). Although this is not an exact match to the mammalian DAP10 signaling motif, this same peptide sequence is encoded by a catfish DAP10 sequence (GenBank DQ114899) and a related sequence, YMNT, is encoded by DAP10 in Takifugu (Guselnikov et al 2003b). It is likely relevant that the YMNV motif encoded by zebrafish and catfish DAP10 is highly similar to the CD28 YMNM motif that, (like DAP10's YINM motif), when tyrosine phosphorylated, binds phosphatidylinositol 3-kinase (PI3K: Cai et al 1995; Prasad et al 1994). The original experiments that led to the definition of the YxxM motif as necessary for PI3K (p85) binding is based on studies using mutated CD28 motifs. When the fourth residue of this motif was mutated from an apolar methionine to a polar cysteine CD28 lost the ability to bind PI3K (Cai et al 1995; Prasad et al 1994). It might be that if this motif is mutated from YxxM to YxxV, (which is a more conserved substitution and matches the zebrafish and catfish DAP10 motif), PI3K binding would be maintained. These observations suggest that the zebrafish DAP10 sequence reported here may signal via PI3K.
Zebrafish FcRγ and FcRγ-like are 44% and 40% identical (72% and 61% similar) to human FcRγ, respectively, and each possesses a single cytoplasmic ITAM (Figure 1C). Zebrafish Cd3ζ and Cd3ζ-like are 34% and 29% identical (57% and 52% similar) to human CD3ζ, respectively. Zebrafish Cd3ζ and Cd3ζ-like possess two cytoplasmic ITAMs and Cd3ζ possesses an additional ITAM-like sequence (Figures 1D and 1E; Underhill and Goodridge 2007).
A phylogenetic analysis was performed in order to further characterize the relationships between all six zebrafish adaptor proteins and their respective orthologs (Figure 2A). The DAP10 and DAP12 proteins form one large family and the CD3ζ and FcRγ proteins form a second family, supporting the hypotheses that DAP10 and DAP12 are evolutionarily derived from the same progenitor gene and that, likewise, CD3ζ and FcRγ are likely derived from a single gene (Guselnikov et al 2003a). The observation that DAP10 from fish and Xenopus group with mammalian DAP12, rather than DAP10 was unexpected (Figure 2A). A second phylogenetic analysis focused only on DAP10 and DAP12 shows the expected grouping for DAP10/DAP12 across all species (Figure 2B), suggesting that the fine-scale organization of Figure 2A is not reliable: this is likely a result of the gaps introduced into the DAP10/DAP12 sequences required to align them with the longer CD3ζ/FcRγ sequences. Of these six adaptors, Dap12, Dap10, FcRγ and Cd3ζ are likely orthologs of their mammalian counterparts, whereas, FcRγL and Cd3ζL are likely paralogs of FcRγ and Cd3ζ, having arisen via a gene duplication event. This interpretation is strongly supported by evidence for a whole-genome duplication event in the lineage leading to the teleosts (Amores et al 1998; Woods et al 2005). In contrast, no evidence is seen for a second copy of either DAP10 or DAP12 in zebrafish (see Section 3.2 for additional discussion).
Zebrafish Dap12 is most closely related to the mammalian DAP12 sequences and groups with these orthologs with ~53% confidence (Figure 2A). To our knowledge, the full-length DAP12 cDNA sequence reported here represents the first from any non-mammalian species although a similar (partial) sequence was predicted from the Takifugu genome that was shown to partner with human KIR2DS2 (Feng et al 2006; Guselnikov et al 2003b) in an in vitro assay. Although the DAP10 family is less conserved, all of the DAP10 proteins possess similar signaling motifs (Figure 1), supporting both this classification and related functions.
Zebrafish FcRγ and FcRγL group with all other FcRγ orthologs with 54% confidence (Figure 2A). A sequence comparison of fish and mammalian FcRγ and FcRγL sequences suggests that the carp cDNA, originally defined as FcRγ (Fujiki et al 2000), is less similar to FcRγ than other fish genes. A second carp FcRγ gene may exist (Figure 2A). Furthermore, this analysis suggests that of the two proposed catfish sequences designated FcRγ, an unpublished FcRγ sequence (GenBank AF538721), here identified as “FcRγ”, is more closely related to the mammalian sequences than to the published FcRγ sequence, here identified as “FcRγL” (Figures (Figures11 and and2A;2A; Shen et al 2003). The CD3ζ family of proteins from mammals, Xenopus, chicken, and multiple fish species cluster with 73% confidence, however, CD3ζL sequences are more divergent from the CD3ζ sequences and from each other (Figure 2A). The three CD3ζL sequences identified here from zebrafish, catfish and trout are not well conserved (Figures (Figures1E1E and and2A)2A) suggesting that less selective pressure has been encountered during the evolution of this gene. The FcRγL and CD3ζL genes may be restricted to bony fish.
Further support for the nomenclature for these genes is provided by BLASTp analyses of mammalian proteins using the peptide sequences of the zebrafish adaptors as queries. Although the similarity is not strong, the mammalian protein with the strongest similarity to zebrafish Dap10 is a primate DAP10 (Table 3) and much of this similarity is focused on the transmembrane domain and cytoplasmic signaling motif (Figure 1B). In contrast to DAP10, DAP12 is much more conserved between fish and mammals (Figures (Figures1A1A and and2),2), suggesting that there has been more selective pressure on DAP12. In addition, this analysis confirms that zebrafish Cd3ζ and Cd3ζL are most similar to mammalian CD3ζ and that zebrafish FcRγ and FcRγL are most similar to mammalian FcRγ. Zebrafish CD3ζ and FcRγ are more closely related to their mammalian orthologs than are CD3ζL and FcRγL (Table 3)
Zebrafish dap10 (hcst) and dap12 (tyrobp) are encoded by BAC DKEY-29H14 (GenBank BX571853) which maps to chromosome 16. Zebrafish FcRγ (fcer1g) is encoded by BAC CH211-286P15 (GenBank BX324220), which maps to chromosome 7 and FcRγL (fcer1gl) is encoded by a chromosome 2 contig (GenBank NW_001513019). Only a single exon of zebrafish cd3ζ (cd247) has been placed on a chromosome 1 contig (GenBank NW_001511947) whereas cd3ζL (cd247l) is encoded by BAC CH211-218D20 (GenBank BX294189), which maps to chromosome 9. The genomic organization of these genes is depicted in Figure 3. Zebrafish dap10 and dap12 are immediately adjacent genes as they are in mammals and Takifugu (Guselnikov et al 2003b).
The presence of two paralogs for a single mammalian gene (e.g. CD3ζ and FcRγ) is not uncommon in zebrafish as much evidence indicates a whole genome duplication event occurred in the lineage leading to the teleosts (Amores et al 1998; Woods et al 2005). The zebrafish paralogs of FcRγ are present on chromosomes 7 and 2 which have been shown to encode other duplicated genes (e.g. ap1g1 and ap1g2). Similarly, the zebrafish paralogs of Cd3ζ are present on chromosomes 1 and 9 which encode other duplicated genes (e.g. dlx2a and dlx2b) and are predicted to be the result of an ancient chromosomal duplication (Woods et al 2005). In contrast the characterization of zebrafish dap10 and dap12 as single copy genes is supported by the estimate that only ~25% of the duplicated genes are currently identifiable within the zebrafish genome (Woods et al 2005). The completion of its genome will conclusively determine if zebrafish possess paralogs for dap10 and dap12.
DAP10 and DAP12 are tightly associated genes in mammals and zebrafish. They are immediately adjacent genes and separated by ~200 and ~ 8,700 bp in human and zebrafish, respectively. In contrast, FcRγ and Cd3ζ, which are encoded by the same chromosome in human and mice, are present on two chromosomes in zebrafish. This difference in physical linkage is likely due to the larger intergenic distance between these genes in mammals (~6,200,000 bp in human) and the extensive number of interchromosomal rearrangements since the ancestral genome of bony vertebrates (Woods et al 2005).
Further evidence supporting the identity of these zebrafish adaptors is provided by the syntenic relationships between the human and zebrafish loci for these genes. Four zebrafish genes which flank each adaptor gene (two genes on each side) were identified and the orthologous genes from human were defined by BLASTp analyses (Table 4). For example, zebrafish FcRγ (fcer1g) is adjacent to zgc:112036 which shares high identity with human NDUFS2: both human FcRγ (FCER1G) and NDUFS2 map to human chromosome 1q23. There is evidence for conserved synteny in all six adaptor gene loci in zebrafish: one or more of the genes which flank each zebrafish adaptor gene maps to the same chromosomal region as the human adaptor gene (Table 4).
The expression of dap10, dap12, FcRγ, FcRγL, cd3ζ, and cd3ζL during zebrafish embryogenesis and in adult tissues (ovary, liver, kidney, spleen and intestine) (Figure 4A) was examined using an RT-PCR strategy. Maternal transcripts of all six adaptors can be detected in the ovary and dap10 transcripts continue to be detected in the 1-cell stage embryo (0 hours post fertilization, hpf). Zebrafish FcRγ is the only adaptor that is detected at later stages of embryogenesis (72 hpf) whereas FcRγ, cd3ζ, and cd3ζL transcripts are detected at the larval stage of 6 days post fertilization (dpf). This may reflect either very low levels of expression, expression in small sets of cells or simply a lack of expression at the early stages of development. Similar expression patterns of dap10, dap12, cd3ζL, FcRγ and FcRγL transcripts can be detected in adult liver, kidney, spleen and intestine, whereas Cd3ζ is primarily detected in ovary and spleen, with lower levels of expression in intestine. Analysis of adaptor expression in zebrafish lymphoid and myeloid cells was completed using flow cytometrically separated cell populations (Figure 4B) and reveals that dap10, dap12 and cd3ζ are restricted to lymphocytes whereas cd3ζL, FcRγ and FcRγL are expressed in both lymphoid and myeloid lineages (Figure 4C). Based on these distributions, it is likely that the adaptor homologs may function in immune signaling systems in zebrafish.
As DAP12, DAP10, CD3ζ and FcRγ are all expressed by mammalian NK cells (Hamerman and Lanier 2006; Wu et al 2000), these data raise the possibility that zebrafish lymphocytes (purified by forward and side scatter) include NK cells. Although clonal NK-like cell lines have been described from catfish (Shen et al 2004), there are currently no reagents for identifying or purifying pure NK cell populations from bony fish. Finally, the observation that zebrafish dap10 and dap12 transcripts are not detected in myeloid cells was unexpected as the mammalian orthologs of these genes are expressed in both lymphoid and myeloid cells (Takaki et al 2006; Wu et al 1999): it is possible that dap10 and dap12 are expressed at very low levels or in a small population of cells within the zebrafish myeloid lineage.
The observation that Takifugu DAP12 can associate with a human KIR in a cell culture system suggests a conserved function for Dap12 (Feng et al 2006); however, KIRs have not been identified in bony fish. Other candidate activating NK receptors, that possess positively charged residues within their transmembrane domains, have been identified from zebrafish including an immunoglobulin-type receptor, Nitr9, and a lectin-type receptor, Illr3 (Panagos et al 2006; Yoder et al 2004): it will be of interest to determine if these receptors physically partner with Dap12. The presence of paralogs for CD3ζ and FcRγ in zebrafish raises additional questions of functional specificity. Do zebrafish TCR and FcεRI utilize a single adaptor for signaling? If so, which one? And what role does the second adaptor paralog play in immune response? Knowledge of the structures of the zebrafish adaptors and the relationships of these structures to each other will be useful for the design of specific reagents and engineering of constructs for examining the specific interactions and functions of these adaptor molecules in the immunological and developmental context of the zebrafish model.
We thank Melissa Haendel and John Hansen for very helpful discussions about gene nomenclature and Barb Pryor for editorial assistance. Zebrafish genes were named in consultation with the ZFIN Nomenclature Committee. Sequence data from this article have been deposited with the GenBank and ZFIN databases under accession numbers EF158445 and ZDB-GENE-061130-1 (Dap10/hcst); EF158446 and ZDB-GENE-061130-2 (Dap12/tyrobp); EF158447 and ZDB-GENE-061130-3 (FcRγ/fcer1g); EF601085 and ZDB-GENE-070502-4 (FcRγ-like/fcer1gl); EF601086 and ZDB-GENE-061130-4 (Cd3ζ/cd247); and EF158448 and ZDB-GENE-070508-2 (Cd3ζ-like/cd247l). This research was supported by NSF grant MCB-0505585 (JAY), NIH grant R01 AI057559 (GWL), and by funding from the All Children's Hospital Foundation (GWL).
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.