A Novel Activating Chicken IgY FcR Is Related to Leukocyte Receptor Complex (LRC) Genes but Is Located on a Chromosomal Region Distinct from the LRC and FcR Gene Clusters1

FcRs have multifaceted roles in the immune system. Chicken FcRs were demonstrated on macrophages decades ago; however, only recently the chicken Ig-like receptor AB1, encoded in the leukocyte receptor complex, was molecularly identified as a high-affinity FcR. The present study was initiated to identify additional receptors with the capability to bind chicken immunoglobulins. Based on database searches, we cloned a novel chicken FcR, designated gallus gallus FcR (ggFcR), which was shown to bind selectively chicken IgY. The receptor consists of four extracellular C2-set Ig domains, followed by a transmembrane region containing arginine as a positively charged amino acid and a short cytoplasmic tail. ggFcR associates with the common γ-chain, indicative for an activating receptor, and real-time RT-PCR revealed high expression on PBMC, thrombocytes, and macrophages. The genomic organization is similar to most Ig-like receptor genes, where each Ig domain is encoded by a separate exon. Additionally, the ggFcR signal peptide is encoded by two exons, the second of which is 36 bp, a hallmark for genes encoded in the leukocyte receptor complex. Phylogenetic analysis also showed a relationship to genes encoded in the leukocyte receptor complex. Surprisingly, ggFcR is not encoded in the leukocyte receptor complex, but it is located as a single isolated gene at the extremity of chicken chromosome 20.

N umerous immunoregulatory cell surface receptors have been discovered in recent years. They form multigene families located on distinct chromosomes that comprise both inhibitory and activating family members (1). The inhibitory receptors display a long cytoplasmic tail with ITIM motifs, whereas activating receptors have only a short cytoplasmic tail, but feature a basic amino acid in the transmembrane region, which interacts with an ITAM-containing adaptor molecule, such as the common ␥-chain, DAP12, or CD3. The extracellular part of the receptor displays either a C-type lectin structure or a different number of Ig domains. Immunoregulatory families with Ig-like domains are broadly distributed on various immune cells both of the innate and adaptive immune systems. In most cases, the binding partners of the respective immunoregulatory Ig-like receptors are still unknown or were discovered only recently. One exception is the family of classical FcRs. They have long been shown to bind the constant region of different immunoglobulins (IgG, IgE, IgA) and are involved in various immune reactions, such as phagocytosis, Ab-dependent cellular cytotoxicity, and immediate hypersensitivity (2). FcRs differ by their ligand preferences, binding affinity, and signal properties: Fc␥RI, the high-affinity receptor for IgG; FcRI, the high-affinity receptor for IgE; and Fc␥RIIIa, a low-affinity receptor for IgG, are activating receptors that signal via the common ␥-chain, whereas Fc␥RII and its isoform are lowaffinity receptors for IgG, which have either ITAMs (Fc␥RIIa and Fc␥RIIc) or ITIMs (e.g., Fc␥RIIb) in their cytoplasmic regions (3). In man, most FcR genes are located on chromosome 1 (2). One exception to this is the human Fc␣R, which is encoded in the leukocyte receptor complex (LRC) 3 on chromosome 19 (4). Another family of immunoregulatory Ig-like receptors, which is closely related to the FcRs, is also located on human chromosome 1. The so-called FcR-like (FCRL) genes are also expressed on a variety of different leukocytes, but, in contrast to the classical FcRs, they have not been shown to bind Ig (5). The functional importance of these proteins is still an unresolved issue.
Recently, an FCRL homolog that is highly related to mammalian FcR and FCRL genes was also identified in the chicken (6,7). To date, this gene is the only FcR/FCRL gene present on chicken chromosome 25, which represents the syntenic region to the mammalian FcR locus. This receptor has not been shown to bind IgY and is therefore proposed to be an FCRL homolog.
Another large immunoregulatory Ig-like receptor family is formed by the chicken Ig-like receptors (CHIR). These genes are located in the LRC on chicken chromosome 31, and thus they are likely homologous to mammalian leukocyte Ig-like receptors (LILR), killer cell Ig-like receptors (KIR), or other LRC-encoded genes (8,9). However, in contrast to their mammalian counterparts, CHIR display Ͼ60 functional receptor genes with extensive haplotypic and allelic variation, including also potentially bifunctional receptors. Recently, we showed that one member, CHIR-AB1, functions as a high-affinity IgY FcR (10). CHIR-AB1 combines features of several mammalian FcRs, such as expression on macrophages, NK cells, and B cells, as well as its potential to signal as inhibitory or activating receptors, based on the presence of a charged transmembrane residue and a cytoplasmic ITIM.
This study was initiated to identify additional FcR-related genes in the chicken based on the analysis of various chicken databases. As shown below, we identified a novel FcR gene in the chicken, designated gallus gallus FcR (ggFcR) (accession no. FM200428), which selectively binds chicken IgY. We demonstrate that it is a potentially activating receptor, which interacts with the common ␥-chain and is mainly expressed on chicken PBMC. Surprisingly, ggFcR is highly related to mammalian and chicken LRC-encoded genes, but it is located on chromosome 20, a location different from the chicken FcR and LRC regions.

Database searches and in silico sequence analysis
Various chicken expressed sequence tag (EST) databases were used to identify FcR genes by key word search. In particular we used chicken EST databases provided by the Biotechnology and Biological Sciences Research Council (www.chick.umist.ac.uk/), by the Delaware Biotechnology Institute (www.chickest.udel.edu/), and by the Bursal Transcript Database (pheasant.gsf.de/DEPARTMENT/DT40/dt40Transcript.html). Additional ESTs were identified by employing the basic local alignment search tool (BLAST) program limited to "gallus gallus" EST databases with a total of 600,000 ESTs deposited and used to obtain a full-length open reading frame of ggFcR. All resulting ESTs were further analyzed using the Lasergene software package (GATC Biotech), which was manually refined and used for primer design of ggFcR. Deduced amino acid sequence was analyzed for structural elements such as signal peptide, Ig domains, transmembrane regions, and secondary structure elements using SMART (smart.embl-heidelberg.de/) (11,12), InterProScan (www.ebi.ac.uk/ InterProScan/) (13), SignalP 3.0 (www.cbs.dtu.dk/services/SignalP/) (14), THMM 2.0 (www.cbs.dtu.dk/services/TMHMM-2.0/), and JPred3 (www. compbio.dundee.ac.uk/ϳwww-jpred/) (15). Corresponding genomic DNA was identified using the cloned ggFcR in a BLAST-like alignment tool search of the chicken genome (genome.ucsc.edu/cgi-bin/hgGateway). For phylogenetic analysis, a neighbor-joining tree was constructed using MEGA4 (16). The HomoloGene database (www.ncbi.nlm.nih.gov/sites/ entrez?dbϭhomologene) was used for further analysis of the chromosomal region.

Radiation hybrid (RH) mapping
The ChickRH6 panel (17) was used to map the ggFcR and the lethal (3) lethal(3)malignant brain tumor-like protein (L3MBTL) gene. Primer sequences and fragment sizes are given in Table I. Chicken and hamster genomic DNAs were used as positive controls whereas TE buffer was used as a negative control. Twenty-five nanograms of each panel DNA was amplified in an 8-l PCR reaction containing 1ϫ PCR Master Mix (ABgene) and 4 M of each of the two primers. PCR reaction was performed in duplicates. PCR conditions were 3 min at 95°C followed by 35 cycles of 30 s at 95°C, 45 s at 55°C, and 60 s at 72°C, followed by a final elongation step at 72°C. Markers were mapped using the CarthaGène program (www.inra.fr/bia/T/CarthaGene/).

Animals and cell preparation
Chickens (line M11, MHC haplotype B2/B2, a kind gift from S. Weigend, Mariensee, Germany) were hatched at the institute and used for experiments at the age of 6 -10 wk.
PBL were prepared by slow-speed centrifugation as described (18), and PBMC were prepared by density centrifugation on Ficoll-Paque (GE Healthcare). PBMC were further used for macrophage preparation by adhering them for at least 48 h on petri dishes (19). Chicken thrombocytes were prepared by short-term cultures of PBMC on petri dishes for 4 h (20). For spleen, liver, and brain RNA preparation, 100 mg of each tissue was taken and directly frozen in liquid nitrogen.

Cloning procedures
Total RNA from PBMC was prepared using the Absolutely RNA RT-PCR miniprep kit (Stratagene), and cDNA synthesis was performed with the ThermoScript RT-PCR system (Invitrogen). Herculase enhanced DNA polymerase (Stratagene) was used for PCR at 2 min of denaturation at 95°C, 35 cycles of 10 s at 95°C, 30 s at 55°C, 2 min at 72°C, and a final extension time of 10 min at 72°C. For cloning of ggFcR, oligonucleotides 922s and 923as (Table II and Fig. 1A) were used. The resulting PCR product was TA cloned into a pCRII TOPO vector (Invitrogen), colonies were screened by PCR and restriction analysis, and plasmids from positive colonies were isolated using the NucleoSpin plasmid kit (Macherey-Nagel) and sequenced (GATC). The full-length ggFcR-FLAG construct was generated using oligonucleotides 1038s and 1229as with EcoRI sites (Table II) as described previously (21). The ggFcR-FLAG-muCD3 construct was generated using oligonucleotides 1038s and 931as (Table II) on ggFcR cDNA as described earlier (10), resulting in an N-terminally FLAG-tagged extracellular region of ggFcR fused to the transmembrane region of chicken CD8␣ and the cytoplasmic domain of murine CD3.

Cell lines, transfections, and cell stainings
Human embryonic kidney 293T cells (22) and the mouse thymoma cell line BWZ.36 (23) were maintained in RPMI 1640 (Biochrom) supplemented with 10% FCS and 1% penicillin/streptomycin in a CO 2 incubator at 37°C. 293T cells were double transfected with the full-length ggFcR-FLAG construct and either the common ␥-chain-V5 construct (FcRI␥) (8) or a mock-V5 control using the Metafectene reagent (Biontex Laboratories) according to the manufacturer's protocol. After 24 h of transient transfection, cells were used for staining with a mouse anti-FLAG mAb (Sigma-Aldrich) followed by an anti-mouse Ig-FITC conjugate (SouthernBiotech), and analyzed with a FACScan (BD Biosciences) using the CellQuest software. For stably expressing reporter cells, 3 ϫ 10 6 BWZ.36 cells were electroporated with 25 g of ggFcR-FLAG-muCD3 construct at 200 V with 950 F capacitance. Cells were selected with 800 g/ml g418  (AppliChem), and single clones were screened for expression by cell staining as described above.

Reporter gene assay
Stably expressing BWZ.36-ggFcR cells were used in reporter gene assays as follows: 3 ϫ 10 5 reporter cells were cultivated for 24 h in 24-well cell culture plates either left uncoated (control) or coated with anti-FLAG mAb (Sigma-Aldrich) or chicken IgA, IgM, and IgY (all 10 g/ml) or stimulated with 10 ng/ml PMA (Sigma-Aldrich) and 0.5 g/ml Ca ionophore (Sigma-Aldrich) as an unspecific positive control. Purified chicken IgM and IgA was kindly provided by B. Kaspers (Munich, Germany). Chicken IgY was purchased from Jackson ImmunoResearch Laboratories. ␤-galactosidase was measured using the high sensitivity ␤-galactosidase assay kit (Stratagene) after an incubation period of 18 h in a 37°C incubator. The activity of ␤-galactosidase (U) was calculated by dividing the amount of chlorophenol red formed (nmol) by a specific time length of incubation (minutes). The amount of chlorophenol red (in nmol) equals the concentration of chlorophenol red formed (nmol/ml) ϫ total assay volume (ml). The concentration of chlorophenol red formed (nmol/ml) is calculated by the OD ϫ 55.

Real-time RT-PCR
Total RNA was extracted from 100 mg of tissue or 1 ϫ 10 7 cells by using TRIzol (Invitrogen). The RNA quality was determined with the 2100 Bioanalyzer (Agilent Technologies). RNA with an integrity number Ͼ8 was then treated with DNase I (Roche), and cDNA synthesis was performed with the SuperScript III first-strand synthesis system (Invitrogen). The 7300 real-time PCR system from Applied Biosystems was used for PCR with the Power SYBR Green RT-PCR reagents kit (Applied Biosystems) using following parameters: 95°C for 10 min, then 40 cycles of 95°C for 15 s and 59°C for 1 min, followed by a melting curve analysis. The cDNA samples were analyzed in triplicates, with oligonucleotides specific for ggFcR (1165s and 1166as, see Table II) and 18S RNA (870s and 871as, Table II) obtaining the cycle thresholds (C t ) for each tissue. The relative amounts of gene-of-interest mRNA were calculated by means of the ⌬⌬C t method as described previously (10).

Cloning of ggFcR, a putative FcR, which displays genomic organization similar to LRC-encoded genes
Initially, various chicken EST databases were searched by using the term "Fc Receptor", and this yielded seven EST clones of the Biotechnology and Biological Sciences Research Council EST database. Analyzing those resulted in two different contigs containing four and three EST clones, respectively. Four EST clones (BU221313, BU232940, BU332675, BU459232) were identified to represent the previously described ggFcR/L gene (6), whereas the other three clones (BU239980, BU422666, BU424877) represented a yet undefined protein.
To obtain a full-length protein, additional EST clones were identified by BLAST search (AJ394599, BU372165, BU450669, BU485512, BX273743, BX273744, CF250847, CV891153, CK607089), and a manually refined consensus sequence of all EST clones was used to design primers for ggFcR. PCR with oligonucleotides 922s and 923as amplified a 1525-bp transcript with an open reading frame of 1428 bp. The deduced protein sequence consists of a signal peptide, four C2-set Ig domains, and a transmembrane region with arginine as a positively charged amino acid followed by a short cytoplasmic tail (Fig. 1A). Note that there are four possible ATG initiation codons encoding methionine residues at positions Ϫ56, Ϫ24, Ϫ23, and Ϫ17 (Fig. 1A). After applying different signal peptide prediction programs, methionine Ϫ24 seems to be most likely the initiation site of ggFcR. The corresponding genomic DNA was identified on chicken chromosome 20, bp 22,577-25,509. The coding sequence of the genomic DNA differed only in two nucleotides from the cloned ggFcR, one of them resulting in a nonsynonymous change (C82 3 R82) in the g strand of the Ig1 domain. The exon/intron organization was obtained by comparison of the cloned ggFcR with the genomic sequence. All exon/intron boundary sequences conformed to the gt-ag rule. The signal peptide was split on two exons ( Fig. 1; exons 1 and 2), the second of which is 36 bp in length. This is typical for all Ig-like receptors in the leukocyte receptor complex, both from chicken and mammals (ggCHIR (8), huKIR (www.ncbi.nlm.nih.gov/books/bookres.fcgi/mono_003/ch1d1.pdf), huLILR (24), and huFcAR (25)), whereas most mammalian FcGR and FCRL genes (5) and the previously identified ggFcR/L gene (6) display 21 bp in the second exon of the signal peptide. Each Ig domain was encoded by separate exons (Fig. 1; exons 3-6), and the stalk, transmembrane, and cytoplasmic regions were encoded by a single exon ( Fig. 1; exon 7). All exons for signal peptides and Ig domains were in phase 1, which is characteristic for Ig-like receptor genes (Fig. 1A). The size of the introns is representative for chicken genes, being only 25% of the size of mammalian genes.
GgFcR is related to LRC-encoded genes, but it is located as a unique gene on chicken chromosome 20 In a next step a phylogenetic tree was constructed to analyze the relationship of ggFcR to various chicken and human LRC-and FcRrelated genes. In this analysis we included the GenBank sequences of human Fc␣R (NM_002000), human Fc␥RIA (NM_000566), the previously characterized ggFcR/L (AM412311), which seems not to bind IgY (6), and ggCHIR-AB1 (AJ745094), a member of the CHIR gene family in the chicken LRC, which was shown to bind IgY with high affinity (10). Since all of them display a different number of C2-set Ig domains, the Ig domains were split according to SMART Ig domain prediction and were used separately in a neighbor-joining tree. The membrane distal Ig domain was termed IG1, followed by IG2, IG3, and IG4. The four Ig domains of ggFcR cluster with the chicken and human LRC-encoded genes huFcAR and ggCHIR-AB1 (Fig. 2, upper branch), whereas the previously identified ggFcR/L is more related to human FcR genes (Fig. 2, lower branch), which was also shown by Taylor et al. (6). The latter is encoded on chicken chromosome 25, which is syntenic to the region on the human chromosome 1 encoding the FcR and FCRL gene families. The human Fc␣R gene and the chicken CHIR family, on the other hand, are encoded in the LRC on human FIGURE 2. Phylogenetic comparison of ggFcR with chicken and human genes encoded in the LRC and FcR region, respectively. Individual Ig domains were aligned by the ClustalW algorithm. The neighbor-joining tree with 1000 bootstrap replicates and pairwise gap deletions was built using MEGA4. An unweighted pair group method with arithmetic mean (UPGMA) and a minimum evolution tree was also constructed, but they were essentially the same as the neighbor-joining tree in the major branching patterns and are not presented here. Accession numbers of genes included in the tree are FM200428 (ggFcR), NM_002000 (huFcAR), NM_000566 (huFcGRIA), AM412311 (ggFcR/L), and AJ745094 (ggCHIR-AB1). chromosome 19 and chicken chromosome 31, respectively. The highly related ggFcR gene, however, is located at the very beginning of chicken chromosome 20 (bp 22,577-25,509), which has by now not been characterized to contain any Ig-like receptor family members. The syntenic region to chicken chromosome 20 is human chromosome 20 (26), so we decided to examine these two chromosomal regions for the presence of Ig-like receptor genes by using the HomoloGene database. The result is shown in Fig. 3. Apart from the ggFcR gene, all genes on the first 1 Mb of chicken chromosome 20 (Fig. 3, FCR to RBM12, boxes 1-8, left panel) are also present on human chromosome 20, but with a different order and orientation (Fig. 3, line-connected boxes 1-6, right panel). Since the ggFcR gene is located adjacent to the 3Ј end of the L3MBTL gene, we were particularly interested in the human genes next to the 3Ј end of this gene. This 5-Mb region on human chromosome 20 (Fig. 3, SGK2 to KCNK15, boxes 7-9, right panel) does not contain any Ig-like receptor genes, but it displays genes present on chicken chromosome 20 (2.4 Mb-5.5 Mb) (Fig. 3, boxes 9 -11, left panel). To confirm the unusual chromosomal location of an LRCrelated FcR gene, we also performed RH mapping using genomic amplicons of the ggFcR gene and the adjacent ggL3MBTL gene, which mapped both on chicken chromosome 20.

GgFcR associates with the common ␥-chain and binds chicken IgY
The positively charged arginine in the beginning of the transmembrane region of ggFcR (Fig. 1A) suggested that this is a potentially activating receptor, which interacts with the ITAM-containing common ␥-chain (FcRI␥). This is similar to the situation in CHIR-A2 (8) and CHIR-AB1 (10), which both contain arginine at the beginning of the transmembrane region and were shown to interact with the common ␥-chain. To test common ␥-chain association, we used a FLAG-tagged full-length construct of ggFcR for cotransfection of 293T cells with either a common ␥-chain-V5 construct or a mock-V5 control. Staining of the transiently transfected cells with anti-V5 mAb showed equal expression of the V5 constructs (data not shown), whereas anti-FLAG staining revealed that ggFcR-FLAG surface expression is only reconstituted if cotransfected with the common ␥-chain (Fig. 4, right panel) but not with mock-V5 control (Fig. 4, left panel).
Since the Ig binding capabilities are an important classification parameter, which distinguishes FcR from FCRL genes, we established a reporter gene assay for ggFcR to address this question. We produced BWZ.36 reporter cells, which are stably expressing a  ggFcR-FLAG-muCD3 construct. Plate-bound chicken IgY, IgA, and IgM were tested as potential ligands for ggFcR reporter cells, and a specific ␤-galactosidase induction was observed only for plate-bound IgY (Fig. 5). The calculated activity of ␤-galactosidase (U) for IgY was 0.01839, which is comparable to the activity of the two positive controls, anti-FLAG (U ϭ 0.02018) and PMA/ Ca-Ionophore (U ϭ 0.02166). In comparison, chicken IgY induced a much higher ␤-galactosidase expression in CHIR-AB1 reporter cells (10), with U ϭ 0.56595, which is more than 30-fold higher than IgY-induced ␤-galactosidase expression in ggFcR reporter cells.
We also used a soluble ggFcR construct in a sandwich ELISA to determine the interaction of ggFcR with the Fc fragment of IgY, as shown for CHIR-AB1, but we could not detect any binding in this assay (data not shown). This is most likely due to the fact that the interaction between ggFcR and IgY is much weaker compared with IgY-CHIR-AB1 and cannot be detected by ELISA.

Expression analysis of ggFcR by real-time RT-PCR
Several immune-relevant tissues and cells were tested for expression of ggFcR by real-time RT-PCR. Oligonucleotides 1165s and 1166as are intron-spanning primers, which amplify 190 bp of the Ig3 and Ig4 domain of ggFcR. The normalized results showed the highest level of ggFcR expression in PBMC (Fig. 6A). Furthermore, in spleen and macrophage preparations we also observed ggFcR expression, whereas in brain and PBL the expression was negligible (Fig. 6A). Chicken PBL were prepared by slow speed centrifugation and comprise a relatively pure fraction of chicken lymphocytes. PBMC, however, were prepared by density centrifugation with Ficoll-Paque and contain, besides lymphocytes, also monocytes and thrombocytes. To further evaluate the high level of ggFcR expression in PBMC, which is not monocyte/macrophage derived (Fig. 6A), we prepared chicken thrombocytes and compared the ggFcR expression of PBMC, PBL, and thrombocytes. The normalized results were calibrated on PBL, and it appears that chicken thrombocytes are the major source of ggFcR-expressing cells in PBMC (Fig. 6B).

Discussion
FcRs represent a typical immunoregulatory Ig-like receptor family, which signal via activating and inhibitory receptors to set thresholds for cell activation, generating a well-balanced immune response. These FcR genes are encoded on a specific region on human chromosome 1 intermingled with the highly related FCRL genes. The pivotal difference between FcR and FCRL receptors is their capability to bind the Ig constant region. Recently, two groups showed that the region on chicken chromosome 25 syntenic to the human FcR locus contains only a single FCR/L gene present, which does not bind chicken immunoglobulins and therefore is considered to be an FCRL homolog (6,7). The present report describes the first cloning and characterization of a novel FcR in chicken, designated ggFcR. Comparison of the newly cloned ggFcR with the recently identified ggFCR/L showed only 21% overall identity on amino acid level. A crucial step to distinguish between an FcR or FCRL homolog is the identification of the ligand-binding properties of ggFcR. This was done using a reporter gene assay on the basis of BWZ.36 cells, which we already applied on the IgY FcR CHIR-AB1 (10). The results clearly show that ggFcR selectively binds chicken IgY, inducing a specific ␤-galactosidase expression in the reporter cells.
The capability of binding Ig constant regions, however, is not restricted to genes encoded in the FcR region on human chromosome 1, but is also present in genes encoded in the LRC. The Fc␣RI gene (CD89) in primates, horses, cows, and rats and the bovine Fc␥2R are two examples in mammals (27,28). Additionally, we have previously shown that a CHIR family member, designated CHIR-AB1, which is also encoded in the LRC, binds IgY with high affinity. Here, we show that ggFcR is an additional IgY FcR present in chicken, which is highly related to genes encoded in the LRC and not to genes in the FcR region. This relationship is validated by various findings. The first refers to the exon/intron organization of ggFcR (Fig. 1B). One aspect of the genomic organization, which distinguishes the FcR family on human chromosome 1 from the LRC-encoded genes on human chromosome 19, is a short mini-exon that encodes the second half of a split signal peptide. For all FcR and FCRL family members except of FCRLA, this mini-exon is 21 bp in size (5). This size also applies to ggFCR/L on the FcR sytenic region on chicken chromosome 25 (6,7). On the other hand, all genes encoded in the LRC, including the human Fc␣R, display also a mini-exon for the split signal peptide, but in contrast to the FcR/FCRL genes, this is 36 bp in size (www.ncbi.nlm.nih.gov/ books/bookres.fcgi/mono_003/ch1d1.pdf and Refs. 24,25). This 36bp mini-exon is also present in all genes encoded in the chicken LRC on chromosome 31 (8). Consequently, the presence of a  36-bp mini-exon in the ggFcR gene clearly indicates a relationship to LRC-encoded genes (Fig. 1B). Additionally, this relationship was also shown by a phylogenetic analysis of the Ig domains of various FcR-and LRC-encoded genes. The analyses demonstrated that the Ig domains of ggFcR are highly related to LRC, but not FcR-encoded genes (Fig. 2). Although all Ig domains of ggFcR clearly cluster with the LRC-encoded genes, the amino acid sequence identities between ggFcR and various CHIR genes range only between 20% and 40% (data not shown), indicating that gg-FcR is distinct from the CHIR gene family, whose members are usually related 65-99%. The third fact, which indicates that ggFcR is mostly related to LRC genes but is distinct from the CHIR family, is the organization of the transmembrane region displaying the amino acids NIVR (Fig. 1A), which interact with the common ␥-chain. As described by Guselnikov et al., the presence of an NxxR motif in the transmembrane region of activating family members is specific for most LRC-encoded genes, but it is missing in FcR genes (29). This NxxR motif appears to be an ancestral element of a primordial FcR/LRC family, which has been lost in CHIR, classical FcR and KIR, but is present in Xenopus FcR-like (XFL)-and LRC-encoded genes, like huLILR, huFc␣R, huGPVI, and muPIR and in the newly cloned ggFcR.
Surprisingly, the ggFcR gene is not encoded in the chicken LRC on chromosome 31, but it is located as a single gene on a unique position on chromosome 20. As depicted in Fig. 3, chicken chromosome 20 represents the syntenic region of human chromosome 20, displaying similar genes in a different order. Detailed analysis of the chromosomal region presented in Ensembl even shows that the ggFcR gene is located on the same contig as the neighboring L3MBTL gene (data not shown). To completely exclude a misassembly of the chicken genome in that area, we also proofed by RH mapping that both the L3MBTL gene and the ggFcR gene are located on chicken chromosome 20. Interestingly, a survey of various genomes from mammals, fish, and amphibians via Ensembl revealed no Ig-like receptor at the 3Ј end of the respective L3MBTL gene (data not shown). In all genomes analyzed, the SGK2 gene was located at this position, which is also present at this location in the human genome (Fig. 3).
The ggFcR gene was already annotated within the Ensembl and HomoloGene databases as a novel chicken gene located on chromosome 20. Due to prediction programs based on closest hits of this gene to mammalian databases, it was predicted to be similar to Ig superfamily member 1 (gene ID 419114) and to venom myotoxin inhibitor DM64 (XM_417301.2). By this report we demonstrated that this gene is predicted wrongly and is a true IgY FcR. Interestingly, the closest hit of a tblastn search to a nonpredicted chicken gene was CHIR-AB3 (AJ879909), clearly indicating again a close relationship to LRC-encoded genes. The origin of the gg-FcR gene on chromosome 20 remains elusive, but the relationship to LRC-encoded genes and the presence of the NxxR motif in the transmembrane region indicate that it is a dispersed gene from an ancient LRC locus.
Further characterization of the expression pattern of the activating ggFcR by real-time RT-PCR showed typical expression on chicken PBMC, macrophages, and spleen cells. Interestingly, gg-FcR is also expressed on thrombocytes, which represent nucleated blood cells phylogenetically closely related to platelets. In various, partly contradictory reports, chicken thrombocytes have long been suspected to have phagocytic activities (20,30,31) and there are also reports from human platelets with phagocytic capabilities. Recently, it was shown that platelets do engulf bacteria, but do not kill them, because a true phagosome is not generated (32). Interestingly, human platelets also express the activating Fc␥RIIA. The function of this receptor on platelets is still not resolved completely. Most of the work was done in transgenic mouse models, since mice lack the genetic equivalent of human Fc␥RIIA. These studies revealed that Fc␥RIIA plays a significant role in the immune clearance of platelets in vivo (33) and that Abs that activate platelets in an Fc␥RIIA-dependent manner lead to thrombosis, shock, and death (34).
In conclusion, ggFcR represents a potentially activating IgY FcR, which is highly related to LRC encoded genes but is located on chicken chromosome 20. It is expressed on chicken PBMC, macrophages, spleen cells, and thrombocytes, but its actual function on these cells still needs to be resolved.