The MHC, the most polymorphic and gene dense region in the vertebrate genome, contains many loci essential to immunity. In mammals, this region spans ∼4 Mb. Studies of avian species have found the MHC to be greatly reduced in size and gene content with an overall locus organization differing from that of mammals. The chicken MHC has been mapped to two distinct regions (MHC-B and -Y) of a single chromosome. MHC-B haplotypes possess tightly linked genes encoding the classical MHC molecules and few other disease resistance genes. Furthermore, chicken haplotypes possess a dominantly expressed class I and class II B locus that have a significant effect on the progression or regression of pathogenic disease. In this study, we present the MHC-B region of the turkey (Meleagris gallopavo) as a similarly constricted locus, with 34 genes identified within a 0.2-Mb region in near-perfect synteny with that of the chicken MHC-B. Notable differences between the two species are three BG and class II B loci in the turkey compared with one BG and two class II B loci in the chicken MHC-B. The relative size and high level of similarity of the turkey MHC in relation to that of the chicken suggest that similar associations with disease susceptibility and resistance may also be found in turkey.
The MHC is a genomic locus found in all jawed vertebrates (Gnathostomes) and is a key component in immune response. Originally identified through tissue graft rejection experiments, the MHC locus has subsequently been found to contain several classes of genes responsible for Ag presentation to the host immune system. Specifically, classical MHC molecules encoded within the MHC possess a highly polymorphic peptide-binding groove to bind peptide Ags through hydrophobic and/or hydrogen bonding and present them to T cell receptors. The MHC class I A genes are expressed in all nucleated cells, interacting with β2-microglobin to present mostly endogenously generated peptides of nine amino acids. MHC class II molecules are heterodimers (α and β genes) primarily expressed on APCs (dendritic cells, B cells, and macrophages) and present mostly exogenously derived peptides of ∼9–11 aa.
The chicken MHC has been defined as two genetically unlinked clusters, the MHC-B and -Y loci, located with the nucleolar organizer region on the same microchromosome (GGA16) (1, 2, 3, 4, 5). The Y locus contains lectin-like and nonclassical MHC genes with varied effects on disease susceptibility (6, 7, 8, 9). At least one class I-like locus is polymorphic and transcribed (10). The chicken MHC-B is subdivided into two regions, the BG and the BL-BF. The BL-BF region contains the classical class I and class II B genes. In contrast to mammalian genomes, which contain on average six paralogous genes for class I, class II A, and class II B genes (11), the chicken BL-BF locus lacks many immune genes present in mammal MHCs (compliments, cytokines, and so on) and contains only two MHC class I and two class II B genes within 50 kb (12, 13). Chicken MHC-B haplotypes predominantly express a single class I and class II B transcript, thereby reducing the diversity of Ags presented (14, 15) and have been described as a “minimal essential MHC.” A single monomorphic class II A gene located 5 cM from the BF-BL region encodes a protein that will dimerize with either class II B product to form the class II molecule (16). Interestingly, unlike mammals, the chicken has two C-type lectin-like genes, one of which is quite similar to NK complex loci (17). The limited repertoire of MHC molecules in the chicken—and the Ags they are able to present—has a remarkable effect on the species’ ability to resist/resolve infectious disease including bacteria, viruses, and parasites (18, 19, 20).
Studies of the closely related quail have identified an expanded set of MHC genes occupying the same genome locus (21). Similar to the chicken, expression was unequal between quail MHC class I and II loci within haplotypes (22). Studies in non-Galliform avian species have identified greater numbers of class I and class II B alleles within individuals compared those seen in the turkey and chicken, suggesting the presence of additional loci (23, 24, 25, 26).
Recent work (27) in the turkey has identified two MHC regions homologous to the chicken B and Y loci. Bacterial artificial chromosome (BAC)4 clones containing portions of these regions were physically mapped to turkey metaphase chromosomes through fluorescent in situ hybridization and genetically mapped by segregation analysis using a resource population. Like the chicken, these two regions were genetically unlinked and located on the same nucleolar organizer region-containing microchromosome (27).
Separated by an estimated 50 million years (28), the genomes of the turkey and chicken have been shown to be highly homologous; chromosomal markers are generally present in both species in syntenic order (29). Gene sequence studies have found most coding and predicted amino acid sequences to be >90% identical (30, 31). However, little is known of the similarities between the two species at the most variable region of the vertebrate genome, the MHC. This work was undertaken to describe the core MHC sequence of the turkey and to compare this most variable genome region to homologous sequences derived from other avian species. The resources available for the chicken (whole-genome sequence, multiple MHC haplotype sequences, and close phylogenetic relationship to turkeys) provide excellent tools for comparative analysis. Results of this study provide genomic resources for the study of the effect of the turkey MHC in disease susceptibility and resistance and present insights into the evolutionary origins of the unique structure of the avian MHC.
Materials and Methods
The CHORI-260 turkey BAC library was generated with DNA from a female (Nici) of a partially inbred Nicholas commercial subline (32). A clone from the library (97E05) containing a portion of the MHC-B region was isolated previously (27). Screening for additional BAC clones was performed as previously described (33) using overlapping oligonucleotide probes based on the end sequences of clone 97E05 (GenBank accession nos. DX922434-5), as well as a PCR product corresponding to the CD1.1 gene (GenBank accession no. EU522671) of the turkey (29). Additional BAC clones were identified and end sequencing of these clones (GenBank accession nos. ET222701-4) anchors them within the 97E05 clone with additional sequence extending further into the 5′ BG region (positions are based on the most recent sequence map of the chicken MHC (34), with “5′” and “upstream” referring to the “BG2-3” region and “3′” and “downstream” referring to “CD1A1-2” region). No clones corresponding to the 3′ MHC-B region were identified.
The BAC clone 97E05 was purified using the Qiagen Large Construct kit (Qiagen), randomly sheared and shotgun subcloned. Plasmid DNA was isolated and sequenced by automated Sanger sequencing at the University of Washington High Throughput Genomics Unit. Sequences were manually edited, aligned, and assembled using Sequencher software (Gene Codes). Southern hybridization confirmed assembly.
Subcloning and sequencing resulted in 3084 groomed reads with a median length of 525 bp. Primer walking on specific subclones assisted to fill minor gaps and/or obtain higher quality sequence in the assembly. Removal of pTARBAC2.1 vector sequence left a 172,697-bp insert of ∼8× coverage (Fig. 1⇓A). As previously reported, this insert terminated in the sixth (last) intron in the tripartite motif (TRIM) 7.2 gene and the first intron of the TAP1 gene (27).
Chicken genome sequence (GenBank accession no. AB268588) from the MHC-B region (34) was used to develop primers for the amplification and sequencing of the 3′ portion of the B locus (sequence not included in the turkey BAC clone insert (TAP1-CenpA). The nascent turkey sequences were aligned with the chicken sequence for further primer design and sequencing (supplemental Table I).5
Gene identification and annotation
Sequences were analyzed with the basic local alignment search tool and Softberry FGENESH (〈http://linux1.softberry.com/all.htm〉) to identify putative transcripts and homologies to known genes. Signal elements were identified with SignalP 3.0 (35). Comparisons between predicted gene sequences, available expressed sequence tags (EST) from poultry species, and the published chicken MHC sequence (12, 34) were performed using Sequencher software (Gene Codes). Repetitive elements were identified using REPEATMASKER and Tandem Repeats Finder (36) (〈http://tandem.bu.edu/trf/trf.basic.submit.html〉), and tRNAs elements were identified using tRNAScan (37). CpG islands were elicited with Softberry CpGfinder (〈http://linux1.softberry.com/all.htm〉), and GC content analysis was performed with 100-bp windows using Isochore (〈www.ebi.ac.uk/Tools/emboss/cpgplot/index.html〉). Identity dot matrix was drawn using PipMaker (〈http://bio.cse.psu.edu/pipmaker〉). Synonymous and nonsynonymous substitutions were identified based on the methods of Nie and Gojobori (38) using SNAP software (39) (〈www.hiv.lanl.gov/content/sequence/SNAP〉).
RT-PCR and cloning
E. coli cells. Plasmid clones were sequenced using vector-specific primers.
Sequencing and assembly of the B locus
A single BAC clone was shotgun sequenced and assembled to derive a majority of the homologous turkey MHC-B region (TRIM7.2 to TAP1). The remaining portion of the turkey core B locus sequence was generated by PCR using genomic DNA (Nici, the DNA source for the BAC library and current whole-genome sequencing (K. M. Reed, personal communication)) as template. This sequence spanned the region between TAP1 and CenpA and overlaps the end of the BAC clone for a combined total of 197 kb of contiguous turkey sequence (GenBank accession no. DQ993255). All sequence generated by PCR was invariant, that is, no polymorphism (single nucleotide or insertion/deletion), suggesting Nici is monomorphic at the MHC-B locus.
The turkey MHC-B region has a high overall GC content of ∼53.6% (Fig. 1⇑B) similar to the 55.5% GC of the chicken. Several repetitive DNA types were identified in the B locus (Fig. 1⇑C), including 12 CR1/long terminal repeats, 30 simple sequence repeats (repeat motif ≥ 5), and 3 complex repeats. A large (∼300 bp) C/T pentameric repeat (assigned microsatellite locus MNT-482) is located at 61.5 kb. This repeat is completely absent in the chicken. Twenty-two tRNA sequences were identified by tRNAScan (Fig. 1⇑D). Analysis of the chicken sequence aligned with the turkey indicates presence of the same tRNA sequences in the same syntenic order.
Gene identification and annotation
On the basis of basic local alignment search tool homologies, FGENESH, and EST analysis, the 197-kb region contained 34 predicted genes (Fig. 1⇑E) compared with 31 genes in the homologous chicken sequence (∼170 kb; Fig. 1⇑F). These genes include seven TRIM-like (TRIM7.1, TRIM7.2, TRIM39.1, TRIM39.2, TRIM27.1, TRIM27.2, and TRIM41), two zinc finger-like (Bzfp1 and Bzfp2), 44G24.1, L-amino acid oxidase-like (LAAO), Hep21, guanidine nucleotide binding like, two butyrophilin like (BTN1 and BTN2), three BG-zipper like (BG1–3), two C-type lectin like (B-NK and Blec), three MHC class II B loci, Tapasin, RING3, DMA, two DMB genes, two MHC class I loci, TAP1 and 2, complement protein C4, and a small portion of the histone gene CenpA.
As shown in Fig. 2⇓, MHC-B similarity is nearly linear between the turkey and chicken. The loci show high sequence homology with nearly perfect syntenic gene order. The turkey sequence contained genes homologous to all of those identified in the chicken haplotypes sequenced to date (12, 34, 40). Of note are the replicated blocks representing the additional BG loci in the turkey (chicken possesses one in the syntenic position), the syntenic and inverted homologies at the two class I genes, the difference in the number of class II B (two located between TAPBP and BRD2 in the turkey compared with one in chicken) and the inversion of the TAPBP gene. The TAPBP gene is in opposite orientation with respect to the chicken (verified by PCR; data not shown).
For each gene identified, the position, predicted coding sequence, and resulting amino acid sequences were determined (Table I⇓). The predicted coding sequences between the turkey and the B21-like haplotype present in the chicken whole-genome sequence (34) share homologies ranging between 85 and 96.5%, with an average 95% nt identity. Aligned amino acid sequences between the two haplotypes are between 73 and 100% identical, with an average similarity of 96% (Table I⇓).
All splice donor and acceptor sequences are the canonical GT/AG. However, analysis of the aligned turkey and chicken sequences suggests alternative predicted protein coding sequences in 11 of the 31 loci identified by Shiina et al. (34). Two instances, BTN1 and C4, were verified through RT-PCR. CpG islands are clearly conserved between these species (vertical arrows; Fig. 1⇑, E and F). Islands are present within the class I and class II B genes as well as TAP1, DMA, and BRD2. CpG islands are also present near TRIM41, the 3′ end containing TRIM7.2, and the two zinc finger genes, Bzfp1 and Bzfp2.
Evidence of expression (EST) in poultry is available for 32 B locus genes. Turkey ESTs were identified for 9 loci (TRIM7.2, GNB2L1, BTN1, BG1–3, RING3, and MHC class I 1 and 2). Two of the 34 genes (TRIM39.1 and BTN2) lack EST/mRNA-based evidence for transcription in any avian species; however, the levels of similarity between turkey and chicken sequences suggest functional genes, with perhaps unique temporal and/or tissue-specific expression patterns. For the remaining genes, a majority—if not all—exons are represented in the EST/mRNA databases.
Examination of the nucleotide substitutions between the two species provides further evidence of gene involvement in immunity, with loci interacting with rapidly evolving pathogens under strong selective pressures and possessing a higher ratio of nonsynonymous (dN) to synonymous (dS) substitutions. Yang and Swanson (41) define a dN/dS ratio > 1, = 1, and < 1 as evidence for positive (diversifying), neutral, and purifying selection, respectively. Axelsson et al. (42) evaluated the substitution rates between turkey and chicken and identified a dN/dS ratio varying from 0.185 to 0.094 depending on chromosome size (macro and micro, respectively). However, most genes in the B region show higher dN/dS ratios (average of 0.259) than genes from other microchromosomes. Ratios for most MHC genes suggested purifying selection, those with very low dN/dS include TRIM7.2, LAAO, TRIM7.1, TRIM41, and BRD2 (Table I⇑). As one might predict, the highly polymorphic class I and class II B genes showed the largest dN/dS values; however, the classical MHC genes failed to show strong evidence of positive selection when the complete coding sequences were analyzed (Table I⇑).
Although the greatest level of nonsynonymous substitutions occur within the class I α1α2 and class II β1 peptide-binding domains, the C-type lectin-like NK cell receptor (Blec2) showed a comparably higher ratio of nonsynonymous to synonymous substitutions between turkey and chicken throughout the coding sequence. Likewise, the level of variation identified among nine chicken haplotypes was similarly skewed to nonsynonymous substitutions (dN/dS = 2.66 (43)), and an expanded analysis with 14 haplotypes showed an even greater ratio (dN/dS = 7.07 (40)). Nearly all polymorphisms identified within the coding region of this gene are nonsynonymous. Similar levels of nonsynonymous polymorphism have been identified in the NK cell receptors of mammals (44).
Phylogenetic analysis of the coding sequence from MHC Ag genes (class I/BF, class II B/BL, and BG (IG to transmembrane domains to obtain alignments)) of the turkey, quail, and two chicken haplotypes (CB/B12 and RJF/B21) was performed using ClustalX (data not shown). The results suggest each gene is monophyletic, originating from a single ancestral locus as similarly reported by Shiina et al. (21).
Three BG loci are located between BTN2 and Blec2 in turkey where only a single locus is present in the chicken. An interesting feature of these BG genes is the organization of tandem repeated 21-bp exons comprising the intracellular coil-coil domains, the significance of which has not been fully identified. BG1, BG2, and BG3 each have a predicted 9, 24, and 14 of these exons, respectively.
Class I genes
A single class I locus flanked by DMB2 and TAP1 was identified in the turkey BAC clone. A second locus was identified in the PCR-amplified region located between TAP2 and C4 (Fig. 1⇑). Both loci were located in the same orientation and position as in the chicken. Class I genes in turkey are comprised of eight exons encoding a signal peptide, α1-3, transmembrane, and cytoplasmic domains similar to those originally identified in the chicken (45). The syntenic loci (e.g., turkey class IA1 and chicken BF1) also posses highly similar 5′ untranslated regions (UTR). The class IA1 locus lacked a recognizable poly(A) signal.
Class II genes
Three class II B loci were identified in the sequenced BAC clone (Fig. 1⇑). One class II B locus is located between Blec1 and TAPBL similar to chicken. Two class II B loci, positioned in the same transcriptional orientation, are flanked by TAPBL and BRD2, in contrast to the single locus observed in the chicken (Fig. 1⇑). Southern hybridization (supplemental Fig. 1),5 locus-specific PCR of genomic DNA, and subsequent resequencing confirmed the presence of three loci in Nici as observed in the BAC clone.
Comparative analysis of this sequence suggests the origin of the turkey class II B2 locus is the result of a recombination/gene conversion event in an ancestral chromosome that involved an inversion. On the basis of the flanking sequence, this event included the 5′ UTR of the ancestral class II B2 (denoted as aClass IIB1 in Fig. 3⇓) locus through the ancestral class II B1 3′ UTR (aClass IIB2), resulting in the observed inversion of the TAPBP gene and the duplication of the ancestral class IIB1 locus (Fig. 3⇓). The turkey class II B1 locus (tClass IIB1 in Fig. 3⇓) is homologous to the chicken BLB1 locus, except it retains a large portion of the 5′ UTR of the BLB2 locus (−300 to TAPBP). The class II B2 locus (tClass IIB2) is fully homologous to the chicken BLB1 locus from 5′ to 3′ UTR. The class II B3 locus (tClass IIB3) is homologous to the chicken BLB2 locus; however, a large portion of the 5′ UTR (−150 to −2500) has significantly diverged from the chicken with no homology to any other known genome sequence. Preliminary studies in the turkey have shown the highest expressed class IIB locus is IIB1. Locus IIB3 is expressed at approximately half the level of IIB1, and class IIB2 appears to be unexpressed, corresponding well with their respective chicken homologues (L. D. Chaves, unpublished observation).
The avian MHC is of significant scientific interest. In the turkey and chicken (and likely quail), it is divided into two distinct regions with the classical MHC genes in the B locus and nonclassical MHC loci in the Y locus. The B locus is tightly compact and very gene dense with the classical class I and class II B loci encompassed within a distance of <50 kb. In contrast, the genetically unlinked MHC-Y locus is much less defined. To date, only a limited amount of sequence data is available for the Y locus and only from the chicken. This region possesses nonclassical MHC loci, lectin-like loci, and additional MHC paralogous loci yet to be defined. Although it lacks classical MHC loci, the chicken Y locus does have an effect on the host response to pathogens (7, 8).
The remarkable similarity between the turkey and the chicken MHC-B loci is unexpected. Despite similar phylogenetic distances from turkey and chicken, the quail contains an expanded set of genes in the region. The sequenced quail haplotype contains 10 class II B, 7 class I, 8 BG-like, 4 NK, and 6 Blec-like genes (21). The duck (Anas platyrhyncho) has at least five class I loci located adjacent to TAP2 (46). Although initial comparisons between chicken and quail suggested the B locus to be rapidly diverging and subjected to extensive selection, the similarity of the turkey locus with that of the chicken seems to contradict this observation. The overall MHC-B region appears largely stable between turkey and chicken, with tremendous conservation of gene content and order. Shiina et al. (21) suggested the rapid divergence between chicken and quail might be due to the increased pathogens that quail might be exposed to as a result of its migratory behavior. In contrast, neither turkey nor chicken migrate, potentially reducing diversity of pathogen exposure.
The ability to compare the turkey and chicken sequences has resulted in improved gene identification. The annotations of 11 genes in the chicken should be amended based on the comparative alignment. Two of those predictions were confirmed by RT-PCR. The BTN1-coding sequence is one-third larger than previously thought, containing the PRY/SPRY domains associated with the TRIM genes as well as a RecF/RecN/SMC domain commonly involved in chromosome maintenance and recombination (47). Less dramatic is the discrepancy in the complement protein C4 where the coding sequence was found to be ∼150 bp larger, resolving a previously incorrect exon prediction. Other genes with minor annotation differences were not verified in this study; however, EST data sets provide added confirmation.
A characteristic of MHC genes is their significant level of variation both within and between species. However, several genes with diverse functions (TRIM7.2, LAAO, TRIM7.1, TRIM41, and BRD2) were highly conserved between turkey and chicken. Although little is known of the avian paralogs, mammalian TRIM genes are members of a large family of genes, some of which have been found to possess immune functions with involvement in disease resistance (48, 49, 50). LAAO is a metabolic enzyme and BRD2 is suggested to be a regulator of transcription (51, 52).
BG genes are unique to avian lineages and appear to be numerous and spread throughout the avian MHC-B. To date, little is known of their biological function. These genes have highly conserved immunoglobin-variable-like and transmembrane domains yet show considerable divergence in their intracellular regions. Three BG loci were identified in this study. Additional BG loci are located upstream of the BF-BL region in the chicken and there is evidence for additional loci upstream in the turkey as well based on BAC end sequences (our unpublished data). The actual number of additional loci, however, is currently not known in either species. As a result of polymorphism and multiple genome copies of BG genes, no expression data were available to confirm the annotations provided in this work.
The putative NK cell receptor Blec2 shows some of the most significant divergence between turkey and chicken. Indeed, even within chicken haplotypes, it is under the highest level of selection (40). This level of selection suggests this gene may have a significant role in immunology and supports the suggestion that an allele of this locus in chickens may be responsible for the genetic resistance to Marek’s disease long associated with the B21 haplotype (12, 17, 43, 53).
The minimal number of avian MHC class I genes and their proximal location within the MHC-B could constrain this gene to coevolve with the dominantly expressed class I gene for proper NK cell surveillance. However, using a reporter gene construct, Viertlboeck et al. (54) found this receptor to be unresponsive in cocultures with unstimulated chicken cells and specific transfectants harboring the class I and Blec genes. Stimulated splenocytes (Con A or PMA) possessed a ligand that the NK cell receptor recognized, but it was not determined if the ligand was allele-specific for each NK cell receptor or if activated cells of an alternative haplotype could also stimulate splenocytes. Further evidence by Rogers and Kaufman (43) suggest the ligand for B-NK is not an MHC molecule, thus confounding the role of Blec2.
Three class II B loci were identified in the turkey, whereas the chicken possesses two. The quail has remarkable flexibility with regards to class II B loci, with between one and three loci occupying the location between TAPBL and BRD2 (22). In a previous study of class II B (β1-domain) sequences in turkeys, Ahmed et al. (55) used PCR to identify three separate domains in genomic DNA from several members of a possibly closed flock. PCR-RFLP found up to three alleles present within a given individual, suggesting turkeys may be polymorphic in both class II B alleles and loci. However, based on the core turkey MHC-B locus, it is likely that two haplotypes sharing alleles (as defined by HinfI digestion) were present in this population. A second contributing factor may have been null amplifications leading to an under estimation of total allele numbers. For example, in the turkey class II B loci identified in the present study and unpublished results, up to three base substitutions occurred at the primer binding site of the sense primer used by Ahmed et al. (55), supporting the possibility of null amplifications.
Similarity at the MHC-B locus between the turkey and chicken should be viewed in light of the limited data. Only one turkey haplotype has been thoroughly examined and the extent of intraspecies variation is not known. Locus-specific PCR on heterogeneous turkey DNA at least confirmed the presence and orientation of classical MHC loci in an additional bird (our unpublished data). Sequence-level characterization of additional haplotypes is needed to verify the general gene content and to identify the overall level of variation within this region. In 14 chicken haplotypes that have been resequenced, overall sequence variability was high (1 SNP/25 bp); however, no variation in gene number or orientation was found (34). Furthermore, recombination within the chicken MHC-B locus has rarely been observed (56, 57, 58, 59, 60).
Haplotypes of the chicken MHC-B have profound influence on resistance or susceptibility to numerous pathogens. Infections of Rous Sarcoma Virus or Marek’s Disease show two of the most dramatic influences of MHC haplotypes on disease (61, 62). In both instances, different haplotypes have strong influences on the progression or regression of disease. However, whereas strong hypotheses suggest a role for the class I locus (61, 63), in neither case is it precisely known which particular gene/allele is responsible for the effects. The effect of MHC haplotypes on disease susceptibility in turkeys is currently unknown.
This study examined the genes present in the core turkey MHC-B locus, the region described as the minimal essential MHC in chickens. Little is known, however, of the distal MHC regions. Examination of overlapping BAC clones has provided the best assembly of the extended chicken B locus (34). Five loci (Bzfp3, Blec3, KIFC, BG2, and BG3) were identified in the chicken upstream of the region included in this study and four loci (TNXB, LTB-4R1, CD1A2, and CD1A1) can be placed downstream (34). The chicken used in whole genome sequencing (RJF 256 from line UCD 001) is an MHC recombinant and is heterozygous upstream of the Bzfp1 gene (L. D. Chaves, unpublished observation). Heterozygosity in addition to the highly similar, polymorphic, and multiple-copy extended BG gene family have made assembly of this region of the chicken genome difficult. Less than 4 kb of overlapping and extending whole genome sequence can be aligned to the B21 haplotype sequenced by Shiina et al. (L. D. Chaves, unpublished results).
Overlapping BAC clones extending the 5′ end of the turkey B-locus have been identified in the CHORI 260 BAC library. However, no clones were found overlapping the 3′ end. On the basis of the genomic sequencing results of the present, the turkey selected for BAC library construction and whole-genome sequencing (Nici) appears monomorphic within the MHC-B locus. This will likely aid the assembly of the turkey extended B locus and may ultimately improve the ordering and assembly of the homologous chicken sequence. Further sequencing of the chicken MHC-B will invariably require additional, overlapping, haplotype-specific BAC/cosmid clones, as well as benefit from future whole genome sequencing in other avian species.
We thank Sue Lamont, Mike Murtaugh, and Mark Rutherford for their helpful discussions in preparing this manuscript and two anonymous reviewers for their comments and helpful suggestions in the presentation of this work.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
↵1 This research was supported by grants from the University of Minnesota Agriculture Experiment Station and the Cooperative State Research, Education, and Extension Service, U.S. Department of Agriculture (2004-35205-14217 and 2009-35205-05302).
↵3 Address correspondence and reprint requests to Dr. Lee D. Chaves at the current Address: National Jewish Health, Division of Allergy and Clinical Immunology, Department of Medicine, Denver, CO 80206. E-mail address:
↵4 Abbreviations used in this paper: BAC, bacterial artificial chromosome; EST, expressed sequence tag; LAAO, L-amino acid oxidase; TRIM, tripartite motif; UTR, untranslated region.
↵5 The online version of this article contains supplemental material.
- Received April 24, 2009.
- Accepted September 15, 2009.
- Copyright © 2009 by The American Association of Immunologists, Inc.