Narrowing the wingless-2 mutation to a 227 kb candidate region on chicken chromosome 12.

Wingless-2 (wg-2) is an autosomal recessive mutation in chicken that results in an embryonic lethal condition. Affected individuals exhibit a multisystem syndrome characterized by absent wings, truncated legs, and craniofacial, kidney, and feather malformations. Previously, work focused on phenotype description, establishing the autosomal recessive pattern of Mendelian inheritance and placing the mutation on an inbred genetic background to create the congenic line UCD Wingless-2.331. The research described in this paper employed the complementary tools of breeding, genetics, and genomics to map the chromosomal location of the mutation and successively narrow the size of the region for analysis of the causative element. Specifically, the wg-2 mutation was initially mapped to a 7 Mb region of chromosome 12 using an Illumina 3 K SNP array. Subsequent SNP genotyping and exon sequencing combined with analysis from improved genome assemblies narrowed the region of interest to a maximum size of 227 kb. Within this region, 3 validated and 3 predicted candidate genes are found, and these are described. The wg-2 mutation is a valuable resource to contribute to an improved understanding of the developmental pathways involved in chicken and avian limb development as well as serving as a model for human development, as the resulting syndrome shares features with human congenital disorders.

INTRODUCTION

The wingless-2 () mutation was originally identified at the University of Connecticut in a flock of chickens synthesized from Rhode Island Red, White Leghorn, Dorking breeds, and Red Jungle Fowl (Zwilling, 1956). The wg-2 mutant phenotype is characterized by the complete absence of forelimbs, severely truncated or absent hindlimbs, craniofacial defects (maxillary region clefting), malformed kidneys, and abnormal plumage (Zwilling, 1956; Somes, 1990; Pisenti et al., 1999). The mutation is inherited in an autosomal recessive pattern and is an embryonic lethal (Somes, 1990; Pisenti, 1995). Reciprocal tissue transplantations between mesoderm and ectoderm suggest that the wg-2 mutation affects ectoderm (Carrington and Fallon, 1984; Hamburger and Hamilton, 1992). Wg-2 was the second wingless mutation characterized in the chicken; the first, now extinct, exhibited a similar phenotype and additionally was described as having incomplete or absent lungs, air sacs, and kidneys (Waters and Bywaters, 1943). An improved understanding of the wg-2 developmental mutation will lead to new knowledge of important developmental pathways in poultry affecting limb, craniofacial (e.g., beak), kidney, and feather development. Further, the wg-2 developmental mutation is an important model system for the investigation of genes and pathways involved in human syndromes with craniofacial (cleft palate), limb (tetra-amelia, phocomelia) and organ (kidney) malformations.

In the late 1980s at the University of California Davis (UCD), a congenic inbred line was established by introgression of wg-2 in order to advance mapping and gene analysis research at a time wherein genomic tools were barely envisioned, much less the advent of the chicken sequencing project to advance poultry and vertebrate biology. The Wingless-2.331 (Wg-2.331) congenic inbred line was derived by first crossing known wg-2 carriers (+/wg-2) into the UCD 331.003 congenic line (Abplanalp et al., 1992; Pisenti et al., 1999). The +/wg-2 carrier progeny were confirmed by test-mating and then used to backcross to the parent line UCD 331.003. The backcross method using confirmed carrier (+/wg-2) progeny was repeated for at least 7 generations, with 2 additional backcrosses after the line was closed to improve fertility and viability, as per available breeding records and personal communications. This labor-intensive multi-year strategy placed the wg-2 locus in a uniform genetic background (99+% inbred) such that only the introgressed region containing the mutation along with other normal polymorphisms differs between normal (+/+ or +/wg-2) and mutant (wg-2/wg-2) individuals. The line was closed (inter se matings only) and perpetuated using +/wg-2 parents as breeders after being verified by test crosses. Over more than 60 yr of breeding, Mendelian ratios fit an autosomal recessive pattern of inheritance for the wg-2 mutation.

Here, we describe genetic and genomic analyses of the congenic inbred line Wg-2.331 that have significantly advanced our knowledge of the genomic region incorporating the wg-2 mutation—work that took place over a decade and spanned 4 chicken genome builds. The work is presented in galGal5 coordinates for the sake of clarity, with references to the state of the build at the time when necessary. At the outset, the chromosomal location and subregion of the wg-2 mutation was mapped by using a custom Illumina 3 K SNP chicken array (Muir et al., 2008a) wherein a tightly linked SNP was identified, which had the advantage of being predictive of carrier status. Subsequently, new SNP were identified and used for fine-mapping, which, combined with exon-based sequencing of candidate genes within the region, significantly narrowed the causative region.

MATERIALS AND METHODS

Genetic Lines

All birds are derived from the UCD Wg-2.331 developmental mutant congenic inbred chicken line. The birds initially carrying the wg-2 mutation were bred into the UCD 331.003 line, which is congenic with UCD 003 [a highly inbred line (F > .99)] except at the major histocompatibility complex (Abplanalp et al., 1992) and a small number of other SNP loci (Robb et al., 2011). Samples throughout this paper are designated as wild type +/+ (normal phenotype), carrier +/wg-2 (normal phenotype), or mutant wg-2/wg-2 (abnormal phenotype). The birds used in this research were cared for by trained staff and according to a protocol approved by the UC Davis Institutional Animal Care and Use Committee (Protocol #18816).

Illumina 3,072 (3 K) SNP Array Analysis

Samples were collected from 9 adult Wg-2.331 birds confirmed as carriers (+/wg-2) by test-mating and 11 phenotype-confirmed mutant embryos (wg-2/wg-2) at 10 d of embryogenesis (10E). DNA was isolated and purified from blood stored in heparin (adults) or sodium citrate (embryos). All DNA samples were isolated using QIAamp® Blood Mini Kit (Qiagen, Valencia, CA) and sent to Illumina, Inc. (San Diego, CA) for analysis. Samples were tested for 3,072 SNP evenly spaced throughout the chicken genome (Gitter, 2006; Muir et al., 2008a; Muir et al., 2008b). SNP linked to wg-2 should be homozygous in all wg-2 mutants, heterozygous in wg-2 carriers, and homozygous for the alternative allele in UCD 331.003.

Genotyping Wg-2.331 Birds Using a Linked SNP

The 3 K SNP analysis found one SNP (rs14034687 aka SNP-16–390-35,608-S-1 and referred to as “SNP390”) in complete linkage disequilibrium (LD) with the mutant phenotype and, thus, had high value in screening for wg-2 status. An Applied Biosystems Custom TaqMan® SNP Genotyping Assay was developed for this purpose (Table 1). Real-time PCR was performed on DNA samples (10 ng/μl) isolated from Wg-2.331 individuals to determine genotypes for SNP390 using TaqMan® Universal PCR Master Mix (Applied Biosystems, Carlsbad, CA) under standard conditions. After 15 min at 95°C to activate the polymerase, samples underwent 45 cycles of 95°C for 15 s, 60°C for 1 min, and a plate read. Genotypes were confirmed by comparison to previously genotyped (SNP390) or phenotype-confirmed samples representing the 3 forms of the SNP (normal +/+, +/wg-2; mutant wg-2/wg-2) with test-mating to affirm carrier status (+/wg-2). Samples were collected for DNA preparation and genotype determination at one of 3 time points, depending on the experiment: adults (blood), one d post-hatch chicks (pin feather), or embryos (blood).

Table 1.

TaqMan® primer-probe sequences for the SNP390 (rs14034687) genotyping assay used to predict UCD Wg-2.331 genotypes: +/+, +/wg-2, wg-2/wg-2.

SNP coordinate[1]	Chr 12: 5,061,391
Amplicon coordinates[1]	Chr 12: 5,061,327 - 5,061,449
Forward primer	5΄ CTCTTTTGTTAGAGCTGTGCTATG- G 3΄
Reverse primer	5΄ CAGGACTGTTGCTTTTGTTCTTTATAGTTTT 3΄
VIC probe (+/+)	5΄ CATGAACACAAAGCC 3΄
FAM probe (wg-2/wg-2)	5΄ CATGAACATAAAGCC 3΄

1Coordinates based on Dec 2015 Gallus_gallus-5.0/galGal5 chicken assembly for GGA 12.

Fine-Mapping

SNP analysis was performed on 79 new individuals to assess the size of the Wg-2.331 causative region segregating in the population. Primer pairs (Supplemental Table S1) were designed (Amplify 1.2, Engels 1993) to amplify 800 to 1,000 bp chromosomal fragments across the region. Eight informative polymorphisms were identified (Supplemental Table S2). DNA was isolated from normal birds (+/+, +/wg-2) and mutant embryos (wg-2/wg-2), and genotypes were determined using SNP390. PCR reactions used GoTaq® (Promega, Madison, WI) under standard conditions. Amplicons were purified by QIAquick® Spin Kit (Qiagen, Valencia, CA) and sequenced by Sanger sequencing on an ABI 3100 (Applied Biosystems, Carlsbad, CA).

Exon Sequencing of Candidate Genes

Exon sequencing was performed on the 6 candidate genes found within the original causative region (WUGSC 1.0/galGal2): PPARG, TSEN2, MKRN2, RAF1, CNBP, and ACAD9 to identify polymorphisms within the exons of each gene. (Figure 1A, Supplemental Table S3). Sequencing was performed using DNA from 6 samples after genotyping with SNP390. Five were from the UCD Wg-2.331 line including one normal homozygote (+/+), 2 normal heterozygotes (+/wg-2), and 2 mutant homozygotes (wg-2/wg-2). The sixth sample was from UCD 001 (Red Jungle Fowl), the line used for the chicken genome reference sequence (International Chicken Genome Sequencing Consortium, 2004). Sanger sequencing was conducted on an ABI 3100 (Applied Biosystems, Carlsbad, CA).

Figure 1.

Refinement of GGA 12 wingless-2 causative region by SNP linkage analysis over multiple generations and genome assemblies. Each gene is denoted by a different shaded box (see Figure 1 Key). The causative region was refined from 842 kb containing 7 annotated genes initially to 227 kb with 3 annotated genes, one partial. The long gray box indicates the causative region. Open triangle heads indicate SNP used for genotyping the margins of a linked region, filled triangle heads indicate the SNP390 genotyping SNP. A dotted line with an arrow indicates that the region boundary exists within a gene (see Supplemental Tables S1–S2, S4–S6 for further information on SNP).

Candidate Gene TSEN2: Haplotype Analysis

Sanger sequencing was conducted on DNA samples from 4 mutant embryos (phenotype-confirmed and SNP390 genotyped as wg-2/wg-2) to investigate their TSEN2 haplotype. Primer pairs were created using Primer3 (Untergasser et al. 2012; see Supplemental Table S4) to amplify the 10 kb TSEN2 gene with 2 kb of flanking 5΄ and 3΄ regions. Amplicons were designed to range in length from 400 to 600 bp and overlapped to resolve poor nucleotide calls. All PCR reactions were performed using Phire™ (Finnzymes, Lafayette, CO) under standard conditions for all primers except AW467–1, -2, and -3, which required 10% DMSO. The TSEN2 SNP status across one 500 bp region also was investigated in an additional 29 genotype-confirmed mutant embryos. Purified amplicons were sequenced (Davis Sequencing, Davis, CA) and aligned by Sequencher (Gene Codes, Ann Arbor, MI). One embryo exhibited evidence of a recombination event and was thus also sequenced across the entire TSEN2 gene.

Linkage Analysis in New Generations of Carriers

After several years of maintaining the line without fine-mapping, the 2014 and 2015 carriers used to perpetuate the line (generating 2015 and 2016 offspring, respectively) were genotyped at up to 19 additional SNP (Supplemental Table S5) to identify recent recombination events. The SNP selected for this purpose were identified from sequence data developed from a capture array experiment employing 2 pooled wg-2 embryos (Robb and Delany, 2012). These SNP were previously observed in other breeds, shown by their listing in the Beijing Genome Institute database (http://chicken.genomics.org.cn/). Using these SNP precluded problems with validation and variant calling confidence, as they were identified by 2 independent sources. The birds were genotyped at SNP390, test-mated, and then examined at the additional SNP loci to find individuals in which linkage between a variant and wg-2 carrier status was lost (lack of association), indicating recombination. Initial SNP genotyping assessed linkage at the ends of the region with only a couple of SNP. In the case in which genotypes did not retain their expected heterozygous condition, additional SNP oriented further inward were assessed to determine the breakpoint of the putative recombination event in these birds. PCR reactions used Phire Hot Start II DNA Polymerase (ThermoFisher Scientific, Waltham, MA), and PCR products were sequenced by Sanger sequencing (Davis Sequencing, Davis, CA, or Quintara Biosciences, Berkeley, CA).

RESULTS

Mapping of the Wingless-2 Mutation

To map the mutation to a chromosome and causative region, SNP were screened for linkage with the mutant phenotype using the Illumina 3 K SNP chicken array based on the WUGSC1.0/galGal2 assembly. Linkage analysis was performed in 20 Wg-2.331 individuals consisting of 9 +/wg-2 carrier adults and 11 wg-2/wg-2 mutant embryos. Of the 3,072 SNP assayed, 2,733 SNP amplified successfully (Muir et al., 2008a). Twelve SNP associated with GGA 12 and linked to wg-2 were identified establishing a 7.16 Mb region of interest between 4,180,275 and 11,344,252 bp (Gitter, 2006; Supplemental Table S6). This region in galGal5 is from 4,121,073 to 11,352,192, establishing a 7.23 Mb region of interest. Of these SNP, SNP390 was in complete LD and was used for genotyping all subsequent samples (Table 1). The GGA 12 causative region was estimated to be a maximum size (CR) of 842 kb after observing recombination involving SNP398 and SNP379 in 3 proven carrier birds (+/wg-2) (Figure 1A). This region based on the galGal5 assembly encodes 7 known genes: PPARG, TSEN2, MKRN2, RAF1, CNBP, RAB43, and ACAD9.

An additional set of 79 Wg-2.331 birds was sequenced across polymorphic regions using new primers (Supplemental Table S1) to enhance marker content across the region and improve linkage analysis. Recombination was observed in 13 individuals. Nine individuals exhibited recombination involving 3 to 4 markers confirming the 5΄ CRmax boundary of the causative region to marker SNP398, located at 4,724,679 bp on galGal5 (Supplemental Table S5). Five individuals (one of these experienced recombination at both ends of the region) indicated a recombination involving 4 to 5 markers, narrowing the region at the 3΄ boundary of the CRmax to marker MK-21 located at 5,194,358 bp, inward from marker SNP379. In summary, the wg-2 causative region was reduced to a CRmax of 470 kb spanning 4,724,679 to 5,194,358 bp on galGal5, which eliminated one gene (ACAD9) from an original list of 7 candidate genes (Figure 1B).

As an alternative approach to identifying the wg-2 mutation, exons of the 5 remaining validated genes at the time were sequenced (Supplemental Table S3) to uncover potentially causative amino acid substitutions. This work was performed while galGal3 was the operative genome build, and 5 genes were included in the study: TSEN2, RAF1, MKRN2, CNBP, and PPARG. Sequencing of 5 Wg-2.331 individuals of confirmed phenotype and SNP390 genotype (one normal +/+, 2 carriers +/wg-2, and 2 mutants wg-2/wg-2) and one Red Jungle Fowl (UCD 001) revealed nucleotide polymorphisms within exons of 3 candidate genes [TSEN2 (n = 25), RAF1 (n = 1), MKRN2 (n = 4)], whereas no polymorphisms were detected in 2 candidates (CNBP and PPARG). Amino acid predictive analysis revealed that of 25 TSEN2 polymorphisms, 7 encoded non-synonymous amino acid substitutions, whereas predictions of the polymorphisms in the other 2 genes indicated synonymous substitutions. In silico alignment (see Figure 2) revealed that 3 of the amino acid substitutions (Y149D, A211T, and A214T) were previously reported in the DT40 cell line (NP_001025765.1 (Baba and Humphries, 1984; Caldwell et al., 2005)), suggesting these would not be likely candidates (Figure 2). One substitution differed among mutant, normal (wild type), and DT40 (M356V) genotypes, and 3 other substitutions were unique to wg-2 mutants (I75V, A232G, and N459S). These results prompted further investigation of TSEN2 specifically.

Figure 2.

Alignment of TSEN2 protein of normal and mutant Wg-2.331 samples in reference to DT40. Seven amino acid substitutions based on sequencing results were identified in wg-2/wg-2 mutant samples. The normal and mutant TSEN2 proteins were then aligned to the NCBI reference protein for TSEN2 (DT40_SEN2) to determine possible similarities. The alignment found that 3 of the substituted amino acids were present in the TSEN2 reference [Y149D, A211T, and A214T, in (a) boxes]. For the remaining substitutions, 3 were unique to the mutant samples [I75V, A232G, and N459S, in (b) boxes], and one was unique in both normal and mutant samples in comparison to the TSEN2 reference [M356V, in (c) box]. One site was polymorphic for amino acid across all 3 samples [in (d) box]. The NCBI reference was sequenced from the DT40 cell line derived from Single Comb White Leghorn bursal lymphocytes (Caldwell et al., 2005, NP_001025765.1).

To confirm the amino acid substitutions observed in TSEN2, 2 additional homozygous normal (+/+) and 2 homozygous mutant (wg-2/wg-2) individuals were sequenced for the entire TSEN2 gene using overlapping primer pairs (Supplemental Table S4). Samples were selected based on their confirmed phenotypes and SNP390 genotypes.

Complete sequencing of the 10 kb TSEN2 gene (and flanking regions) revealed 2 distinct genotype-specific haplotypes. The mutant (wg-2/wg-2) haplotype retained the polymorphisms observed in the original sequencing study along with newly identified variants. Surprisingly, one homozygous normal (SNP390 +/+) sample was heterozygous for the 2 TSEN2 haplotypes. Thus, it seemed likely that recombination had occurred between TSEN2 and marker SNP390. To verify this putative recombination event, 29 additional samples were sequenced using one primer pair (AW-45, see Supplemental Table S4).

Of the 29 samples (normal, carrier, and mutant birds) sequenced at this smaller section, 6 exhibited evidence of a recombination event (Supplemental Table S7). One of these, 2A1, was then sequenced across TSEN2 in full to determine if this event indicated loss of linkage between the phenotype and the previously identified mutant-associated TSEN2 polymorphisms. Sequence from this sample, a phenotypically confirmed normal embryo heterozygous for SNP390 (+/wg-2), exhibited homozygous polymorphisms in TSEN2 matching those in the mutant individuals (Figure 2). The presence of these polymorphisms in a phenotypically normal embryo indicates that they are not linked to the mutant phenotype. Also based on these results, the wg-2 causative region was reduced to a CRmax of 267 kb, from a TSEN2 SNP in the terminal exon of the gene (referred to as AW-45, amplified by primer set AW-45) to SNP MK-21, or 4,927,330 to 5,194,358 as per the galGal5 assembly (Figure 1C), and therefore 2 of the candidate genes, TSEN2 and PPARG (5΄ of the recombination event), were eliminated from the wg-2 region of interest leaving 4 known genes (MKRN2, RAF1, RAB43, and CNBP).

Over several years subsequent to the TSEN2 analysis, the population was maintained by the breeding of carriers determined by SNP390 genotyping. We therefore decided to examine the ends of the linked region in SNP390 +/wg-2 individuals starting in 2014, to assess if recombination events had occurred in the intervening years. Several putative recombination events were identified in the 2014 and then 2015 carrier birds, and these were bred to perpetuate the Wg-2.331 line. An additional 19 SNP were genotyped to assess the recombination breakpoints (Supplemental Table S5). These SNP were derived from a capture array (Robb and Delany, 2012) and independently observed by other researchers as shown by their listing in the Beijing Genome Institute database. A new genome assembly (galGal5 in 2015) became available, such that the prior causative region was shown to span a 267 kb region and included a fourth validated gene, RAB43. The new recombination events observed in carriers of the 2014 and 2015 generations indicated a CRmax of 226,637 bp in the galGal5 assembly, from marker IY-13 to IY-3 (Figure 1D). This region excluded MKRN2, limiting the list of validated candidate genes to 3 (RAF1, CNBP, and RAB43, Figure 1D).

DISCUSSION

Prior to the use of the Illumina 3 K SNP chicken array (Muir et al., 2008a; Muir et al., 2008b) and application of other genomics tools described here, the wg-2 mutation was genetically characterized only by its autosomal recessive mode of inheritance (Somes, 1990). By employing the chicken 3 K SNP array we established an initial region of interest of ca. 7 Mb mapping to GGA 12 and affirmed the autosomal mode of inheritance (Gitter, 2006). An important additional benefit of these results was the discovery of marker SNP390, which to date continues to exist in complete LD with the wg-2 mutation. SNP390 was incorporated in the Wg-2.331 congenic inbred line breeding program as a predictor of wg-2 genotype. This marker is a significant aid for the long-term perpetuation of the line, as the predictive genotyping conserves limited resources by reducing animal numbers, staff time, and the related costs. However, caution must be exercised in using a single or even several markers alone in predicting carriers; such predicted carriers must be test-mated routinely to ensure recombination does not disrupt the association, which, if it occurs without awareness, could result in loss of the mutation. The alternative dilemma is that by using only a single marker (e.g., SNP390) in selecting carriers for breeding during the early years of a project such as this, the breeders chosen carry a region surrounding the polymorphism that is not disrupted by recombination. Thus, use of markers for breeding schemes must be part of a comprehensive and balanced approach depending on the scientific, genetic, and breeding goals.

In this study, several complementary methods were employed toward our long-term aim of identifying the causative element of the wg-2 phenotype. SNP genotyping coupled with fine-mapping was used to take advantage of naturally occurring recombination events to reduce the region, while direct study of gene sequences prioritizes mutations that may be highly disruptive. Direct sequencing, SNP genotyping, predicting amino acid substitutions resulting from polymorphisms, and comparing polymorphisms among different organisms have all served to refine the potential causative region to a maximal size of 227 kb, containing only 3 confirmed genes as per the latest assembly (galGal5). An important consideration for studies such as this one and for candidate gene analysis in general is the quality of the genome assembly. The GGA 12 region of interest as per the galGal5 assembly has no evidence of sequencing gaps and exhibits conserved synteny for PPARG, TSEN2, MKRN2, and RAF1 with the human (GRCh37/hg19) and mouse (GRCm38/mm10) assemblies (http://genome.ucsc.edu/) of HSA 3 and MMU 6, respectively. CNBP and RAB43 are also syntenic in both human and mouse, but are not as proximal to the other genes as in the chicken assembly. While the first 4 genes map to the p arm of HSA 3 in human, CNBP and RAB43 are on the q arm; in mouse Cnbp and Rab43 are actually 5΄ of the other genes in MMU 6. There are also 3 genes that the UCSC Genome Browser displays on the Ensembl Gene Prediction track that may become fully validated in a future chicken genome assembly, as occurred with RAB43 during the course of this study. The predicted genes have sequence homology with known genes in other organisms, but have not been fully reviewed and validated in chicken. However, all 3 predicted genes have EST and are therefore likely to be real. Table 2 compares the known genomic attributes of the 3 validated genes and the 3 predicted genes in chicken, mouse, and human. Both validated and predicted genes of the wg-2 causative region are discussed briefly below with a comparative overview as to developmental function(s).

Table 2.

Comparative genomic attributes of genes in the 227 kb chicken causative region of wingless-2.

	Start Location (Mb)			Gene Size (bp)
Validated genes	Chicken, GGA 12[1]	Human, HSA 3[2]	Mouse, MMU 6[3]	Chicken	Human	Mouse
RAF1	4.954	12.58	115.62	27,932	80,626	58,063
CNBP	5.04	129.17	87.84	9269	14,439	8492
RAB43	5.09	129.09	87.79	34,582	34,582	22,927
Predicted Genes
ISY1	5.05	129.13	87.82	9211	31,674	20,312
CECR5	5.07	17.14 (HSA 22)	120.51	7163	27,767	21,806
EFCC1/CCDC48	5.15	129	87.73	42,751	39,113	25,043

1Gallus_gallus-5.0/galGal5.

2GRCh38/hg38.

3CRCm38/mm10.

4Only 13.5 kb of RAF1 are included in the causative region (see Figure 1D).

RAF1 (v-raf-1 murine leukemia viral oncogene homolog) is a serine-threonine kinase that belongs to the RAS-MAPK signaling pathway (http://www.uniprot.org/uniprot/P04049). Mutations in RAF1 are associated with 2 human developmental disorders, LEOPARD syndrome and Noonan syndrome, which are both characterized by cardiac and facial abnormalities with additional anomalies unique to each disorder (Sarkozy et al., 2008; Tidyman and Rauen, 2008). RAF1 regulates a member of the fibroblast growth factor (FGF) signaling family (MacNicol et al., 1993; Fongang and Kudlicki 2013). Notably, FGF8 is an inducer of chick limb development (Crossley et al., 1996).

Cellular nucleic acid binding protein () is hypothesized to be involved in tissue remodeling with a controlling role in both cell death and proliferation rates (Calcaterra et al., 2010). CNBP expression in a gastrula stage (about 24 h of incubation) chick embryo was reported to localize to the neuroectoderm of the prospective forebrain (Abe et al., 2006). By ca. 3 to 3.5 d of incubation, CNBP expression localized to the forebrain and midbrain with low expression in the tail, heart, and wing buds (Abe et al., 2006). CNBP knockdown caused truncation of the forebrain in chick (Abe et al., 2006) and mouse (Chen et al., 2003) embryos. Knockdown of this binding protein in zebrafish established its role in pharyngeal and craniofacial cartilaginous structures derived from the neural crest (NC). Specifically, morpholino-dependent depletion of the protein caused an increase in NC cell apoptosis and a decrease in proliferation (Weiner et al., 2011).

RAB43 is a member of the RAS oncogene family of small GTPases, which regulate intracellular membrane trafficking involving vesicles (formation, movement, tethering, and fusion). Human RAB43 is required for maintaining the structure/function of the Golgi complex (Haas et al., 2007; http://www.uniprot.org/uniprot/Q86YS6). Rab43 knockout mice are relatively normal, and no developmental impacts were reported (Kretzer et al., 2016).

The 3 predicted genes within the 227 kb wg-2 region of GGA 12 are ISY1 [NM_020701.2], CECR5 [NM_017829.5], and CCDC48 [NM_024768.2]. In Saccharomyces cerevisiae, ISY1 (Interactor of Syf1p) was identified as part of a spliceosome complex (http://www.uniprot.org/uniprot/Q9ULR0). A knockout (isy1Δ) resulted in lower pre-mRNA splicing efficiency indicating a function in pre-mRNA splicing (Dix et al., 1999). This predicted gene includes the SNP390 marker within a predicted intron. CECR5 (Cat Eye Syndrome Chromosome Region, Candidate 5) is a candidate gene for the human Schmid-Fraccaro Syndrome (Cat Eye Syndrome) resulting from partial duplication or trisomy of HSA 22. The affected individuals express variable phenotypes involving skeletal, craniofacial, and organ systems (Footz et al., 2001; http://www.uniprot.org/uniprot/Q9BXW7). The predicted gene CCDC48 (Coiled-coil domain containing 48), also known as EFCC1 (EF-Hand and coiled-coil domain containing 1), is a protein-coding gene with Gene Ontology annotation related to calcium ion binding, with little known about the gene otherwise (http://www.uniprot.org/uniprot/Q9HA90).

In summary, the wg-2 mutation was mapped to GGA 12 within a maximal 227 kb candidate region. This is a conservative estimate of the region, as the edges are defined by the outermost SNP that have been observed to be unlinked to the mutation. The distance between the unlinked SNP and the next instances of linkage is included in the CRmax to ensure that the causative mutation is included. The current size of the region is very small given the estimated recombination rate for chicken chromosomes (median value of 2.8 cM/Mb on macro- and 6.4 cM/Mb on microchromosomes, International Chicken Genome Sequencing Consortium, 2004). Although fine-mapping is unlikely to provide anything more than marginal reductions in the size of the causative region in the future, we continue to genotype at the margins to detect any recombination that might occur. Over the course of our work, which included fine-mapping and sequencing along with availability of improved assemblies, we eliminated 4 candidate genes from an original list of 7 genes. Thus, we have RAF1, CNBP, RAB43, and the 3 predicted genes for experimental consideration. As the predicted genes have reported EST, they are likely real and cannot be excluded. Our current research involves the development and analysis of a new 300 kb region capture array designed to obtain high-quality sequence reads for the GGA 12 wg-2 region of interest and small flanking regions across all coding and non-coding DNA. Pooled DNA from mutants was sequenced using this capture array technology to provide an initial assessment of overall variation in the region (Robb and Delany, 2012). The new array employs 2 wild type, 4 carrier, and 18 mutant individuals to improve our ability to distinguish potentially causative variation from the high degree of polymorphisms observed. A causative variant must be exclusive to and shared by all wg-2 mutant homozygotes, found heterozygous in carriers and absent in normal homozygotes. This new capture array effort will produce a comprehensive understanding of the wg-2 region by elucidating all variation, including SNP, indels, gaps, duplications, and any structural rearrangements, even in non-genic regions (e.g., promoters, enhancers, introns, and intergenic regions). Going forward, this dataset will enumerate the comprehensive list of those variants unique to, and potentially causative of, wg-2. Ultimately, to establish the cause of the phenotype, sequence-based approaches will be paired with study of gene expression differences between wg-2 mutant and normal siblings.

SUPPLEMENTARY DATA

Supplementary data are available at online.

Supplemental Table S1. PCR primers utilized for fine mapping the 2 Mb wingless-2 causative region to 842 kb of GGA 12.

Supplemental Table S2. Informative SNP derived from sequencing (Supplemental Table S1) showing linkage to Wg-2.331.

Supplemental Table S3. PCR primers for exons of candidate genes (PPARG, TSEN2, MKRN2, RAF1, CNBP) in 842 kb causative region (on GGA 12).

Supplemental Table S4. PCR primers for TSEN-2 (on GGA 12) sequencing of UCD Wg-2.331 individuals.

Supplemental Table S5. PCR primers to fine map the 267 kb wingless-2 causative region to 227 kb of GGA 12.

Supplemental Table S6. Informative SNP in showing linkage to Wg-2.331 from a 3 K array (Muir et al., 2008a).

Supplemental Table S7. Evidence for recombination between TSEN2 and SNP390. TSEN2 sequencing was conducted in 29 individuals (tables below), resulting in the identification of a recombination between one SNP locus (AW-45) of TSEN2 and marker SNP390 in 6 normal phenotype individuals (table at right). The pedigree of those individuals shows they were from the same family, suggesting inheritance of a single recombined chromosome present in one of the parents (parents were not sequenced).

Click here for additional data file.