Whole-Genome Resequencing Reveals Genetic Indels of Feathered-Leg Traits in Domestic Chickens

Journal of Genetics (2019) 98:47 © Indian Academy of Sciences https://doi.org/10.1007/s12041-019-1083-4

RESEARCH ARTICLE

Whole-genome resequencing reveals genetic indels of feathered-leg traits in domestic chickens

SHAOHUA YANG, ZHAOYUAN SHI, XIAOQIAN OU and GUOQING LIU∗

College of Food Science and Bioengineering, Hefei University of Technology, 193 Tunxi Road, Hefei 230009, Anhui, People’s Republic of China *For correspondence. E-mail: [email protected].

Received 22 November 2017; revised 28 August 2018; accepted 19 December 2018; published online 9 May 2019

Abstract. Whole-genome resequencing provides the opportunity to explore the genomic variations and pave way for further functional assays to map the economical trait loci. In this study, we sequenced the genomes of mixed chicken samples from a full-sib family, with feathered and unfeathered legs at an average effective depth of 4.43×, using Illumina Hiseq 2000 instruments. Over 2.1 million nonredundant short indels (1–71 bp) were obtained. Among them, 16,375 common indels that were polymorphic between the comparison groups were revealed for further analysis. The majority of the common differential indels (76.52%) were novel. Follow-up validation assays confirmed that 80% randomly selected indels represented true variations. The indels were annotated based on the chicken genome sequence assembly. As a result, 16,375 indels were found to be located within 2756 annotated genes, with only 33 (0.202%) located in exons. By integrated analysis of the 2756 genes with gene function and known quantitative trait loci, we identified a total of 24 promising candidate genes potentially affecting feathered-leg trait, i.e. FGF1, FGF4, FGF10, FGFR1, FRZB, WNT1, WNT3A, WNT11, PCDH1, PCDH10, PCDH19, SOX3, BMP2, NOTCH2, TGF-β2, DLX5, REPS2, SCN3B, TCF20, FGF3, FSTL1, WNT7B, ELOVL2 and FGF8. Our findings provide a basis for further study and reveal key genes for feathered-leg trait in chickens.

Keywords. whole-genome resequencing; indels; feathered-leg trait; chicken.

Introduction of genetic variants by controlling morphology, physiology and behaviour in chickens has helped us to understand the There are hundreds of domestic chicken breeds in the genetic basis of both simple and complex traits of chick- world (Eric and Michael 2017). As the first domesti- ens (Mou et al. 2011; Imsland et al. 2012; Wang et al. cated bird, the domestic chicken is considered variable for 2012). unravelling the molecular and genetic basis of phenotypic With the establishment of next-generation sequencing variations. Most importantly, chickens and their eggs have (NGS) technology, whole-genome sequencing makes it been the essential supply of animal protein for human pop- feasible to determine the potential of genomics, greatly ulations. In addition, physical appearance traits, such as facilitating detection of genomic variations. The availabil- feather-crested head, beard and feathered feet, are usually ity of massive data analyses and draft genome sequence used to distinguish different chicken breeds and also may provides us with various genomic variations accurately and be associated with some important economic traits. efficiently, such as a large numbers of single-nucleotide Identification of genetic and genomics variations over polymorphisms (SNPs), short insertions and deletions the entire genome is a fundamental task for pheno- (indels) and large structural variations (SVs) (Frazer et al. typic diversity and causal variant discovery. As mutations 2009; Zhan et al. 2011). Especially, a large number of these for biologically important traits, they can provide us an genetic variations based on NGS data were elucidated to exceptional opportunity to discover novel functions of be functionally associated with phenotypic diversity and specific genes (Johnsson et al. 2012). The identification important traits in many species (Kang et al. 2015). Among domestic chicken breeds, the Guangde chicken Shaohua Yang, Zhaoyuan Shi and Xiaoqian Ou contributed equally exhibits several phenotypic variations compared with to this work. other chicken breeds, especially its feathered legs and feet.

1 47 Page 2 of 8 Shaohua Yang et al.

This breed is remarkably popular in Anhui, because of converted into sequence data using Illumina CASAVA its fame for the unique feathered-leg feature in few local v1.8.2 and then subjected to the quality control (QC) chickens, which bring them better flying characteristics. A procedure using QC Toolkit v2.3 and PICARD (http:// recent study using genomewide association study (GWAS) picard.sourceforge.net/) to discard unusable reads (e.g. has directly reported that a nonsense mutation in FGF20 reads with adapters, reads with unknown bases and low causes a lack of almost all body feathers in the scale- quality reads). less chicken line (Wells et al. 2012). Despite the major breakthrough of identifying the scaleless mutation, the Read mapping underlying genetic mechanism of feathers only for the feathered-leg trait of chicken is still unknown. Hence, our The high-quality reads were mapped on to chicken refer- purpose here is to identify the trait-related gene and com- ence genome, together with annotation of genes, which was plete functional verification. Whole-genome resequencing retrieved from the Ensembl Genome Browser (http://www. is widely utilized for large-scale identification of molecu- ensembl.org/). The usable reads mapped to multiple chro- lar markers in mapping of economical trait loci. Further, mosomal positions and unmapped reads were removed. the data of high-throughput DNA sequencing provide the For consensus assembly, sequence reads were aligned opportunity to explore the genomic variations and pave using BWA software v0.5.8. FastQC v0.10.1 (http://www. way for further functional assays to clarify the regulatory bioinformatics.babraham.ac.uk/projects/fastqc/) was used mechanisms of related traits. for the quality filtration of these nucleotide bases.

Methods Variant detection and annotation

Animals and sampling The filtered reads were applied for identification of indels using SAMTools v0.1.18 by individually comparing the The mixed samples of genomic DNA were obtained from target genomes with the reference genome. The detected four feathered and four unfeathered chickens (male) owned SNPs and indels were filtered with stringent parameters by a standardized farm (Anhui Kirin Mountain Agricul- such as the minimum read depth of 4, the variant frequency ture, Xuancheng, China). Meanwhile, the two kinds of >90% and an average quality for the novel allele of >20. chickens were from a full-sib family to minimize differences Here, loci with homozygous and heterozygous genotypes in genetic conditions. According to the manufacturer’s in a separate sample but different from the reference base protocol, genomic DNA was isolated from whole blood were distinguished as SNPs. Additionally, BWAv0.5.8 was using a DP (318) Blood DNA kit (Tiangen Biotech, also used to estimate the read depth, which influences the Beijing, China). The quality and quantity of the puri- accuracy of SNP calling by filtering the candidate SNPs. fied DNAs were ascertained by gel electrophoresis and The chicken SNP database (http://www.ncbi.nlm.nih.gov/ NanoDrop (Thermo Fisher Scientific, Waltham, USA). projects/SNP) was used to reveal the putative SNPs we All procedures for animal handling were reviewed and detected. approved by the Institutional Animal Care and Use Com- The ANNOVAR tool was used to annotate variants. The mittee (IACUC) of the Hefei University of Technology online websites Ensembl (http://www.ensembl.org/) and (permit number: DK838). UCSC (https://genome.ucsc.edu/) were used for gene and region annotations. Additionally, the genomic position Whole-genome resequencing and distribution of SNPs and indels on each chromosome were visualized for further analysis using Circos software High-quality genomic DNAs were sampled for library (http://circos.ca/). Of which, the effects of synonymous preparation using the paired-end TruSeq DNA sample and nonsynonymous variants on genes were predicted by prep kit (Illumina, San Diego, USA), following the manu- using SnpEff v3.1h (http://snpeff.sourceforge.net/). Fur- facturer’s protocol. The genomic DNA was sheared to the ther, Chicken QTLdb (http://www.animalgenome.org/cgi- 200–400 bp range with a major peak at 350 bp, followed bin/QTLdb/GG/index) was used to locate candidate genes. by end repair, A-tailing and PE-adaptor ligation. Library quality was determined using a Bioanalyzer 2100 (Agilent SNP and indel validation Technologies, Santa Clara, USA). Sample concentrations were detected using a Qubit (Invitrogen, Carlsbad, USA), Polymerase chain reaction (PCR) was carried out to val- and DNA was diluted to a final concentration of 2 nM idate the presence of SNPs and small indels by Sanger for the sequencing step. Subsequently, PE sequencing was sequencing. Approximately 200–500 bp flanking sequences performed using Illumina HiSeq 2000 platform (Novogene of the selected SNPs were extracted according to the Bioinformatics, Beijing, China). To obtain high-quality reference genome sequence of the Gallus gallus and then reads for each chicken breed, the original image data were the primers were designed using PRIMER3 software Whole-genome resequencing for feathered-leg traits Page 3 of 8 47

(http://frodo.wi.mit.edu/primer3). The PCR amplification consistent with our resequencing results, indicating a high reactions were performed in a final reaction volume of accuracy of resequencing (table 2). 50 μL, containing 10 ng of genomic DNA, 20 pmol of each primer, 10 mM of dNTP mixture, 5 μLof10× PCR Indel identification buffer and 1 U Taq DNA polymerase enzyme (TaKaRa Biotechnology, Dalian, China). Reaction conditions were In total, over 2.1 million nonredundant short indels (1– as follows: initial denaturation at 94◦C for 10 min, fol- 71 bp) were obtained from two groups after filtering using lowedby32cyclesof30sat94◦C, annealing at 58–62◦C SAMtools. Among them, we focussed on the unique indels for 30 s, extension at 72◦C for 30 s and a final extension at that were polymorphic between the feathered and unfeath- 72◦C for 10 min. Then, 40 μL PCR product of each frag- ered chickens, which were inferred to be relevant to the ment were purified using DNA Purification kit (TaKaRa physical appearance trait. As a result, we obtained 16,375 Biotechnology, Dalian, China) and sequenced by using an common indels that were polymorphic between the com- ABI3730XL sequencer (Applied Biosystems, Foster City, parison groups. We further compared the results with USA). Both forward and reverse sequences were aligned the variants in the SNP database and found that the using DNAMAN software to determine the presence of majority of the common differential indels (76.52%) were genetic polymorphisms. novel.

Integrated analysis of resequencing, quantitative trait loci Functional annotation and genomic distribution (QTL) and known pathways Among the 16,375 common indels, the largest indel was We specifically focussed on those differential variations 71 bp, but most (94.33%) of them were less than 10 bp. between the feathered-leg and unfeathered-leg groups Single-bp indels (40.6%) were the most common, with located in the flanking regions and genic regions to identify nearly equal number of insertions and deletions. The dis- candidate genes and casual mutations related to morphol- tribution of indel length is shown in figure 1. The indels ogy traits. Further, we located the positions of genes con- were mainly located in autosomes, while the Z chromo- taining these variations using QTL mapper v.2.019 (www. some was found to have fewer indels (2.01 versus 97.99%, animalgenome.org/cgi-bin/QTLdb, updated on April 27, figure 2). 2017). In addition, we performed gene ontology and Kyoto Functional annotation of the 16,375 common differen- Encyclopedia of Genes and Genomes pathway analysis tial indels was performed using ANNOVAR (table 3). using the DAVID (Huang et al. 2008) and KOBAS tools As a result, the majority (40.373%) of indels were found (Xie et al. 2011), with the P value < 0.05 determined by to be located in the intergenic regions, and the remain- Fisher’s exact test and fold enrichment of more than 1.5 as ing indels were within 2756 genes, including in introns the criteria for significance. (56.048%), exons (0.202%), untranslated regions (UTRs) Results (0.274%) and other regions (ncRNAs, splicing sites etc.). Additionally, there were 17 frame-shift indels (six frame- Read mapping and coverage of reference chicken genome shift deletions and 11 frame-shift insertions), and two stop gain indels, which caused the loss of a stop codon. Genomic sequencing libraries were size-selected and paired-end sequencing was carried out with an Illumina HiSeq Integrated analysis with known QTLs 2000 instrument. For the purpose of quality check, FASTQ analysis was conducted to remove low-quality sequences Firstly, we compared the positions of the 1246 indels and an average of 32,100,132 high-quality sequencing located in 815 genes with QTLs known to be associ- reads were considered for further analysis (as shown in ated with feather growth traits in the QTL database. We table 1). We collected an average of 32,100,132 reads. found that 96 indels, corresponding to 42 unique genes, After mapping to chicken reference genome using BWA, overlapped with the confidence intervals of 32 known ∼97.85% mapped reads with an average coverage of QTLs. Of these 96 indels, 76 were notably located in 24 97.84% were obtained. Additionally, the average depth of functional genes that were close to the known QTLs with mapping was 4.43×. The alignments between the unique a genetic distance of less than 5 cM. Hence, these 24 individual and the reference genome were applied for clas- functional genes in these two groups were retained for sifying the genetic and genomic variations detected in this further analyses. study. To estimate the accuracy and reliability of resequencing Identification of candidate genes results, PCR and Sanger sequencing were applied to validate 10 randomly selected SNPs of the eight samples. We Based on the known QTLs and the biological functions found that the sequences of the eight SNPs were absolutely of genes, eight genes (REPS2, SCN3B, TCF20, FGF3, 47 Page 4 of 8 Shaohua Yang et al.

Table 1. Summary of sequencing, mapping statistics, and indel count for individuals.

Sample Raw reads Mapped reads Mapped rate (%) Genome coverage (%) Sequencing depth (×) Indel

Feathered 33,719,832 32,997,174 97.86 99.48 4.54 220,632 Unfeathered 31,901,392 31,203,090 97.81 99.40 4.31 214,735

Table 2. The results of sequencing for the eight SNPs.

Chromosome Position Ref. Indel called Illumina Sanger Type

1 193738514 C CT Yes Yes Hom 1 193738652 A AT Yes No Hom 1 194701472 AGAGGGG AG Yes Yes Hom 10 8174215 T TTC Yes Yes Het 10 8789116 A AA Yes Yes Hom 12 13935462 GT G Yes Yes Het 12 19458339 AGC A Yes No Het 17 2485327 GG G Yes Yes Het 17 6630776 AA A Yes Yes Hom 17 8375212 AT A Yes Yes Het

Hom, homozygous; het, heterozygous.

Figure 1. The distribution of indel length (bp). The indel represents the common differential one between the comparison groups.

FSTL1, WNT7B, ELOVL2 and FGF8) were considered as Discussion the novel and promising candidates for an association with feathered-leg trait. The indel genotypes in these eight genes With the rapid development of bioinformatics tools and are shown in table 4. In addition, the remaining 16 known sequencing technologies, it is now feasible to sequence genes (FGF1, FGF4, FGF10, FGFR1, FRZB, WNT1, the targeted regions of many species in the genome WNT3A, WNT11, PCDH1, PCDH10, PCDH19, SOX3, more economically and efﬁciently (Mamanova et al. 2010; BMP2, NOTCH2, TGF-β2 and DLX5) affecting feather Mertes et al. 2011). The NGS have been used suc- bud formation and feather development were also revealed. cessfully to detect causal variants of human disease in Whole-genome resequencing for feathered-leg traits Page 5 of 8 47

Figure 2. Different indels between the comparison groups. Indels derived from difference between the comparison groups are shown in blue (deletion), red (duplication) and green (both).

some cases (Ku et al. 2011; Gilissen et al. 2012) and 16,375 common indels that were polymorphic between also have been useful for mechanism research in ani- the comparison groups. To the best of our knowledge, mals (Frazer et al. 2007; Koboldt et al. 2013). Compared this is the first study involving whole-genome resequenc- with targeted sequencing of only specific regions, whole- ing for feathered-leg trait in domestic chicken breeds. genome sequencing can also detect SVs, giving us a Based on these indels, we identified a few candidate more comprehensive view of genetic variations in the genes potentially associated with the physical appearance genome. Moreover, it can identify much more numer- traits. ous rare variants of complex traits than GWAS (Metzker Although we mixed the samples for sequencing, we 2010). Applying NGS, we obtained the first draft genome could still identify some important genes coincide with sequences of Guangde chicken and identified a total of previous studies. For instance, 16 known genes (FGF1, 47 Page 6 of 8 Shaohua Yang et al.

Table 3. Statistics of indels in the genomes of the selected indi- Indels in a gene coding region, especially in frame-shift, viduals. played important roles in gene function (Ng et al. 2008; Hu and Ng 2012). Among the eight candidate genes, the a Category Indel Indel (%) REPS2 gene (also known as POB1) with one frame-shift indel in exon 8 encodes a protein complex that regulates All 16,375 Intergenic 6611 40.373 the endocytosis of growth factor receptors and plays a key Upstream 198 1.210 role in endocytic membrane transport events and actin Downstream 259 1.582 dynamics (Royle 2013). Additionally, the expression of Upstream; downstreamb 45 0.275 the REPS2 gene can negatively affect receptor internal- UTR5 13 0.079 ization and inhibit growth factor signalling (Oosterhoff . UTR3 32 0 195 et al. 2003). The REPS2 gene might be involved in feather Splicing 4 0.024 ncRNA_intronic 2 0.012 bud formation through a negative regulatory mechanism. Intronic 9178 56.048 SCN3B with one exonic indel (exon 11) and one intronic Exonic 0.202 indel is a component of complexes composed of a large Nonframe-shift deletion 5 alpha subunit and some regulatory beta subunits and is Nonframe-shift insertion 9 responsible for the cell growth regulation and differentia- Frame-shift deletion 6 Frame-shift insertion 11 tion of action potentials in neurons and muscle (Li et al. Stop gain 2 2013). Thus, SCN3B may play a vital role in feather formation and development. aPercentage was calculated based on total annotated indels. TCF20 with one indel in the 3-UTR was found to be b Variant located in both upstream and downstream regions. a transcriptional coactivator which enhances the activ- ity of transcription factors such as JUN, SP1, PAX6 and ETS1, and regulates the transcription of genes related FGF4, FGF10, FGFR1, FRZB, WNT1, WNT3A, WNT11, to cell differentiation and proliferation (Schäfgen et al. PCDH1, PCDH10, PCDH19, SOX3, BMP2, NOTCH2, 2016). Meanwhile, the mutations in this gene were asso- TGF-β2 and DLX5) affecting feather bud formation and ciated with autism spectrum disorders (Darvekar et al. feather development revealed in this study were similar to 2013). Competitive signalling of feather promoting and those previous ﬁndings (Yue et al. 2006, 2012; Rishikaysh inhibitory factors can prompt the establishment of the et al. 2014). In addition, eight genes (REPS2, SCN3B, periodic pattern of the feather tract. Here, we speculated TCF20, FGF3, FSTL1, WNT7B, ELOVL2 and FGF8) that the TCF20 transcription factors may be a candidate were considered as the novel and promising candidates to molecule upstream of the described feather patterning sig- be associated with feathered-leg trait. nals.

Table 4. The indel genotypes in eight genes between the feathered and unfeathered groups.

Indel genotypes in the feathered group Indel genotypes in the unfeathered group Gene name Indel Location Indel sequence Sample 1 Sample 2 Sample 3 Sample 4 Sample 5 Sample 6 Sample 7 Sample 8

REPS2 5Nins Intron GTCTT del/del del/del del/del del/del ins/ins ins/ins ins/ins ins/ins 5Ndel Intron ACCAC ins/del ins/del ins/del ins/del del/del del/del del/del ins/del 2Ndel Intron GA ins/ins ins/ins ins/ins ins/ins del/del del/del del/del ins/del 2Ndel Intron CC ins/del ins/del ins/del ins/del ins/ins ins/ins ins/ins ins/ins 4Nins Exon8 AAGT del/del del/del del/del del/del ins/del ins/del ins/del ins/del SCN3B 2Ndel Intron CA ins/ins ins/ins ins/ins ins/ins del/del del/del del/del del/del 6Nins Exon11 TCTCCA ins/del ins/ins ins/ins ins/ins ins/del ins/del ins/del ins/del TCF20 2Nins 3-UTR AA del/del ins/del del/del del/del ins/del ins/ins ins/ins ins/ins FGF3 1Nins Intron A ins/ins ins/ins ins/ins ins/ins del/del del/del del/del del/del FSTL1 6Ndel Intron TATACA ins/ins ins/ins ins/del ins/ins ins/del ins/del ins/del ins/del 2Nins Intron AA del/del del/del del/del del/del ins/ins ins/ins ins/ins ins/ins 1Nins Intron C del/del ins/del del/del del/del ins/ins ins/del ins/del ins/del 11Ndel Intron GCTGCAGGCAG ins/ins ins/ins ins/ins ins/ins ins/del ins/del ins/del ins/del WNT7B 1Ndel Intron T ins/ins ins/ins ins/del ins/ins ins/del del/del del/del del/del 2Nins Intron CT ins/del ins/del del/del ins/del ins/ins ins/ins ins/ins ins/ins ELOVL2 5Nins Intron TCACT ins/del ins/del ins/del ins/del ins/ins ins/ins ins/ins ins/ins FGF8 6Nins Intron GTGTGT del/del del/del del/del del/del ins/ins ins/ins ins/ins ins/ins 3Ndel Intron GGA ins/ins ins/ins ins/ins ins/del del/del del/del del/del del/del

N, nucleotide; ins, insertion; del, deletion. Whole-genome resequencing for feathered-leg traits Page 7 of 8 47

FGF3 with an intronic indel and FGF8 with two intronic 8.27 million SNPs in inbred mouse strains. Nature 448, 1050– indels are members of the fibroblast growth factor (FGF) 1053. family, which serve as positive feather promoting sig- Frazer K. A., Murray S. S., Schork N. J. and Topol E. J. 2009 Human genetic variation and its contribution to complex nals during the early stages of bud development and are traits. Nat. Rev. Genet. 10, 241–251. involved in a variety of biological processes including Gilissen C., Hoischen A., Han G. B. and Veltman J. A. 2012 embryonic development, morphogenesis, cell growth and Disease gene identification strategies for exome sequencing. tissue repair (Bei and Maas 1998). FGF receptors are Eur. J. Hum. Genet. 20, 490–497. expressed in the ectoderm and dermis before feather bud Hu J. and Ng P. C. 2012 Predicting the effects of frameshifting indels. Genome Biol. 13, R9. initiation in a diffuse manner. Tao et al. (2002) identified Huang D. W., Sherman B. T. and Lempicki R. A. 2008 System- that FGF signalling could promote feather development. atic and integrative analysis of large gene lists using DAVID Thus, FGF3 and FGF8 were considered as major genes of bioinformatics resources. Nat. Protoc. 4, 44–57. affecting feather development. Imsland F., Feng C., Boije H., Bed’hom B., Fillon V., Dorshorst FSTL1 with four indels in intron is a member of B. et al. 2012 The rose-comb mutation in chickens constitutes a structural rearrangement causing both altered comb morphol- the follistatin-like proteins and played an important role ogy and defective sperm motility. PLoS Genet. 8, e1002775. in inhibiting myoblast proliferation and differentiation Johnsson M., Gustafson I., Rubin C. J., Sahlqvist A. S., Jonsson (Tania et al. 2017). Van den Berg et al. (2010) showed that K. B., Kerje S. et al. 2012 A sexual ornament in chickens is FSTL1 locally antagonized the action of the BMPs and affected by pleiotropic alleles at HAO1 and BMP2, selected was involved in the development of ectopic feather buds. during domestication. PLoS Genet. 8, e1002914. Kang H., Wang H., Fan Z., Zhao P., Khan A., Yin Z. et al. 2015 WNT7B with two intronic indels also located within a QTL Resequencing diverse Chinese indigenous breeds to enrich the region on chromosome 1 linked to feather bud traits and its map of genomic variations in swine. Genomics 106, 286–294. biological function and genetic effects revealed that it was Koboldt D. C., Steinberg K. M., Larson D. E., Wilson R. K. and a key candidate gene associated with such traits (Arens- Mardis E. R. 2013 The next-generation sequencing revolution man et al. 2014). ELOVL2 with one intronic indel encodes and its impact on genomics. Cell 155, 27–38. Ku C. S., Naidoo N. and Pawitan Y. 2011 Revisiting Mendelian for a transmembrane protein involved in the synthesis of disorders through exome sequencing. Hum. Genet. 129, 351– long polyunsaturated fatty acids. To date, ELOVL2 has 370. not been reported to be associated with feather devel- Li R. G., Wang Q., Xu Y. J., Zhang M., Qu X. K., Liu X. et al. opment. Therefore, it might be a major gene by other 2013 Mutations of the SCN4B-encoded sodium channel β 4 signalling pathways. Although we identified eight can- subunit in familial atrial fibrillation. Int. J. Mol. Med. 32, 144– 150. didate genes based on the indels in the genic and gene Mamanova L., Coffey A. J., Scott C. E., Kozarewa I., Turner flanking regions that were polymorphic between the feath- E. H., Kumar A. et al. 2010 Target-enrichment strategies for ered and unfeathered groups, further association analysis next-generation sequencing. Nat. Methods 7, 111–118. should be performed to validate functionally important Mertes F., Elsharawy A., Sauer S., van Helvoort J. M., van indels and to characterize more important molecular der Zaag P. J., Franke A. et al. 2011 Targeted enrichment of genomic DNA regions for next-generation sequencing. Brief markers of feather bud formation and development. Funct. Genomics 10, 374–386. Metzker M. L. 2010 Sequencing technologies-the next genera- Acknowledgements tion. Nat. Rev. Genet. 11, 31–46. Mou C., Pitel F., Gourichon D., Vignoles F., Tzika A., Tato P. et al. 2011 Cryptic patterning of avian skin confers a devel- This work was supported by the National Key R&D Program opmental facility for loss of neck feathering. PLoS Biol. 9, of China (2018YFD0502100) and National Science Foundation e1001028. Project of Anhui (1908085MC63). Ng P. C., Levy S., Huang J., Stockwell T. B., Walenz B. P., Li K. et al. 2008 Genetic variation in an individual human exome. PLoS Genet. 4, e1000160. References Oosterhoff J. K., Penninkhof F., Brinkmann A. O., Anton G. J. and Blok L. J. 2003 REPS2/POB1 is downregulated during Arensman M. D., Kovochich A. N., Kulikauskas R. M., Lay A. human prostate cancer progression and inhibits growth factor R., Yang P. T., Li X. et al. 2014 WNT7B mediates autocrine signalling in prostate cancer cells. Oncogene 22, 2920–2925. Wnt/β-catenin signaling and anchorage-independent growth Rishikaysh P., Dev K., Diaz D., Qureshi W. M., Stanislav F. and in pancreatic adenocarcinoma. Oncogene 33, 899–908. Mokry J. 2014 Signaling involved in hair follicle morphogen- Bei M. and Maas R. 1998 FGFs and BMP4 induce both Msx1- esis and development. Int. J. Mol. Sci. 15, 1647–1670. independent and Msx1-dependent signaling pathways in early Royle S. J. 2013 Protein adaptation: mitotic functions for mem- tooth development. Development 125, 4325–4333. brane trafficking proteins. Nat.Rev.Mol.CellBiol.14, 592– Darvekar S., Rekdal C., Johansen T. and Sjøttem E. 2013 A phy- 599. logenetic study of SPBP and RAI1: evolutionary conservation Schäfgen J., Cremer K., Becker J., Wieland T., Zink A. M., Kim of chromatin binding modules. PLoS One 8, e78907. S. et al. 2016 De novo nonsense and frameshift variants of Eric T. D. and Michael D. S. 2017 Pigeonetics takes flight: evolu- TCF20 in individuals with intellectual disability and postnatal tion, development, and genetics of intraspecific variation. Dev. overgrowth. Eur. J. Hum. Genet. 24, 1739–1745. Biol. 427, 241–250. TaniaN.P.,MaarsinghH.,TBosI.S.,MattiottiA.,Prakash Frazer K. A., Eskin E., Kang H. M., Bogue M. A., Hinds D. A., S., Timens W. et al. 2017 Endothelial follistatin-like-1 reg- Beilharz E. J. et al. 2007 A sequence-based variation map of ulates the postnatal development of the pulmonary vascu- 47 Page 8 of 8 Shaohua Yang et al.

lature by modulating BMP/Smad signaling. Pulm. Circ. 7, DNA reveals nonsense mutation in FGF20 in the scaleless line 219–231. of featherless chickens. BMC Genomics 13, 257. Tao H., Yoshimoto Y., Yoshioka H., Nohno T., Noji S. and Xie C., Mao X., Huang J., Ding Y., Wu J., Dong S. et al. 2011 Ohuchi H. 2002 FGF10 is a mesenchymally derived stimu- KOBAS 2.0: a web server for annotation and identiﬁcation of lator for epidermal development in the chick embryonic skin. enriched pathways and diseases. Nucleic Acids Res. 39, W316– Mech. Dev. 116, 39–49. W322. Van den Berg G., Somi S., Bufﬁng A. A., Moorman A. F. Yue Z., Jiang T. X., Widelitz R. B. and Chuong C. M. 2006 and van den Hoff M. J. 2010 Patterns of expression of Wnt3a gradient converts radial to bilateral feather symmetry the follistatin and follistatin-like 1 genes during chicken via topological arrangement of epithelia. Proc. Natl. Acad. Sci. heart development: a potential role in valvulogenesis and USA 103, 951–955. late heart muscle cell formation. Anat. Rec. (Hoboken) 290, Yue Z., Jiang T.X., Wu P., Widelitz R. B. and Chuong C. M. 2012 783–787. Sprouty/FGF signaling regulates the proximal-distal feather Wang Y., Gao Y., Imsland F., Gu X., Feng C., Liu R. et morphology and the size of dermal papillae. Dev. Biol. 372, al. 2012 The crest phenotype in chicken is associated with 45–54. ectopic expression of HOXC8 in cranial skin. PLoS One 7, Zhan B., Fadista J., Thomsen B., Hedegaard J., Panitz F. and e34012. Bendixen C. 2011 Global assessment of genomic variation in Wells K. L., Hadad Y., Ben-Avraham D., Hillel J., Cahaner A. cattle by genome resequencing and high-throughput genotyp- and Headon D. J. 2012 Genome-wide SNP scan of pooled ing. BMC Genomics 12, 557.

Corresponding editor: N. G. Prasad