Genome-wide selection reveals candidate genes associated with multiple teats in Hu sheep

Abstract

Increasing the number of teats in sheep helps to improve the survival rate of sheep lambs after birth. In order to analyze the candidate genes related to the formation of multiple teats in Hu sheep, the present study was conducted to investigate the genetic pattern of multiple teats in Hu sheep. In this study, based on genome-wide data from 157 Hu sheep, Fst, xp-EHH, Pi and iHS signaling were performed, and the top 5% signal regions of each analyzed result were annotated based on the Oar_v4.0 for sheep. The results show that a total of 142 SNP loci were selected. We found that PTPRG, TMEM117 and LRP1B genes were closely associated with polypodium formation in Hu sheep, in addition, among the candidate genes related to polypodium we found genes such as TMEM117, SLC25A21 and NCKAP5 related to milk traits. The present study screened out candidate genes for the formation of multiple teats at the genomic level in Hu sheep.

Keywords:

• Hu sheep
• multiple teats
• candidate genes
• selection signals

Introduction

Sheep is an important breed of livestock After thousands of years of domestication, artificial selection has led to the diversity of sheep populations, the Hu sheep is China’s famous Taihu breed, with high fertility, year-round estrus, its lactation function is excellent, and its nutritional value is worth studying, Chen et al. found that the Hu sheep’s milk contains a unique lactic acid bacterium, which has vertical transmission advantages.¹ Traditionally, the greater the number of teats in a mammal, the greater the number of children it produces. In sheep breeding, sheep udder traits are correlated with growth rates of weaned lambs.² Optimal udder and teat traits in ewes can significantly improve lamb survival and growth.³ The optimal udder and teat traits in ewes significantly improve lamb survival and growth.³ In sheep, polydactylous rams and ewes generally have 2–6 teats with a small anterior and large posterior distribution. Lactation performance studies have shown that some polydactyls can lactate normally.⁴ Some studies have found that some polydactyls can lactate normally.⁵ A study on removing multiple teats in sheep found that lactation yield was significantly reduced, but there was no difference in lactation yield between 2 and 4 teat.⁶ The results showed a significant reduction in milk production, but no difference in milk production between 2 and 4 teats.⁶ The localization of functional genes for multiple teats in sheep will facilitate the pace of scientific research on new lines of multiple teats in sheep and goats, as well as breeding selection. In sheep, teats are a highly heritable trait, and the development of extra teats is a complex multigenic trait.⁷ In pig production, teat number influences growth productivity.⁷ And teat number of pigs can be improved by selective breeding.⁸ This trait can be utilized to increase the number of teats in sheep to enhance the milk production performance of sheep.

Mammalian teats can be classified into functional and nonfunctional teats. Zhao et al. found that LHFP, DPYSL2 and TDP-43 genes were associated with the formation of teat number in Hu sheep through a GWAS study of HD chip data from populations with different teat counts.⁸ Peng et al. conducted GWAS analysis of HD SNP chip data from sheep and screened out BBX, CD47 and CASK genes associated with the formation of paramecium in sheep, and found that Wnt, Oxt ytcoin, MAPK and axon guidance pathway were important signaling pathways affecting paramecium formation in sheep.⁴ Laura et al. showed that sheep redundant teats have similar immune cells and similar structures such as lymphoid follicles as normal teats, but there is a significant increase in the length and throughput diameter of the redundant teats compared to normal teats.⁷ In a study of commercial sows, Audrey et al. found that elevated numbers of functional teats increased the number of weaned piglets.⁹ The genes associated with teat number are different in different animals. BOVO et al. found that NUDT and NAR6A1 genes were associated with teat number in African rabbits through GWAS study and Fst analysis of 50K chip data in African rabbits, which differed from the candidate genes associated with teat number in pigs.¹⁰ GWAS analysis of 60K SNP chip data from three pig species by Tang et al. found that the VRTN and KDM6B genes were closely related to teat number formation in pigs.¹¹ Yang et al. GWAS analysis of 50K chip data from pigs revealed that VRTN and ABCD4 genes influenced teat number formation in pigs.¹²

The extreme imbalance between high fertility and low lactation in Hu sheep results in lambs with light birth quality, low survival rate and growth retardation, which becomes a bottleneck restricting the breed’s ability to maximize farming profitability, and death in the form of lambs is the greatest loss to sheep farming enterprises. Currently, most studies on mammalian teat number use GWAS and other analytical methods. In this study, we analyzed the selection signals of HD chip data of 157 Hu sheep from a multiple teats population and a two-teat population in order to screen out candidate genes related to the formation of multiple teat number in Hu sheep through different analytical methods in anticipation of providing data references for the selection and breeding of Hu sheep.

Materials and methods

Experimental materials

In this study, genome-wide data of 157 Hu sheep from a normal group of 76 two-teaters (Figure 1) and a group of 81 with extra teats were used (Figure 2), which were obtained from https://figshare.com/s/197c20e3229490ccb5d6. Phenotypic data were obtained from https://figshare.com/s/0b326ab8632bbd1c82d6.⁸ The data set was filtered by the following criteria: (i) absence of SNPs for chromosomal and physical location; (ii) SNPs >0.1 for genotypic deletions; (iii) SNPs <0.01 for minor allele frequencies; (iv) individuals with genotyping rates <90%; (v) p-values for Fisher’s exact test Hardy–Weinberg equilibrium <0.00001.

Figure 1. Pictures of normal teats in Hu sheep. Pictures of Hu sheep with two teats.

Figure 2. Pictures of Hu sheep with extra teats. (a) The picture shows a photograph of three teats of a lake sheep. (b) The picture shows the four teats of a Hu sheep.

Test methods

Different biological analysis tools have different results, but no single method can completely detect all biological results in genome-wide selective scanning analysis.¹³ In order to capture genomic selection signals more accurately, multiple methods are often required.¹³ For this purpose, we chose four complementary statistical analysis tests: Fst (fixation index), xp-EHH (cross population extended haplotype homozygosity), iHS (integrated haplotype homozygosity score) and Pi (Population nucleic acid diversity). In this study, Fst, xp-EHH were analyzed between the two-teat and multi-teat populations, and Pi, iHS were analyzed separately for the multi-teat population.

F_st analysis

Calculation of F_st (Fixation Index) for two-teat and multi-teat groups using VCFTOOLS.¹⁴ It is a measure of genetic differences between populations and is used to determine the degree of genetic differentiation between populations; higher F_st values mean greater genetic differences between populations and vice versa. F_st analysis allows for the study of genetic diversity between populations or subspecies. We used VCFTOOLS for F_st analysis, set the sliding window to 50,000 bp and the step size to 12,500 bp. The formula for F_st analysis is: $$F_{\mathit{\text{st}}}=\frac{\mathit{\text{MSP}}-\mathit{\text{MSG}}}{\mathit{\text{MSP}}+(N_C-1)\mathit{\text{MSG}}}$$

MSG is the mean square of error for detected within-population loci, MSP is the mean squared deviation of loci between tested populations, and N_C denotes the corrected mean between overall sample sizes.

PI analysis

The multiple teat population of Hu sheep was analyzed for population nucleic acid diversity (Pi) using VCFTOOLS. We set the sliding window of the Pi analysis to 500 kb set the step size to 50 kb, and the SNP loci with 5% of the final Pi value were considered as significant candidates¹⁵ that were used to analyze polymorphism in the multiple teat groups in Hu sheep.

xpEHH analysis

We used rehh Hu sheep genomic data to do heterozygous extended haplotype purity tests (xpEHH),¹⁶ Rehh (R package for detecting recent positive selection using extended haplotype homozygosity) is a statistical tool for detecting selection signals based on extended haplotype homozygosity (EHH). It can be used to identify positive selection events occurring in genetic variants in the human genome.

The xpEHH statistic was calculated by comparing the two-teat group to the multiple-teat group with the formula: $$\mathit{\text{xpEHH}}=\frac{\mathit{\text{unxpEHH}}-\mathit{\text{mean}}\left(\mathit{\text{unxpEHH}}\right)}{\mathit{\text{sd}}\left(\mathit{\text{unxpEHH}}\right)}$$ where unxpEHH is set: $${\mathit{\text{unxpEHH}}}_{\mathit{\text{scores}}}=\mathit{\text{In}}\left(\frac{{\mathit{\text{iES}}}_{\mathit{\text{pop}}1}}{{\mathit{\text{iES}}}_{\mathit{\text{pop}}2}}\right)$$ iES_pop1 integrates EHH genetic distance statistics, and iES_pop2 integrates genetic distances for EHH statistics for multi-teat populations.

iHS analysis

Separate polypodium clusters for integrated haplotype purity scoring (iHS),¹⁷ a single marker locus was used to replace the core haplotypes in the EHH statistic, defining them as core loci. And their ratios were calculated to select the signal detection statistic used to complement the F_st analysis, calculated as: $$\mathit{\text{iHS}}=\frac{\mathit{\text{uniHS}}-\mathit{\text{mean}}\left(\mathit{\text{uniHS}}\right|p_s)}{\mathit{\text{sd}}\left(\mathit{\text{uniHS}}\right|p_s)}$$ where uniHS is set: $$\mathit{\text{uniHS}}=\mathit{\text{In}}\left(\frac{{\mathit{\text{iHH}}}_A}{{\mathit{\text{iHH}}}_D}\right)$$

IHH is the integration of genetic distances of EHH (integrated EHH), a represents the ancestral (progenitor) allele, and D represents the new mutant (derived) allele.

SNP annotation and analysis

For each scan the results were visualized using-R. We took the first 5% of SNP sites for F_st, xpEHH, and iHS values, and only the last 5% of SNP sites for Pi analysis were taken as positively selected sites to identify the selected regions in the genome, and all the obtained SNPs were annotated based on Sheep Oar_v4.0 (https://www.sheephapmap.org/). Genes were functionally analyzed with reference to the NCBI database (http://www.ncbi.nlm.nih.gov/gene), and visual intersection analysis was performed on individual analytical annotations using Draw Venn Diagram (Draw Venn Diagram (ugent.be)), using the DAVID tool (http://david.Abcc.ncifcrf.gov/) for gene ontology database (GO, http://geneontology.org)¹⁸ and Kyoto Encyclopedia of Genes and Genomes (KEGG, Kyoto Encyclopedia of Genes and Genomes)¹⁹ analysis. Related candidate genes for protein interactions analyzed on the STRING website (https://cn.string-db.org/).

Results

F_st analysis results

Multiple teats are a special phenotypic feature of Hu sheep compared to two teats. Multiple teats population favors the survival of lambs with multiple births. Using PLINK 19.0 software, F_st analysis was performed on the two-teat and multi-teat populations of Hu sheep, and the top 5% of F_st values. And the regions with F_st values >0.012 were considered as candidate regions for the multi-teat population and its significant loci were mainly distributed on chromosomes 3, 14, 2 and 8. A total of 10,278 loci were selected, and gene annotation was performed on the selected regions (Figure 3), resulting in 2088 genes.

Figure 3. Manhattan plot of F_st analysis. The different colors in the figure represent the different chromosomes 1–27, and the thin red line in the figure represents the value of F_st = 0.115419, which is the value of the top 5% of the F_st analysis.

Results of Pi analyses

Nucleic acid polymorphisms were analyzed in the multi-teat population of Hu sheep, and the top 5% of the Pi values, with a Pi value >0.000106168, were considered as the selected regions, and 5140 SNP loci were selected (Figure 4), and annotation analyses were performed on the selected regions, and 3,557 genes were obtained.

Figure 4. Manhattan plot of Pi analysis results. This figure shows the pi analyzed values for a multiple teats population of Hu sheep. Different colors in the figure represent different chromosomes and the thin red line represents the top 5% threshold line.

Results of xpEHH analyses

The xpEHH analyses were performed using VCFTOOLS on the multi-teat and two-teat populations of Hu sheep by comparing the extended haplotype purity differences between the two populations, with the two-teat population serving as the reference population and the multi-teat population serving as the test population for the selection signals analyses. As shown in Figure 5, the top 5% of the absolute value of the xpEHH analysis results were selected as the loci in the multiple teat population, and 1779 SNP loci were found in the multi-teat population, and 1779 genes were obtained after annotation analysis of the relevant loci.

Figure 5. Manhattan plot of xpEHH analysis results. Different chromosomes are represented by different colors, red line in the figure represent candidate regions selected for xpEHH analysis.

Results of iHS analyses

iHS detects strong selection signals in favor of alleles that have not yet reached fixation.¹⁷ In this study, iHS analysis was performed individually on the Hu sheep multiple teat population, as shown in Figure 6. The top 5% of the absolute value of the iHS results were considered as the region of selection, and 23,558 were the selected loci. Chromosomes 1, 7, 10, 20, and 21 showed peaks indicating loci on these chromosomes that were partially not yet fixed. Annotation analysis of the significant candidate loci yielded 3968 genes.

Figure 6. iHS Analysis of the manhattan diagram notes. The red line in the graph represents the top 5% threshold line.

Veen analysis and gene enrichment

The sheep 4.0 genome was used as a reference to annotate the candidate-selected regions analyzed by F_st, Pi, xpEHH and iHS, as shown in Figure 7(a), and 142 genes were overlapped (Table 1). For further functional analysis of the overlapped genes, we used DAVID online software to analyze the GO and KEGG pathways of the 142 genes, as shown in Figure 7(b,c).

Figure 7. Veen diagram and enrichment analysis. Figure 5(a) Shows the common genes annotated from the candidate SNP loci for the four analyses. The 0.05F_ST in blue in the figure represents the genes annotated for the top 5% of SNPs loci analyzed by F_st. The light red part 0.05 iHS analyzed for the first 5% of SNPs loci, 0.05 PI in green in the figure represents the genes annotated from the first 5% of SNPs loci analyzed by Pi, and 0.05 xpEHH in yellow in the figure represents the genes annotated from the first 5% of SNPs loci analyzed by xpEHH. (b) Shows the bubble plot of GO enrichment analysis of candidate genes. (c) Shows the bubble plot of KEGG enrichment analysis of candidate genes.

Table 1. Candidate gene correlation analysis values. The genes presented in the table are candidate genes common to the four selection signal analyses in the text.

Candidate gene analysis values
Fst	xpEHH	Pi	iHS	Chrom	Gene name
0.0125134	0.0969546	0.000137341	0.169807	10	FARP1
0.0141406	0.102782	0.000138787	0.155412	5	MYO9B
0.0136369	0.0971815	0.000129846	0.272623	3	ATXN10
0.0198374	0.0892936	0.000108544	0.141152	4	DNAH11
0.0140788	0.13531	0.000126995	0.096153	3	LONRF2
0.0180752	0.114316	0.000117083	0.201886	1	IL23R
0.0143596	0.0905919	0.000113971	0.223994	6	PPARGC1A
0.0128181	0.0957009	0.000112079	0.326865	2	ASTN2
0.0211912	0.166071	0.000116878	0.449407	9	CSMD3
0.0140285	0.0998928	0.000128737	0.203701	23	DLGAP1
0.0359107	0.106669	0.000111978	0.233394	2	LRP1B
0.017715	0.0904149	0.000111504	0.244635	1	NEGR1
0.0167893	0.0979947	0.000121609	0.238596	3	TRHDE
0.0250087	0.140692	0.000116135	0.287968	3	NELL2
0.028746	0.116311	0.000130173	0.210642	23	LAMA3
0.0300968	0.116607	0.000126524	0.115656	4	DPP6
0.0121405	0.125584	0.000124213	0.202213	25	GRID1
0.0159519	0.0855406	0.000106478	0.174637	3	COMMD1
0.0126131	0.0807562	0.00015858	0.211059	11	ABCA6
0.0198632	0.123574	0.000109622	0.0873149	19	PTPRG
0.0226703	0.106955	0.000132225	0.205564	1	FILIP1L
0.0127494	0.0833948	0.000114888	0.280595	9	KCNQ3
0.020226	0.0984031	0.000107705	0.292845	22	ATRNL1
0.0200469	0.105022	0.0001104	0.158133	8	UST
0.0202424	0.152509	0.000112369	0.210909	2	SH3GL2
0.0287782	0.15284	0.000108731	0.278738	1	DOCK7
0.0189957	0.138567	0.000127119	0.286006	1	ROR1
0.0247168	0.14545	0.000110012	0.179432	3	AFF3
0.0236198	0.0980806	0.00010693	0.228344	2	ERBB4
0.0208061	0.179315	0.000106989	0.107739	7	EPB41L4A
0.0133874	0.105651	0.000161158	0.000161158	22	PTPRE
0.0167454	0.0864918	0.00010626	0.133672	14	PRSS54
0.0144944	0.102314	0.000119282	0.254217	2	SPAG16
0.0159174	0.0857668	0.000118954	0.315966	2	ZRANB3
0.0247274	0.101588	0.000116345	0.215574	19	CNTN4
0.0234435	0.122773	0.00010872	0.19239	4	SND1
0.0203703	0.0994272	0.000108947	0.11489	12	ENAH
0.016176	0.0901188	0.000181069	0.17421	6	ANK2
0.0136707	0.0977661	0.000126762	0.184579	22	CPN1
0.0116329	0.0987518	0.000124808	0.280894	16	SLIT3
0.01365	0.0875218	0.000109289	0.211522	1	ROBO1
0.0153057	0.111301	0.000108326	0.202917	19	ITPR1
0.0176753	0.149455	0.000109367	0.150048	19	IL17RB
0.0127748	0.103659	0.000179457	0.152553	11	FTSJ3
0.0306481	0.0861801	0.000111007	0.24795	3	ITPR2
0.0488301	0.196328	0.000114648	0.494689	4	JAZF1
0.0119866	0.107459	0.000107881	0.101285	4	PPP1R3A
0.0161019	0.0924775	0.000114488	0.226897	5	EDIL3
0.0138765	0.145718	0.000151757	0.0999466	15	TCN1
0.0273828	0.229155	0.000112944	0.202219	1	KCNH8
0.0247217	0.132741	0.000110542	0.409258	3	DIP2B
0.0266611	0.111026	0.000117511	0.129541	9	COLEC10
0.0167149	0.108635	0.000122642	0.115138	6	TBC1D1
0.0187447	0.117845	0.000114142	0.21125	24	SHISA9
0.0118631	0.0846617	0.000119908	0.138795	11	MSI2
0.0514379	0.14704	0.000117351	0.166087	4	DYNC1I1
0.0142223	0.114591	0.000113487	0.0927003	11	TNRC6C
0.0184914	0.132091	0.000112948	0.247652	19	SYNPR
0.0157646	0.0994011	0.000116483	0.134906	1	ATP13A5
0.0304602	0.123526	0.000108335	0.305165	19	RBMS3
0.0131376	0.0970959	0.000125877	0.160061	21	ME3
0.0151188	0.138775	0.000111976	0.147204	26	ZMAT4
0.0187297	0.089567	0.000106193	0.198316	11	WNT9B
0.0155832	0.103169	0.000272938	0.456257	12	CENPF
0.017966	0.136204	0.000120394	0.124856	22	TACC2
0.0185585	0.116945	0.000114885	0.142182	4	IGF2BP3
0.016965	0.111521	0.000114206	0.161755	8	UTRN
0.0353733	0.131086	0.000127401	0.214695	1	AGBL4
0.0205176	0.126865	0.000108182	0.185674	24	RBFOX1
0.0377021	0.108956	0.000159571	0.143001	24	ZNF75A
0.0128484	0.119562	0.000107236	0.153584	3	TMEM117
0.0127333	0.089255	0.000107953	0.193532	11	WNT3
0.0247217	0.0925592	0.000110542	0.137656	3	ATF1
0.0125574	0.0873669	0.000129163	0.262501	18	SLC25A21
0.0182351	0.100485	0.000139822	0.107074	2	AKNA
0.0122888	0.080822	0.000131315	0.151024	21	NTM
0.0125437	0.0906824	0.00011277	0.176868	17	RIMBP2
0.0153819	0.134034	0.000126995	0.1129	3	CHST10
0.0193773	0.0987052	0.000126995	0.245499	19	FHIT
0.0184189	0.136008	0.000107743	0.130126	9	TOX
0.0182553	0.119871	0.000109367	0.15323	19	CHDH
0.0229908	0.130526	0.000115598	0.239359	22	PCDH15
0.0120501	0.0833012	0.000116726	0.350292	19	ERC2
0.0238652	0.0898589	0.000107236	0.16631	9	RIMS1
0.0171986	0.104021	0.000114169	0.173474	19	NEK10
0.0127695	0.212083	0.000159236	0.334631	12	PTPRC
0.0159637	0.104666	0.000108702	0.204044	8	SASH1
0.0226703	0.1191	0.000132225	0.205564	1	CMSS1
0.0204131	0.0989284	0.000144231	0.103201	21	GTF2H1
0.0126131	0.0845532	0.000172712	0.284135	11	ABCA10
0.014493	0.121616	0.000109893	0.0916305	6	LIAS
0.0307172	0.0871735	0.000114002	0.200442	4	POU6F2
0.0161839	0.100632	0.000118706	0.286596	21	OPCML
0.0291673	0.0810811	0.000118184	0.127225	2	ATF2
0.0143623	0.0855181	0.000119798	0.160472	2	ACVR1
0.0166242	0.110417	0.000130349	0.176624	21	SYTL2
0.0151202	0.116888	0.000109073	0.169802	13	CELF2
0.014949	0.101819	0.000108395	0.212098	5	SGCD
0.0286155	0.133724	0.000144605	0.227797	1	PHTF1
0.0126866	0.0829231	0.000111916	0.179042	4	ICA1
0.0201986	0.0844209	0.000128055	0.113099	22	LOXL4
0.0326285	0.138967	0.000121828	0.215216	11	NLRP1
0.0221557	0.119844	0.000122121	0.232353	7	NRXN3
0.0186338	0.0796293	0.000133901	0.135295	2	BSPRY
0.011987	0.123649	0.000117114	0.360086	13	MACROD2
0.020053	0.10266	0.000135996	0.109809	17	DCHS2
0.0158716	0.133275	0.000131231	0.165556	16	RAB3C
0.0336139	0.130163	0.000107625	0.294293	14	CDH13
0.0128935	0.108603	0.000119741	0.339256	1	DAB1
0.014642	0.0850342	0.000129501	0.2363	3	ALK
0.0266508	0.12241	0.000112674	0.120674	3	SLC2A3
0.0230233	0.123364	0.000144605	0.227797	1	MAGI3
0.0166619	0.115759	0.000115855	0.166924	4	VPS41
0.0245185	0.131403	0.000180443	0.106776	23	ZNF407
0.0234623	0.111376	0.000107665	0.300443	5	FSTL4
0.0223086	0.0802089	0.000108544	0.170743	4	CDCA7L
0.0185054	0.0893213	0.000145745	0.200327	25	WDFY4
0.0151169	0.116915	0.000111692	0.389165	6	COL25A1
0.0202696	0.114126	0.00011648	0.194774	2	THSD7B
0.0157284	0.0915644	0.000110284	0.160306	6	PGM2
0.0175023	0.143651	0.000115178	0.242156	13	SLCO4A1
0.0178244	0.095394	0.000123052	0.206526	22	SORCS1
0.0270416	0.122419	0.000116006	0.328892	21	DLG2
0.0144588	0.113309	0.000126162	0.484322	2	NCKAP5
0.0169597	0.143806	0.000106533	0.138327	2	SCARA5
0.0124323	0.112974	0.000124546	0.164256	16	LIFR
0.0166625	0.0927614	0.00011998	0.00011998	17	TTC28
0.0133908	0.107662	0.000107254	0.228132	2	GALNTL6
0.0136874	0.118105	0.000126995	0.152868	11	ASIC2
0.0144645	0.0978911	0.000137953	0.148899	15	LDLRAD3
0.0165791	0.0852525	0.000123718	0.232836	21	FAT3
0.0191899	0.110809	0.000125693	0.272377	4	CHRM2
0.0208294	0.141616	0.000166036	0.177876	25	CDH23
0.0129701	0.140614	0.000113007	0.230124	5	CHSY3
0.0231262	0.130589	0.000107305	0.234443	3	CTNNA2
0.0171521	0.139332	0.000113059	0.215997	2	PARD3B
0.0159097	0.091068	0.000120543	0.232896	11	CEP112
0.0208248	0.0861008	0.000118445	0.25845	11	PRKCA
0.0167998	0.097787	0.000153543	0.526849	16	DNAH5
0.0362153	0.167257	0.000116439	0.0972408	26	LRRC3B
0.0164443	0.124258	0.000112898	0.118093	9	FAM49B
0.0184089	0.0908601	0.000126343	0.225265	1	CFAP45

The bolded font in this table represents important candidate genes.

Peng et al. showed that paraplasia in sheep may be mediated by regulating cell proliferation, and that the selective signaling pathway of its occurrence is involved in breast cancer development.⁶ In this study, calcium binding pathway (GO:0005509) was shown and NELL2, CDH23, PCDH15, CDH13, SLIT3, FAT3, EDIL3, LRP1B, FSTL4 genes were shown. In order to screen out the candidate genes among the further candidate genes related to polydactyly in Hu sheep, we carried out gene interactions analyses on the related genes. As shown in Figure 8, the PTPRG, TMEM117 and LRP1B genes were shown to be associated with multiple genes.

Figure 8. Related candidate genes protein interaction diagram. The genes identified by the red boxes in the figure are thought to be strongly correlated with the formation of the polydactyl trait in Hu sheep.

Discussion

Detection of genomic regions under positive selection specific to a population is essential to reveal the genetic basis of variation in localized adaptive traits.²⁰ In this study, we genotyped genomic data between a multiple teat population and a two teats population of Hu sheep for F_st analysis. Fixation indices analysis is a measure of whether the actual frequency of genotypes in a population deviates from the theoretical proportion of genetic equilibrium (Harwin’s equilibrium). The genetic structure did not show a large genetic differentiation between the multi-teat population and the two-teat population of Hu sheep. In order to reduce the interfering factors of F_st analysis and increase the accuracy of selection signal detection. Screened the selected genes for the two populations of Hu sheep based on the theory of linkage disequilibrium as well as the extended haplotype purity test for the positive characteristics of the haplotypes. Analyzed the allele frequency analysis between the multi-teat and two-teat populations of Hu sheep using xp-EHH analysis as well as the pure haplotypes of multi-teat populations of Hu sheep for iHS analyses were complemented. The correlation analysis of its haplotypes revealed that the xpEHH analysis of its haplotypes revealed that the candidate xpEHH values ranged from 0.08 to 0.25 and the highest xpEHH value was found on chromosome 9. The same iHS analysis showed that the locus associated with multi-teat was fixed on chromosome 9. Its nucleotide diversity showed that loci with higher diversity were mainly located on chromosome 9. In addition, Pi analysis was used to screen the genomic selection signals of the multi-teat population of the Hu sheep. The highest Pi value was 0.014, which revealed that the subpopulation diversity of the polyposis population of the Hu sheep was not high. And based on the above analyses, we considered that the heritability of the multi-teat population of the Hu sheep was high.

In order to further identify the candidate genes related to multi-teat formation in Hu sheep, we annotated the overlapping regions of the four analysis methods, and among the 143 candidate genes screened for association with multi-teat. Based on the results of the gene interactions map we identified the mammary gland-related PTPRG,²¹ TMEM117²² and LRP1B genes.

Protein tyrosine phosphatase receptor type G (PTPRG) gene, located on sheep chromosome 19 at position NC_056072.1 (39155807.39930537, complement). This gene has been reported to inhibit mammary epithelial cell motility²³ and has been associated with growth traits in sheep.²⁴ In addition PTPRG has been identified as a tumor suppressor gene in breast cancer. The expression of PTPRG is reduced during breast carcinogenesis and is particularly low in high malignancy grade breast cancer.²⁵ PTPRG is involved in the control of FGFR1 activity and influences the sensitivity of sarcoma cells to FGFR kinase inhibitors, and is important for cytostasis in breast cancer.²⁶

The transmembrane protein 117 (TMEM117) gene, located on sheep chromosome 3 NC_056056.1 (142302614.142937834, complement), was reported to be associated with posterior udder attachment width in hesitant cows.²² Zhu et al.²⁷ also conducted a GWAS study and found that the TMEM117 gene was associated with saturated fatty acid composition in Simmental cows. Xia et al. find gene linked to beef traits²⁸ the gene has also been reported to be associated with fatty acid formation in Ethiopian Indigenous Goat. This gene has also been reported to be associated with fatty acid formation in Ethiopian Indigenous Goat.²⁹ The gene was also reported to be associated with fatty acid formation in Ethiopian Indigenous Goat.²⁹

LDL receptor related protein 1B (LRP1B) gene, located on chromosome 2 of sheep at position NC_056055.1 (168072439.170268670). was reported to reduce malignant tumor cells in breast cancer cell lines,³⁰ in a study by Peng et al. on sheep paramammary glands, LRP1B was shown in GWAS analysis and was the significant locus.⁶ This gene has also been reported to be associated with litter size in Greater Montenegrin sheep,³¹ LRP1B may contribute to birth weight variability in pigs³² and is involved in smooth muscle cell migration in chickens.³³

Extra teats in Hu sheep have similar immune cell and lymphatic structures as normal teat formation, and it has been shown that part of the extra teats in Hu sheep have the function of milk production, and the screening of individuals with extra teats in Hu sheep has certain benefits for lamb survival and production traits in Hu sheep populations.⁷ In this study, we analyzed the selection signals of excess teats in Hu sheep, hoping to provide a molecular basis for the production as well as screening of excess teats in Hu sheep through genetic aspects. In this study, we found that the genes PTPRG, TMEM117 and LRP1B were closely associated with the production of surplus teats in Hu sheep.

In addition, this study identified genes associated with production growth among the relevant candidate genes. Genes such as DNAH11,³⁴ NELL2,³⁵ ANK2,³⁶ SLIT3,³⁷ DIP2B,³⁸ NEK10,³⁹ GRID1,⁴⁰ and PHTF1⁴¹ associated with reproductive traits. And SH3GL2,⁴² DOCK7,⁴³ ENAH,⁴⁴ SHISA9,⁴⁵ AGBL4,⁴⁶ SGCD,⁴⁷ COL25A1,^48,49 DLG2,⁵ DOCK7,⁵⁰ and other genes associated with milk production traits. And TMEM117,²² SLC25A21,⁵¹ NCKAP5,⁵² PRKCA,⁵³ ITPR2,⁵⁴ NEGR1,⁵⁵ and other genes, and ASTN2,⁵⁶ GRID1,⁵⁷ FILIP1L,⁵⁸ ATRNL1,⁵⁹ and ERC2,⁶⁰ which are associated with stress tolerance.

Conclusions

In this study, candidate genes related to multi-teat in Hu sheep, such as TMEM117, PTPRG and LRP1B genes, were screened by selective signal analysis of HD chip data. These genes provide data for breed selection of Hu sheep, and also provide reference for the study of other multi-teat.

Author contributions

Conceptualization, W.Z., X.L. (Xiaopeng Li) and Z.H. (Zhipeng Han); methodology, X.L. (Xiaopeng Li), C.Z. (Cheng-long Zhang); software, X.L. (Xiaopeng Li), and R.Y. (Ruizhi Yang); validation, Z.H. (Zhipeng Han); formal analysis, W.Z.; investigation, Z.H. (Zhipeng Han) and L.Z. (Lulu Zhang) ; resources, W.Z.; data curation, W.Z.; writing—original draft preparation, W.Z.; writing—review and editing, W.Z., C.L. (Chunjie Liu); visualization, W.Z.; supervision, R.W. (Ruotong Wang) and J.W. (Jieru Wang); project administration, S.L. (Shudong Liu); All authors have read and agreed to the published version of the manuscript.

Acknowledgments

We are thankful to all lab members for their support during the study.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This study was funded by grants from the Natural Science Foundation of China [No: 32060743], Bingtuan Science and Technology Program [No: 2022CB001-09], Science and Technology Action for Industrial Development for Rural Revitalization in the Autonomous Region [No: 2022NC110] and Autonomous Region Agricultural Area High-efficiency Mutton Sheep Breeding and Promotion Technology System Project [No: xjnqry-g-2310].