TONG Shi-feng ,ZHU Mo ,XIE Rui ,LI Dong-feng ,ZHANG Li-fanLIU Yang

1 Department of Animal Genetics,Breeding and Reproduction,College of Animal Science and Technology,Nanjing Agricultural University,Nanjing 210095,P.R.China

2 Key Laboratory of Urban Agriculture,Ministry of Agriculture and Rural Affairs,Shanghai 200240,P.R.China

Abstract Erhualian (E),Meishan (MS) and Mi (MI) pigs are excellent indigenous pig breeds in Chinese Taihu Basin,which have made great contributions to the genetic improvement of commercial pigs. Investigation of the genetic structure and inbreeding level of the 3 pig breeds is of great significance for the sustainable breeding of commercial pigs. The length and number of runs of homozygosity (ROH) as well as the frequency of genomes covered by ROH can be used as indicators to evaluate the level of inbreeding and the origin of the population. In this study,the ROH characteristics of E,MS,MI and Landrace (L) pigs were analyzed by SLAF-seq data,and the inbreeding coefficient based on ROH (FROH) was calculated. In addition,we have identified candidate genes in the genomic regions associated with ROH. A total of 10 568 ROH were detected in 116 individuals of 4 pig breeds. The analysis showed that there were significant differences in genetic structure between 3 Taihu Basin pig breeds and L,and the genetic structure of E and MI was similar. The results of FROH showed that the inbreeding level of MS was the highest (0.25±0.07),while E and MI were lower than L. Compared with the other 3 pig populations,MS showed a higher frequency of long ROH (＞5 Mb),indicating higher inbreeding in MS in recent generations. A large number of candidate genes related to reproductive traits are located in the genomic regions with a high frequency of ROH,and these genes are expected to be used as candidate genes in marker-assisted selection (MAS) breeding programs. Our findings can provide theoretical support for genetic conservation and genetic improvement of 3 pig breeds in Chinese Taihu Basin.

Keywords: runs of homozygosity,inbreeding coefficient,pig,candidate gene

1.Introduction

Taihu pigs are distributed around Taihu Lake region in the lower Yangtze River valley of China. They are known as their high prolificacy,which have made an important contribution to the genetic improvement of commercial pigs (Bosseet al.2014). Due to their different performances and traits (Wanget al.2015),Taihu pigs are mainly classified into 7 breeds,such as Erhualian pigs (E),Meishan pigs (MS),Mi pigs (MI) and so on.Protecting the genetic diversity of Taihu pig breeds and avoiding inbreeding are beneficial to their sustainable breeding.Therefore,it is of great significance to investigate the genetic diversity of Taihu pigs and strengthen the monitoring of inbreeding degree of Taihu pigs. However,there are few studies to analyze the genetic structure and inbreeding level of the existing Taihu pig population by genome-wide detection for runs of homozygosity (ROH).

ROH is a continuous homozygous segment of DNA sequence which are the same ancestral identical haplotypes inherited from the parents. The length,number and frequency of ROH can be used as indicators to evaluate the level of inbreeding and the origin of the population (Ceballoset al.2018). The calculation of population inbreeding coefficient based on ROH (FROH)provides another criterion for evaluating the inbreeding level of a population,and it is considered as one of the most accurate estimates of the level of autozygosity of an animal genome (Peripolliet al.2017). The traditional inbreeding coefficient is calculated based on pedigree information (FPED),and the accuracy of estimating FPEDdepends to a large extent on the integrity and accuracy of pedigree data (Wanget al.2013). However,due to incomplete and incorrect records of pedigree data in reality,the accuracy of FPEDis reduced. FROHcan circumvent the defects of FPEDfor routine monitoring of inbreeding,especially in indigenous pig breeds (Schiavoet al.2021). Genome-wide detection of ROH has been widely used to study the genetic structure,inbreeding level and evolution process of different livestock populations (Britoet al.2017;Liuet al.2021;Schiavoet al.2021). And genomic regions with high ROH frequency can be used to identify the candidate genes with economically important traits (Xieet al.2019;Shiet al.2020).

High-throughput single nucleotide polymorphism (SNP)genotyping technology provides a new opportunity to estimate inbreeding level using genome-wide information directly (Schiavoet al.2021). Based on ROH,evaluating inbreeding level is the most accurate method that uses genome-wide data to estimate inbreeding level (Ceballoset al.2018). Specific-locus amplified fragment sequencing(SLAF-seq) is a simplified genome sequencing technique for efficient SNP identification and large-scale genotyping.Compared with the high-cost whole-genome sequencing technology,SLAF-seq can effectively reduce the cost of sequencing,and it has the advantages of high accuracy,no need to refer to genome sequence and polymorphism information,it also can avoid repetitive sequences (Sunet al.2013). SLAF-seq has been widely used in the whole genome research of animals and plants (Liet al.2017,2018;Zhanget al.2020).

With the import of a large number of western commercial pig breeds into China,the population size of Chinese indigenous pig breeds have decreased,and the reduction of population size have led to an increase in the probability of inbreeding. Therefore,we hypothesized that with the reduction of Taihu pig breeds population size,the inbreeding level of Taihu pig breeds population should be increased. In this study,SLAFseq data were used to analyze the ROH patterns of 4 pig breeds,including 3 native pig breeds (E,MI and MS)from Chinese Taihu basin and Landrace pigs (L). And the genetic structure,genetic relationship and inbreeding level of 4 pig breeds were compared and analyzed. The results may provide valuable information for maintaining the genetic diversity of 3 indigenous pig breeds in Chinese Taihu basin.

2.Materials and methods

2.1.SNP genotyping and quality control

A total of 107 individuals from 4 pig breeds were included in our study: 28 E,23 MI,27 MS and 29 L. Genomic DNA was extracted from ear tissue and genotyped with SLAFseq method by Illumina HiSeq 2500 System (Illumina,San Diego,CA,USA). Demultiplexed reads were filtered using the Trimmomatic Software version 0.39. Pairedend reads were mapped to theSus scrofa11.1 using BWA 0.7.17 Software. SNP calling was performed by both GATK 4.1.8.1 and samtools 1.10 analysis. The biallelic SNPs that passed the following filters were used for maximum missing data of 30%. We only focused on SNPs on the autosomes for further analyses. The software PLINK (v1.90) (Purcellet al.2007) was used for quality control of the data and the following standards were set: (i) Removal of SNP loci with a call rate of less than 0.95;(ii) removal of SNP loci with a minor allele frequency (MAF) of less than 0.05;and (iii) discarding of individuals with a call rate of less than 0.90.

2.2.Population structure

Principal component analysis (PCA) was performed using PLINK1.9 (Purcellet al.2007),and R (v4.0.2) was used to generate the PCA figure and the 3D PCA picture.

2.3.Runs of homozygosity detection

ROH was detected with the detectRUNS package of R Software (www.r-project.org). We defined ROH according to the following criteria: (i) The minimum number of SNPs in a sliding window was 50;(ii) one heterozygous genotype and no more than two missing SNPs were allowed per window;(iii) the minimum ROH length was set to 1 Mb to eliminate the impact of strong linkage disequilibrium (LD);(iv) the minimum SNP density was 1 SNP per 500 kb and the maximum gap between consecutive SNPs was set to 1 Mb to avoid affecting the length of ROH with a low SNP density;and (v) to minimize the number of the false-positive ROH,the minimum number of SNPs that constituted the ROH (l)was calculated with the method proposed by Lenczet al.(2007),l=,whereαis the percentage of falsepositive ROH (set to 0.05 in the present study),nsis the number of SNPs per individual,niis the number of individuals andhetis the proportion of heterozygosity across all SNPs. After calculation,the minimum number of SNPs constituting an ROH was set to 80,93,77,80 for E,L,MI,and MS,respectively.

In this study,we classified ROH into 5 different categories according to their physical length: 1-5 Mb,5-10 Mb,10-20 Mb,20-40 Mb,and ＞40 Mb. For each length category,we computed the frequency of ROH numbers and the average length of an ROH per breed.

2.4.lnbreeding coefficient

To evaluate the inbreeding degree of different pig breeds,the genomic inbreeding coefficients FROHfor different pig breeds were calculated based on ROH. The inbreeding coefficient based on the proportion of autosomes covered in runs of homozygosity per individual (FROH) was determined. FROHwas calculated as follows: FROH=LROH/LAUTO,where LROHis the total length of ROH on autosomes and LAUTOis the total length of the autosomes covered by SNPs.

2.5.Detection of common runs of homozygosity

To identify genomic regions with a high frequency of ROH,we calculated the percentage of the occurrence of SNPs in ROH by counting the times of a SNP detected in a particular ROH across individuals and selected the top 1% of SNPs observed in an ROH. The SNP regions showed a percentage higher than 35.71,55.17,34.78,and 55.56% as genomic regions with a high frequency of ROH for subsequent analyses for E,L,MI,and MS,respectively. The function of these genes was annotated at ensemble biomart (http://asia.ensembl.org/biomart/martview/244b07db6f169a19f1e0362778df6ab5).Moreover,we conducted an extensive literature search.

3.Results

3.1.Population stratification assessment

As seen in Fig.1,E,MI,MS,and L were separated by PC1 (Fig.1-A) and all breeds were separated by 3 PCs(Fig.1-B).

3.2.Distribution of runs of homozygosity

The sequencing procedure yielded the average of 4 million sequencing paired-end reads for Chinese indigenous pig breeds and 9 million sequencing pairedend reads for L. After the read filtering procedures,the average of 3.5 million paired-end reads for Chinese indigenous pig breeds and 8 million paired-end reads for L were mapped to theSus scrofa11.1 genome. After filtering,144 180,260 685,154 177,and 160 594 SNPs and 28,29,23,and 27 individuals were respectively retained from the E,L,MI,and MS. A total of 10 568 ROH were identified in 107 individuals. Among all identified ROH,the lengths of 8 118 ROH were shorter than 5 Mb,while those of 1 525 ROH ranged from 5 to 10 Mb,those of 660 ROH from 10 to 20 Mb,those of 203 ROH from 20 to 40 Mb and 62 ROH were longer than 40 Mb (Table 1).

The average number and length of ROH in 4 pig breeds were summarized in Table 2. Among all ROH,the longest ROH was 107.86 Mb,which consisted of 2 815 SNPs and presented on chromosome 1 in MS. The individual in which the largest number of ROH (202 ROH)was detected in L and the individual in which the lowest number of ROH (33 ROH) was detected in MI. The pig breed that had the longest average ROH length was MS ((564.92±167.40) Mb);the pig breed with the lowest average ROH length was E ((310.45±120.79) Mb).

The frequency of ROH numbers within the 5 categories of ROH length (1-5 Mb,5-10 Mb,10-20 Mb,20-40 Mb,and ＞40 Mb) is illustrated (Fig.2). L exhibited a higher frequency of ROH than other 3 pig populations in 1-5 Mb but the highest frequency of ROH was found in MS in the 5-10 Mb,10-20 Mb,and 20-40 Mb categories.

The relationship between the total genomic length covered by ROH per individual and the total number of ROH per individual is plotted in Fig.3. L and MS exhibited a larger number of ROH than E and MI. L and MS also presented some extreme individuals with a length of ROH that covered more than 750 Mb.

3.3.lnbreeding coefficient of runs of homozygosity(FROH)

The average inbreeding coefficient,variation range and distribution of 4 pig breeds are summarized in Table 2.The average FROHof MS was the highest among these 4 pig breed populations,at 0.25. The average FROHof E (0.14) was the lowest. The average FROHof L and MI was between those of E and MS. At the individual level,the individual with the highest FROHappeared in L (0.40),and the individual with the lowest FROHappeared in MI(0.04) (Fig.4).

Table 1 Summary of the number of runs of homozygosity (ROH)in different categories in each breed

Table 2 Descriptive statistics for runs of homozygosity and inbreeding coefficients (FROH) within each breed

3.4.Genomic regions with a high frequency of ROH

The genomic regions that were most commonly associated with ROH were identified in the 4 pig breeds,and we assessed the proportion of SNPs in ROH by calculating the frequency of SNPs occurring in those ROH across all individuals. The result was plotted against the position of the SNP along the chromosome (Fig.5). A total of 106 regions were detected as genomic regions with a high frequency of ROH,and a total of 1 216 genes were identified.

4.Discussion

The distribution and frequency of ROH in the genome of 4 pig populations (E,L,MI,and MS) were analyzed by SLAF-seq. The results showed that the frequency of long ROH (＞5 Mb) of MS was the highest among the 4 pig breed populations,and the frequency of long ROH of MS and L was significantly higher than E and MI. The average FROHof MS was the highest,and the average FROHof E was the lowest. In addition,the total genomic length covered by ROH and the total number of ROH of MS were close to L,but significantly higher than E and MI. To some extent,these results confirmed our hypothesis that with the reduction of population size,MS showed a high level of inbreeding,while the inbreeding level of E and MI is relatively lower,which may be due to the fact that E and MI have not undergone intensive artificial selection. These findings provided a reference basis for maintaining genetic diversity and preventing inbreeding decline of 3 indigenous pig breeds in Chinese Taihu Basin.

The abundance,length and genomic distribution of ROH are good indicators to reflect the inbreeding level and historical origin of the population (Bosseet al.2012).PCA plots showed that there were significant differences in genetic relationship between 3 Taihu Basin pig breeds and L,and the genetic relationship between E and MI was similar,which was consistent with the study of Wanget al.(2015),but there were also significant differences between MS and E or MI.

In theory,since Chinese indigenous pigs have not undergone intensive artificial selection like western pigs,the genetic variability of Chinese indigenous pigs should be more abundant (Wanget al.2018). In this study,the average length and number of ROH of E and MI were lower than L,which indicated that the genetic diversities of E and MI were richer than L. This result was consistent with the above point of view. However,the average ROH length and FROHof MS were higher than L,which may be due to the high inbreeding degree of MS,and the decrease in genetic variation. It is consistent with the previous study by Zhanget al.(2018). Animals with the same cumulative ROH length show different ROH numbers,because different individuals in the population have different distances from the last common ancestor(Mastrangeloet al.2017). In general,the total length of ROH is proportional to the total number,and the higher the level of inbreeding,the larger the total length and total number of ROH (Xieet al.2019). In this study,through the analysis with the relationship between the total length and total number of ROH,it was found that the total length and total number of ROH of MS and L were significantly higher than those of E and MI,and the inbreeding level of MS was significantly higher than that of E and MI.

Long fragments of DNA are disturbed by generational(or meiotic) recombination events,so short ROH (＜5 Mb)may originate from distant common ancestors or the generation of long ROH (＞5 Mb) comes from more recent ancestors (Ceballoset al.2018;Zhanget al.2018). The frequency of short ROH of L was the highest,indicating that L had a higher level of inbreeding in distant common ancestors. The frequency of long ROH of MS was significantly higher than E and MI,which was similar to L in number,indicating that both MS and L had recent inbreeding. The number of ROH (＞40 Mb) of E was the highest,which was close to that of L. It may be indicated that E had a higher level of inbreeding in the closest generations.

In this study,a large number of candidate genes were identified to be associated with reproductive traits based on the genomic regions with a high frequency of ROH (Table 3).COMTis expressed in porcine granulosa cells,so the substrate (2-hydroxyestrogen) and product (2-methoxyestrogen) ofCOMTcan regulate the proliferation of granulosa cells and affect the development of follicles (Karpetaet al.2012).DGCR8affects the development of porcine oocytes,which has a potential role in maternal recognition of pregnancy and embryo implantation (Liuet al.2017). Insufficient expression ofMAPK1reduces trophoblast cell proliferation and blocks signal pathways,which has an important impact on thedevelopment of embryo and placenta (Martinezet al.2020).MIFshows high activity in maternal tissue during embryogenesis and plays an important role in embryonic receptivity and epitheliochorial placentation (Paulesuet al.2005).ARVCFis necessary for animal embryogenesis and makes an impact on the formation of ectoderm,the deletion ofARVCFleads to disrupt gastrulation (Fanget al.2004). The adaptor moleculeCRKLparticipates in a variety of signal transduction pathways and acts on the whole window of implantation in the uterus and embryo(Nautiyalet al.2004). The deletion ofHIRAcan inhibit the formation of male-female pronucleus,HIRAmutant oocytes are incapable of developing parthenogenetically(Linet al.2014).SMARCB1plays a role in the maintenance and differentiation of embryonic stem cells by mediating chromatin landscapes remodeling,and the deletion ofSMARCB1can cause embryonic lethality at the peri-implantation stage (Sakakuraet al.2019). Small nucleolar RNA (snRNA):U2andU6,play a role in snRNP assembly and pre-mRNA splicing in oocytes (Zhao and Yu 2004).MOSmay be involved in regulating the maturation of porcine oocytes (Daiet al.2005).RPS20can be used as internal standard to study the gene expression pattern of porcine endometrium on the 12th d of pregnancy (Wanget al.2011).SOX17is a critical regulator of human primordial germ cells(HPGCs) fate (Kobayashiet al.2017),and may involve in early differentiation of the principal body axes (Hassounet al.2009).EPHB2may play an important role in porcine embryo implantation and may contribute to litter size (Firat-Karalaret al.2014).PADI6is mainly expressed in the ovary,and plays a vital role in oocyte growth,fertilization and early embryo development (Xiaet al.2016).GDF9is expressed in porcine oocyte,granulosa,and theca cells of developing preovulatory follicles (Paradiset al.2009).VDAC1presents in the porcine oocyte plasma membrane and around the cortical area,and it is synthesized in germinal vesicle (GV) and meiosis II (MII) stage porcine oocytes (Cassaráet al.2009).TPPP3plays a significant role in embryo decidualization (Shuklaet al.2019).DPF3is identified as a new candidate gene related to the number of nipples in Duroc pigs in recent studies (Liet al.2021).NMNAT2plays a role in controlling redox homeostasis during oocyte maturation,and NMNAT2-NAD+-SIRT1 is an important pathway mediating the effects of maternal age on oocyte developmental competence (Wuet al.2019).

Table 3 Candidate genes located in genomic regions with a high frequency of runs of homozygosity (ROH) associated with pig reproduction traits

5.Conclusion

In this study,we used SLAF-seq data to analyze the ROH patterns of 3 indigenous pig breeds,Erhualian (E),Meishan (MS) and Mi (MI) pigs,in Chinese Taihu Basin and Landrace (L) pigs. The genetic structure of the 3 Taihu Basin pig breeds was significantly different from that of L.The genetic relationship between E and MI was close,the inbreeding level of E and MI was lower than that of L,while the inbreeding level of MS was the highest among the 4 pig breeds. Moreover,MS had the higher inbreeding level in recent generations. A large number of potential candidate genes related to reproductive traits were identified in the genomic regions with a high frequency of ROH. Our results can provide a theoretical basis for the maintenance of genetic diversity and genome selection of indigenous pig breeds in Chinese Taihu Basin.

Acknowledgements

This research was supported by the Jiangsu Agricultural Science and Technology Innovation Found,China (SCX(20) 3290),the Open Foundation of Key Laboratory of Urban Agriculture,Ministry of Agriculture and Rural Affairs of China (201906),and the Joint Research Project on Pig Breeding in Anhui Province,China(340000211260001000431).

Declaration of competing interest

The authors declare that they have no conflict of interest.

Ethical approval

All the experimental programs were completely carried out following the Management Guide (PZ2019105) of the Animal Ethical and Welfare Committee (AEWC) of Nanjing Agricultural University,China. The sampling procedures complied with the “Guidelines on Ethical Treatment of Experimental Animals” (2006) No.398 set by the Ministry of Science and Technology of China.