Association of SNPs within PTPN3 gene with wool production and growth traits in a dual-purpose sheep population

Abstract Protein tyrosine phosphatase non-receptor type 3 (PTPN3), a member of the membrane-associated non-receptor protein tyrosine phosphatase (PTP) family, plays significant roles in the cytoplasm and affects the development and growth of skin and hair. A recent study identified the PTPN3 as the potential gene related to sheep wool quality. To detect single-nucleotide polymorphisms (SNPs) of PTPN3 and elucidate its association with wool production and growth traits in fine wool sheep a total of 644 healthy SG (South African mutton merino♂ × Gansu alpine fine-wool sheep♀, SG) and SSG (South African mutton merino♂ × SG♀, SSG) hybrid sheep were selected. Pooled-DNA sequencing and SNPscan methods were used to scan and genotype SNPs within PTPN3. Association analyses between SNPs and wool production and growth traits were implemented. Consequently, the results revealed that PTPN3 has six SNPs (two missense mutations, one synonymous mutation, and three intron mutations), of which four loci (SNP2, SNP3, SNP4, and SNP5) were significantly positively correlated with growth and wool traits (p < 0.05). SNP4 was significantly (p < 0.05) linked with thigh wool length, and SNP6 was significantly (p < 0.05) associated with abdomen wool length. Moreover, one strongly linked SNP block was identified to be correlated with wool production and growth traits (body weight and body size). The significant SNPs founded by this study could serve as useful genetic markers for breeding fine-wool sheep.®


Introduction
Sheep (Ovis aries) provides a lot of products and byproducts, e.g., meat and wool for human beings. The wool industry produces $1,160 million kilograms of clean wool every year from a global herd of over a billion sheep. 1 The price of meat has been raised with the global demand increasing. To meet the growing needs of the market, the sheep industry has developed from wool to meat or meat-wool dual-purpose sheep gradually. 2 Therefore, it is important to select desired individuals with high production of wool and meat. Marker-assisted selection is an efficient way to select desired individuals if genetic markers linked with meat and wool production are identified.
It is an effective way to identify genetic markers through scanning SNPs in a gene with known function. Tyrosine phosphatase non-receptor type 3 (PTPN3), as a member of the membrane-associated non-receptor protein tyrosine phosphatase family, plays an important role in the cytoplasm. 3,4 PTPN3 is located on chromosome 2 in sheep and has a high level of expression in the adult hypothalamus. 5 A previous study has reported that PTPN3 is related to Epidermal Growth Factor Receptor (EGFR) endocytosis, degradation, and signaling pathways. 6 A recent genome-wide association study (GWAS) found that S54656.1, located in sheep the PTPN3 gene, was significantly associated with wool crimp. 7 According to this published information, we hypothesize that the PTPN3 could have a potential effect on wool characteristics and growth traits.
In this study, we aim to identify potential SNP loci within PTPN3 and analyze its correlation with wool production and growth traits in a dual-purpose hybrid sheep population. Our results could provide useful genetic markers for breeding dualpurpose sheep breeds.

Sample collection and DNA extraction
A total of 644 healthy SG (South African mutton merino# Â Gansu alpine fine-wool sheep$, SG) and SSG (South African mutton merino# Â SG$, SSG) were randomly selected from five sheep farms in Gansu, China. The summarized information of these samples was documented in Table S1. For growth traits, body weight, chest circumference, body height, body length, and shin circumference were measured. For wool traits, the weight of greasy wool, wool length, curl elasticity, fineness, single fiber strength were recorded. The detailed measure methods for both meat and wool traits can be found in a thesis. 8 The blood samples of 644 sheep were obtained and stored at À20 C. Genomic DNA was extracted using a phenolchloroform method 9 and diluted to 50 ng/mL after being assayed by Thermo Nandrop2000 spectrophotometer (NanoDrop 2000 Spectrophotometer, Massachusetts, USA).

PCR amplification and SNPscan genotyping
A total of 32 DNA pools were mixed and each pool was composed of 20 individuals. According to the sequence of PTPN3 in the NCBI database (GenBank No. NC_040253.1), 14 pairs of primers were designed by Premier 6.0 to amplify the full length. The sequence of 14 paired primers was documented in Table S2. These primers were diluted with double-distilled water to 10 lmol/mL. The PCR reaction was performed in a 25 mL volume system, including 1 mL of genomic DNA, 12.5 mL of 2 Â Easy Taq PCR Super MIX (TranGen), 0.5 mL (5nM) of each primer, and 10.5 mL of double-distilled water. The PCR conditions were run with an initial denaturation for 5 min at 95 C; 33 cycles of denaturation at 94 C for 30 s, annealing at Tm for 30 s; and extension at 72 C for 30 s$60s; and a final extension at 72 C for 5 min, with a final cooling at 4 C. After that, PCR products were directly sequenced by the Liuhe Huada Gene Technology Company (Beijing, China) and Aoke Dingsheng Biotechnology Company (Beijing, China).
The DNAstar (DNAstar, USA) and chromas (Technelysium Pty Ltd., South Brisbane, Queensland, Australia) software were used for sequence alignments and screening candidate SNPs. The qualified SNPs were genotyped by the SNPscan TM method by a commercial company (Shanghai Tianhao Biotechnology Co., Ltd.). Five DNA duplicate samples and 1 blank pair (DdH 2 O) were selected randomly for quality control, 10,11 taking Genotyping call rate, Minor allele frequency (MAF), and Hardy Weinberg equilibrium (HWE) into consideration.

Statistical analysis
Microsoft Excel 2013 was used to evaluate the genotype frequencies, allele frequencies, and genetic parameters. Six genetic parameters were calculated the following formulas: homozygosity Ho ¼ P n i¼0 P 2 i , heterozygosity Where n represented the number of alleles, Pi and Pj were the frequency of the i-th and j-th alleles, respectively; Ei was the theoretical value, Oi was the observation value. The association analysis between SNPs and phenotypes were implemented in R 3.6.1 following the formula: Where Y referred to the phenotype, m represented the population mean value, G was the genotype effect, s was the gender effect, p was the batch effect, b was the breed effect, and e was the random error. Differences in mean litter size among genotypes were tested using the LSD test in the agricolae R package. 12 p < 0.05 was to be considered significant. p < 0.01 was considered to be highly significant.
Haplotype (HP) and linkage disequilibrium (LD) were analyzed by the Haploview 4.2 software (Broad Institute, MIT and Harvard, USA). The statistical model: Y ¼ m þ c þ s þ p þ b þ e was established for the combined genotype analysis. where Y referred to the measured value of phenotypic traits, m represented the population mean, c was the combined genotypes, s was the gender effect, p was the batch effect, b was the variety effect, and e was the random error. The association analyses between combined genotypes and phenotypes were implemented by agricolae R package. 12

Analysis of genetic polymorphism and genetic structure in the population
In the study, six SNPs within PTPN3 were identified by pooled-DNA sequencing, the He, Ho, PIC were summarized in (Table 1, Fig. 1). Among them, SNP1 was a synonymous mutation located in exon 2. SNP4 and SNP6 were missense mutations located on exon 25 and exon 27, respectively. SNP2 and SNP3 were identified in intron 23, and SNP5 was identified in intron 25. Except for SNP2, SNP3, and SNP4, the other three loci were in line with HWE (p > 0.05) ( Table 1).

Correlation analysis between SNPs and wool traits
The correlation analysis between the six SNPs and wool traits were shown in Table 2. The results show that the SNP1 was significantly associated with greasy wool weight (p < 0.05). The SNP4 was significantly associated with thigh wool length, while SNP6 and SNP1 were significantly associated with abdomen wool length (p < 0.05) and mean fiber diameter (MFD) (p < 0.01), respectively (Table 3). Furthermore, SNP2 and SNP3 were significantly associated with crimp ratio and SNP1 was significantly associated with crimp elastic rate ( Table 2). SNP6 locus was significantly associated with breaking strength (p < 0.05) ( Table 2). In addition, the LD analysis indicated that SNP2 and SNP3 were tightly linked and two haplotypes of TC and CT were identified (Fig. S1, Fig. 2, Table 4). According to the results of LD analysis, the three combined genotypes (SNP2 and SNP3) were significantly correlated with the crimp ratio (p < 0.05) ( Table 5).

Correlation analysis of SNPs and growth traits
The correlation analysis between the six loci and the growth traits was performed. The results were shown        in Table 6. SNP2, SNP3, SNP4, and SNP5 were significant for body weight, chest girth, and shin circumference (p < 0.05). Moreover, LD analysis implied that SNP2 and SNP3 were tightly linked, and two haplotypes of TC and CT were constructed with frequencies of 0.555 and 0.442, respectively (Figs. 2, 3, Table 4). Additionally, the association between combination genotypes and growth traits was carried. These combined genotypes of SNP2 and SNP3 had a significant effect on body weight (p < 0.05) and shin circumference (p < 0.01), as shown in Table 7.

Discussion
PTP plays an important role in many cellular processes, including cell differentiation, activation, cycle progression, apoptosis, cytoskeleton rearrangement, movement, matrix adhesion, adhesion, metabolic homeostasis. As of now, in animal husbandry, only one piece of literature has reported that the PTPN3 affects sheep's wool traits. 7 Meanwhile, it is related to a few human diseases, such as cancer, cardiovascular, neurological, metabolic, immune, and infectious diseases. [13][14][15] Hu et al. used immunohistochemical to detect its expression in colorectal cancer tissues and analyzed the regulatory effect of miR-95 on PTPN3. The results demonstrated that the PTPN3 gene was down-regulated by a major regulatory factor miR-95. 16 Wang et al. revealed one locus which was located in sheep PTPN3, and it was significant for wool crimp. 7 This study found that SNP2, SNP3, SNP4, and SNP5 were significant for growth traits. SNP1 was significant for wool fiber diameter and its standard deviation, the weight of greasy wool, and crimp elasticity, which were in accord with the previous studies. SNP6 was significant for abdomen wool length and wool fiber breaking strength. Among these loci, SNP2 was a synonymous mutation and does not cause changes in encoded amino acid while both SNP4 and SNP6 were missense mutations that might exert significant influence on phenotype by affecting the expression of related genes mRNA and the spatial structure of proteins. In this experiment, SNP4 is located in exon 25 of PTPN3, this base mutation causes the change of encoding amino acid Met!Thr. In the process of wool growth, tyrosine forms melanin particles in melanocytes; 17 SNP6 is located in exon 27, causing the change of encoding amino acid Arg!Cys, the correlation analysis between the genotype and phenotype showed that the wool length of mutant genotype individuals at these two loci and their significantly related parts were higher than that of the wild genotype. Whereas, the mechanism of action of these loci needs to be further studied.
Sheep wool and growth traits were complex traits and were controlled by multiple genes. The interaction among genes or multiple SNP loci of the same gene may also have a more significant effect on wool and growth traits. Zeng et al. 18 analyzed the combined genotypes of candidate genes for wool traits and growth traits of Chinese Merino sheep and found that sheep individuals with the combination of BB and AA (BB-AA) had the largest effects on chest girth, body oblique length, and body weight. Additionally, the BB-AA combined genotype has a large positive genetic contribution to body oblique length, chest circumference, and body weight, but a small genetic contribution to wool fineness. Previous studies have also shown that many SNP loci were not the causative mutations that directly cause phenotypic changes, instead, it interacts with the causative mutation by being in a linkage disequilibrium state together with the causative mutation and indirectly affects the change of amino acids and even proteins, ultimately leads to the change of phenotype. 19 In this experiment, we found that the combined genotype of SNP2 and SNP3 locus was significant for curl, bodyweight, and shin circumference. These results suggested that we can select individuals with combination genotypes to improve selection efficiency in future molecular breeding work.

Conclusion
In this study, we detected 6 SNPs in PTPN3. Among them, SNP2, SNP3, SNP4, and SNP5 were significantly associated with growth traits (p < 0.05). SNP4 was significantly associated with thigh wool length and SNP6 was significantly associated with abdomen wool length (p < 0.05). One strongly linked SNP block (SNP2 and SNP3) was identified and it was correlated with wool production and growth traits. Summarily, the results provided a molecular basis for molecular breeding, but its mechanism needs further indepth study.