Introduction
Sow reproductive performance is affected by a number of factors including ovulation rate and embryo survival rate. Current technologies enable scientists to improve the efficiency of traditional selection methods by applying genetic markers through marker-assisted selection. Two main approaches have been successfully considered to choose the candidate genes of reproductive traits. One is based on their physiological roles in reproduction and the other is on their location with regard to quantitative trait loci (QTL) for the reproductive trait of interest (reviewed by Spotter and Distl, Reference Spotter and Distl2006). In this study, we have focused on identifying associations between different reproductive phenotypes of pigs with the genotypes at single-nucleotide polymorphisms (SNPs) in two genes, namely the follicle-stimulating hormone receptor (FSHR) and the zona pellucida 3 glycoprotein (ZP3) genes. FSHR in pig is a candidate gene which influences sow reproductive traits (Remy et al., Reference Remy, Lahbib-Mansais, Yerle, Bozon, Couture, Pajot, Greber and Salesse1995; Zhang et al., Reference Zhang, Xiong, Deng, Zeng, Qiu, Xiao, Xia, Xu and Ruan2002). ZP3 in pig is an extracellular matrix protein which surrounds eggs and early embryos and is critical for normal fertilisation and pre-implantation development (David, Reference David1997). Braundmeier et al. (Reference Braundmeier, Demers, Shanks and Miller2004) previously reported that zona-binding ability was correlated with average litter size. We have recently reported the association between ZP3 genes variations and reproductive traits (the effective teat number, TN and number of piglets born alive, NBA) (Yuan et al., Reference Yuan, Moaeen-ud-Din, Gong, Peng, Yang, Feng, Liu, Hu, Affara, Jafe and Zhang2007) and results showed the significant association between mutation (the 18 bp deletion/insertion from intron2) and NBA. These results have increased our interest in investigating the role of variations in the FSHR and ZP3 genes. Thus, the objective of this study was to look for possible variations of DNA sequence in 10th exon of FSHR gene along with two SNPs (G261A and T302C) of the ZP3 gene, to genotype a number of pig breeds by using SNP chips and to score nucleotide variation that may be associated with various reproductive traits including the effective TN and NBA.
Material and methods
Pig populations and traits
Genomic DNA was isolated from 612 sows from six purebred lines, which included three Chinese domestic purebred lines (42 Small Meishan, 46 Qingping and 41 Jinhua sows) from the corresponding National Conserved Farmers in China and three European purebred lines (225 Duroc, 195 Large White and 65 Landrace) from Hubei Tianzhong Boar Co. Ltd., Sanli Town, Hubei Province, China. The nutrition of the Chinese breeds was varied depending on the physiological characteristics and maternity status. The management and the health programme followed the European pig breeding company’s recommendations. Gilts were mated by artificial inseminations (AI) on the second or later observed oestrus depending upon the body weight (majority were mated at the third oestrus). Phenotypic data were obtained for TN and NBA per litter along with family structure for the dams within breed. The selected sows had at least two or more parities.
DNA extraction and PCR-SSCP
Pig DNA was extracted from peripheral leucocytes using standard procedures as reviewed by Sambrook and Russell (Reference Sambrook and Russell2001). PCR primers (forward primer: 5′ ctctgaccttcatccaa 3′; reverse primer: 5′ agtatagcagccacaga 3′) were designed from 10th exon of FSHR gene (AF025377). The reaction reagent was amplified by PCR and carried out by using 20 to 40 ng genomic DNA as template based on the methods described by Lee et al. (Reference Lee, Lo and Choo1992). PCR conditions were 94°C for 10 min, followed by 94°C for 40 s, annealing temperature 57.0°C for 50 s and 72°C for 45 s for 30 cycles and extending at 72°C for 10 min.
SSCP (single-strand conformation polymorphism) analysis was performed on the basis of the methods described by Lee et al. (Reference Lee, Lo and Choo1992). SSCP gel electrophoresis was carried out in 10% polyacrylamide gel (39 acrylamide: 1 bisacrylamide) that were made up in 1× TBE (Tris borate EDTA) buffer, which contained 5% glycerol. Heat-denatured PCR products (2 μl) mixed with 6 μl of sequence-stop solution containing formamide were applied to the gel and electrophoresis was carried out at a constant power of 140 V for 16 to 20 h.
SNP discovery by sequencing of PCR products
PCR products that showed different bands with SSCP were purified by using Qiagen columns (Qiagen, Shanghai, China), according to the manufacturer protocol (Qiagen catalogue no. 282010). Purified products were sequenced by using an ABI 377 automatic sequencer for identification of SNPs.
Microarray based genotyping of SNPs in the different pig breeds
PCR for microarray-based genotyping was carried out in 10 μl reaction volumes containing 10 to 30 ng of porcine genomic DNA, 1 μl dNTPs (250 μmol/l each), 10 pmol of each primer, 1 μl of 10× buffer and 0.1 μl of dynazyme DNA polymerase (Dynazyme). Amplification conditions were 5 min at 94°C, followed by first to 22nd cycles of 30 s at 94°C, 30 s at 65°C, 30 s at 72°C and touchdown to 52°C by decreasing the annealing temperature by 1°C every cycle, and finally 23 cycles of 30 s at 94°C, 30 s at 52°C, 30 s at 72°C and a final extension of 10 min at 72°C.
The genotyping of the samples was performed according to the protocol (Ji et al., Reference Ji, Hou, Li, He and Lu2004) with some modification. Briefly, samples were amplified in a 96-well plate by using primers that covered the SNP region. The average PCR product was 100 bp. After purifying the PCR product by using QIAquick columns (Qiagen), the products were resuspended in 3× SSC to a final concentration of 300 ng/μl and transferred to 384-well plates (Applied Biosystems, Foster City, CA, USA; no. 4309849).
The purified products were printed on GAP slides (Corning no. 40005) using a BioRobotics MicroGrid II arrayer (Genomic Solutions, Ann Arbor, MI, USA) and baked for 2 h at 80°C. There are 9000 spots of 3000 samples due to three repeats of one sample on slide. After baking the slides, the bound DNA was denatured by 2 min incubation in distilled water at 95°C. The slides were then transferred into a bath of 95% ethanol at room temperature, rinsed, and then spun dry in a clinical centrifuge. Slides were stored in a closed box at room temperature until used.
Hybridisation mixes consisted of Cy3- and Cy5-labelled oligonucleotide probes capable of detecting the two different SNP alleles mixed in equimolar amounts. The hybridisation mix contained 2 μl of salmon testis DNA, 10 μl of 20× SSC, 6 μl of 1 μmol each labelled probes, 2 μl of 10% SDS and 14 μl distilled water in 40 μl volume. Hybridisation was conducted in a moist sealed hybridisation chamber (TeleChem, Sunnyvale, CA, USA) under a cover slip at 37°C for 30 min. After hybridisation, the slides were rinsed and washed at room temperature with 5× SSC−0.1% SDS, 5× SSC and distilled water (5 min for each wash), and subsequently dried by centrifugation at 600 rpm for 5 min.
The hybridised slides were scanned using the Axon 4100A scanner (Molecular Devices, Sunnyvale, CA, USA). Analysis and genotype calling of the images acquired by the scanner were performed using Blue fuse software (Bluegnome, Cambridge, UK).
Statistical analysis
Genotype frequencies for the FSHR and ZP3 SNPs were estimated within the six purebred breeds. The χ 2 test was employed to study Hardy–Weinberg equilibrium in each population by using Statistical Packages for the Social Sciences (2002). The following linear model (Groeneveld, Reference Groeneveld1990) was employed for the analysis of variance in six purebreds:
where Y was observed trait; Xi was fixed effect of genotype i; Wj was fixed effects of breed j; Sk was fixed effect of sire term k; Fl was fixed effect of litters l; XWij was fixed interaction effects of genotype i with breed j; XFil was fixed interaction effects of genotype i with litter l; XSik was fixed interaction effects of genotype i with sire term k; WFjl was fixed interaction effects of breed number with litter l, XWSFijkl was fixed interaction effects genotype i, breed j, sire term k and litter l; e was error. The multivariable analysis of variance with approximate F statistic is provided as well as the univariate analysis of variance for each dependent variable. After an overall F test of the genotype effect had shown significance (P < 0.05), a GLM procedure was applied to evaluate differences among least-squares means for genotype groups.
Results
Discovery of new SNPs in the FSHR and ZP3 genes
Two new non-synonymous SNPs were discovered in the FSHR gene by sequencing PCR products: one SNP (C1166T) was located at position 1166 in the first transmembrane region, and resulted in a substitution of isoleucine (Ile) to threonine (thr) at residue 377. The other SNP (G1190A) was located at position 1190 in the topological domain, and resulted in a substitution of valine (Val) for alanine (Ala) at residue 396.
Allele frequencies
Table 1 shows the allele frequencies for these four SNPs in six purebred lines. The TT1166 genotype frequency for the FSHR gene position 1166 SNP (C1166 T) was higher than the CC1166 genotype frequency except for Qinping population. The position 1190 SNP (G1190A) of the FSHR gene showed a lower GG1190 frequency than it was expected from the Hardy–Weinberg equilibrium.
The frequency of the AA genotype at position 261 SNP (G261A) of the ZP3 gene was higher than the GG genotype frequency of our studied population except for Landrace. Moreover, no GG261 genotype was found in Duroc and Jinhua populations. The overall frequency of the TT genotype at position 302 SNP (T302C) of the ZP3 gene was higher than the CC genotype frequency except for Duroc population.
Association of SNPs with TN
TN association analysis was done by using the established model based on breed effect, genotype effect, interaction effects and random factors. The analyses showed that only breed effect and the genotype for position 261 SNP (G/A) of the ZP3 gene had a significant difference (P < 0.05), and TN of sows with the GG genotype (14.83 ± 0.18) was significantly more than that of sows with the AA genotype (14.26 ± 0.09) (P < 0.05). However, no significant difference was observed for the other three SNPs in the ZP3 and FSHR gene (Table 2).
a,b,cValues in the same column with same superscripts significantly differ at P < 0.05; A,BValues in the same column with same superscripts significantly differ at P < 0.01.
Association of SNPs with NBA
Association analysis between genotypes and NBA was conducted by using the established model based on breed effect, genotype effect, litter effect, sire terms within line, interactions among effects and random factors. F-test showed breed effect, litter effect, genotype effect, interaction effects of genotype with litter number, and interaction effects of genotype with breed were significant for the position 261 SNP (G261A) of ZP3 gene. Thus these effects were considered when we evaluated least-squares means in genotype effect at this position. Sows with GA genotype in NBA had more piglets (9.84 ± 0.16) than that of the GG genotype (8.78 ± 0.25) (P < 0.01), and the AA genotype (9.22 ± 0.13) (P < 0.01). In addition, sows with GA genotype in NBA2+ had more piglets (9.94 ± 0.19) than that of the GG genotype (8.81 ± 0.34) (P < 0.01), and the AA genotype (9.38 ± 0.18) (P < 0.05), whereas no significant difference was observed in NBA1 (Table 2). The same procedure were used at the position 302 SNP (T302C) of the same gene, GLM results showed NBA2+ of sows with CC genotype had significantly (P < 0.05) more piglets (9.77 ± 0.33) than sows with CC genotype (9.21 ± 0.17). While no significant difference was found among the three genotypes in NBA and NBA1.
F-test showed only breed effect, litter effect, and interaction effects of genotype with litter number had significance at the position 1166 SNP (C1166 T) and (G1190A) of the FSHR gene and no significant difference in genotype effect was found in our tested population (Table 2).
Discussion
At present, candidate genes for reproductive traits in pigs can be chosen by considering the following four factors or a combination of them: (1) the physiological role of a gene in reproduction (Rothschild et al., Reference Rothschild, Jacobson, Vaske, Tuggle, Wang, Short, Eckardt, Sasaki, Vincent, Mileham and Plastow1996); (2) genes located or closely linked with possible reproductive relative QTL (positional candidate genes); (3) using information on orthologous genes in syntenic chromosomal regions of other species (positional comparative candidate genes) (Haley, Reference Haley1999); and (4) differential expression of genes in the tissue under specific investigation (Liang and Pardee (Reference Liang and Pardee1992); reviewed by Spotter and Dist (Reference Spotter and Distl2006)).
ZP is laid down during oogenesis and increases in thickness as oocytes increase in diameter and it plays important roles during mammalian oogenesis, fertilisation, pre-implantation development (Wassarman, Reference Wassarman1999; Wassarman et al., Reference Wassarman, Jovine, Litscher, Qi and Williams2004; Herrick et al., Reference Herrick, Lschweissinger, Kim, Bayan, Mann and Cardullo2005) and passage through the oviduct (Rankin et al., Reference Rankin, Soyal and Dean2000). During the stages of pre-implantation development, the embryo remains encased in the ZP, without which the very early embryo is unable to traverse the oviduct (Bronson and McLaren, Reference Bronson and McLaren1970). Just prior to implantation, the embryo escapes from the zona pellucida by mechanisms not yet fully described. ZP effectuates the individuality of the embryo by preventing aggregation of individual embryos and facilitates survival in the oviduct (Kolbe and Holtz, Reference Kolbe and Holtz2005). Braundmeier et al. (Reference Braundmeier, Demers, Shanks and Miller2004) previously reported zona-binding ability was correlated with fertility when estimated by average litter size (r = 0.64, P < 0.05). Thus, the mutations of the ZP3 gene possibly have an effect on reproduction traits. Unfortunately, the map location of the gene in the pig remains unknown. However, Yuan et al. (Reference Yuan, Moaeen-ud-Din, Gong, Peng, Yang, Feng, Liu, Hu, Affara, Jafe and Zhang2007) have discovered 13 new SNPs in the exon1 of ZP3 gene, and four of them, which have the restriction enzyme position, were chosen to genotype and associate with reproductive traits in the European breeds (Duroc, Large White and Landrace). Here, we selected another two SNPs under the requirement of SNP chips and genotyped in three Chinese domestic purebreds and three European purebreds and analysed the association of these mutations and haplotypes with NBA and TN. Association analysis of SNPs indicated that these two mutations may influence the process of fertilisation and pre-implantation development, and finally may affect the observed phenotypic differences in reproductive traits performance. Certainly, the relationship between mutations and phenotypic data is still poorly understood and needs to be explored by further experimentation.
The follicle-stimulating hormone (FSH) is essential for normal gametogenesis. In females FSH is required for ovarian development and follicle maturation whereas in males FSH determines sertoli cell number and qualitatively normal spermatogenesis. As the receptor of FSH, the FSH-receptor (FSHR) mediates action of FSH on ovarian follicular granulosa cells. Recently, SNPs in the human FSHR gene have been identified and severely affects the process of gametogenesis and infertility (Gromoll et al., Reference Gromoll, Simoni, Nordhoff, Behre, De Geyter and Nieschlag1996; Ahda et al., Reference Ahda, Gromoll, Wunsch, Asatiani, Zitzmann, Nieschlag and Simoni2005; Wunsch et al., Reference Wunsch, Ahda, Banaz-Yasar, Sonntag, Nieschlag, Simoni and Gromoll2005; De Leener et al., Reference De Leener, Montanelli, Van Durme, Chae, Smits, Vassart and Costagliola2006). Although the FSHR gene was first investigated as a potential candidate gene influencing NBA in pig in 1995 (Remy et al., Reference Remy, Lahbib-Mansais, Yerle, Bozon, Couture, Pajot, Greber and Salesse1995), only a few researchers have done further investigation on SNPs of this gene. Zhang et al. (Reference Zhang, Xiong, Deng, Zeng, Qiu, Xiao, Xia, Xu and Ruan2002) performed an association study of SNP located at exon7 in the FSHR gene with size of reproductive organ and revealed sows that could provide more piglets with bigger reproductive organs usually carried BB genotype on the site of FSHR gene. Our findings demonstrated the trends that the sows with the AA genotype at position 1190 SNP (G/A) of 10th exon had higher NBA2+ than the sows with the GA and GG genotypes, although it was not significant (Table 2). Two other factors could affect our results: (1) the worst performing sows would be screened out after one or two parities, and therefore potentially useful genetic variation for the purposes of the study would be lost; (2) although we could model the effects of breeds, litters, the sire term effect and the interact effects among breeds, genotype and parities successfully, other effects such as nutrition and reproductive diseases were not considered here.
For animal breeding we must get reliable information. Due to the limited number of samples we could not analyse the association between genotype and phenotypic traits within the single breeds. Thus to confirm our finding, further studies are required using more samples and thorough examination of potential association of these four mutations with reproduction traits in all breeds.
Acknowledgments
Financial assistance from 863 High Technology and Development Project of China (2002AA211041, 2006AA10Z195), Hubei project (2003AA211041, 2006CA023), Wuhan project (20067003111-06) and National Project of China (04EFN214200206) is greatly appreciated.