Association study between single nucleotide polymorphisms in candidate genes and reproduction traits in Italian Large White sows

Livestock Science 155 (2013) 172–179

Contents lists available at SciVerse ScienceDirect

Livestock Science journal homepage: www.elsevier.com/locate/livsci

Association study between single nucleotide polymorphisms in candidate genes and reproduction traits in Italian Large White sows S. Dall'Olio a,n, L. Fontanesi a,b, L. Buttazzoni c, C. Baiocco d, M. Gallo d, V. Russo a a

Department of Agricultural and Food Sciences, Division of Animal Science, University of Bologna, Viale Fanin 46, 40127 Bologna, Italy Centre for Genome Biology, University of Bologna, 40126 Bologna, Italy Consiglio per la Ricerca e la Sperimentazione in Agricoltura (CRA-PCM), Via Salaria 31, 00015 Monterotondo, Roma, Italy d Associazione Nazionale Allevatori Suini (ANAS), Via Lazzaro Spallanzani 4/6, 00161 Roma, Italy b c

a r t i c l e in f o

abstract

Article history: Received 16 January 2013 Received in revised form 8 May 2013 Accepted 9 May 2013

In this study we evaluated the effects of candidate single nucleotide polymorphisms (SNPs) on reproduction traits of Italian Large White (ITLW) purebred sows at first parity. We genotyped DNA samples extracted from hair roots of 1,548 sows distributed across six farms located in Northern Italy. A total of 21 SNPs in 20 genes, selected from literature, were genotyped. Seventeen SNPs showed minor allele frequencies 40.05 in the ITLW breed. Association analyses were performed with the following traits: number of piglets born alive (NBA1), number of stillborn piglets (NSB1), total number born (TNB1), NBA1 estimated breeding values (EBVs) and NBA1 random residuals (RRs). SNPs in BMPR1B, FUT1, GPX5, RBP4 and TGFBR1 genes showed significant association (P nominal o0.003) with NBA1 EBVs. Additional three markers (EPOR, GDF9:c.1806T4C and STAT5B) presented suggestive effects (P nominal o0.03) on NBA1 EBVs. The BMPR1B SNP showed a trend (P nominal from 0.01 to 0.05) on NBA1, NBA1 RRs and TNB1. This gene could be a promising candidate for additional investigations aiming to confirm evidences reported in this study. & 2013 Elsevier B.V. All rights reserved.

Keywords: Pigs Genes Single nucleotide polymorphisms Association analyses Litter size First litter

1. Introduction Reproductive efficiency of sows (measured as piglets weaned per sow per year) is an important factor affecting pig farming productivity and profitability. Reproductive efficiency of sows is determined by the number of piglets born alive (NBA), survival at weaning and the number of farrowings per sow per year. NBA is a complex phenotype influenced by numerous parameters including ovulation rate (OR), number of embryos, embryonic and fetal survival, and uterine capacity (UC) (Vallet et al., 2010). Most of the traits determining litter size variability (i.e., OR and UC) are difficult, time-consuming and expensive to

n Corresponding author. Tel.: +39 05 120 965 95; fax: +39 051209 659 6. E-mail address: [email protected] (S. Dall'Olio).

1871-1413/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.livsci.2013.05.012

measure on a large number of sows, whereas NBA is easily recorded in pig farms. Estimated Breeding Values for NBA at first parity (NBA1) have been routinely calculated in the Italian Large White (ITLW) breed and used to improve the number of viable offspring since the late 1990s. Other coselected traits in this heavy pig breed are growth rate, feed conversion efficiency, lean tissue deposition (lean cuts), fatness (back-fat thickness), ham weight and ham salting losses. At present, a specific goal of selection for the ITLW breed is to maintain an appropriate back-fat thickness, as the production of high quality dry-cured hams needs green thighs with at least 2 cm of fat coverage (Bosi and Russo, 2004). In comparison with performance and carcass and meat production traits, genetic progress for NBA as well as other reproduction traits (e.g. number of stillborns piglets, NSB; and total number of piglets born, TNB) is slower, because these traits are sex-limited and have low heritability (Bidanel, 2011).

S. Dall'Olio et al. / Livestock Science 155 (2013) 172–179

173

pedigree and date of birth were recorded (December, 2010). Then, total number of piglets born at first parity (TNB1) was computed as the sum of NBA1 and NSB1. After data editing (e.g. presence of contemporary sows, age at first farrowing between 280 and 490 d), records of 1548 sows (85.9% of the initial collected samples) were used for all subsequent analyses (Table 1). Means of NBA1 and TNB1 in this dataset were 9.7772.33 and 10.3472.34, respectively. NBA1 records were used to calculate NBA1 EBVs using a BLUPsingle trait-animal model including environmental effects of herd–year–month at birth (contemporary groups), age at first farrowing, inbreeding coefficient of sows and mating type (artificial insemination or natural service). In addition, random residuals (RRs) for NBA1 data were calculated by using a linear fixed model including the same environmental factors used to calculate NBA1 EBVs. RRs include genetic factors, permanent environmental factors, unknown environmental factors and measurement errors (Fontanesi et al., 2010a). Both NBA1 EBVs and NBA1 RRs were expressed in standard deviation units around the rolling average of data of sows farrowing since 1990. NBA1 EBVs and NBA1 RR means were +0.69 71.23 (range from −3.25 to +5.39) and 0.00 72.27 (range from −7.20 to +8.54), respectively (Table 1). Additional hair root samples were collected from Italian Duroc (ITDU, n¼91) and Italian Landrace (ITLA, n ¼50) pigs for allele frequency evaluation in different breeds.

Several studies have been carried out to identify markers associated with reproduction traits in different pig breeds and populations in order to use them in marker assisted selection (MAS) programs (Bidanel, 2011; Distl, 2007; Rempel et al., 2010; Sironen et al., 2010; Spötter and Distl, 2006). The first marker resulting to be associated with litter size in pigs was the ESR1-PvuII restriction fragment length polymorphism (RFLP; Rothschild et al., 1994). However, the effects of this polymorphism on litter size were not consistent across populations (Distl, 2007). The ESR1-PvuII RFLP and a single nucleotide polymorphism (SNP) in the ESR2 gene were recently tested for association with litter size in the ITLW breed (Dall′Olio et al., 2011). The obtained results did not confirm their putative effects on prolificacy, as previously reported in other populations. Because of relevant potential impact of DNA markers, it is important to verify in the target population claimed effects of markers before their implementation in actual breeding plans. The objective of this study was to investigate 21 candidate SNPs for reproduction traits selected from literature and to evaluate their association with litter size in a large population of first parity ITLW sows. 2. Materials and methods 2.1. Animals and phenotypes Hair root samples were collected from 1803 ITLW sows registered in the Herd Book held by the National Association of Pig Breeders (Associazione Nazionale Allevatori Suini, ANAS, Italy; http://www.anas.it). These sows were raised in six farms (referred to as farm 1 to farm 6, Table 1) having at least 120 sows each, located in Emilia Romagna and Lombardia regions (North of Italy). For each sow NBA1 and NSB (including mummified fetuses at birth) at first parity (NSB1),

2.2. Genotypes Genomic DNA was extracted from hair roots using a standard protocol with proteinase K. Twenty-one candidate SNPs in 20 genes (ADRB2, AFP, BMPR1B, CTSL, CYP21A2, EDNRA, EPOR, FUT1, GDF9, GNRHR, GPX5, IGF1R, LEPR, MAN2B2, PIT1, PNAS4, RBP4, SCG2, STAT5B and TGFBR1) were selected

Table 1. Number of samples and descriptive statisticsa of litter size traits of the first litter/farrowing farm. Farms No. of sows No. of sows sampled after data editing

NBA1b Mean SD

NSB1c

TNB1d

Range Mean SD

NBA1 EBVe

Range Mean SD

Range Mean

1

691

674

9.60 2.23 3;16

0.58

0.98 0;6

10.18

2

427

307

10.02 2.13 3;16

0.67

0.99 0;6

10.69 2.33 3;17

3

102

92

9.71 2.85 3;16

0.96

1.47 0;7

10.66 2.66 3;16

4

323

228

9.55 2.59 3;18

0.22

0.79 0;7

9.77 2.53 3;18

5

126

123

9.97 2.43 3;15

0.90

1.46 0;7

10.87 2.49 3;16

6

134

124

10.30 2.12 3;16

0.33

0.67 0;3

10.63 2.06 3;16

All

1803

1548

9.77 2.33 3;18

0.57

1.04 0;7

10.34 2.34 3;18

a

2.20 3;17

SD

NBA1 RRf Range

+0.453 1.243 −3.247; +4.799 +1.091 0.855 −1.040; +4.405 +0.882 1.362 −1.916; +5.395 +0.265 1.403 −2.961; +4.112 +0.951 0.904 −0.939; +4.337 +1.319 1.202 −0.768; +4.469 +0.686 1.232 −3.247; +5.395

Mean SD

Range

0.000 2.218 −6.937; +6.497 0.000 2.103 −7.199; +5.566 0.000 2.736 −7.109; +4.852 0.000 2.575 −7.075; +8.538 0.000 2.258 −6.435; +4.681 0.000 2.049 −7.054; +4.869 0.000 2.272 −7.200; +8.538

Mean, standard deviation (SD) and range (minimum;maximum values). NBA1¼ number of piglets born alive at first parity. c NSB1¼ number of stillborn piglets at first parity. d TNB1¼ total number of piglets born at first parity. e NBA1 EBVs¼ estimated breeding values for number of piglets born alive at first parity, expressed in standard deviation units around the rolling average of data of sows farrowing since 1990. f NBA1 RRs ¼ random residuals for number of piglets born alive at first parity, expressed in standard deviation units around the rolling average of data of sows farrowing since 1990. b

174

S. Dall'Olio et al. / Livestock Science 155 (2013) 172–179

Table 2. Details of the investigated SNPs. Gene symbol

Gene name

SSCa

GenBank accession number: genotyped SNPb

SNP location within the genec References (amino acid substitution)

IGF1R

Insulin-like growth factor 1 receptor Transforming growth factor beta type 1 receptor Adrenergic receptor beta 2 Erythropoietin receptor Growth and differentiation factor 9 Growth and differentiation factor 9 Fucosyltransferase 1 (galactoside 2-alpha-Lfucosyltransferase, H blood group) Leptin receptor Cytochrome P450, family 21, subfamily A, polypepide 2 Glutathione peroxidase 5 Alpha-fetoprotein Bone morphogenetic protein receptor type 1 beta Endothelin receptor type A Gonadotropin releasing hormone receptor Mannosidase alpha class 2 B member 2 Cathepsin L Peptidase domain containing 1 Signal transducer and activator of transcription 5B POU class 1 homeobox 1 Retinol binding protein 4 Secretogranin II

1

AJ491314:g.145T4C

I9

Kopečný et al. (2002)

1

DQ519377:g.64939A 4G

E 7 (p.Ile417Val)d

Shimanuki et al. (2005)

2 2 2

AF000134:g.673C 4T EU407778:g.2373C 4T AY649444:c.908C 4T

P I4 E 2 (p.Ala39Val)

Muráni et al. (2009) Vallet et al. (2005) Zhang et al. (2008)

2

AY649444:c.1806T4Ce

E2

Zhang et al. (2008)

6

U70883:g.915G 4A

E 1 (p.Ala103Thr)

Meijerink et al. (1997)

6 7

AF092422:c.221T 4C M83939:g.3462G 4 A

E 4 (p.Met70Thr) 3′–UTR

Mackowski et al. (2005) Buske et al. (2006a)

7 8 8

AF124818:c.n1897C4T AY120900:g.547T4C AY65994:c.960C 4T

3′–UTR I 10 E 10

Buske et al. (2006b) Kim et al. (2002) Campbell et al. (2008)

8 8

AY540998:g.371C 4T AF227686:g.1721C 4G

I6 3′–UTR

Kim et al. (2004) Jang et al. (2001)

8

D28521:c.1421A 4Gf

E 10 (p.Met475Val)

Campbell et al. (2008)

10 10

AJ315771:g.5325C 4T DQ406743:g.1401T 4C

E 5 (silent) I2

Fontanesi et al. (2010b) Mo et al. (2008)

12

DQ144238:g.191A 4 G

I 14

Ballester et al. (2006)

13 14 15

DQ485156:g.538C 4G DQ344026:g.447G 4Cg AY870646:c.622A4G

I5 I4 E (p.Thr211Ala)

Franco et al. (2005) Muñoz et al. (2010) Du et al. (2008)

TGFBR1 ADRB2 EPOR GDF9 GDF9 FUT1

LEPR CYP21A2 GPX5 AFP BMPR1B EDNRA GNRHR MAN2B2 CTSL PNAS4 STAT5B PIT1 RBP4 SCG2 a

Porcine chromosome. SNP nomenclature based on rules reported in the http://www.genomic.unimelb.edu.au/mdi/mutnomen/ web site. c E¼ exon, I¼ intron, P¼ promoter, 3′–UTR¼ 3′ untranslated region. d Or p.Ile413Val for the intermediate isoform (Shimanuki et al., 2005; Chen et al., 2006). In particular, two different long cDNA sequences, giving two isoforms of 503 and 499 amino acids (delete 112–116 amino acids residues), were identified for the alternative splicing of the first 12 nucleotides of the exon 3 (Shimanuki et al., 2005; Chen et al., 2006). In addition, alternative splicing of 125 bp of exon 7, giving the shorter isoform of 379 residues, was identified in the Berkshire breed (Chen et al., 2006). e In linkage disequilibrium with the GDF9:AY649444:c.1801C 4T SNP causing the p.Phe337Leu substitution (Zhang et al., 2008). f Previously referred to as D28521:c.1574A 4G (Campbell et al., 2008). g Alleles DQ344026:g.447G and DQ344026:g.447C of the RBP4 gene (Muñoz et al, 2010) correspond to the alleles 1 and 2 of RBP4 MspI PCR-RFLP analysis, respectively (Rothschild et al., 2000). b

for genotyping (Table 2). SNPs were identified from previously published studies in genes involved in physiological mechanisms related to reproduction (Table S1) and/or located in QTL regions known to be relevant for reproduction traits. In particular, five genes (AFP, BMPR1B, EDNRA, GNRHR, and MAN2B2) map on porcine chromosome 8 (SSC8) where QTLs for OR, number of corpora lutea and UC were identified (Hu et al., 2010; http://www.animalgenome.org/QTLdb/). Most of the analyzed genes (BMPR1B, CYP21A2, EPOR, FUT1, GDF9, GNRHR, GPX5, IGF1R, LEPR, MAN2B2, PIT1, RBP4, and TGFBR1) have been previously investigated in associated studies with reproduction traits of sows (Buske et al. 2006a, 2006b; Campbell et al., 2008; Chen et al., 2004; Franco et al., 2005; Jang et al., 2001; Terman, 2011; Vallet et al., 2005; Zhang et al., 2008). SNPs in five genes (ADRB2, CTSL, PNAS4, SCG2 and STAT5B) have been investigated for the first time in

association studies with reproduction traits in this trial, as these genes are involved in obesity and energy reserves for pregnancy (Muráni et al., 2009), blastocyst implantation (Spötter et al., 2001), embryonic development (Mo et al., 2008), ovulation rate (Du et al., 2008) and pre-implantation development (Nakasato et al., 2006), respectively. Genotyping was carried out using a primer extension assay on Sequenom MassArray platform and i-Plex™ Gold reagents by GeneSeek (Lincoln, NE, USA). Quality control and SNP evaluation criteria were the following: (i) 1.5% of the DNA samples were genotyped in duplicate and a concordance 499% between genotypes across duplicates was set as the threshold for validation; (ii) Hardy–Weinberg equilibrium (HWE) was tested in the analyzed markers; and (iii) only SNP with minor allele frequency (MAF) 45% were used in the association analyses.

S. Dall'Olio et al. / Livestock Science 155 (2013) 172–179

2.3. Statistical analyses Single marker association analyses were performed by using the general linear model (GLM) procedure of SAS (version 9.02). For NBA1 EBVs and NBA1 RRs, the initial model included the fixed effects of genotype for each marker analyzed individually and the herd (1–6). The herd was a significant factor of NBA1 EBVs variability (P o 0.0001), whereas it did not have a significant effect on NBA1 RRs, so for this trait the herd was removed from the model. For NBA1 and TNB1, the initial model included the fixed effects of herds (1–6), year (1 and 2), season of birth (1–4; starting with January to March), age at first farrowing (1–4 classes: o350 d, from 351 to 400 d, from 401 to 450 d and 4450 d), level of inbreeding of sows (1–5 classes:o0.0049, from 0.005 to 0.020, from 0.021 to 0.035, from 0.036 to 0.049 and 40.050), boar origin (four classes: ITLW, ITLA, ITDU, and unknown), and the SNP genotype (two or three classes). The level of sow inbreeding (average of 0.03570.037, range from 0.000 to 0.287) and boar origin were not significant (P 40.05) and were removed from the models. Additive effects of markers were estimated as half of the difference between the two homozygous groups. Dominance effects were estimated as the difference between the heterozygotes and the mean of the two homozygous genotypes. Significant differences from zero of the estimates of effects were tested by t-test. Haplotypes (SNPs of the GDF9 gene) were inferred by using the PHASE program v. 2.0 (Stephens et al., 2001). Haplotype substitution effects were estimated using the REG procedure of SAS (version 9.02) with a model including the number of copies (0, 1, 2) of three haplotypes. For NSB1 (ranging from 0 to 7, average of 0.5771.04; Table 1), data were categorized into two groups based on absence (group 1: 1038) and presence (group 2: 510) of NSB1. Then, for each SNP, differences in allele and genotype frequencies between these two groups were tested using the FREQ procedure of SAS (version 9.2), calculating Armitage's trend tests. Bonferroni correction was used to adjust P-nominal values in single marker analyses: threshold for significance was 0.05/17 ¼0.003 for 17 SNPs with MAF45%. Results with P-nominal values ranging from 0.003 to 0.05 were considered as suggestive evidence of associations.

3. Results 3.1. Polymorphic markers and allele frequencies Seventeen out of 21 genotyped SNPs had MAF 45% and did not deviate from the Hardy–Weinberg equilibrium in the genotyped ITLW sows (Table S2). These markers were used in association analyses with litter size traits at first parity. Four markers were excluded from the association study because they were monomorphic (EDNRA) or had 0%oMAFo4% (ADRB2, IGF1R, and PIT1) (Table S2). Several markers among the 17 polymorphic SNPs in the ITLW were not informative in the other investigated breeds (ITDU and ITLA; Table S2).

175

Haplotypes were inferred for the two GDF9 SNPs (AY649444:c.908C 4T and AY649444:c.1806T4C): three (ITLW and ITDU breeds) or two (ITLA) haplotypes were identified and their frequencies are reported in Table S3. 3.2. SNPs associated with reproduction traits in Italian Large White sows Significant and suggestively significant results obtained in the association analysis between 17 SNPs and five traits (raw phenotypes for NBA1, TNB1 and NSB1; EBV for NBA1 and RR for NBA1) are shown in Table 3 (all other results are reported in Tables S4 and S5). Most significant or suggestively significant results were obtained for NBA1 EBV (8 out of 12), one for NBA1 RR, one for NBA1, one for TNB1 and one for NSB1. Consistent results were obtained for the BMPR1B: c.960C4T SNP, for which the three different measures of NBA1 (NBA1 EBV, NBA1 RR, and NBA1) and TNB1 gave significant or suggestively significant results across all criteria used to consider litter size at first parity (P-nominal values¼0.0012, 0.0392, 0.0331 and 0.0096, respectively). These results strongly suggest that this polymorphism is associated with reproduction traits in the ITLW breed. SNPs in the FUT1, GPX5, RBP4 and TGFBR1 genes were significantly associated with NBA1 EBV (Po0.003). It is worth mentioning that only eight sows had the FUT1: g.915AA genotype (Table 3), and excluding this less frequent genotypic class, the result for this polymorphism was less significant, but still below P o0.05. Three markers (EPOR, GDF9:c.1806T 4C and STAT5B) were suggestively associated (P o0.05) with NBA1 EBV (Table 3). Six markers (BMPR1B, EPOR, FUT1, GPX5, RBP4 and TGFBR1) with P o0.05 showed additive genetic effects (Table 4). Haplotype substitution analysis indicated that GDF9 haplotype [T908:T1806] was associated with higher NBA1 EBV (P o0.05) (Table S3). The PNAS4 SNP showed suggestive differences of allele frequencies for NSB1 (Group 1: MAF¼0.18; Group 2: MAF¼0.15; P ¼0.0233; Tables 3 and S5). Armitage's trend test for the same marker and comparison was P ¼0.0218. 4. Discussion In our exploratory study, we used candidate SNPs selected from literature to identify markers associated with reproduction traits at first parity of ITLW sows, like TNB1, NSB1, and three different phenotypes of the same trait (NBA1 EBV, NBA1 RR and NBA1). This last approach was chosen to maximize the possibilities to identify true associations and overcome potential biases derived by the different ways in which the traits are recorded and/or calculated, considering their low heritability. Properties of EBV in association studies have been mainly evaluated in dairy cattle for which a lower or equivalent power has been reported as compared to raw or adjusted phenotypic parameters (Israel and Wellert, 2002). However, in a simulation study based on a commercial pig population, Ekine et al. (2010) indicated that using EBVs in association studies could result in a higher level of false positives for traits with low

176

S. Dall'Olio et al. / Livestock Science 155 (2013) 172–179

Table 3. Significant or suggestively significant results in the association analysis between the genotypes of 17 SNPs and reproduction traits. Gene symbol

SNP

Estimated least square means 7 SE (no. of sows)

Trait

P-value

Genotype 11

Genotype 12

Genotype 22

BMPR1B

c.960C 4T

NBA1 EBV NBA1 RR NBA1 TNB1

0.727 70.057 (511) −0.116 70.101 (511) 9.90 70.27 (511) 10.40 7 0.26 (511)

0.8377 0.051 (646) −0.0147 0.090 (646) 10.02 7 0.27 (646) 10.57 70.26 (646)

1.0707 0.079 (246) 0.3307 0.145 (646) 10.39 7 0.30 (646) 10.977 0.30 (646)

0.0012 0.0392 0.0331 0.0096

EPOR FUT1 GDF9 GPX5 PNAS4 RBP4 STAT5B TGFBR1

g.2373C4 T g.915G 4A c.1806T4C c.1939C4T g.1401T 4C g.447G4C g.191A 4 G g.64939A 4G

NBA1 EBV NBA1 EBV NBA1 EBV NBA1 EBV NSB1c NBA1 EBV NBA1 EBV NBA1 EBV

0.8737 0.045 (932) 0.8757 0.042 (1037) 0.8057 0.042 (1109) 0.9837 0.058 (557) – 1.06770.082 (238) 0.8797 0.044 (1003) 0.9657 0.056 (490)

0.754 70.062 (431) 0.6607 0.081 (237) 1.020 7 0.082 (262) 0.7267 0.054 (615) – 0.845 7 0.057 (610) 0.693 70.059 (440) 0.748 7 0.050 (713)

0.4647 0.160 (57) −0.628 70.421 (8) 0.478 70.331 (13) 0.649 7 0.094 (168) – 0.68570.073 (321) 0.826 70.158 (56) 0.6797 0.074 (288)

0.0160 0.0001b 0.0209 0.0002 0.0233 0.0010 0.0219 0.0012

a

NBA1¼number of piglets born alive at first parity; NSB1 ¼ number of stillborn piglets at first parity; TNB1¼ total number of piglets born at first parity; NBA1 EBVs¼ estimated breeding values for number of piglets born alive at first parity; NBA1 RRs ¼ random residuals for number of piglets born alive at first parity. a Significant associations at the Bonferroni adjusted alpha level (P o0.003) are in bold. b P¼ 0.0122 not including 8 homozygous FUT1:g.915AA genotypes. 4Only two GDF9:c.908TT samples were identified and not considered for statistical analysis. c Estimated by differences of allele frequencies between two groups of sows categorized based on absence and presence of NSB1 (Table S4).

Table 4. Additive and dominance effects 7 standard error (SE) of reproduction traits for six markers with P o0.05 in single marker analysis. Gene symbol

SNP

Trait

Additive-effect 7 SE

P-value

Dominance effect 7 SE

P-value

BMPR1B

c.960C 4T

NBA1 EBV NBA1 RR NBA1 TNB1

−0.1717 0.047 −0.223 7 0.088 −0.2427 0.093 −0.284 7 0.093

0.0003 0.0120 0.0098 0.0023

−0.0627 0.065 −0.1227 0.126 −0.1277 0.129 −0.1147 0.129

0.3447 0.3340 0.3263 0.3765

EPOR FUT1 GPX5 RBP4 TGFBR1

g.2373C4T g.915G 4A c.1939C 4T g.447G 4C g.64939A 4G

NBA1 NBA1 NBA1 NBA1 NBA1

0.2047 0.081 0.7517 0.211 0.1667 0.054 0.1917 0.051 0.1437 0.044

0.0117 0.0004 0.0020 0.0002 0.0012

0.0857 0.099 0.5377 0.224 −0.0907 0.070 −0.0317 0.069 −0.0747 0.062

0.3913 0.0169 0.1991 0.6567 0.2298

EBV EBV EBV EBV EBV

NBA1¼number of piglets born alive at first parity; TNB1 ¼total number of piglets born at first parity; NBA1 EBVs ¼ estimated breeding values for number of piglets born alive at first parity; NBA1 RRs ¼ random residuals for number of piglets born alive at first parity.

heritability. In a previous study in which we compared results of association analyses for production and carcass traits obtained with EBV and RR, we observed that EBVs tended to overestimate marker effects, raising the risk of false positive results in these analyses (Fontanesi et al., 2012). Based on these observations, only markers with consistent significant or suggestively significant results for more than one criterion could be deemed as potentially associated with litter size in the ITLW breed. All other results might be considered with less confidence and should be taken with caution if not confirmed in other studies. Therefore, the results of this trial indicated that the BMPR1B:c.960C4T SNP, showing a P-nominal valueo0.05 for NBA1 EBV, NBA1 RR, NBA1 and TNB1, is associated with litter size in the ITLW breed, with a beneficial additive effect of the minor allele (c.960T). In the ovine species, BMPR1B gene is a major gene (originally named FecB) affecting ovulation rate and prolificacy associated with the Booroola hyperprolific phenotype (Souza et al., 2001). The porcine BMPR1B gene maps in a region of SSC8 close to QTL for OR, litter size and prenatal

survival (Hu et al., 2010; http://www.animalgenome.org/ QTLdb/). A suggestive association of the BMPR1B gene on NBA1 was already reported in Iberian  Meishan F2 sows (Tomas et al., 2006). BMPR1B expression in the porcine granulosa and theca cells is temporally regulated and positively correlated with plasma oestradiol suggesting that its regulation by estrogen may be implicated in normal folliculogenesis (Paradis et al., 2009). Sun et al. (2011) showed that this gene was differentially expressed in ovarian follicles at the pre-ovulatory stage in Chinese Taihu and Large White cycling sows and its level of expression was associated with litter size traits in these pigs. As we investigated a BMPR1B synonymous polymorphism, the analyzed SNP could be in close linkage disequilibrium with a causative mutation in this gene or in a close gene. Most of the other genes (EPOR, FUT1, GDF9, PNAS4, RBP4, STAT5B and TGFBR1), for which SNPs were associated or suggestively associated with one or the other litter size measure (mostly NBA1 EBV), play important biological roles directly or indirectly related with reproduction

S. Dall'Olio et al. / Livestock Science 155 (2013) 172–179

(Table S1). For a few of them (such as GDF9 and PNAS4), other studies conducted on pigs have already demonstrated a functional relevance on litter size or differences in expression in key reproductive tissues (Mo et al., 2008; Zhang et al., 2008). Results obtained in ITLW sows suggested a favorable additive effect of the g.447G RBP4 allele on NBA1 EBVs, in accordance with the effect of the corresponding allele 1 of the MspI polymorphism on NBA1 raw data found in other studies (Muñoz et al., 2010; Rothschild et al., 2000). Spötter et al. (2009) identified an effect of allele 1 on NBA of all parities in German Landrace, but not in German Large White sows. Contrasting results have been obtained by others (Blowe et al., 2006; Drögemüller et al., 2001; Linville et al., 2001). The FUT1:g.915G 4A SNP was the most significant marker associated with NBA1 EBV in ITLW sows (Table 3). FUT1 is associated with F18+ Escherichia coli infections, causing post-weaning diarrhea and/or edema disease in young pigs (Meijerink et al., 1997, 2000). FUT1: g.915AA genotype is considered the resistant genotype, whereas susceptible pigs have either the g.915AG and the g.915GG genotypes (Meijerink et al., 2000). In our data set, allele g.915A (p.103Thr) was the less frequent one in all investigated breeds (from 0.02 in the PI to 0.29 in the ITDU) and the resistant genotype (g.915AA) was present only in ITLW (0.01) and in ITDU (0.06). Our association results indicated a negative effect of the g.915AA genotype on NBA1 EBVs. A similar effect of this genotype was found in a local Czech breed (Horak et al., 2005). If results obtained in ITLW sows are confirmed, the negative effect of allele g.915A on NBA1 could explain its low frequency in this Italian heavy pig breed as well as in the hyper-prolific Meishan and almost all native Chinese breeds (Bao et al., 2008). Considering its antagonist association with postweaning diarrhea resistance, as a consequence, selection for increased litter size could decrease, at least in part, disease resistance. This gene might represent a good example of the problems that pig breeding and in general selection programs have to face. A similar opposite effect on disease resistance and production traits have been recently demonstrated for another marker (MUC4: g.8227C 4G) associated with susceptibility enterotoxigenic Escherichia coli K88 causing diarrhea and death of piglets before or just after weaning (Fontanesi et al., 2012). SNPs in two genes (GDF9:c.1806T4C and TGFBR1), which are involved in signaling systems that regulate ovarian follicular development and maturity (Paradis et al., 2009), were suggestively or significantly associated with NBA1 EBVs. A preliminary study of association between a SNP in pig GDF9 gene and litter size, based on just a few sows, did not show any effect on reproduction traits, even if that report indicated differences of expression in the pituitary tissue between Duroc and hyperprolific Erhualian sows (Zhang et al., 2008). The TGFBR1 gene was recently reported to be associated with growth, carcass and number of corpora lutea (Chen et al., 2012). STAT5B is a key component of the cytokine signaling pathway involved in diverse biological processes including adiposity, energy homeostasis and regulation of pre-

177

implantation development (Ballester et al., 2006; Nakasato et al., 2006). The STAT5B:g.191A 4G SNP was suggestively associated with NBA1 EBV (P o0.05), providing a possible link between energy metabolism and reproduction performances in sows as identified by Rempel et al. (2010). Results obtained for the EPOR SNP, a candidate gene for fetal survival, revealed a putative negative effect of the minor allele (g.2373T). The analyzed polymorphism, creating an extra hematopoietic transcription factor GATA-1 site when the allele g.2373T occurs, was already associated with UC in a four-way cross population (Bidanel, 2011; Vallet et al., 2005). In addition, our association results indicated a favorable additive effect of the major GPX5:c.n1897C allele on NBA1 EBVs. We previously identified an association of the same GPX5 SNP with number of functional teats, a trait related to the mothering ability of the sows affecting piglets survival and reproductive efficiency (Dall'Olio et al., 2012). GPX5, encoding an antioxidant enzyme that protects sperm plasma membranes and DNA integrity, was already investigated as candidate for litter size for its location in a region of SSC7 containing QTLs for uterine capacity, ovulation rate and litter size (Buske et al., 2006b). No difference of allele frequency was identified in two extreme groups for litter size, based on just a few sows (Buske et al., 2006b). PNAS4 SNP was suggestively associated with NSB1 (P o0.05). Variability in the porcine PNAS4 gene was shown to affect fat deposition in Chinese and European pig breeds (Mo et al., 2008) providing a potential link between fatness and reproduction traits in ITLW sows. It is worth mentioning that at present the ITLW breed is under co-selection for both prolificacy and an appropriate backfat thickness for the production for high quality processed products (Bosi and Russo, 2004). This peculiarity might indirectly influence reproduction performances, due to its positive relationship with back-fat thickness (Estany et al., 2002).

5. Conclusion The current pilot study provided interesting results. Consistent evidences indicated that variability in the BMPR1B gene, or in another close gene, may be responsible for a fraction of the genetic variability for reproduction traits of ITLW sows at first farrowing. Additional studies should be carried out to better characterize this gene and its variability and to confirm its involvement in affecting female reproduction traits. For several other genes, results should be considered as preliminary and further studies are necessary. Results will be further exploited using new sequencing technologies and genome wide association approaches, as already done in other pig lines (Onteru et al., 2012; Uimari et al., 2011)

Conflict of interest statement This data has not been published and/or sent to other journals before. The authors have no conflict of interest.

178

S. Dall'Olio et al. / Livestock Science 155 (2013) 172–179

Acknowledgments We thank ANAS for providing data. This work was supported by the Italian MiPAAF SelMol and Innovagen projects.

Appendix A. Supporting information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.livsci. 2013.05.012.

References Ballester, M., Sardina, M.T., Folch, J.M., 2006. Polymorphism and chromosomal localization of the porcine signal transducer and activator of transcription 5B gene (STAT5B). J. Anim. Breed. Genet. 123, 284–287. Bao, W.B., Wu, S.L., Musa, H.H., Zhu, G.D., Chen, G.H., 2008. Genetic variation at the alpha-1-fucosyltransferase (FUT1) gene in Asian wild boar and Chinese and Western commercial pig breeds. J. Anim. Breed. Genet. 125, 427–430. Bidanel, J-P., 2011. Biology and genetics of reproduction. In: Rothschild, M.F., Ruvinsky, A. (Eds.), The Genetics of the Pigs, CABI Cambridge, MA, USA, pp. 218–241. Blowe, C.D., Boyette, K.E., Ashwell, M.S., Eisen, E.J., Robison, O.W., Cassady, J.P., 2006. Characterization of a line of pigs previously selected for increased litter size for RBP4 and follistatin. J. Anim. Breed. Genet. 123, 389–395. Bosi, P., Russo, V., 2004. The production of the heavy pigs for high quality processed products. Ital. J. Anim. Sci. 3, 309–321. Buske, B., Sternstein, I., Reißmann, M., Brockmann, G., 2006a. Detection of novel single-nucleotide polymorphisms (SNPs) in the CYP21 gene and association analysis of two SNPs for CYP21 and ESR2 with litter size in a commercial sow population. J. Anim. Breed. Genet. 123, 343–348. Buske, B., Sternstein, I., Reißmann, M., Reinecke, P., Brockmann, G., 2006b. Analysis of association of GPX5, FUT1 and ESR2 genotypes with litter size in a commercial pig cross population. Arch. Tierz. Dummerstorf 49, 259–268. Campbell, E.M., Nonneman, D.J., Kuehn, L.A., Rohrer, G.A., 2008. Genetic variation in the mannosidase 2B2 gene and its association with ovulation rate in pigs. Anim. Genet. 39, 515–519. Chen, K., Hawken, R., Flickinger, G.H., Rodriguez-Zas, S.L., Rund, L.A., Wheeler, M.B., Abrahamsen, M., Rutherford, M.S., Beever, J.E., Schook, L.B., 2012. Association of the porcine transforming growth factor beta type I receptor (TGFBR1) gene with growth and carcass traits. Anim. Biotechnol. 23, 43–63. Chen, C., Chang, T., Su, H.Y., 2004. Characterization of porcine leptin receptor polymorphisms and their association with reproduction and production traits. Anim. Biotechnol. 15, 89–102. Chen, K., Rund, L.A., Beever, J.E., Schook, L.B., 2006. Isolation and molecular characterization of the porcine transforming growth factor beta type I receptor (TGFBR1) gene. Gene 384, 62–72. Dall'Olio, S., Fontanesi, L., Tognazzi, L., Buttazzoni, L., Gallo, M., Russo, V., 2011. ESR1 and ESR2 gene markers are not associated with number of piglets born alive in Italian Large White sows. Ital. J. Anim. Sci. 10, e35. Dall'Olio, S., Fontanesi, L., Tognazzi, L., Buttazzoni, L., Gallo, M., Russo, V., 2012. Association analysis between DNA markers and number of functional teats in Italian Large White pigs. Veterinary Science. Current Aspects in Biology, Animal Pathology, Clinic and Food Hygiene. Springer-Verlag pp. 181–185. Distl, O., 2007. Mechanisms of regulation of litter size in pigs on the genome level. Reprod. Domest. Anim. 42, 10–16. Drögemüller, C., Hamann, H., Distl, O., 2001. Candidate gene makers for litter size in different German pigs lines. J. Anim. Sci. 79, 2565–2570. Du, H.L., Chen, J., Zhang, Y.S., Zhang, X.Q., 2008. Molecular cloning, mapping, and polymorphism of the porcine SCG2 gene. Biochem. Genet. 46, 369–379. Ekine, C.C., Rowe, S.J., Bishop S.C., de Koning D.J., 2010. What is the best phenotype for genome-wide association studies in data with defined pedigrees? In: Proceedings of the 9th World Congress on Genetics Applied to Livestock Production, Leipzig, Germany E0263.

Estany, J., Villalb, D., Tor, M., Cubiló, D., Noguera, J.L., 2002. Correlated response to selection for litter size in pigs: II. Carcass, meat, and fat quality traits. J. Anim. Sci. 80, 2566–2573. Fontanesi, L., Speroni, C., Buttazzoni, L., Scotti, E., Dall'Olio, S., Nanni Costa, L., Davoli, R., Russo, V., 2010a. The IGF2 intron3-g.3072G 4A polymorphism is not the only Sus scrofa chromosome 2p mutation affecting meat production and carcass traits in pigs: evidences from the effects of a cathepsin D (CTSD) gene polymorphism. J. Anim. Sci. 88, 2235–2245. Fontanesi, L., Bertolini, F., Dall'Olio, S., Buttazzoni, L., Gallo, M., Russo, V., 2012a. Analysis of association between the MUC4:g.8227C4G polymorphism and production traits in Italian heavy pigs using a selective genotyping approach. Anim. Biotechnol. 23, 147–155. Fontanesi, L., Speroni, C., Buttazzoni, L., Scotti, E., Nanni Costa, L., Davoli, R., Russo, V., 2010b. Association between cathepsin L (CTSL) and cathepsin S (CTSS) polymorphisms and meat production and carcass traits in Italian Large White pigs. Meat Sci. 85, 331–338. Franco, M.M., Antunes, R.C., Silva, H.D., Goulart, L.R., 2005. Association of PIT1, GH and GHRH polymorphisms with performance and carcass traits in Landrace pigs. J. Appl. Genet. 46, 195–200. Horak, P., Urban, T., Dvořák, J., 2005. The FUT1 and ESR genes: their variability and associations with reproduction in Přeštice Black-Pied sows. J. Anim. Breed. Genet. 122, 210–213. Hu, Z.-L., Park, C.A., Fritz, E.R., Reecy, J.M., 2010. QTLdb: a comprehensive database tool building bridges between genotypes and phenotypes. In: Proceedings of the 9th World Congress on Genetics Applied to Livestock Production, Leipzig, Germany, August 2010 1–6. Israel, C., Wellert, J.I., 2002. Estimation of quantitative trait loci effects in dairy cattle populations. J. Dairy Sci. 85, 1285–1297. Jang, Z., Gibson, J.P., Archibald, A.L., Haley, C.S., 2001. The porcine gonadotropin-releasing hormone receptor gene (GNRHR): genomic organization, polymorphisms, and association with the number of corpora lutea. Genome 44, 7–12. Kim, J.G., Nonneman, D., Vallet, J.L., Christenson, R.K., 2002. Mapping of the porcine alpha-fetoprotein (AFP) gene to swine chromosome 8. Anim. Genet. 33, 471–472. Kim, J.G., Rohrer, G.A., Vallet, J.L., Christenson, R.K., Nonneman, D., 2004. Addition of 14 anchored loci to the porcine chromosome 8 comparative map. Anim. Genet. 35, 474–475. Kopečný, M., Stratil, A., Bartenschlager, H., Peelman, L.J., Van Poucke, M., Geldermann, H., 2002. Linkage and radiation hybrid mapping of the porcine IGF1R and TPM2 genes to chromosome 1. Anim. Genet. 33, 398–399. Linville, R.C., Pomp, D., Johnson, R.K., Rothshild, M.F., 2001. Candidate genes analysis for loci affecting litter size and ovulation rate in pigs. J. Anim. Sci. 79, 60–67. Mackowski, M., Szymoniak, K., Szydlowski, M., Kamyczek, M., Eckert, R., Rozycki, M., Switonski, M., 2005. Missense mutations in exon 4 of the porcine LEPR gene encoding extracellular domain and their association with fatness traits. Anim. Genet. 36, 135–137. Meijerink, E., Fries, R., Vögeli, P., Masabanda, J., Wigger, G., Stricker, C., Neuenschwander, S., Bertschinger, H.U., Stranzinger, G., 1997. Two α (1,2) fucosyltransferase genes on porcine Chromosome 6q11 are closely linked to the blood group inhibitor (S) and Escherichia coli F18 receptor (ECF18R) loci. Mamm. Genome. 8, 736–741. Meijerink, E., Neuenschwander, S., Fries, R., Dinter, A., Bertschinger, H.U., Stranzinger, G., Vögeli, P., 2000. A DNA polymorphism influencing α (1,2)fucosyltransferase activity of the pig FUT1 enzyme determines susceptibility of small intestinal epithelium to Escherichia coli F18 adhesion. Immunogenetics 52, 129–136. Mo, D., Zhu, Z., Te Pas, M.F.W., Li, X., Yang, S., Wang, H., Wang, H., Li, K., 2008. Characterization, expression profiles, intracellular distribution and association analysis of porcine PNAS-4 gene with production traits. BMC Genet. 9, 40. Muñoz, M., Fernández, A.I., Ovilo, C., Muñoz, G., Rodriguez, C., Fernandez, A., Alves, E., Silió, L., 2010. Non-additive effects of RBP4, ESR1 and IGF2 polymorphisms on litter size at different parities in a Chinese– European porcine line. Genet. Sel. Evol. 42, 23. Muráni, E., Ponsuksili, S., Srikanchai, T., Maak, S., Wimmers, K., 2009. Expression of the porcine adrenergic receptor beta 2 gene in longissimus dorsi muscle is affected by cis-regulatory DNA variation. Anim. Genet. 40, 80–89. Nakasato, M, Shirakura, Y., Ooga, M., Iwatsuki, M., Ito, M., Kageyama, S., Sakai, S., Nagata, M., Aoki, F., 2006. Involvement of the STAT5 signaling pathway in the regulation of mouse preimplantation development. Biol. Reprod. 75, 508–517. Onteru, S.K., Fan, B., Du, Z-Q., Garrick, D.J., Stalder, K.J., Rothschild, M-F., 2012. A whole-genome association study for pig reproductive traits. Anim. Genet. 43, 18–26.

S. Dall'Olio et al. / Livestock Science 155 (2013) 172–179

Paradis, F., Novak, S., Murdoch, G.K., Dyck, M.K., Dixon, W.T., Foxcroft, G.R., 2009. Temporal regulation of BMP2, BMP6, BMP15, GDF9, BMPR1A, BMPR1B, BMPR2 and TGFBR1 mRNA expression in the oocyte, granulosa and theca cells of developing preovulatory follicles in the pigs. Reproduction 138, 115–129. Rempel, L.A., Nonneman, D.J., Wise, T.H., Erkens, T., Peelman, L.J., Rohrer, G.A., 2010. Association analyses of candidate single nucleotide polymorphisms on reproductive traits in swine. J. Anim. Sci. 88, 1–15. Rothschild, M., Jacobson, C., Vaske, A., Tuggle, C., Short, T.H., Sasaki, S., Eckardt, G.R., McLaren, D.G., 1994. A major gene for litter size in pigs. Proc 5th World Congr. Genet. Appl. Livest. Prod. 21, 225–228. Rothschild, M.F., Messer, L., Day, A., Wales, R., Short, T., Southwood, O., Plastow, G., 2000. Investigation of the retinol-binding protein 4 (RBP4) gene as a candidate gene for increased litter size in pigs. Mamm. Genome 11, 75–77. Shimanuki, S., Mikawa, A., Miyake, Y., Hamasima, N., Mikawa, S., Awata, T., 2005. Structure and polymorphism analysis of transforming growth factor beta receptor 1 (TGFBR1) in pigs. Biochem. Genet. 43, 491–500. Sironen, A.I., Uimari, P., Serenius, T., Mote, B., Rothschild, M., Vilkki, J., 2010. Effect of polymorphisms in candidate genes on reproduction traits in Finnish pig populations. J. Anim. Sci. 88, 821–827. Souza, C.J.H., MacDougall, C., Campbell, B.K., McNeilly, A.S., Baird, D.T., 2001. The Booroola (FecB) phenotype is associated with a mutation in the bone morphogenetic receptor type 1 B (BMPR1B) gene. J. Endocrinol. 169, R1–R6. Spötter, A., Müller, S., Hamann, H., Distl, O., 2009. Effect of polymorphisms in the genes for LIF and RBP4 on litter size in two German pig lines. Reprod. Domest. Anim. 44, 100–105. Spötter, A., Distl, O., 2006. Genetic approaches to the improvement of fertility traits in the pigs. Vet. J. 172, 234–247.

179

Spötter, A., Drögemüller, C., Kuiper, H., Brenig, B., Leeb, T., Distl, O., 2001. Characterization and comparative mapping of the porcine CTSL gene indicates a novel synteny between HSA9q21–4q22 and SSC10q11– 4q12. Cytogenet. Cell. Genet. 95, 92–96. Stephens, M., Smith, N.J., Donnelly, P., 2001. A new statistical method for haplotype reconstruction from population data. Am. J. Hum. Genet. 68, 978–989. Sun, X., Mei, S., Tao, H., Wang, G., Su, L., Jang, S., Deng, C., Xiong, Y., Li, F., 2011. Microarray profiling for differential gene expression in PMSGhCG stimulated preovulatory ovarian follicles of Chinese Taihu and Large White sows. BMC Genomics 12, 111. Terman, A., 2011. The IGF1R gene: a new marker for reproductive performance traits in sows? Acta Agric. Scand. A 61, 67–71. Tomas, A., Frigo, E., Casellas, J., Ramírez, Ovilo, C.O., Noguera, J.L., Sanchez, A., 2006. An association study between polymorphisms of the porcine bone morphogenetic protein receptor type 1b (BMPR1B) and reproductive performance of Iberian  Meishan F2 sows. Anim. Genet. 37, 297–298. Uimari, P., Sironen, A., Sevón-Aimonen, M.L., 2011. Whole-genome SNP association analysis of reproduction traits in the Finnish Landrace pig breed. Genet. Sel. Evol. 43, 42. Vallet, J.L., Freking, B.A., Leymaster, K.A., Christenson, R.K., 2005. Allelic variation in the erythropoietin receptor gene is associated with uterine capacity and litter size in swine. Anim. Genet. 36, 97–103. Vallet, J.L., Nonneman D.J., Kuehn L.A., 2010. Quantitative genomics and female reproduction. In: Zhihua Jang, Troy L. Ott (Eds.). Reproductive Genomics in Domestic Animals pp. 23–52. Zhang, Y., Du, H., Chen, J., Yang, G., Zhang, X., 2008. Porcine growth differentiation factor 9 gene polymorphisms and their association with litter size. J. Genet. Genomics 35, 163–169.