Relationship of Polymorphisms Within ZBED6 Gene and Growth Traits in Beef Cattle

GENE-38577; No. of pages: 5; 4C: Gene xxx (2013) xxx–xxx

Contents lists available at SciVerse ScienceDirect

Gene

journal homepage: www.elsevier.com/locate/gene

Relationship of polymorphisms within ZBED6 gene and growth traits in beef cattle

Yong-Zhen Huang a, Hua He a, Zhao-Yang Zhan a, Yu-Jia Sun a, Ming-Xun Li a, Xian-Yong Lan a, Chu-Zhao Lei a, Chun-Lei Zhang b, Hong Chen a,⁎ a College of Animal Science and Technology, Northwest A&F University, Shaanxi Key Laboratory of Molecular Biology for Agriculture, Yangling, Shaanxi 712100, China b Institute of Cellular and Molecular Biology, Jiangsu Normal University, Xuzhou, Jiangsu 221116, China article info abstract

Article history: ZBED6 is a novel transcription factor that was identified and shown to act as a repressor of IGF2 transcription Accepted 15 April 2013 in skeletal muscle. The aim of this study was to examine the association of the ZBED6 polymorphism with Available online xxxx growth traits in beef cattle breed. Three single nucleotide polymorphisms (SNPs) were identified in the bovine ZBED6 by sequencing pooled DNA samples (Pool-Seq) and forced polymerase chain reaction–restriction Keywords: fragment length polymorphism (Forced PCR–RFLP) methods. Overall, we reported one mutation (SNP1) in the Cattle promoter region and two missense mutations (SNP2 and 3) in the coding region (single exon) within the bovine ZBED6 Association analysis ZBED6 gene, and the haplotype variability and extent of linkage disequilibrium (LD) in 1522 individuals Growth traits representing four main cattle breeds from China (Nanyang, NY; Qinchuan, QC; Jiaxian, JX; and Chinese Holstein, CH). We also investigated haplotype frequencies and linkage disequilibrium coefficients for three SNPs in all study populations. LD and haplotype structure of ZBED6 were different between breeds. The result of haplotype analysis of three SNPs showed that eight different haplotypes were identified in all breeds. The wild-type haplotype (Hap 1: GCA) and mutant-type haplotype (Hap 8: AGG) shared by all four populations accounted for 15.1, 29.8, 21.7, 57.5% and 9.5, 8.6, 16.7, 0% of all haplotypes were observed in NY, QC, JX and CH, respectively. The statistical analyses indicated that three SNPs were significantly associated with growth traits in NY cattle population (P b 0.05 or P b 0.01) at five different ages. The results of this study suggest that the ZBED6 gene possibly is a strong candidate gene that affects growth traits in beef cattle breeding program. © 2013 Elsevier B.V. All rights reserved.

1. Introduction for a repressor and leads to a threefold up-regulation of IGF2 expression in skeletal muscle. The mutation in IGF2 abrogates in vitro interaction Zinc finger, BED-type containing 6 (ZBED6), is a novel transcription with a repressor, which leads to a threefold increase of IGF2 expression factor that was identified and shown to act as a repressor of IGF2 tran- in postnatal muscle. The mutation has major effects on muscle growth, scription in skeletal muscle myogenesis and development (Markljung heart size, and fat deposition (Van Laere et al., 2003). et al., 2009). The ZBED6 protein is highly conserved among 26 different species, The IGF2 mutation, which causes a single nucleotide substitution, including the pig (Markljung et al., 2009). The ZBED6 has a broad namely an IGF2 intron 3-G3072A transition abrogates a binding site tissue distribution by northern blot and quantitative real-time PCR (qPCR) (Markljung et al., 2009). Electrophoretic mobility shift assay (EMSA) showed that ZBED6 bound with the wild-type q sequence, proposed that ZBED6 is a bona fide repressor of IGF2 transcription Abbreviations: bp, base pair(s); BW, body weight; CH, Chinese Holstein; EMSA, that interacts with quantitative trait nucleotide (QTN) region of IGF2 electrophoretic mobility shift assay; GLM, general linear models; He, heterozygosity; HWE, Hardy–Weinberg equilibrium; IGF2, Insulin-like growth factor 2; JX, Jiaxian cattle; (Markljung et al., 2009). The qPCR showed that ZBED6 mRNA was LD, linkage disequilibrium; LSM, least square means estimates; Ne, effective allele decreased to 70% in ZBED6 silenced C2C12 cells. Luciferase activity numbers; NY, Nanyang; PIC, Polymorphism information content; QC, Qinchuan; QTN, was reduced to a large extent in wild type q construct when compared quantitative trait nucleotide; SNPs, single nucleotide polymorphisms; SPSS, statistical with construct containing P3 alone. These results suggested that ZBED6 product and service solutions; ZBED6, Zinc finger, BED-type containing 6. ⁎ Corresponding author at: No.22 Xinong Road, College of Animal Science and Technology, represses expression of IGF2 promoter 3 after binding with the QTN Northwest A&F University, Yangling, Shaanxi 712100, China. Tel.: +86 29 87092012; (Markljung et al., 2009). fax: +86 29 87092164. The purpose of this study was to identify single nucleotide polymor- E-mail addresses: [email protected] (Y.-Z. Huang), phisms (SNPs) in the bovine ZBED6 gene, and to carry out haplotype [email protected] (H. He), [email protected] (Z.-Y. Zhan), construction and association analysis so as to contribute to the under- [email protected] (Y.-J. Sun), [email protected] (M.-X. Li), [email protected] (X.-Y. Lan), [email protected] (C.-Z. Lei), standing of the role of ZBED6 in variation of growth traits in cattle, [email protected] (C.-L. Zhang), [email protected] (H. Chen). which possibly contributed to animal breeding and genetics.

2. Materials and methods 2.3. Genotyping

2.1. Animal source, DNA preparation and growth data PCR primers were redesigned to facilitate genotyping of the three SNPs using DNA Pool-Seq and Forced PCR–RFLP techniques in four We investigated to explore the genetic variation in bovine ZBED6, study populations. an initial investigation of 1522 individuals representing four cattle The three SNPs (SNP1: ZBED6 Promoter G-826A; SNP2: ZBED6 Exon breeds in China: Nanyang cattle (NY, n = 265), Qinchuan cattle 1C680G;SNP3:ZBED6 Exon 1 A1043G) of the ZBED6 gene were geno- (QC, n = 723), Jiaxian cattle (JX, n = 440), and Chinese Holstein typed by forced PCR–RFLP methods with one of the primers containing (CH, n = 94). Nanyang, Qinchuan, and Jiaxian are three important mismatches of one or two nucleotides, which enable the use of restric- breeds for beef production in China; whereas Chinese Holstein is a tion enzymes for discriminating sequence variations (Huang et al., dairy breed. These four breeds are the main breeds of China and 2011a). Primers, restriction enzymes selected (MBI, Vilnius, Lithuania), they are reared in the provinces of Henan and Shaanxi, respectively. and fragment sizes are given in Table S1. Calves were weaned on average at 6 months of age and raised from Then, aliquots of 10 μL PCR products were digested with 10 U Hae weaning to slaughter on a diet of corn and corn silage. The animals of III/Hinf I/Mlu I for 8 h at 37 °C. The digested products were detected each breed were selected to be unrelated for at least three generations, by electrophoresis with about 1–2 h in the 4.0% agarose gel stained with the aim of having diverse lineages within each breed. The Nanyang with ethidium bromide. animals were from the Nanyang Cattle breeding center (Nanyang City, Henan Province, PR China); the Jiaxian animals were from the Jiaxian 2.4. Data analyses Cattle breeding farm (Jiaxian County, Henan Province, PR China); Qinchuan cattle were from the reserve farm (Fufeng County, Shaanxi 1. Gene frequencies were determined for each breed by direct province, PR China) and the Qinchuan Cattle fineness breeding center counting. Hardy–Weinberg equilibrium (HWE) had been tested (Yangling County, Shaanxi province, PR China); the Chinese Holstein based on likelihood ratio for different locus-population combinations animals were from a milk cattle breeding farm (Xi'an City, Shaanxi and the number of observed and effective alleles by POPGENE Province, PR China). software (Version 3.2; Yeh et al., 1999). Genomic DNA of 1522 cattle was isolated from 2% heparin-treated 2. Population genetic indexes, such as He (gene heterozygosity), Ho blood samples and stored at −80 °C, following the standard procedures (gene homozygosity; Ho + He = 1), and Ne (effective allele num- (Sambrook and Russell, 2002). The content of DNA was estimated spec- bers; reciprocal of homozygosity) were computed by POPGENE trophotometrically, and then the genomic DNA was diluted to 50 ng/μL. software; He and Ho are a measure of genic variation of a population All DNA samples were stored at −20 °C for subsequent analysis. and were calculated according to Nei's (1973) methods; Polymor- We quantified the growth traits of 265 Nanyang and 143 Jiaxian phism information content (PIC) was calculated according to cattle. The animals were weaned at an average of 6 months of age Botstein et al.'s (1980) methods. and raised from weaning to slaughter on a corn–corn silage diet. The 3. The linkage disequilibrium (LD) structure as measured by D′ and r2 traits under study were the body weight at birth (BW0), 6 months was performed with the HAPLOVIEW software (Version 3.32) (BW6), 12 months (BW12), 18 months (BW18), and 24 months (Barrett et al., 2005). The most commonly used measures of LD are (BW24). Statistical analysis was performed as described in Gilbert D′ and r2. Several studies have shown that r2 is not as sensitive as et al. (1993). D′ to allele frequencies, therefore, r2 was used as a pairwise measure of LD (Amandine et al., 2010; Huang et al., 2011b; Zhao et al., 2007). 4. Haplotypes were obtained for each animal using the PHASE com- 2.2. Variants discovery (primer design, PCR amplification and commercially puter program (Version 2.1) (Huang et al., 2013; Stephens et al., sequenced) 2001). 5. The association analyses between single SNP marker genotypes Primers used to amplify the bovine ZBED6 gene were designed and growth traits were analyzed using the general linear model from a published gene sequence (Ensembl: ENSBTAG00000045702; (GLM) procedure of SPSS software (Version 18.0) (Derecka et al., NCBI: AC_000173.1). Primers, restriction enzymes selected (MBI, 2009; Holzer and Precht, 1992; Huang et al., 2010a). Vilnius, Lithuania), and fragment sizes are given in Table S1. The detec- Statistical analysis was performed on records of growth traits in tion results of allelic variation at the SNPs were based on the electro- NY cattle breed. All analyses were done in two steps, first using a phoretic pattern of the restriction enzyme-treated PCR products. full animal model and then using a reduced animal model. The full an- PCR was performed in 25 μL of reaction volume, containing 50– imal model included fixed effects of marker genotype, birth year, sea- 100 ng genomic DNA, 10 pM of each primer, 1 × buffer (including son of birth (spring vs. fall), age of dam, sire, farm, sex and random 1.5 mM MgCl ), 200 μM dNTPs and 1.5 units of Taq DNA polymerase 2 effects (permanent environment, animal, and residual). The effect asso- (MBI, Vilnius, Lithuania). ciated with season of birth, age of dam, sire, farm and sex was not PCR reactions were carried out using a PCR thermal cycler system matched in the linear model, as the preliminary statistical analyses indi- (Bio-Rad, USA). An initial denaturation for 3 min at 95 °C; 35 cycles of cated that these effects did not have a significant influence on variability 94 °C for 30 s; annealing at Annealing temperature (AT) (Table S1) for of traits in the analyzed populations. 35 s; primer extension at 72 °C for 45 s. The final extension was The reduced model was used in the final analysis (Boldman et al., performed at 72 °C for 10 min. 1993; Henderson, 1986; Hickford et al., 2009). The SPSS software Five pairs of primers were designed for PCR amplification of the (Version 18.0) was used to analyze the relationship between the ZBED6 gene from cattle genomic DNA. PCR products were commercially genotypes and traits in cattle (Huang et al., 2010b, 2010c). The reduced sequenced for genetic variants discovery. Generally, PCR products am- linear model included fixed effects of age and genotype. The linear plified from genomic DNA were directly sequenced in both directions model: (Sangon, Shanghai, China). In an effort to discover SNPs in a cost-effective manner, SNP discovery ¼ μ þ þ þ ; was implemented by sequencing pooled DNA samples (Pool-Seq), Yijk Ai Gj Eijk which were amplified from DNA of founder animals from the study populations. The sequences were imported into the BioXM software where Yijk = the trait measured on each of the ijkth animal; μ =the (Version 2.6) and were analyzed and searched for SNPs. overall population mean; Ai = fixed effect due to the ith age; Gj =the

fixed effect associated with jth genotype; Eijk = the random error. The genotype GG (Table S1; Fig. 1c). The frequency of allele SNP3-A least square means with standard errors for different genotypes and (NY: 70.45%; QC: 72.50%; JX: 66.47%; CH: 100%) has shown a high growth traits were used. In this model, age and marker genotypes prevalence in the four cattle populations and AA genotype was were considered as fixed effects, dam (litter) as a random effect and more frequent than other genotypes (NY: 54.55%, QC: 56.50%, JX: growth traits as the dependent variables. 52.02% and CH: 100.00%). The genotypic frequencies of the SNP1 locus in NY and SNP3 locus in NY, QC, and JX cattle populations, the genotypic frequencies were 3. Results deviated from the Hardy–Weinberg equilibrium (P > 0.05), which showed that there were not a dynamic equilibrium even in artificial 3.1. SNPs identified selection, migration, and genetic drift function in the these cattle populations. The bovine ZBED6 gene is located on chromosome 16, contains only a single exon, and encodes 980 amino acids. In the present study, genomic DNA of all four cattle breeds was successfully amplified using 3.2. Diversity analyses primer pairs for the ZBED6 gene (Table S1; Fig. 1). The genotyping of the three SNPs was successfully implemented Genetic indices (Ho, He, Ne and PIC) in these four Chinese cattle using Forced PCR–RFLP in these animals (Tables S1 and S3). It was populations were presented in Table S3. The values of the difference firstly found that three novel mutations could be detected by Hae between expected and observed He (gene heterozygosity) were III/Hinf I/Mlu I endonucleases restriction site for these SNPs in bovine approaching 0.5. The values of Ne (effective allele numbers) were ZBED6 gene. It is a very useful strategy to scan large sample size approaching 2. The maximum and minimum PIC values were 0.2810 sequence variants with DNA sequencing and Forced PCR–RFLP methods, and 0.3725 except the CH cattle populations in SNP1-5 and 7 loci. which will overcome the inaccuracy, the complicated technical We are according to the classification of PIC (PIC value b 0.25, low demands, slow speed and unstable reproducibility. Interestingly, in polymorphism; 0.25 ≤ PIC value ≤ 0.5, intermediate polymorphism; this study, the Forced PCR–RFLP method was successfully carried out and PIC value > 0.5, high polymorphism). At the SNP1-7, the NY, QC to accurately detect the polymorphism of the bovine ZBED6 gene and JX cattle populations belonged to intermediate genetic diversity. (Fig. 1). This reflected that there was not a very high genetic diversity within At the SNP1 — Hae III locus (−826G > A), digestion of the 328 bp Chinese bovine IGF2 and ZBED6 gene in the analyzed populations. PCR fragment of ZBED6 promoter with Hae III resulted in fragment Genetic diversity is essential for preservation of adaptive potential of lengths of 297 and 31 bp band for genotype GG individual (homozy- species and improvement of production of potentially high selected gous); GA (heterozygous) showed 328, 297, and 31 bp bands; AA breeds. (homozygous) showed 328 bp bands (Table S1; Fig. 1a). The frequencies of genotype and allele were calculated in the four Chinese bovine populations (Table S3). The frequency of allele SNP1-G (NY: 63.16%; 3.3. Linkage disequilibrium and haplotype analysis QC: 78.45%; JX: 70.83%; CH: 100%) has shown a high prevalence in the four breeds and GG genotype (NY: 38.00%; QC: 60.34%; JX: 45.83%; Linkage disequilibrium between polymorphism pairs and haplotype CH: 100%) was more frequent than other genotypes. structure analysis of the IGF2 and ZBED6 gene in all cattle populations At the SNP2 — Hinf I locus (680C > G), digestion of the 281 bp PCR are shown in Tables S4 and S5. fragment of ZBED6 exon 1 with Hinf I resulted in fragment lengths of The linkage disequilibrium between the 7 SNPs in all cattle popu- 281 bp for genotype CC; 281, 243, 38 bp for genotype CG and 243, lations was estimated, which indicated that the D′ values ranged from 38 bp for genotype GG (Table S1; Fig. 1b). The frequencies of genotype 0.00 to 0.206; the r2 values were from 0.00 to 0.038; the mean r2 and allele were calculated in the four Chinese bovine populations. between adjacent SNPs was 0.001, 0.038, 0.033 and 0.000 for NY, The frequency of allele SNP2-C has shown a high prevalence in all QC, JX, and CH, respectively. These indicated that the three SNPs populations (NY: 55.50%; QC: 55.00%; JX: 59.31%; CH: 57.50%) and CG were having little linkage disequilibrium. The possibility is that genotype was more frequent in most of the populations (NY: 45.00%, recombination will be high and LD will be low in genovariation — QC: 54.00%, JX: 44.15% and CH: 45.00%). dense regions. At the SNP3 — Mlu I locus (1043A > G), digestion of the 227 bp The result of haplotype analysis of three SNPs showed that eight PCR fragment with Mlu I resulted in fragment lengths of 227 bp for different haplotypes were identified in these animals. Hap 1 (GCA) genotype AA; 227, 207, 20 bp for genotype AG and 207, 20 bp for had the highest haplotype frequencies in QC (29.80%), JX (21.70%),

Fig. 1. Forced PCR–RFLP detection results of ZBED6 gene PCR product. Gel = 4% agarose gel electrophoretic patterns. M = DNA molecular weight marker is Marker I (600, 500, 400, 300, 200, 100 bp); (a) SNP1 −826G > A genotype: GG: 297 bp + 31 bp; GA: 328 bp + 297 bp + 31 bp; AA: 328 bp; (b) SNP2 680C > G genotype: CC: 281 bp, CG: 281 bp + 243 bp + 38 bp; GG: 243 bp + 38 bp; (c) SNP3 1043A > G genotype: AA: 227 bp; AG: 227 bp + 207 bp + 20; GG: 207 bp + 20; It is difﬁcult to see the 31, 38 and 20 bp DNA fragment on 4% agarose gel.

Please cite this article as: Huang, Y.-Z., et al., Relationship of polymorphisms within ZBED6 gene and growth traits in beef cattle, Gene (2013), http://dx.doi.org/10.1016/j.gene.2013.04.049 4 Y.-Z. Huang et al. / Gene xxx (2013) xxx–xxx and CH (57.50%); Hap 4 (GGG) had the highest haplotype frequencies We found 8 haplotypes present in the ZBED6 gene: 8 haplotypes in NY (23.20%). in NY, QC, and JX, 2 haplotypes (Hap 1 and 7) in CH, with only 2 haplotypes (Hap 1 and 7) shared by all three populations (Table S5). The most frequent haplotypes are present in all the breeds studied, 3.4. Association analysis whereas the rare haplotypes are exclusive to specific breeds. The high-frequency haplotypes have probably been present in the popula- We analyzed the associations of these three SNPs with growth traits tion for a long time. Consequently, most of the new mutants are derived in NY cattle breed. The results of the association analyses between sin- from common haplotypes, implying that rarer variants represent more gle markers and growth traits are shown in Table 1. recent mutations and are more likely to be related to common Statistical results showed that the animals with SNP1-AA haplotypes than to other rare variants (Posada and Crandall, (−826G > A) genotype had significantly greater body weight than 2001). Geographic location information also provides the rela- those with genotypes SNP1-GG and GA (P b 0.01 or P b 0.05) at tionships among haplotypes, as the immediate descendents of a birth, 12 and 24 months in NY (Table 1), demonstrating that the allele new mutation are more likely to remain in the original population SNP1-A might be associated with an increased in body weight at than to emigrate to some distant population, unless high levels of birth, 12 and 24 months in NY cattle population. The animals with gene flow occur between those populations (Slatkin and Maddison, SNP2-GG (680C > G) genotype had significantly greater body weight 1989). than those with genotypes SNP2-CC and CG (P b 0.01 or P b 0.05) at 6 The above statistical results showed that the individuals with the and 12 months in NY (Table 1), demonstrating that the allele SNP2-G SNP1-AA, SNP2-GG and SNP3-GG related with greater body weight at might be associated with an increased in body weight at 6 and different ages (Table 1). These results suggested that the cattle with ge- 12 months old of age in the NY cattle population. The animals with notype SNP1-AA, SNP2-GG and SNP3-GG could be selected to obtain SNP3-GG (1043A > G) genotype had significantly greater body greater body weight. The SNP2-GG and SNP3-GG are two missense mu- weight than those with genotypes SNP3-AA and AG (P b 0.01 or tations that exist in the coding region of the bovine ZBED6 gene; thus, P b 0.05) at birth, 6, and 24 months in NY (Table 1), demonstrating we postulate that these mutations change the amino acid sequence of that the allele SNP3-G might be associated with an increased in proteins, which may affect the level of translation efficiency of ZBED6 it- body weight at 6 and 12 months old of age in the NY cattle self, thereby altering their functions, which appear to be a trend to con- population. trol body weight in each stage. As shown in Table 1, the growth trait value of heterozygotic geno- The ZBED is an important transcription factor affecting development, type was lower than that of homozygotic genotype in three SNPs of cell proliferation, and growth in placental mammals (Markljung et al., bovine ZBED6 gene. In the NY cattle population, the SNP1-AA (birth, 2009). More recently, we described the identification of the polymor- 12 and 24 months), SNP2-GG (6 and 12 months) and SNP3-GG phism in the ZBED6 gene that was significantly associated with growth (birth, 6 and 24 months) were highly significantly associated with traits in bovine. body weight at five different ages. That suggests that these mutations indirectly contribute to body weight characteristics at different times in cattle. 5. Conclusions

4. Discussion In this study, we analyzed three SNPs association with body weight in NY cattle population. These data strongly suggest that ZBED6 poly- Compared with previously reported sequence, three SNPs including morphisms may be used as a genetic marker which could be used for one non-coding mutation in promoter (SNP1: −826G > A) and two the breeding of new breeds of beef cattle. However, we conclude that missense mutations (SNP2: 680C > G and SNP3: 1043A > G) in CDS further research and validation of the various allelic effects and func- were identiﬁed in these animals (Table S2). The SNP2 and 3 are mis- tional mechanisms are needed in an independent sample prior to sense mutation p. Ala 227 Gly and p. His 348 Arg. All three SNPs were claiming that the ZBED6 variants identiﬁed or others can be used for novel and were submitted to the NCBI dbSNP (Table S2). marker-assisted selection.

Table 1 Associations between three SNPs and phenotypic body weight traits in Nanyang cattle: Mean ± SE (kg).

SNPs Genotypes Number Body weight traitsa (Total = 265) BW0 BW6 BW12 BW18 BW24

1 GG 98 28.69 ± 0.51B 156.71 ± 3.90 210.83 ± 4.38B 298.88 ± 6.14 353.71 ± 7.33C GA 139 30.01 ± 0.38A 159.23 ± 2.91 224.05 ± 3.27AB 298.84 ± 4.59 370.54 ± 5.48B AA 28 30.95 ± 0.44A 162.94 ± 3.37 232.75 ± 3.79A 300.75 ± 5.32 394.78 ± 6.35A P 0.0051 (b0.01) 0.4675 0.0013 (b0.01) 0.9577 0.0000 (b0.01)

2 CC 88 29.19 ± 0.44 147.09 ± 2.64B 218.77 ± 3.67B 300.65 ± 5.13 372.35 ± 6.70 CG 119 30.27 ± 0.41 159.77 ± 2.47B 218.59 ± 3.42B 301.77 ± 4.80 378.80 ± 6.26 GG 58 30.64 ± 0.51 176.54 ± 3.01A 237.65 ± 4.19A 294.46 ± 5.87 370.08 ± 7.67 P 0.0665 0.0000 (b0.01) 0.0010 (b0.01) 0.6047 0.6378

3 AA 145 29.57 ± 0.37B 154.70 ± 2.66b 219.48 ± 3.32 300.11 ± 4.43 365.50 ± 5.47b AG 84 29.00 ± 0.59B 163.50 ± 4.26a 222.50 ± 5.31 297.83 ± 7.09 375.11 ± 8.75ab GG 36 31.07 ± 0.43A 165.51 ± 3.06a 229.74 ± 3.81 299.46 ± 5.08 388.29 ± 6.27a P 0.0068 (b0.01) 0.0226 (b0.05) 0.1286 0.9636 0.0271 (b0.05)

The data are expressed as least square means ± standard errors (Mean ± SE). Values with different superscripts within the same column differ signiﬁcantly at P b 0.05 (a, b, ab) and P b 0.01 (A, B, AB). a Traits under study were birth body weight (BW0), body weight at 6 months old (BW6), body weight at 12 months old (BW12), body weight at 18 months old (BW18) and body weight at 24 months old (BW24).

Conflict of interests Henderson, C.R., 1986. Estimation of variances in animal model and reduced animal model for single traits and single records. J. Dairy Sci. 69, 1394–1402. Hickford, J.G.H., et al., 2009. Polymorphisms in the ovine myostatin gene (MSTN) and The authors declare that they have no conflict of interests. their association with growth and carcass traits in New Zealand Romney sheep. Anim. Genet. 41, 64–72. Holzer, C., Precht, M., 1992. Multiple comparison procedures for normally distributed Acknowledgments ANOVA models in SAS, SPSS, BMDP, and MINITAB. Comput. Stat. Data Anal. 13, 351–358. This study was supported by the National Natural Science Foundation Huang, Y.Z., et al., 2010. Six novel coding SNPs of the NPM1 gene and their associations – of China (Grant No. 31272408), Agricultural Science and Technology with growth traits in bovine. Arch. Anim. Breed. 5 (33), 368 371. Huang, Y.Z., et al., 2010. Two novel SNPs of SREBP1 gene are associated with body Innovation Projects of Shaanxi Province (No. 2012NKC01-13), Program weight and average daily gain in bovine. Anim. Biotechnol. 21, 170–178. of National Beef Cattle and Yak Industrial Technology System Huang, Y.Z., et al., 2010. Novel 12-bp deletion in the coding region of the bovine NPM1 – (CARS-38). The funders had no role in study design, data collection and gene affects growth traits. J. Appl. Genet. 51, 199 202. Huang, Y.Z., et al., 2011. Sequence variants in the bovine nucleophosmin 1 gene, their analysis, decision to publish, or preparation of the manuscript. linkage and their associations with body weight in native cattle breeds in China. Anim. Genet. 42, 556–559. Appendix A. Supplementary data Huang, Y.Z., et al., 2011. Haplotype combination of SREBP-1c gene sequence variants is associated with growth traits in cattle. Genome 54 (6), 507–516. Huang, Y.Z., et al., 2013. Haplotype distribution in the GLI3 gene and their associations Supplementary data to this article can be found online at http:// with growth traits in cattle. Gene 513, 141–146. dx.doi.org/10.1016/j.gene.2013.04.049. Markljung, E., et al., 2009. ZBED6, a novel transcription factor derived from a domesti- cated DNA transposon regulates IGF2 expression and muscle growth. PLoS Biol. 7 (12), e1000256. http://dx.doi.org/10.1371/journal.pbio10 00256. References Nei, M., 1973. Analysis of gene diversity in subdivided populations. Proc. Natl. Acad. Sci. U. S. A. 70, 3321–3323. Amandine, M., Yves, A., Bertrand, S., Gilles, R., Hubert, L., Dominique, R., 2010. Genetic Posada, D., Crandall, K.A., 2001. Intraspecific gene genealogies: trees grafting into variability and linkage disequilibrium patterns in the bovine DNAJA1 gene. Mol. networks. Trends Ecol. Evol. 16, 37–45. Biotechnol. 44, 190–197. Sambrook, J., & Russell, D.W., (Translated by Huang, P.T.), 2002. Molecular cloning A Barrett, J.E., Fry, B., Maller, J., Daly, M.J., 2005. HAPLOVIEW: analysis and visualization of Laboratory Manual. Science, Beijing. LD and haplotype maps. Bioinformatics 21, 263–265. Slatkin, M., Maddison, W.P., 1989. A cladistic measure of gene flow inferred from the Boldman, K.G., Kriese, L.A., Van Vleck, L.D., Kachman, S.D., 1993. A Manual for Use of phylogenies of alleles. Genetics 123 (3), 603–613. MTDFREML: A Set of Programs to Obtain Estimates of Variances and Covariances. Stephens, M., Smith, N., Donnelly, P., 2001. A new statistical method for haplotype ARS. USDA, Washington, DC. reconstruction from population data. Am. J. Hum. Genet. 69, 978–989. Botstein, D., White, R.L., Skolnick, M., Davis, R.W., 1980. Construction of a genetic link- Van Laere, A.S., et al., 2003. A regulatory mutation in IGF2 causes a major QTL effect on age map in man using restriction fragment length polymorphisms. Am. J. Hum. muscle growth in the pig. Nature 425, 832-826. Genet. 32, 314–331. Yeh, F.C., Yang, R., Boyle, T., 1999. POPGENE Version 3.1, Microsoft Window-Based Derecka, K., et al., 2009. Sequence variants in the bovine gonadotrophin releasing Freeware for Population Genetic Analysis. University of Alberta, Edmonton, AB, hormone receptor gene and their associations with fertility. Anim. Genet. 41, 329–331. Canada. Gilbert, R.P., Bailey, D.R., Shannon, N.H., 1993. Linear body measurements of cattle Zhao, H., Nettleton, D., Dekkers, J.C.M., 2007. Evaluation of linkage disequilibrium before and after twenty years of selection for postweaning gain when fed two measures between multi-allelic markers as predictors of linkage disequilibrium different diets. J. Anim. Sci. 71, 1712–1720. between single nucleotide polymorphisms. Genet. Res. 89, 1–6.

Please cite this article as: Huang, Y.-Z., et al., Relationship of polymorphisms within ZBED6 gene and growth traits in beef cattle, Gene (2013), http://dx.doi.org/10.1016/j.gene.2013.04.049