Genetic polymorphisms related to meat traits in purebred and crossbred Nelore cattle

The objective of this work was to estimate the allelic and genotypic frequencies of CAST/XmnI, a calpastatin gene polymorphism, and CAPN530, a calpain 1 large subunit gene polymorphism, in different beef genetic groups (Nelore and Nelore x Bos taurus), and to investigate associations between these polymorphisms and carcass and meat traits. Three hundred animals – comprising 114 Nelore, 67 Angus x Nelore, 44 Rubia Gallega x Nelore, 41 Canchim, 19 Brangus three-way cross and 15 Braunvieh three-way cross – were genotyped by PCR-RFLP and phenotyped for rib-eye area (REA), back-fat thickness (BT), intramuscular fat (IF), shear force (SF) and myofi brillar fragmentation index (MFI). The occurrence of the two alleles of the CAST/XmnI and CAPN530 single nucleotide polymorphisms (SNPs) in a B. indicus breed, which permitted association studies in purebred and crossbred Nelore cattle, was fi rst shown in the present work. No relationship was found between the CAST or CAPN1 SNPs and growth-related traits (REA) or fat deposition (BT and IF), since calpastatin and μ-calpain are not physiologically involved with these traits. Moreover, the association results between genotypes and aged meat tenderness (assessed by SF and MFI) showed that these markers are useless in assisted selection for purebred Nelore and their crosses with B. taurus.


Introduction
Tenderness is very relevant to beef consumers and is, therefore, a trait of interest for animal selection.The breakdown of myofi brillar proteins controlled by the μ-calpain enzyme and modulated by its inhibitor, calpastatin, is the main mechanism of post-mortem meat tenderization (Delgado et al., 2001).When calpastatin activity increases, it reduces μ-calpain activity, with Pesq. agropec. bras., Brasília, v.44, n.12, p.1660-1666, dez. 2009 a negative effect on meat tenderness (Pringle et al., 1997).Therefore, genes encoding calpastatin (CAST) and µ-calpain (CAPN1) are considered important functional candidates for meat tenderness in livestock.
The CAST gene was mapped to chromosome BTA7 (Bishop et al., 1993), and the CAPN1 gene, to the telomeric end of the BTA29 chromosome, the same position verifi ed for meat tenderness QTL (Smith et al., 2000).Chung et al. (2001a) detected genetic variants at intron 6 of bovine CAST, which can be identifi ed by PCR-RFLP using the XmnI restriction enzyme.They also found some evidence indicating that genotyping this polymorphism could help identify animals with distinct calpastatin enzymatic activity.Two non-synonymous single nucleotide polymorphisms (SNPs) -CAPN316 (located in exon 9) and CAPN530 (located in exon 14) in the gene CAPN1 have been associated with meat traits in Bos taurus beef breeds (Page et al., 2002(Page et al., , 2004;;Corva et al., 2007).Rincon & Medrano (2006) developed a PCR-RFLP method for genotyping the CAPN530 polymorphism using the PsyI enzyme.However, this marker failed to show segregation in B. indicus cattle of the Brahman breed (Casas et al., 2005), discouraging association studies between this SNP and meat quality in other zebu-infl uenced breeds.
According to Casas et al. (2005), polymorphisms that segregate and associate with traits of interest in B. taurus may or may not segregate in B. indicus.Thus, it is necessary to develop and use other markers in candidate genes or chromosomal regions, to allow association studies in B. indicus.It is also important to consider major variations in allelic frequencies between breeds, even within subspecies.Furthermore, association results for molecular markers and production traits obtained in B. taurus populations are not directly applicable to B. indicus populations, because allelic substitution effects are specifi c to each population and its environment.
The objective of this work was to estimate the allelic and genotypic frequencies of the CAST/XmnI and CAPN530 polymorphisms in different beef genetic groups (Nelore and Nelore x B. taurus), as well as to determine the occurrence of associations between these polymorphisms and carcass and meat traits.After slaughter in collaborating abattoirs, the carcasses were identifi ed and chilled for 24 hours.Two 2.54-cm thick samples of the longissimus dorsi muscle were then removed from an area between the 11 th and 13 th ribs of the left half of each carcass.

Materials and Methods
Samples collected between the 12 th and 13 th ribs provided rib-eye area (REA), back-fat thickness (BT) and shear force (SF) measurements.Samples collected between the 11 th and 12 th ribs were used to measure the myofi brillar fragmentation index (MFI) and intramuscular fat (IF, percentage of total lipids), as well as for the extraction of genomic DNA.The REA was measured by the quadrant-point method and the BT was determined with the aid of a ruler, both according to the USDA Quality Grade protocol (United States Department of Agriculture, 1997).After these initial measurements, carried out in the slaughterhouse, the samples of longissimus dorsi were deboned, vacuum wrapped and aged under controlled temperature (between 1 and 2°C) for 14 days, followed by freezing at -20°C.The other phenotypic traits, SF, MFI and IF, were determined in the laboratory following methods described by Wheeler et al. (1995), Culler et al. (1978) and Bligh & Dyer (1959) respectively.Pesq. agropec. bras., Brasília, v.44, n.12, p.1660-1666, dez. 2009 Genomic DNA was extracted from meat samples (250 mg) by the non-phenolic method, with digestion with proteinase K and precipitation with NaCl and alcohol (Sambroock et al., 1989).
CAST and CAPN1 genotyping was done by the PCR-RFLP method.To determine the A and B alleles of the CAST SNP, a fragment of approximately 2,000 base pairs located at intron 6 was amplifi ed with forward 5' -AGC AGC CAC CAT CAG AGA AA -3' and reverse 5' -TCA GCT GGT TCG GCA GAT -3' primers and digested by the XmnI restriction enzyme (Chung et al., 2001a).To identify the A and G alleles of CAPN530 polymorphism, a fragment of 341 base pairs located at exon 14 was amplifi ed with forward 5' -CGT TTC TTC TCA GAG AAG AGC GCA GG G A -3' and reverse 5' -CCT GCG CCA TTA CTA TCG ATC GCA AAG T -3' primers and digested by the PsyI restriction enzyme (Rincon & Medrano, 2006).After digestion of the amplifi cation products, the DNA fragments of CAST and CAPN1 genes were separated in agarose gel (2 and 3% respectively).
The individual genotypes were determined by DNA fragment size analysis (measured in base pairs, bp), by comparison with a standard molecular weight of 100 bp.The allelic and genotypic frequencies were calculated for each polymorphism according to Weir (1996).The differences in allele frequency within and between the genetic groups studied was calculated using contingency tables (Curi & Moraes, 1981) adapted from Goodman (1965).
For the association studies, the traits of interest were analyzed using the general linear model (GLM) of the Statistical Analysis System (SAS Institute, 2004) program, and the least square means of the genotypes were compared by the Tukey's test.The Bonferroni correction was applied to analyses involving multiple comparisons.
The linear model for adjustment of quantitative variables included the genotype and the contemporary group effects as follows: Y ijk = µ + G i + CG j + e ijk , where Y ijk is the trait of interest, µ is the general mean, G i is the fi xed effect of i th genotype (i = 1, ..., 3), CG j is the fi xed effect of j th contemporary group (j = 1, ..., 13), and e ijk is the random error.The defi nition of contemporary groups included variations of genetic group, sex, age of slaughter, feedlot year and farm of origin.These variations could not be considered separately in the model, since there is an important confounding among them.The bull effect was not included in the model, since the number of genotyped offspring of individual bulls was very small.Thus, due to the large number of parents, the possibility of confounding between the genotype and bull effects on the evaluated traits was diluted.The interactions between SNPs and the studied genetic groups were not signifi cant, and were not included in the fi nal analysis.

Results and Discussion
The CAST/XmnI polymorphism segregated in the six genetic groups studied with a signifi cantly higher frequency of the A allele in all analyses (Table 1).The A allele frequency was signifi cantly higher in the Angus x Nelore and Rubia Gallega x Nelore groups than in the Nelore and Canchim groups.Brangus and Braunvieh three-way crosses had an intermediate A allele frequency in comparison to the groups that differed statistically.Homozygous AA followed by heterozygous AB were the most frequent genotypes in all but the Canchim group, in which the homozygous BB was the second most frequent genotype.
No previous studies with B. indicus were found to adequately compare the results for allelic frequencies (1) Means followed by equal letters, uppercase among genetic groups and lowercase within genetic groups, do not differ by the Goodman test, at 5% probability. (2)The number between parentheses indicates the number of animals in each genetic group.
Table 1.Allelic and genotypic frequencies for the CAST/XmnI polymorphism in the different genetic groups and in the total animal sample (1) .Pesq. agropec. bras., Brasília, v.44, n.12, p.1660-1666, dez. 2009 of CAST/XmnI.Nevertheless, the present fi ndings and the frequencies of 0.75 and 0.25 for alleles A and B, described in B. taurus animals of the Angus breed by Chung et al. (2001a), suggest that the A allele frequency is higher than the B allele frequency in B. taurus and in B. indicus.Consequently, the absence of differences in CAST/XmnI allele frequencies between B. indicus and B. taurus suggests a lack of association between this polymorphism and meat tenderness, a trait that differs between the subspecies (Wheeler et al., 1994).
Results for association analyses between CAST/ XmnI genotypes and studied traits (Table 2) were not signifi cant (p>0.05).In spite of the CAST/XmnI being a SNP in a non-coding region, it could be linked or in linkage disequilibrium with causative polymorphisms associated to phenotype differences.Also, no codifying RNA transcript from intronic regions are involved in a number of biological processes, such as controlling transcriptional and post-transcriptional levels of gene expression (Nakaya et al., 2007).Thus, intronic polymorphisms have gained importance as possible causative SNP and are not to be overlooked.However, as in Chung et al. (2001b), the results of the present trial show no association between the CAST/XmnI polymorphism and meat tenderness based on shear force or myofi brillar fragmentation index.Even with a small number of animals genotyped, 47 B. taurus animals of the Angus breed, Chung et al. (2001b) found that genotypes had signifi cant effect on calpastatin activity, although they could not explain the observed variation in meat tenderness.Chung et al. (2001b) and the present work showed greater meat tenderness in animals with two copies of the A allele, that were, however, not statistically significant.AA animals had lower shear force and higher myofibrillar fragmentation index when comparing minimum square means.These results may be explained by the low CAST/ XmnI effect or by the weak disequilibrium linkage between this polymorphism and the causative mutation.Considering the size of the effect, the association between CAST/XmnI and the evaluated traits may have been signifi cant with more animals in the sample.The weak linkage disequilibrium hypothesis showed that other polymorphisms described in the CAST gene were closely related to variation in meat tenderness, namely the SNP A2959G (access number AF159246), at region 3' UTR, and G/C (nucleotide 282 of access number AY008267), at intron 5, identifi ed by Barendse (2004) and Schenkel et al. (2006) respectively. Curi et al. (2008) reported the genotyping of A2959G (AF159246) SNP of bovine CAST gene by PCR-RFLP technique for the fi rst time.The accuracy of the method was confi rmed through the direct sequencing of PCR products of nine individuals.The lack of connection between allelic forms of the CAST gene and growth related traits (such as REA) and fat (BT and IF) was expected since, in theory, calpastatin is not involved in their physiology.However, it may be possible to fi nd linkage disequilibrium between this marker and functional polymorphisms in surrounding genes, as well as the possibility of pleiotropic effects.Chung et al. (2001b) found signifi cant associations between CAST polymorphisms and percentage of kidney, pelvic and heart fat, and Schenkel et al. ( 2006) associated the gene with carcass fat yield.
The CAPN530 polymorphism segregated in all genetic groups studied (Table 3).The A allele frequency was lower than the G allele frequency in all analyses.There were no signifi cant differences in allelic frequencies between genetic groups.The AA genotype was less frequent in all groups.Higher GG frequency was prevalent in all but one group, the Rubia Gallega x Nelore, in which the AG genotype was most prominent.In contrast with the results presented by Casas et al. (2005), which showed absence of the A allele in a B. indicus population of the Brahman breed, this study revealed a 21.9% frequency in Nelore B. indicus.Thus, for the fi rst time, it was possible to study the association between this polymorphism and traits of economical interest in B. indicus bovines.The (1) The numbers between parentheses indicate the number of animals from each genotype.
Table 2. Least square means and standard errors of the rib-eye area (REA), back-fat thickness (BT), intramuscular fat (IF), shear force (SF) and myofi brillar fragmentation index (MFI) determined for genotypes of the CAST/XmnI polymorphism.Pesq. agropec. bras., Brasília, v.44, n.12, p.1660-1666, dez. 2009 results presented here and elsewhere (Page et al., 2004;White et al., 2005;Corva et al., 2007) suggest that the A allele frequency of CAPN530 is lower then the G frequency in both B. taurus and B. indicus animals, and that the A allele occurrence is not extremely different between breeds.In two different B. taurus populations -a commercial herd, Simmental x Angus, and a research herd, GPE Cycle 7 -, Page et al. (2004) showed frequencies of 37 and 28% for the A allele.White et al. (2005) found a 14% frequency in a study with animals that combined B. taurus and B. indicus infl uence.Finally, Corva et al. (2007) described frequencies between 2 and 18% in four genetic groups formed by the crosses of Angus, Hereford and Limousin breeds (all B. taurus).Thus, like for CAST/ XmnI, it appears there is not great difference between B. taurus e B. indicus for the allele frequencies of the CAPN530, and this may indicate a lack of association between polymorphism and meat tenderness.
The analyses also showed no association (p>0.05) between genotypes of CAPN530 and phenotypes (Table 4).The CAPN530 polymorphism in the CAPN1 gene is a guanine (G allele) to adenine (A allele) substitution that causes a valine to isoleucine change at codon 530, domain III of the μ-calpain enzyme.The alteration between two apolar aminoacids is relatively conservative, although it may change protein stability and assemblage, thus affecting proteolytic activity (Page al., 2002).Page et al. (2002Page et al. ( , 2004) ) and Corva et al. (2007) found evidence of the association between polymorphism and meat tenderness.Interestingly, while Page et al. (2002Page et al. ( , 2004) report a favorable effect of the G allele, Corva et al. (2007) dispute the superior meat tenderness -smaller shear force values -of the AG genotype in comparison with the GG genotype.Analogous to conclusions reported by White et al. (2005), the present work failed to show association between genotypes of CAPN530 and meat tenderness.In addition to the environment-genotype interaction, which may produce divergent results even with direct markers within a breed, this disagreement can be due to differences: in disequilibrium or in the linkage phase between markers and quantitative trait locus (QTL); in epistatic interactions affecting the candidate gene; and in the magnitude of the effect of the candidate gene over the phenotype (Curi et al., 2005).These differences often underpin comparisons amongst populations and breeds, and in bovines they normally concur to the diversity of two subspecies.Hence, the contrast between the fi ndings of Page et al. (2002Page et al. ( , 2004)), the present results and those of White et al. ( 2005) may be understood, as the fi rst works mentioned only used Table 3. Allelic and genotypic frequencies for the CAPN530 polymorphism in the different genetic groups and in the total animal sample (1) .
(1) Means followed by equal letters, uppercase among genetic groups and lowercase within genetic groups, do not differ by the Goodman test, at 5% probability. (2)The number between parentheses indicates the number of animals in each genetic group. (1)The number between parentheses indicates the number of animals from each genotype.Table 4. Least square means and standard errors of the rib-eye area (REA), back-fat thickness (BT), intramuscular fat (IF), shear force (SF) and myofi brillar fragmentation index (MFI) determined for genotypes of the CAPN530 polymorphism.Pesq. agropec. bras., Brasília, v.44, n.12, p.1660-1666, dez. 2009 B. taurus animals, while the last two also included B. indicus animals.Overall, these results indicate that the CAPN530 SNP is not a functional marker for variation in beef tenderness.Also, in B. indicus populations it is not linked or in strong linkage disequilibrium with the causative polymorphism, which confi rms that polymorphism is useless in marker-assisted selection for beef herds with a broad genetic background.However, a relatively novel CAPN1 polymorphism, the CAPN4751 SNP, has shown great potential as a marker for meat tenderness selection in B. taurus, B. indicus and B. indicus x B. taurus beef herds (White et al., 2005;Van Eenennaam et al., 2007).As expected, no relation between allelic forms of the CAPN1 gene and the growth and fat traits (REA, BT and IF) was found.On the other hand, with the possibility of linkage disequilibrium between the CAPN1 and surrounding genes or of pleiotropic effects, positive associations between variants of the gene and carcass traits were described (Juszczuk-Kubiak et al., 2004;Casas et al., 2005;Cheong et al., 2008).

Conclusions
1. Despite novel demonstration of the occurrence of the two alleles of CAST/XmnI and CAPN30 polymorphisms in Bos indicus animals, there is no association between genotypes of these markers and meat tenderness.
2. The analyzed polymorphisms are useless in marker-assisted selection programs for Nelore and their crosses with Bos taurus.