|
|
||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
ANIMAL GENETICS |
USDA, ARS, US Meat Animal Research Center, Clay Center, NE 68933-0166
| Abstract |
|---|
|
|
|---|
Key Words: genetic marker marker-assisted selection osteopontin twinning
| INTRODUCTION |
|---|
|
|
|---|
Recently, Schnabel et al. (2005)
proposed that a SNP upstream of osteopontin (SPP1) was a positional candidate polymorphism explaining a QTL on chromosome 6 that affected milk traits. White et al. (2007)
was able to detect significant effects in 2 populations for post-weaning growth; birth weight and weaning weight, however, showed no significant associations. This lack of association was somewhat puzzling given the known biology of SPP1 in terms of tissue growth (Rangaswami et al., 2006
) and embryonic growth (Weintraub et al., 2004
).
The US Meat Animal Research Center (USMARC) Twinning population has been selected for 25 yr on ovulation and twinning rate. Previous work has demonstrated that remarkable progress can be made by selection despite the low heritability of twinning rate (Gregory et al., 1997
).
The objectives of this study were to evaluate the SPP1 marker in the USMARC Twinning population, confirm SPP1 as a marker for postweaning growth, evaluate the potential influence of SPP1 on birth weight, and evaluate the genetic parameters between twin- and single-born animals for each growth trait. In addition, the Twinning population supported analysis of the effects of SPP1 on reproductive phenotypes of twinning and ovulation rates. A subobjective of this study was to demonstrate the efficacy of detecting associations using a model with polygenic effects and genotypic probabilities.
| MATERIALS AND METHODS |
|---|
|
|
|---|
Traits analyzed in the current study included birth weight (BWT), weaning weight adjusted to 205 d of age (WW205), yearling weight adjusted to 365 d of age (YW365), postweaning gain (PWG; weight gain between WW205 and YW365), twinning rate (TWN; number of calves per parturition), and ovulation rate (OR) as an indicator of the number of ova ovulated per estrous cycle as measured in heifers. Ovulation rate was determined by counting the number of corpora lutea by palpation or ultrasound or both in successive estrous cycles. Fostered animals, triplets, and quadruplets were removed from the data set.
Genotyping
The DNA extracted from ear notch or semen samples was genotyped for the SPP1 polymorphism from 309 sires representing all of the males that had produced progeny, 116 males born in 2005 with available DNA, and 2,276 females born from 1988 to 2004. Additional nontwinner genotyping included the sire of the USMARC Belgian Blue x MARC III QTL population (Casas et al., 2000
). This was done to evaluate the SPP1 polymorphism in relation to previously discovered QTL for BWT and postweaning growth traits segregating in the region as a result of this sire.
The assay design for the SPP1 polymorphism was previously described by White et al. (2007)
. Briefly, a polymorphism assay for a single base pair in length was developed to observe the alleles for the deletion/insertion polymorphism SPP1, previously described as OPN3907 (Schnabel et al., 2005
). Alleles designated as T9 and T10 describe 9 or 10 consecutive thymidine bases on the sense strand of SPP1, respectively. Microsatellite genotyping procedures were previously described by Kappes et al. (2000)
, and map positions were used directly from the USMARC linkage map (Snelling et al., 2005
). Previously, Kappes et al. (2000)
had genotyped 181 progeny-tested sires from this population for 12 microsatellites on BTA6. An additional 128 progeny-tested sires were genotyped for 5 of the 12 microsatellites (Table 1
).
|
|
![]() |
where pT10/T10 is the probability that j is homozygous for allele T10, pT9/T9 is the probability that j is homozygous for allele T9, pT10/T9 is the probability that j inherited allele T10 from its dam and allele T9 from its sire, and pT9/T10 is the probability of the heterozygote with the opposite parental inheritance. The regressors are xA, xD, and xI for the additive, dominance, and genomic imprinting components of gene action, respectively. Components that were not statistically significant were dropped from the model.
Twinning rate and OR were fit together in a 2-trait, repeated-records analysis, using MTDFREML (Boldman et al., 1995
) and the model described by Van Vleck and Gregory (1996)
. The model for TWN included the fixed effects of year of parturition, season of parturition, age at parturition, and genotypic regressions of SPP1. The model for OR included fixed effects of birth-year season, age at ovulation, month of ovulation, and genotypic regressions of SPP1. Random effects for both traits included breeding value with full relationships accounted for and environmental variance common to the repeated records of the animal in addition to the residual. Covariances between breeding values of the 2 traits were included, as were covariances between the common environmental effects of the 2 traits, but the residual covariance was set to 0 because there was no correspondence between ovulation records and twinning (parturition) records. (Co)variances were estimated by allowing MTDFREML to iterate until a convergence criteria of 1014 was obtained and until the iterations resulted in no change in the parameter estimates upon restarting.
Two-trait analyses of growth traits were conducted using MTDFREML, with records from single births as 1 trait and records from twin births as the second trait to estimate genetic parameters for each growth trait. Four 2-trait analyses were run for the following combinations: BWT single births (BWT-S) BWT twin births (BWT-T), WW205 single births (WW205-S) WW205 twin births (WW205-T), PWG single births (PWG-S) PWG twin births (PWG-T), and YW365 single births (YW365-S) YW365 twin births (YW365-T).
For each trait, the fixed effects included a contemporary group effect, age of dam as linear and quadratic covariates, and individual and maternal genotypic regressions of SPP1. Dams greater than 10 yr of age were treated as 10-yr-olds. For all twin traits, birth order was added as a fixed effect. For BWT, the contemporary group included year-season of birth, type of birth (single, twin), type of rearing (animal raised as single, twin, or artificially), and sex (bull, heifer, or freemartin). For WW205, the contemporary group definition included type of rearing in addition to the factors for BWT. For PWG and YW365, the contemporary group definition included whether males were castrated shortly after weaning or not, in addition to the factors for WW205. Breeding value was included as a random effect for all 3 traits, again with a covariance structure defined by the numerator relationship matrix. For BWT and WW205, maternal breeding value and maternal permanent environment effects were fit as random effects in the model. All covariances among direct and maternal breeding values of the 3 traits were included, as were the covariances between the permanent environmental effects of the 2 maternal traits and all residual covariances. Variance component estimates were obtained using the same criteria as above.
The males with SPP1 genotypes represented a small and highly selected proportion of the population. Furthermore, there were no SPP1 genotypes on freemartins. Because of the potential for selection to bias the results, only female growth phenotypes were allowed to contribute to the estimates of genetic regressions for SPP1. This was accomplished by setting the SPP1 regressors to zero for all male and freemartin records; this was effective because sex class was part of the contemporary group definition. Prior analyses on female data had shown that the predicted genotypes yielded similar estimates to the genotyped animals while increasing the power to detect effects. Retaining the records on male and freemartin calves in the analysis substantially improved the estimation of variance components, breeding values, and maternal permanent environment effects.
| RESULTS |
|---|
|
|
|---|
The SPP1 polymorphism was segregating in the USMARC Twinning population with a minor allele (T9) frequency in founders of 5.4% from the GenoProb analysis. Of the 2,701 animals genotyped, the minor allele frequency was 5.2%, with only 6 T9/T9 animals identified. Only 2 allele sizes, T9 and T10, were observed segregating in this population. The minor allele can be traced back to founders of Angus, Hereford, Holstein, Norwegian Red and Simmental decent. The Belgian Blue x MARCIII sire was heterozygous for the SPP1 polymorphism.
Significant additive associations of SPP1 were detected for animals of single births for BWT-S, WW205-S, and YW365-S with effects of 1.14, 5.16, and 7.89 kg, respectively (Table 3
). Two-trait TWN and OR analysis resulted in no statistically significant association with SPP1 marker genotype. No significant effects were identified for dominance and imprinting or for maternal genotype of SPP1 in this population. These effects were subsequently dropped from the analyses. Heritability estimates for BWT-S and BWT-T were 0.5 and 0.33, respectively (Table 4
). Proportion of permanent environmental variance was 0.015 for BWT-S and 0.14 for BWT-T. Genetic correlations among single and twin births for all growth traits were high with a range of rg = 0.86 to 0.96 (Table 5
). Additionally, the genetic correlation between TWN and OR was high (rg = 0.77).
|
|
|
| DISCUSSION |
|---|
|
|
|---|
No significant association was detected for SPP1 on PWG. The estimated effects suggest that most of the impact of SPP1 is realized prior to weaning (which is a component of YW365). The lack of significance for BWT and WW205 in White et al. (2007)
does not support this conjecture. However, there are 2 possible explanations for differences in results between these 2 studies. First, BWT and WW205 were analyzed using a model with maternal effects in this study. This model accounted for genetic and permanent environmental variation in both of these traits and thereby decreased the residual error and increased the power to detect genotypic differences. Second, the calves analyzed in White et al. (2007)
may have had reduced nutrient availability relative to this population due to the milking potential in the Twinning herd. Therefore, the White et al. (2007)
calves may not have expressed their full growth potential prior to weaning and likely grew more postweaning in compensation. The direction of the effect is still toward increased growth in PWG for T10/T10 Twinning females.
Analysis of TWN and OR resulted in no significant association of SPP1 with either trait. Frequency of the minor T9 allele was lower in the USMARC Twinning population (5.2%) than was observed in GPE7 (13.1%) and GPE8 (13.5%). Estimates of genetic parameters for BWT, WW205, YW365, TWN, and OR were consistent with previously published analyses (Gregory et al., 1997
). However, our current analyses were the first to divide each growth trait into 2 separate traits (single and twin births). In BWT, the permanent environmental variance, and in WW205, the maternal heritability, were estimated higher in twin data than for single birth data. Uterine environment (e.g., capacity, placental nutrients) may be more important when rearing twins because of potential limitations thereby increasing the permanent environmental variance in twin BWT. The increase in WW205 may be explained by the large proportion of dairy influence in this herd (Gregory et al., 1997
). Females rearing single calves might not have had the opportunity to express their genetic potential for milking ability. This is a result of 2 calves having a higher capacity to consume milk than 1. Thus, the maternal or permanent environmental variances may be underestimated for the single birth animals.
Twin data has traditionally been ignored in sire evaluation for growth traits. Differences in direct heritability estimates for singles and twins for BWT suggest that twin animals should not be included in analyses with limited numbers of twin births. Lack of differences between estimates of heritability for weaning and post-weaning growth traits suggests that twins may be included in the analyses. However, inclusion of twin data in genetic evaluation programs is likely dependent on having enough twin records to estimate genetic parameters.
The association of the SPP1 SNP in this region of BTA6 is consistent with previously discovered QTL for growth traits in beef cattle. Early studies reported QTL for BWT on BTA6 mapping directly to the region containing SPP1 (Davis et al., 1998
; Casas et al., 2000
). The estimated phenotypic difference between T10 and T9 alleles was 1.14 kg of BWT in this study, about one-half the size of the effect for BWT reported in each of the previously listed QTL scans. A possible explanation for the difference may be that the region of BTA6 contains more than 1 gene having an effect on BWT. This also is supported by a BWT QTL just proximal to OPN reported by Kneeland et al. (2004)
. Additional QTL for correlated traits, calving ease and stillbirth, have been found in dairy cattle analyses by Schrooten et al. (2000)
and Kühn et al. (2003)
, respectively, in the same region. It is possible that the effects for BWT QTL from the genomic scans may be overestimated, a common problem in QTL analyses (Bogdan and Doerge, 2005
). Another possibility is that the size of effects can vary because of different genetic backgrounds.
In addition to the BWT QTL, Casas et al. (2000)
detected QTL associated with postnatal growth traits including yearling weight, HCW, and longissimus muscle area in this region in a Belgian Blue x MARC III sired population at USMARC. The Belgian Blue x MARC III sire genotyped T10/T9 adding additional support for the SPP1 indel as positional mutation candidate for growth. However, it should be noted that the SPP1 marker may be only in linkage with the functional mutation(s) for the growth QTL in this region. Another possibility is the indel could be functional for milk traits and an additional mutation in or near SPP1 may be segregating for growth in this animal used to create the QTL mapping family. Additional evidence for post-weaning ADG QTL has been mapped to this chromosome in the M1 Beefbooster line and in Hanwoo (Korean cattle) populations (Kim et al., 2003
; Kneeland et al., 2004
). In the M1 Beefbooster line, preweaning QTL have been mapped proximal and distal to the region containing SPP1 on BTA6.
Schnabel et al. (2005)
first described SPP1 as a positional candidate for milk production traits in a Holstein population resulting from a QTL scan of BTA6. Previous research had indicated a role for SPP1 in mammary gland development and lactation in a tissue-specific transgenic antisense RNA mouse model (Nemir et al., 2000
). Schnabel et al. (2005)
were able to show that the SPP1 indel polymorphism was in concordance with all the sires segregating the QTL. When SPP1 genotype was included as a fixed effect in the QTL analysis, evidence for the QTL in that region was completely eliminated. The authors concluded that the indel polymorphism lies upstream of the promoter of SPP1 in a region that has been shown to contain tissue-specific regulatory elements, making it a strong candidate as the functional mutation for milk traits.
Osteopontin (SPP1, OPN, and Eta-1) is a secreted glycoprotein that plays a role in many different mammalian biological functions. It has been shown to be involved in cell adhesion, tissue remodeling, inflammation, cell survival, embryo implantation, and maintenance of pregnancy (Denhardt et al., 2001
; Johnson et al., 2003
). The role of SPP1 in growth pathways has been studied extensively in bone tissue growth and cancer progression (Standal et al., 2004
; Rangaswami et al., 2006
). More recently, SPP1 has been shown to increase in a myoblast proliferation and differentiation in vitro model (Ishibashi et al., 2005
). Additional evidence of altered growth has been shown in SPP1 knockout mice by examining embryo size in utero at 3 stages of gestation using magnetic resonance microscopy. Homozygous knockout mice had smaller embryos at all 3 stages of gestation with no differences in litter size when compared with the controls (Weintraub et al., 2004
). The known biology and function of SPP1 makes it an excellent positional candidate. However, we agree that much more work to investigate the functional biology of SPP1 polymorphism is needed to consider the mutation as the causative polymorphism (de Koning, 2006
).
| IMPLICATIONS |
|---|
|
|
|---|
| Footnotes |
|---|
2 The authors thank L. Flathman, K. Beavers, and D. Sypherd for technical assistance and D. Griess and J. Byrkit for secretarial support. ![]()
4 Current address: USDA, ARS, Animal Disease Research, Pullman, WA 99164. ![]()
3 Corresponding author: allan{at}email.marc.usda.gov
Received for publication July 11, 2006. Accepted for publication September 22, 2006.
| LITERATURE CITED |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
R. A. Cushman, M. F. Allan, L. A. Kuehn, W. M. Snelling, A. S. Cupp, and H. C. Freetly Evaluation of antral follicle count and ovarian morphology in crossbred beef cows: Investigation of influence of stage of the estrous cycle, age, and birth weight J Anim Sci, June 1, 2009; 87(6): 1971 - 1980. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. J. Alexander, L. A. Kuehn, T. P. L. Smith, L. K. Matukumalli, B. Mote, J. E. Koltes, J. Reecy, T. W. Geary, D. C. Rule, and M. D. MacNeil A Limousin specific myostatin allele affects longissimus muscle area and fatty acid profiles in a Wagyu-Limousin F2 population J Anim Sci, May 1, 2009; 87(5): 1576 - 1581. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. Gutierrez-Gil, J. L. Williams, D. Homer, D. Burton, C. S. Haley, and P. Wiener Search for quantitative trait loci affecting growth and carcass traits in a cross population of beef and dairy cattle J Anim Sci, January 1, 2009; 87(1): 24 - 36. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. F. Allan, L. A. Kuehn, R. A. Cushman, W. M. Snelling, S. E. Echternkamp, and R. M. Thallman Confirmation of quantitative trait loci using a low-density single nucleotide polymorphism map for twinning and ovulation rate on bovine chromosome 5 J Anim Sci, January 1, 2009; 87(1): 46 - 56. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. E. Echternkamp, R. M. Thallman, R. A. Cushman, M. F. Allan, and K. E. Gregory Increased calf production in cattle selected for twin ovulations J Anim Sci, December 1, 2007; 85(12): 3239 - 3248. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |