| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
,5
* Department of Animal Sciences, The Ohio State University, Columbus 43210-1095 and
and
Department of Animal Science, University of Florida, Gainesville 32611-0901
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResults and DiscussionImplicationsLiterature Cited |
|---|
Key Words: Beef Cattle • Body Composition • Genetic Parameters • Insulin-Like Growth Factor • Ultrasound • Selection
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionImplicationsLiterature Cited |
|---|
reported that IGF-I concentration was negatively correlated with percentage of carcass fat, carcass fat accretion rate, total carcass fat, and fat thickness, and that it was positively correlated with percentage of carcass protein in Simmental crossbred bulls. Conversely, Luxford et al. (1998) found positive genetic and phenotypic correlations of ultrasonic fat depth at the last rib with plasma concentrations of IGF-I at 5 wk of age in pigs. In addition, Johnston et al. (2001) reported that plasma IGF-I was positively genetically correlated with various measures of fatness, both on live steers and heifers and on carcasses. Hayden et al. (1993) observed that plasma concentrations of IGF-I were highly positively correlated with empty body fat accretion and with empty body weight gain and protein deposition in steers exhibiting compensatory growth. In Angus bulls and heifers, Bishop et al. (1989) found that phenotypic correlations of serum IGF-I concentrations with carcass characteristics were low, but favorable, for the last two 28-d periods of a 140-d postweaning test. Davis and Simmen (2000) reported that bulls with lower IGF-I concentrations had higher marbling scores and quality grades, but also had higher backfat thickness and yield grades. However, in that study, carcass data were only available for bulls not saved for breeding. In addition, little is known about genetic relationships of IGF-I concentration with ultrasound measures of body composition in live cattle. Therefore, the objective of the present study was to determine genetic, phenotypic, and environmental correlations of IGF-I with ultrasound measures of backfat thickness and longissimus muscle area of bulls and heifers when the ultrasound data were adjusted to age- or weight-constant endpoints. |
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionImplicationsLiterature Cited |
|---|
Selection procedures and the mating scheme for this experiment have been described previously (Davis et al., 1995; Davis and Simmen, 1997). In brief, the four bull calves with the highest and the four with the lowest residuals (adjusted for age of calf and age of dam) for IGF-I concentrations were saved each year for breeding within the respective selection lines. Selection was based on the mean IGF-I concentration of three blood samples (taken at d 28, 42, and 56 of the 140-d postweaning test). Selected bulls were used for breeding as yearlings and then sold.
Approximately eight cows were culled from each line each year (based on physical unsoundness, reproductive failure, and oldest age) and replaced with approximately eight pregnant heifers with the highest or lowest residuals (adjusted for age of calf and age of dam) for serum IGF-I concentrations. All available heifers were bred and selections were made among heifers that conceived. Selection of heifers was based on the mean IGF-I concentration of three blood samples collected at the same time as for bulls.
Management Procedures
Spring-born calves were reared by their dams without creep feed until weaning at approximately 7 mo of age. Following weaning, bull calves were given ad libitum access to a corn–soybean meal-based diet (39.9% rolled shelled corn, 25% whole oats, 10% alfalfa meal [17% CP], 10% wheat middlings, and 10% soybean meal [44% CP] on a DM basis) plus grass hay (2.3 kg•bull-1•d-1). Heifers were fed a corn–soybean meal (65.8% ear corn, 10% alfalfa meal [17% CP], and 20% soybean meal [44% CP] on a DM basis) and hay diet intended to yield postweaning gains of approximately 0.75 kg/d. Bulls and heifers were fed in separate three-sided barns with adjoining exercise lots located at EORDC.
Fall-born calves were weaned at an average age of 140 d and then fed a corn–soybean meal diet formulated to yield gains of approximately 0.9 kg/d, plus grass hay, in drylot for 112 d. Following the 112-d growing period, bull and heifer calves remained at EORDC and were fed for the 140-d postweaning test period in the same manner as spring-born bulls and heifers. The postweaning periods of spring- and fall-born calves fell at different times of the year, in accordance with their different seasons of birth.
Data Collection
Real-time ultrasound measurements of backfat thickness (BF) and longissimus muscle area (LMA) were collected on d 56 and 140 of the 140-d postweaning period using an Aloka 500V real-time ultrasound machine fitted with a 3.5-MHz, 17.2-cm linear probe (Corometrics Medical Systems, Wallingford, CT). Measurements were recorded and interpreted by two faculty members from the Department of Animal Sciences at The Ohio State University, both of whom previously participated in the ultrasound technician training program provided by Iowa State University. The faculty member collecting the data on a given date varied depending on availability. Measurements were recorded on all bulls and heifers, whereas in the previous study by Davis and Simmen (2000), carcass data were only available for bulls not saved for breeding. Age of calves at d 56 of the postweaning period ranged from 260 to 358 d, with a mean of 314 d and a standard deviation of 17 d (Table 1). Age at the conclusion of the 140-d period ranged from 343 to 443 d, with a mean and standard deviation of 398 and 17 d, respectively.
|
Serum Samples
Approximately 25 mL of blood was collected into sterile 16- x 150-mm glass tubes via jugular venipuncture of each animal. The blood was allowed to clot for 24 h at 4°C. Serum was obtained by centrifugation (1,800 x g for 20 min) and frozen at -20°C until it was assayed.
Radioimmunoassay for Insulin-Like Growth Factor I
The RIA for IGF-I concentration was performed at the University of Florida using antiserum raised against human IGF-I in rabbits (UBK487), following procedures described by Bishop et al. (1989).
Statistical Analysis
Data were analyzed using animal models with a set of multiple-trait, derivative-free, restricted maximum likelihood (MTDFREML) computer programs (Boldman et al., 1993). Three-generation pedigrees of base population animals were used along with pedigrees of animals born subsequent to the base population to create the numerator relationship matrix (A). Included in the analysis were 2,427 animals in the A-1 matrix, 1,521, 636, and 561 of which had records for mean IGF-I concentration, ultrasound data on d 56, and ultrasound data on d 140 of the postweaning period, respectively. First, each IGF-I and ultrasound trait was analyzed singly to obtain estimates of direct heritability (
), maternal heritability (
), the proportion of phenotypic variance due to permanent environmental effects of dam (c2), and the correlation between direct genetic and maternal genetic effects (ram). The statistical model for this analysis included fixed effects of birth year of calf (1989 through 1999 for IGF-I measurements, and 1995 through 1999 for ultrasound data), season of birth (spring vs. fall), IGF-I selection line (high vs. low), sex of calf (bull vs. heifer), and age of dam (2, 3, 4, 5 to 9, 
10 yr), random animal, maternal genetic, and maternal permanent environmental effects, and a linear covariate for age of calf (i.e., on-test age for IGF-I measurements and age at d 56 or 140 for ultrasound data). Fixed effects were fitted as separate main effects. Potential interactions among the fixed effects and of the covariates with fixed effects were ignored. Analysis of ultrasound data was repeated using weight at d 56 or 140 as the linear covariate in place of age of calf. Estimates of maternal heritability were 0.01 for all measures of IGF-I other than IGF28 and were 0.01 or less for BF and LMA, regardless of the covariate used. In addition, the proportion of phenotypic variance due to permanent environmental effect of dam was zero for all measures of IGF-I and was <0.10 for BF and LMA. Either maternal genetic and permanent environmental effects were unimportant in these data or else the data were not adequate in size or structure (records per cow and number of generations with data) to allow accurate estimation of such effects. Therefore, maternal and permanent environmental effects were deleted from the final model. Analysis of the ultrasound data using the reduced model was again repeated with either age or weight of calf as the linear covariate.
Secondly, IGF-I concentration on d 28, 42, and 56 of the postweaning test, as well as mean IGF-I concentration, was paired with the ultrasound measurements of BF and LMA in a series of bivariate analyses using the reduced model for each trait to estimate the additive genetic (rA1A2), environmental (rE1E2), and phenotypic (rP1P2) correlations among the traits. Separate analyses were again performed using the age and weight covariates for BF and LMA. Serum IGF-I concentration was adjusted for on-test age in all bivariate analyses.
In both the univariate and bivariate analyses, convergence was defined as the point where the variance of the simplex was less than 10-9. In an attempt to ensure that the log likelihood was the global, and not a local, maximum, "cold restarts" of the MTDFREML programs were performed using the converged values. Restarts were performed until -2 times the log likelihoods used in the simplex search algorithm did not change to the second decimal place from one restart to the next. Approximate standard errors were obtained for using the formula of Swiger et al. (1964). Approximate standard errors for the genetic correlations were derived using the formula presented by Falconer and Mackay (1996).
|
Results and Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionImplicationsLiterature Cited |
|---|
.
Estimates of heritabilities obtained for serum IGF-I concentration are shown in Table 2. Direct heritability for IGF-I concentration at d 28, 42, and 56 of the postweaning test and for mean IGF-I concentration was 0.26 ± 0.07, 0.32 ± 0.08, 0.26 ± 0.07, and 0.32 ± 0.08, respectively. These estimates generally agree with those obtained by Davis and Simmen (2000) using data from earlier years of the same selection experiment. The estimates also agree well with the values of 0.31 observed by Herd et al. (1995) at birth, 0.32 by Johnston et al. (2001) at 8 to 10 mo of age, and 0.34 and 0.43 by Johnston et al. (2002) at 9 and 22 mo of age in cattle fed at two different locations in Australia.
|
) ranged from 0.17 ± 0.11 to 0.28 ± 0.12 depending on the covariate (i.e., age or weight) and time of measurement (d 56 or 140 of the postweaning period), values that generally were slightly smaller than the estimate of 0.28 obtained by Davis and Simmen (2000) for carcass BF of bulls from the same experiment. Direct heritabilities for ultrasound estimates of LMA (Table 2) ranged from 0.20 ± 0.10 to 0.36 ± 0.12. These values agree well with those of Rouse and Wilson (1997), who reported estimates of 0.26 and 0.25 for carcass LMA and BF, respectively, for the Angus breed. Koots et al. (1994) reported average heritabilities from a large number of studies to be 0.44 and 0.42, respectively, for carcass BF and LMA.
Direct heritabilities of ultrasound measurements were generally similar at age- and weight-constant endpoints (Table 2). Arnold et al. (1991), Johnson et al. (1993), Robinson et al. (1993), and Devitt and Wilton (2001) also found heritability estimates for ultrasound BF and LMA to be similar at age- and weight-constant endpoints. Koots et al. (1994) found that there were no consistent differences among unadjusted, age-adjusted, and live weight-adjusted heritability estimates for carcass traits in the large number of studies that they analyzed.
Additive genetic correlations of IGF-I concentrations with BF at d 56 of the postweaning period were near zero when age at d 56 was used as the covariate, whereas the genetic correlation of IGF-I with BF at d 140 averaged 0.20 when age at d 140 was used as the covariate for BF (Table 3). Davis and Simmen (2000) reported a genetic correlation of -0.26 between IGF-I concentration and carcass BF of bulls from the same selection experiment. The different genetic correlations with BF obtained in these two studies were likely due to growth pattern differences between the bulls used by Davis and Simmen (2000) and the heifers that were included in the present study, as well as to differences in diets between the bulls and heifers. In addition, in the study of Davis and Simmen (2000), carcass data were only available for bulls not selected for breeding, whereas in the present study, ultrasound data were available for all bulls and heifers.
|
for IGF-I and carcass LMA of bulls. In all cases, genetic correlations were larger and more positive at d 140 than at d 56, indicating that genetic relationships of serum IGF-I with BF and LMA were more fully expressed at older ages. Calves with higher serum IGF-I concentrations had higher BF and LMA at the conclusion of the postweaning period.
Environmental correlations of the various IGF-I measurements with BF at d 56 and 140, and with LMA at the same time points, averaged 0.27, 0.31, 0.35, and 0.48, respectively, (Table 3) when ultrasound data were adjusted to an age-constant endpoint, indicating that environmental effects that resulted in increases in serum IGF-I concentration also resulted in increases in BF and LMA. Davis and Simmen (2000) reported an environmental correlation of IGF-I concentration with both carcass BF and LMA of 0.09.
Phenotypic correlations of IGF-I measurements with BF at d 56 and 140 and with LMA at d 56 and 140 averaged 0.20, 0.28, 0.21, and 0.40, respectively (Table 3), when the age-constant endpoint was used, indicating moderate phenotypic associations of endocrine IGF-I with these traits. In all instances, phenotypic correlations were larger on d 140 than on d 56 due to the older ages allowing more time for expression of phenotypic relationships of serum IGF-I with BF and LMA. Davis and Simmen (2000) reported phenotypic correlations of IGF-I concentrations with BF and LMA that averaged 0.00 and 0.12, respectively, indicating little phenotypic association of serum IGF-I with these traits when measured on the carcasses of bulls.
Additive genetic correlations of serum IGF-I concentration with BF averaged 0.05 at d 56 and -0.07 at d 140 (Table 4) when the backfat data were adjusted for weight. The d-56 estimates were similar to those obtained using age as a covariate, whereas the weight-adjusted d-140 estimates were considerably smaller than the age-adjusted estimates. Additive genetic correlations of IGF-I concentration with LMA on d 56 and 140 averaged -0.22 and -0.04, respectively. Therefore, adjustment of LMA for weight rather than age resulted in genetic correlations with IGF-I that were more strongly negative at d 56 and slightly negative, rather than moderately positive, at d 140. Devitt and Wilton (2001) pointed out that weight-constant LMA is a closer measure of actual muscling of the carcass than age-constant LMA, because adjustment for weight removes the effect of growth rate and body size. Robinson et al. (1993) found that 50% of the variation in LMA was explained by weight at time of ultrasound measurement. Comparison of the age- and weight-adjusted genetic correlations indicate that the positive correlation of IGF-I with BF and LMA at d 140 (Table 3) was partially mediated by weight. Davis and Simmen (1997) reported genetic correlation estimates of -0.43 and -0.31 between mean IGF-I concentration and weight at d 56 and 140, respectively, using data from the same divergent IGF-I selection experiment.
|
Thus, additive genetic correlations of serum IGF-I concentration with BF and LMA were small on d 56 and moderately positive on d 140 when age of calf was used as the covariate for the ultrasound measurements. Bishop et al. (1989) found that phenotypic correlations of serum IGF-I concentrations on d 0 and 140 of a postweaning test with LMA reflected a moderately positive degree of association (0.31 and 0.24, respectively). Anderson et al. (1988) reported that IGF-I concentration was negatively correlated with percentage of carcass fat, carcass fat accretion rate, total carcass fat, and fat thickness, and that it was positively correlated with percentage of carcass protein in Simmental crossbred bulls. Results of Anderson et al. (1988) agree with those of other studies that have shown that IGF-I stimulates protein synthesis and inhibits protein breakdown in skeletal muscle in vitro (Roeder and Hossner, 1986; Roeder et al., 1988) and in vivo (Douglas et al., 1991; Hossner et al., 1997). The IGF have been shown to stimulate myoblast proliferation, as well to maintain muscle fiber differentiation (Bass et al., 1999). The IGF axis is an important component of muscle development, as evidenced by the severe disruption of skeletal muscle development that results from deletion of the IGF-I or IGF-II genes (Bass et al., 1999). Unlike myostatin, which seems to act predominantly on muscle growth, the IGF gene deletions appear to influence total overall growth (Bass et al., 1999). White et al. (2003) suggested that increased muscle IGF-I levels stimulate satellite cell proliferation and maintain a high number of proliferating satellite cells at a point in the growth curve where satellite cell numbers and activity normally decline, resulting in the increased muscle growth observed in Revalor-S implanted steers.
Conversely, Luxford et al. (1998) reported positive genetic and phenotypic correlations of ultrasonic fat depth at the last rib with plasma concentrations of IGF-I at 5 wk of age in pigs. In addition, Herd et al. (2002) observed positive coefficients for the regression of circulating levels of IGF-I on EBV for subcutaneous fat depth at the 12th/13th rib and P8 rump sites of cattle measured ultrasonically during the postweaning period. Johnston et al. (2001) also found that IGF-I was positively genetically correlated with various measures of fatness, both on live cattle (r = 0.62 and 0.72 for ultrasound estimates of P8 rump and rib fat thicknesses, respectively) and carcasses (r = 0.47 for intramuscular fat percentage and 0.38 for P8 fat depth). Cameron and Cienfuegos-Rivas (1994) found phenotypic correlations of serum IGF-I concentration with ultrasonic measures of longissimus muscle depth and subcutaneous fat depth that ranged from 0.20 to 0.42 when lambs were fed for a time-constant postweaning period, estimates that agree well with those of the present study. Hayden et al. (1993) observed that plasma concentrations of IGF-I were highly positively correlated with empty body fat accretion and with empty body weight gain and protein deposition in steers undergoing compensatory gain after dietary energy restriction. Ballard et al. (1993) showed that administration of exogenous IGF-I to rats resulted in marked increases in nitrogen retention.
In the present study, environmental correlations of IGF-I concentration with ultrasound measurements were positive and generally of moderate magnitude, indicating that environmental effects that resulted in increases in serum IGF-I concentrations also resulted in increases in BF and LMA. Phenotypic correlations of IGF-I concentration with BF and LMA area were positive, and were of greater magnitude at the age-constant than at the weight-constant endpoint.
|
Implications |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionImplicationsLiterature Cited |
|---|
|
Footnotes |
|---|
2 This experiment was a contributing project to North Central Regional Project NC-196, "The Genetics of Body Composition in Beef Cattle."
3 The authors wish to thank W. D. Shriver, J. D. Wells, and F. J. Michel for their excellent technical assistance.
5 Current address: Department of Physiology and Biophysics, University of Arkansas for Medical Sciences and Senior Investigator in Developmental Biology, Arkansas Children’s Nutrition Center, 1120 Marshall St., R-2027, Slot 512, Little Rock, Arkansas 72202.
4 Correspondence— phone: 614-292-4984; fax: 614-292-7116; E-mail: Davis.28{at}osu.edu.
Received for publication October 23, 2002. Accepted for publication May 8, 2003.
|
Literature Cited |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionImplicationsLiterature Cited |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. A. Afolayan and N. M. Fogarty Genetic variation of plasma insulin-like growth factor-1 in young crossbred ewes and its relationship with their maintenance feed intake at maturity and production traits J Anim Sci, September 1, 2008; 86(9): 2068 - 2075. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Estany, M. Tor, D. Villalba, L. Bosch, D. Gallardo, N. Jimenez, L. Altet, J. L. Noguera, J. Reixach, M. Amills, et al. Association of CA repeat polymorphism at intron 1 of insulin-like growth factor (IGF-I) gene with circulating IGF-I concentration, growth, and fatness in swine Physiol Genomics, October 19, 2007; 31(2): 236 - 243. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Yilmaz, M. E. Davis, and R. C. M. Simmen Estimation of (co)variance components for reproductive traits in Angus beef cattle divergently selected for blood serum IGF-I concentration J Anim Sci, August 1, 2004; 82(8): 2285 - 2292. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) and heritability of direct effects (