Background 

Anabolic implants increase muscle growth through IGF-I and myostatin pathways activating satellite cells and are known to negatively affect marbling score, which may be due to changes in IGF-I and myostatin pathways. Interestingly, the terminal implant administered late in the finishing phase has the greatest negative impact on marbling score. Implant programs that eliminate the late-finishing implant do not significantly reduce quality grade compared to non-implanted cattle even when cattle are implanted during suckling, growing, and early-finishing periods. Therefore, administration of implants closer to the end of the finishing phase has a greater negative impact on quality grades, whereas administration of implants early in the production cycle may not negatively influence quality grades. Presumably, implants cause similar changes in IGF-I and myostatin pathways regardless of implant timing (growing or finishing phase), which suggests that intramuscular adipose tissue reacts differently to these changes during the finishing phase compared with the growing phase.  

Few studies have evaluated the effects of muscle growth and metabolism on marbling development. In a previous project, we observed that IGF-I and myostatin expression in muscle were differentially related with receptor expression in marbling and rib fat in the stocker versus finishing phase. In the study, we utilized anabolic implants to alter IGF-I and myostatin pathways in skeletal muscle and determined the influence on marbling development. Presumably anabolic implants cause similar changes in skeletal muscle when administered during the growing and finishing phases. However, based on published research, implantation during the late stages of the finishing phase has a greater impact on final marbling score than implantation during the growing phase. Together these data suggest that intramuscular adipose tissue reacts differently to changes in IGF-I and myostatin pathways in muscle during the finishing phase compared with the growing phase.  

We hypothesized that implanting steers once at the beginning of the finishing phase would not negatively impact marbling compared with non-implanted controls, whereas implanting steers late in the finishing phase would negatively impact marbling. In addition, we hypothesized that both early and late administration of an implant would increase expression of IGF-I and decrease expression of myostatin in skeletal muscle. The objectives of the current experiments were to determine gene expression profiles related to adipocyte lipogenesis and adipogenesis of intramuscular and subcutaneous fat, and genes involved in aerobic/anaerobic metabolism and markers of satellite cell activation in muscle. In addition, we determined expression of genes involved in IGF-I and myostatin intercellular signaling mechanisms, and angiogenic growth factors in muscle and adipose tissue. 

Methodology 

In Exp. 1, predominantly Angus heifers (n = 187; BW = 361 kg) were fed during a 122-d finishing period. Heifers were randomly assigned to 1 of 3 implant programs at the Willard Sparks Beef Cattle Research Center at Oklahoma State University. Implant programs included: 1) no implant during the growing or finishing phase (Control; CON); 2) a single trenbolone acetate/estradiol implant administered at the beginning of the finishing phase (d 0; early implant; Early); and 3) a single trenbolone acetate/estradiol implant administered during the late stage of the finishing phase (d 56; late implant; Late). Heifers were allotted to feedlot pens (7 pens per treatment; 8-9 heifers/pen) and adapted to a common finishing diet before initiation of the study. Heifers on all treatments were fed a common high concentrate finishing ration throughout the finishing phase. A subset of heifers from the no implant and early implant treatments (6 heifers per treatment) were harvested on d 28 and 29 (28 and 29 days after early implant) for collection of longissimus dorsi and adipose tissue. An additional subset of heifers from all three treatments (6 heifers per treatment) were harvested on d 84 and 85 (28 and 29 d after the late implant) for the same tissue collections. Heifers were harvested following administration of the implant to collect tissue samples during the period when estradiol and trenbolone acetate levels are the greatest. Each subgroup of heifers (N = 30) were harvested at the Oklahoma State University Food and Agricultural Products Center to determine carcass characteristics (marbling score, 12th-rib fat thickness, ribeye area, USDA Quality and Yield Grade) and carcass composition (specific gravity), and to collect samples of longissimus muscle, subcutaneous fat over the 12th-rib and perirenal fat. The remaining heifers (n = 157) were fed to a common 12th-rib fat thickness of 1.27 cm and harvested at a commercial abattoir to collect carcass data.  

Longissimus dorsi muscle and adipose samples were collected from 30 heifers that were subject to one of the three implant programs. Tissue samples were stored at -80˚C in RNAlater (Invitrogen) until later dissected under magnification. Intramuscular fat was identified in small cross-sections of muscle, separated from the muscle tissue, and stored separately. Total RNA was isolated from both longissimus muscle and fat samples using the RNA isolation procedure for TRIzol Reagent (Ambion). cDNA was synthesized according to the manufactures protocol using QuantiTect Reverse Transcription Kit (Qiagen). Quantitative real-time PCR (qRT-PCR) was used to determine the mRNA expression level of specific genes using RT2 SYBR Green Fluor Mastermix (Qiagen) and gene-specific primers.  

In Exp. 2, tissue samples were collected from steers harvested in an intermediate harvest following a growing phase. Longissimus dorsi samples were dissected, under magnification, according to the maturity of the intramuscular fat. Intramuscular fat was identified in cross-sections of longissimus dorsi muscle samples and sorted, based on visual assessment (Figure 1), into one of three categories: immature (MM), intermediate (ME), and mature (MA). Intramuscular fat was removed from muscle tissue and stored separately according to maturity. Muscle fibers lying immediately adjacent to intramuscular depots was collected and stored separately according to the maturity category of the intramuscular fat it was associated with. Muscle fibers not associated with any intramuscular adipose tissue development were also collected and stored separately.  

Performance and carcass data were analyzed using a general linear model that included the fixed effect of treatment with pen as the experimental unit. Gene expression data were analyzed using a general linear model that included the fixed effects of treatment, tissue and treatment × tissue interaction with animal as the experimental unit. All data were analyzed using SAS (SAS Inst., Inc., Cary, NC). LSmeans were separated using Fisher Protected LSD with alpha 0.10. Treatment differences were declared significant when P ≤ 0.05. Pearson’s correlations were also computed between the gene expression in muscle and its related intramuscular adipose tissue for each stage of maturity using the Proc Corr procedure of SAS. 

Findings 

Feedlot Performance  

Feedlot performance data are summarized in Table 1. The estradiol-TBA implant (20 mg of estradiol and 200 mg of trenbolone acetate) administered on d 0 or 56 improved overall feedlot performance and efficiency. From d 0 to 56 cattle that had received an implant had a 28% improvement in ADG versus non-implanted cattle (P < 0.01). During the final period (d 56-122), all treatments were significantly different for ADG. Early-implanted cattle outperformed the controls by 18% (P < 0.01), while the late-implant group outperformed the controls by 40% (P < 0.01). Overall, an 18% improvement in ADG was reported for implanted heifers versus control (P < 0.01), and early-implanted heifers tended to have a 5.8% improvement over late-implanted heifers (P = 0.10).  

During each period and overall, there were no treatment effects of implant on DMI (P ≥ 0.17; Table 1). During the initial 56 days on feed, the implanted heifers were 28% more efficient (P < 0.01) than non-implanted heifers (Table 1). From d 56-122, the late-implanted cattle were 42% more efficient than the non-implanted cattle (P < 0.01), and 28% more efficient than the early-implanted cattle (P < 0.01). Overall, implanting heifers improved their efficiency by more than 19% (P < 0.01), and time on implant (early vs. late) did not affect overall feed efficiency (P = 0.56).  

On a carcass adjusted basis (Table 1), implanting heifers increased ADG by 24% and feed efficiency by 25% (P < 0.01), independent of time of implant (P ≥ 0.16).  

Carcass Characteristics  

All carcass measurements and grading distributions are presented in Table 2. Implanting heifers increased HCW (P < 0.01) and dressing percentage (P = 0.04) versus non-implanted heifers, independent of implantation time. Rib-fat thickness was not affected by treatment (P = 0.25). Implanting did increase LM area (P = 0.01) by 7.5%. Implanting heifers late in the finishing period tended to improve yield grade (P = 0.07), while early implantation was not different from the control. Numerically, marbling score was greatest for the non-implanted heifers, but no statistical differences were observed (P = 0.18). Our study revealed no significant differences in quality grade or yield grade distributions (P > 0.20; Table 2). Non-implanted cattle numerically were greater in percent total choice and lower in percent of yield grade 1 carcasses. Late implantation numerically decreased the percentage of yield grade 3 carcasses.  

Longissimus dorsi Gene Expression  

All gene expression data for the early and late harvest are summarized in Table 3 and Table 4, respectively. In longissimus dorsi muscle, the mRNA expression of NADH dehydrogenase 2 (ND2) was significantly greater for non-implanted cattle at harvest 1, but significantly lower at harvest 2. This resulted in a treatment × harvest interaction (P < 0.01). This interaction revealed that cattle that received an implant on d 0 had greater mRNA expression of ND2 at harvest 2 than at harvest 1 (56 d earlier; P < 0.01). This suggests that longer exposure to a combination implant increases the aerobic muscle metabolism in heifers. During harvest 2 cattle from the late implant group had greater mRNA expression of ND2 than heifers from the early implant group that were harvested 56 d earlier (harvest 1; P = 0.01). This indicates that stage of production (growing vs. finishing) also affects the mRNA expression of ND2 in implanted heifers. At harvest 2, the mRNA expression of COX3 was greater in implanted cattle, specifically the early-implantation group (P = 0.03). This generated a treatment × harvest interaction (P = 0.03), specifically identifying a harvest effect for the early-implanted heifers (P = 0.05). During harvest 2, LDHA was not affected by treatment. 

The mRNA expression of genes related to longissimus dorsi angiogenesis are presented in Table 3 and Table 4, relative to harvest 1 and 2, respectively. The mRNA expression of vascular endothelial growth factor A (VEGFA) was not affected by treatment during either harvest period (P > 0.85). Expression of angiopoietin-1(ANGPT1) was unaffected during the first harvest, but at harvest 2, implanted heifers showed greater mRNA expression than control heifers. Overall, cattle at harvest 2 revealed lower expression than cattle at harvest 1 (P < 0.01), suggesting an overall decrease in vascular remodeling as heifers mature. Another gene involved with angiogenesis is angiopoietin-2 (ANGPT2). The mRNA expression of ANGPT2 was greater in implanted cattle during the first harvest than non-implanted cattle. Suggesting that additional vascular remodeling was associated with increased muscle development and aerobic metabolism. During harvest 2 ANGPT2 was not affected by treatment, which resulted in a treatment × harvest interaction (P = 0.03). Specifically, there was a harvest effect for the early implanted heifers, indicating that implant duration can reduce the mRNA expression of ANGPT2 (P < 0.01). Heifers in the second harvest that were implanted late had decreased expression of ANGPT2 than heifers in the first harvest that were implanted early. This would signify that stage of production (maturity) alters the expression of ANGPT2 in implanted heifers.  

Myostatin (MSTN) and insulin-like growth factor-1 (IGF-1) are both involved in intracellular signaling pathways. The mRNA expression of MSTN and IGF-1 were not affected by implantation during the first harvest (P > 0.18). During the second harvest, MSTN was significantly reduced in heifers that had been implanted, which would be expected in heifers that were growing more rapidly. There was no treatment effect for IGF-1 expression, but the average expression of the implanted cattle tended to be less than the non-implanted cattle. Overall there was a significant harvest effect (P = 0.05). Heifers had lower mRNA expression levels of IGF-1 during the finish stage harvest than during the first harvest when they were still at a more immature production state (P = 0.05).  

The mRNA expression of myogenic differentiation 1 (MyoD), myogenic factor 5 (Myf5) and myogenin (MyoG) were not affected my implant treatment during either harvest (P > 0.32). The expression of paired box 7 (Pax7) was increased in cattle that had received an implant during the first harvest (P = 0.05). A contrast revealed that cattle implanted early and harvest during the first harvest had greater mRNA expression of Pax7 than heifers implanted late and harvested during the second harvest (P = 0.02). This indicates that cattle implanted late in the finishing period have less myogenic activity than cattle implanted during the growing stage of production.  

Muscle Gene Expression  

For Exp. 2, expression of genes evaluated in longissimus dorsi muscle is reported in Table 5. Genes associated with metabolism, COX3 and ND2, showed no change among different adipose tissue maturity categories. There was no change in fiber type expression among the differing maturity levels of adipose tissue. Satellite cell activity was evaluated in muscle tissue by measuring the expression of PAX7, MYOG, and MYF5. The results show no change in expression in muscle associated with more mature adipose depots indicating that satellite cell activity does not affect intramuscular fat development. VEGFA, ANGPT1, and ANGPT2 all genes involved in angiogenesis, expression were not different as adipose tissue matured. Extracellular matrix changes were detected with increasing maturity and changes in collagen expression. Levels of COL1A1 (P = 0.01), COL1A2 (P = 0.004), and COL6A2 (P = 0.04) were all increased in muscle associated with immature intramuscular adipose development above the levels of the other maturity categories. However, genes responsible for the remodeling of ECM, MMP2 and TIMP4 remained constant throughout maturity stages. It was attempted to evaluate MMP9 expression; however, mRNA levels were not detectable in skeletal muscle tissue (data not shown).  

Intramuscular Adipose Tissue Gene Expression  

Expression of genes evaluated in intramuscular adipose tissue is reported in Table 6. As intramuscular fat increased with visible maturity, PREF1 levels decreased (P = 0.002) while PPARγ (P = 0.02) and FABP4 (P < 0.0001) increased with maturity. These results show an increase in the number of differentiated adipocytes indicating that we were successful in separating intramuscular adipose tissue into different stages of maturity based on visual assessment under the microscope. FASN levels didn’t change and G3PDH (P = 0.004) was lower with more mature adipose tissue, which is contrary to what was expected. Angiogenic genes VEGFA and ANGPT1-2 remained constant as adipose tissue matured. The effects of extracellular signaling pathways on adipogenesis were also evaluated. WNT5B (P < 0.0001) mRNA expression in intramuscular adipose tissue was evaluated and had the highest expression in mature intramuscular fat with no differences between immature and intermediate stages of development. It was attempted to evaluate WNT10B but mRNA expression was undetectable in intramuscular adipose tissue.  

Intercellular Signaling Pathways Gene Expression  

MSTN and IGF1 expression remained constant in muscle tissue regardless of adipose tissue maturity category. Myostatin’s receptor, ACVR2B (P= 0.002), and FST (P = 0.01) measured in intramuscular adipose tissue did differ with maturity being expressed highest in immature adipose tissue. Changes in IGF1 receptor (IGF1R) mRNA expression in adipose tissue were detected with lower expression (P = 0.03) in intermediate than immature and mature categories. mRNA expression of IGFBPs were different among adipose tissue, but changes were not consistent between IGFBPs. IGFBP-2 (P < 0.0001) and -3 (P < 0.0001) both increased with increasing maturity; conversely, IGFBP-6 decreased (P < 0.0001) with increasing maturity. mRNA expression of IGFBP-1, -4, and -5 was consistent among maturity categories. Adipokines and their receptors present in muscle were also evaluated. ADIPOQ mRNA expression remained constant as adipose matured, which is contrary to what was expected. Expression level of the ADIPOQ receptors (ADIPOR) had differing expression patterns. ADIPOR1 expression didn’t change with increasing maturity, but ADIPOR2 mRNA expression increased (P < 0.0001) as muscle became associated with more mature adipose tissue. As expected, leptin mRNA levels increased (P = 0.05) as number of differentiated adipocytes increased. LEP receptor (LEPR) remained constant along maturity categories. An attempt was made to evaluate the resistin signaling pathway; however, resistin mRNA in intramuscular tissue and resistin receptor mRNA in muscle were not detectable (data not shown).  

Pearson’s Correlations between Muscle and Intramuscular Adipose Tissue  

Correlations were calculated between intramuscular adipose tissue and the muscle tissue that corresponded with each maturity category (Table 7). A strong positive correlation (r = 0.69 to 0.94) was observed between mRNA expression of VEGFA, ANGPT1 and ANGPT2, genes involved in angiogenesis, in longissimus dorsi muscle and the expression of ANGPT1 and ANGPT2 in intramuscular adipose tissue. mRNA expression of genes involved with adipogenesis measured in adipose tissue, FASN and PPARγ, were also shown to be strongly correlated (r = 0.89 to 0.96) with angiogenic genes measured in muscle tissue. In addition, the adipokines ADIPOQ and LEP showed a moderate to strong correlation (r = 0.56 to 0.94) with angiogenic genes mRNA expression in muscle tissue. Interestingly, these correlations were observed only in immature intramuscular development. These data suggest that there is a highly coordinated set of changes that occur between skeletal muscle and intramuscular adipose tissue during the early development of intramuscular adipose tissue.  

Implications 

Results of this study confirm that utilizing an estradiol-trenbolone acetate implant can significantly improve growth rate and feed efficiency in the feedlot phase of beef cattle production. In addition, carcass weight, dressing percentage, and longissimus dorsi muscle area were improved by implanting, without negatively affecting carcass quality. Time of implantation did not greatly alter any performance or carcass results. Results suggest that using a combined estradiol–trenbolone acetate implant, especially early in the finishing phase, does not have adverse effects on the development of marbling in heifers.  

Development of intramuscular fat is a complex process that is influenced by a variety of signals. The close proximity of intramuscular fat to muscle tissue during development indicates that intercellular signaling between these two tissues is crucial for marbling development. Experiment 2 shows that early in the development of intramuscular adipose tissue, remodeling of the extracellular matrix occurs along with angiogenesis which is critical for development of marbling. The strong correlation between angiogenic growth factors in longissimus dorsi muscle with angiogenic growth factors and markers of adipocyte differentiation in immature intramuscular fat development suggests that there is a highly coordinated change that occurs between skeletal muscle and intramuscular fat during the early stage of adipose development.  

Further understanding the interactions between skeletal muscle and adipose tissue during intramuscular development could allow for development of management strategies that reduce waste fat and optimize carcass quality. Optimization of the development of muscle and adipose tissue will allow for the efficient production of high quality beef that will meet the demands of the consumer.