INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

THE CONDITION OF DOUBLE MUSCLING (DM) in Belgian Blue cattle stems from an 11-bp deletion in the myostatin gene, also known as GDF-8, that gives rise to a premature stop codon and a severely truncated myostatin protein (7, 11). On the basis of structural similarities, it appears that myostatin belongs to the greater transforming growth factor- (TGF-) family and has been identified as an important negative regulator of muscle development in a mouse model of gene deletion (16). In the absence of myostatin, the skeletal musculature of mice is two to three times greater in mass than that of wild-type mice (16). In parallel with this, the skeletal musculature of DM Belgian Blue cattle is ~20% greater than in normal-muscled (NM) cattle (21).

The enlarged musculature of the myostatin-deficient mouse results from effects on both early (hyperplasia) and late (hypertrophy) myogenic processes (16), and mRNA is expressed at different levels in various postnatal muscles (11, 16). Such observations suggest that myostatin may influence muscle development at a number of stages. The aim of this experiment was to measure the fetal expression of myostatin in relation to skeletal muscle-specific transcription factors (MRFs) in normal muscle and the myostatin-deficient DM condition, to help determine a role for myostatin in the physiology of muscle formation.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Animals. Fetuses with an expected DM phenotype were produced using standard superovulation and embryo transfer techniques (9). Donor Belgian Blue cows were inseminated with semen from Belgian Blue bulls. Embryos were recovered 7 days after insemination and transferred to Hereford × Friesian recipient heifers. Fetuses with an expected NM phenotype were produced either by artificial insemination of heifers from the Hereford × Friesian recipient herd or by embryo transfer into recipients from this herd. Using in vitro techniques (15), we generated embryos from ovaries collected from beef and dairy cows of NM phenotype. Semen used to generate NM fetuses was from Friesian bulls. The recipients were grazed in a single herd and slaughtered at 50, 70, 90, 120, 160, 210, and 260 days of gestation.

RNA isolation and RT-PCR. Total RNA was isolated from hindlimb and muscle samples using Trizol (GIBCO BRL, Grand Island, NY) according to the manufacturer's instructions and treated with RNAse-free DNAse 1 (Roche Diagnostics, Indianapolis, IN). First-strand cDNA was synthesized in a 20-µl RT reaction from 2 µg of total RNA, using a Superscript Preamplification kit (GIBCO BRL), and semiquantitative, multiplex RT-PCR was carried out with 2 µl of the RT reaction as template. Additional RT reactions were produced for each RNA sample (n = 53) as required. Pairs of primers for ubiquitin-activating enzyme (UAE) were included in PCR reactions as an internal control to correct for cDNA and loading differences in the Southern analysis. All PCR reactions included primers for both the candidate and control gene, apart from myostatin, which was amplified independently from UAE. In this situation, reactions were carried out at the same time, with duplicates of cDNA template from the same stock and differing only by the addition of primers. Q-solution (20%, QIAGEN, Valencia, CA) was also added to the PCR reaction for MyoD/UAE.

View larger version (39K): [in this window] [in a new window]   Fig. 1.   Optimization of cycling conditions for RT-PCR using RNA from 70-day hindlimb. The Southern radiograph (A) and graph (B) show linearity of signal detection for each pair of primers tested. Fourteen cycles were selected for the experimental detection of myostatin and 18 cycles for MyoD, myogenin, and ubiquitin activating enzyme (UAE). OD, optical density.

Southern analysis. PCR samples were run on a 2% agarose gel, transferred overnight onto Hybond N+ membrane (Amersham, Bucks, UK), and immobilized by ultraviolet cross-linking (Stratalinker, Stratagene, La Jolla, CA). Membranes were prehybridized at 42°C in 5× sodium chloride-sodium citrate (SSC), 50% formamide, and 5× Denhardt's solution with 1% SDS and 0.25 mg/ml of salmon sperm DNA for 2 h, then hybridized in the same solution overnight with the appropriate random-primed, ³²P-labeled cDNA probes. After hybridization, membranes were washed three times at 50°C for 15 min each with 2× SSC and 0.1% SDS. They were exposed against X-Omat AR10 film (Eastman Kodak, Rochester, NY) for varying periods of time from 1 to 48 h at 80°C (Fig. 1). For quantitation, radiographs were scanned with a Bio-Rad 670 Densitometer, and band densities were estimated using Molecular Analyst (Bio-Rad Laboratories, Hercules, CA).

Northern analysis. Differences in expression of genes, as determined by semiquantitative PCR, between NM and DM underwent further measurement by Northern analysis. A limited number of RNA samples from 120-day samples of muscle (n = 2) from NM and DM were used for Northern hybridization with ³²P-labeled cDNA probes made by PCR amplification of the entire coding sequence of myostatin or MyoD and a ³²P end-labeled oligonucleotide probe for 28S ribosomal RNA (AACGATCAGAGTAGTGGTATTTCACC). RNA (12 µg for myostatin and 20 µg for MyoD) was run on a 1.2% formaldehyde-agarose gel, transferred, and cross-linked as described earlier. Membranes were prehybridized at 60°C in Church and Gilbert buffer (1 mM EDTA, 0.5 M NaHPO₄, pH 7.2, 7% SDS) for 2 h, then hybridized in the same solution overnight with the appropriate ³²P-labeled probe. After hybridization, membranes were washed as previously described, then washed with 1× SSC and 0.1% SDS at 50°C for 15 min. Membranes were exposed against X-Omat AR10 for varying periods of time; band densities on radiographs were estimated as described earlier and then stripped and rehybridized with the ³²P end-labeled 28S probe for estimation of RNA loading.

Statistics. For each gene of interest, an average optical density value for UAE was calculated, and individual UAE values were expressed as a percentage of the average. Values of optical density for the specific gene of interest were then multiplied by the UAE percentages and used in an ANOVA. Data were log-transformed when not normal in distribution, and age and breed were tested as main effects using ANOVA. Comparisons were also made between the breeds at each gestational age.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Results are presented with the significance of main effects from ANOVA indicated in Figs. 2 and 3. Differences between NM and DM at each gestational age are indicated by asterisks if significant.

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Measurement of crown-rump length (A) and weight of M. semitendinosus (B) in normal (NM)- and double-muscled (DM) cattle fetuses of different gestational ages. Means + SE. *P < 0.05; ***P < 0.001; NS, not significant.

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Optical density values of myostatin (A), MyoD (B), and myogenin (C) expression in NM and DM cattle fetuses of different gestational ages after adjustment for expression of UAE, as determined by semiquantitative RT-PCR. Means + SE. *P < 0.05; ***P < 0.001.

Skeletal and muscle growth. During fetal development, there was no difference in skeletal growth between DM and NM fetuses, as determined by measurement of crown-rump length (Fig. 2A). However, in musculus semitendinosus that was dissected from animals of 120 days' gestation and older, muscle weights were always greater in DM (Fig. 2B, P < 0.001).

Semiquantitative RT-PCR and Southern analysis. Expression of myostatin mRNA (Fig. 3A) varied with developmental age (P < 0.001) and showed a peak of expression at 90 days of gestation. Levels were also greater in DM than in NM (P < 0.001). Myostatin expression returned to relatively low levels in NM at 120 days, whereas expression in DM remained elevated until 210 days. By 260 days, there was no difference between the breeds.

Northern analysis. A comparison of NM and DM expression at 120 days (Fig. 4) supports the observations made by RT-PCR with increased expression of myostatin and MyoD in DM.

View larger version (108K): [in this window] [in a new window]   Fig. 4.   Northern radiographs of myostatin and MyoD mRNA in M. semitendinosus from NM and DM cattle fetuses at 120 days gestation, with corresponding detection of 28S ribosomal protein RNA as a loading control.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The final phase of determining numbers of muscle fibers in a muscle occurs during the terminal differentiation and fusion of proliferating myoblasts into primary and secondary myotubes and myofibers. It is generally considered that heterogeneous populations of myoblasts separately give rise to successive generations of muscle fibers (20). In cattle, differentiation of early myoblasts into primary myotubes in the hindlimb is initiated at 39 days of gestation (17), whereas differentiation of late myoblasts into secondary myofibers is established at 90 days (18). Numbers of muscle fibers then increase until 240 days of gestation (18), which is ~40 days before birth (4). The condition of DM in cattle is typified by an increase in the number of muscle fibers in all skeletal muscles and increased mass in those muscles that are located superficially on the body (21). Recently, it has been established that DM in Belgian Blue cattle arises from a deletion in the myostatin gene, which gives rise to the production of a truncated protein (11).

Peak levels of myostatin expression in NM coincide with established primary fiber development and the onset of secondary fiber formation in the hindlimb. Increased expression is not extended throughout 120 to 210 days suggesting that myostatin is unlikely to be a principal regulatory control of the ongoing development and maturation of secondary myofibers. Other members of the TGF- family, TGF-s1-3, are differentially expressed in both developing skeletal muscle and transformed myoblasts (3, 14), and the pattern of myostatin expression that we report is most similar to that of TGF-1. The effects of TGF- on myoblasts are variable, promoting differentiation when myoblasts are relatively mature (25), otherwise inhibiting myoblast proliferation (1). It is tempting to speculate that myostatin may also differentially affect myoblasts, possibly stimulating the induction of terminal differentiation and fusion of early myoblasts while downregulating proliferation in late myoblasts. Such a suggestion has support from other studies that demonstrate that TGF- has variable inhibitory effects on proliferation of early and late myoblasts (6) and fetal and postnatal satellite cells (10).

The expression of myostatin mRNA in DM is similar to NM, but levels are relatively higher throughout myogenesis. A number of possibilities present themselves as explanations for the elevated expression of mutant myostatin. A functional myostatin protein is almost certainly not produced in DM (11), so a component of feedback regulation of myostatin expression may be missing. The regulatory component may be myostatin itself as for TGF- (12, 14). Alternatively, an intracellular/autocrine feedback loop may operate, as suggested for interactions between insulin-like growth factors and TGF-s (3), or an endocrine growth factor stimulated by myostatin may then regulate expression of myostatin mRNA. A further possibility is that breed differences in the cis or trans factors regulating myostatin mRNA are reflected in relative levels of expression.

In NM, the greatest levels of MyoD mRNA expression occurred at 70-120 days, during terminal differentiation of early and late myoblasts. MyoD and myogenin are both integral components of the process of myoblast withdrawal from proliferation and subsequent commitment to differentiation and myofiber formation. Proliferating myoblasts contain basal amounts of MyoD and myogenin (22, 25), and induction of MyoD is associated with the arrest of proliferation in cells (8). However, coexpression of myogenin and p21, which are both MyoD dependent (8, 22), is required to completely arrest the cell cycle (2). These studies indicate that not only proliferation, but cell cycle withdrawal, is a discrete process from further myoblast differentiation and that MyoD is likely to regulate downstream events in the myogenic cascade. Developmental changes in the NM expression of myogenin mRNA were broadly similar to the pattern for MyoD, as might be expected if the expression of myogenin during differentiation is dependent on earlier induction of MyoD. Our data fit well with culture studies where myogenin expression is low in proliferating myoblasts, peaks during differentiation, and declines in myofibers (22).

In DM, MyoD mRNA was increased relative to NM. This difference may be a consequence of myostatin deficiency, having arisen either as a result of increased numbers of myoblasts from DM expressing MyoD at the same developmental age or a greater induction of MyoD above a critical threshold level in a population of terminally differentiating myoblasts. The manner by which myostatin might regulate MyoD expression may be similar to the action of TGF-s on MyoD and myogenin. TGF- treatment of myogenic cultures results in an inhibition of differentiation that is associated with reduced expression of MyoD mRNA (23). Although the downstream cascade of events is likely to include a reduction of MyoD-stimulated expression of myogenin, there is also a direct inhibitory effect of TGF- on the transcriptional activity of both myogenin and MyoD (5, 13). The transition of myoblasts from states of proliferation to differentiation is delayed in MyoD-deficient satellite cells (19, 24), illustrating the importance of tightly regulated and coordinated expression of MyoD during terminal differentiation and myofiber formation.

In conclusion, we propose that one role played by myostatin in normal muscle development is to regulate MyoD expression during terminal differentiation of primary and secondary myoblasts. In the myostatin-deficient condition, it is likely that the elevated expression of MyoD promotes the increased formation of muscle fibers throughout fetal myogenesis. Furthermore, we have also provided evidence that there is a feedback loop for regulation of myostatin expression in NM animals, the pathway of which is currently unknown.

Perspectives

Future research will undoubtedly focus on the pathways by which myostatin controls myoblast and satellite cell division and how this interacts with the biologically more significant myogenic factors that determine whether or not muscle formation takes place.