INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

MYOSTATIN (GDF-8), a recently identified member of the transforming growth factor (TGF)- superfamily of growth factors (15), is expressed predominantly in skeletal muscle and may be a key regulator of skeletal muscle development and growth. Mice in which the myostatin gene is nonfunctional (i.e., knocked out) have considerably more skeletal muscle than the wild-type controls; individual muscles from homozygous mutants weigh two to three times more than those from wild-type mice (15). Furthermore, the "double-muscle" phenotype of three breeds of cattle [Belgium Blue, Piedmontese (8, 11, 16), and Asturiana de los Valles (6)] has been linked to nucleotide deletions, transitions, or transversions within the coding region of the myostatin gene. These mutations likely compromise the biological activity of the protein, which leads to increased muscling via hyperplasia and hypertrophy.

It is not known whether myostatin influences only myofiber formation or perhaps muscle growth and metabolism in association with conditions causing hypertrophy, arrested growth, or atrophy. In the present study, we evaluated the expression of myostatin in multiple porcine tissues and established the ontogeny of myostatin expression with regard to fetal and postnatal growth. Additionally, we evaluated the impact of growth hormone administration on myostatin expression in longissimus and deep and superficial semitendinosus muscle in growing pigs and determined whether the low-birth-weight (runt) piglet syndrome could be associated with differences in myostatin expression. Finally, we determined the impact of food deprivation on myostatin mRNA abundance.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Animal care and use. All procedures were reviewed and approved by the Purina Mills Animal Care and Use Committee. When required, euthanasia was accomplished by exsanguination after mechanical or electrical stunning. Surgical biopsies were performed under local (skeletal muscle) or general (mammary gland) anesthesia according to procedures established by the corporate veterinary staff.

RT-PCR. Standard protocols (21) for RT-PCR were followed. Two sets of primers for the PCR were synthesized based on the mouse myostatin cDNA sequence (15). The first set is described as follows: the sense primer (5'-TCT TGC TGT AAC CTT CCC-3') was designed based on the region between nucleotides 814 and 831, and the antisense primer (5'-GGA TCC TAA TAC GAC TCA CTA TAG GGA GGA ACA CAA ATT CAC ACT CTC C-3') based on the region between nucleotides 1037 and 1056. For the second set, the sense primer (5'-GAA GTC AAG GTG ACA GAC ACA C-3') corresponded to nucleotides 866-887, and the antisense primer (5'-CAT GGC TGG AAT TTT CCC-3') corresponded to nucleotides 1190-1207. The antisense primers contained a 29-bp bacteriophage T7 promoter sequence synthetically added to the 5' end for in vitro transcription. Single-stranded cDNA template was produced for PCR by reverse transcription of porcine longissimus muscle and mammary gland total RNA. Sequence analysis of the PCR products (243 and 342 bp for primer sets 1 and 2, respectively; Fig. 1) indicated that the products obtained with skeletal muscle and mammary gland RNA are identical and have 99% homology (excluding the T7 promoter sequence) with a porcine myostatin cDNA sequence available now in GenBank (accession number AF019623).

View larger version (118K): [in this window] [in a new window]   Fig. 1.   RT-PCR amplification of myostatin cDNA from total RNA extracted from porcine skeletal muscle (lanes 3 and 6) and nonlactating (lanes 4 and 7) and lactating (lanes 5 and 8) mammary gland. Lane 1 contains the 100-bp DNA ladder. Lane 2 is blank. Primer sets generating the smaller (243 bp) and larger (342 bp) PCR products are described in MATERIALS AND METHODS.

Northern blot analysis. Total RNA was extracted from muscle samples using the acidic guanidine thiocyanate method described by Chomczynski and Sacchi (4) and poly(A⁺) mRNA purified using oligo(dT)_12-18 agarose. Northern blot analysis was performed with a radiolabeled probe generated from the 243-bp PCR product via random priming. Muscle poly(A⁺) mRNA (6.2 µg for semitendinosus muscle and 9.4 µg for longissimus muscle) was separated on a 1.0% formaldehyde-agarose gel and transferred to a nylon membrane. After hybridization, the membranes were washed to a final stringency of 0.2× SSC (0.03 M NaCl, 0.003 M sodium citrate, pH 7.0) and 0.1% SDS at 60°C.

In situ hybridization. The antisense riboprobes used for in situ hybridization were generated using the porcine myostatin PCR product (243 bp) as the template (with T7 promoter sequence to the 5' end of the antisense primer) and a commercially available in vitro transcription kit (Promega, Madison, WI). The sense control riboprobe was generated similarly with the bacteriophage T7 promoter sequence added to the sense strand of the PCR product. Riboprobes were radiolabeled with [-³⁵S]UTP. Muscle (longissimus and hindlimb muscle from a fetus at 59 days of gestation) and lactating mammary gland (collected ~24 h postpartum) samples were frozen immediately in 2-methylbutane prechilled in liquid nitrogen. Protocols for the in situ hybridization procedure have been described previously (14). Sections (10 µm) were mounted on siliconized glass slides and submerged in a series of solutions as follows: 3.7% formaldehyde in PBS (pH 7.0), 15 min; proteinase K (2 µg/ml in PBS), 7.5 min; 0.1 M triethanolamine and 0.25% acetic anhydride, 10 min. The slides were then prehybridized at 42°C for 6 h in a solution containing 50% formamide, 0.3 M NaCl, 20 mM Tris · HCl (pH 7.4), 5 mM EDTA, 10 mM NaH₂PO₄, 10% dextran sulfate, 1× Denhardt's solution, and 0.5 mg/ml yeast RNA. The slides were subsequently washed in 2× SSC (0.3 M NaCl, 0.03 M sodium citrate, pH 7.0) and hybridized at 45°C for 20 h in the same solution containing the ³⁵S-labeled antisense riboprobe (750,000 cpm per slide). For control slides, the ³⁵S-labeled sense riboprobe was used under the same conditions. After hybridization, slides were washed as follows (single washes except as noted): with 4× SSC containing 10 mM dithiothreitol, 30 min; twice with 4× SSC, 10 min; twice with a wash buffer (0.4 M NaCl, 10 mM Tris · HCl at pH 7.5, 5 mM EDTA) at 37°C; wash buffer containing 20 µg RNase A/ml, 30 min at 37°C; wash buffer at 37°C, 5 min; 2× SSC, 37°C, 15 min; and finally with 0.1× SSC at 42°C for 45 min. Microautoradiography was completed by coating slides with an autoradiography emulsion solution (type NTB-2, Eastman Kodak, Rochester, NY) at 45°C and allowing a 1-wk exposure in a dark box.

Ribonuclease protection assay. A ribonuclease protection assay (RPA) was used for evaluating the tissue specificity of expression and for relative quantification of myostatin mRNA in porcine tissues. The myostatin PCR product (243 bp) was used as the template (T7 promoter sequence added to the antisense primer) for the in vitro transcription to generate the riboprobe. A commercially available transcription kit (MaxiScript II; Ambion, Austin, TX) was used with [-³²P]UTP (800 Ci/mmol). The RPA was performed with 20 µg total RNA and 100,000 cpm of gel-purified riboprobe using a commercial kit (RPA II, Ambion) according to the manufacturer's protocol. Hybridization of the riboprobe with total RNA resulted in a protected fragment of ~243 bp (see Fig. 3). The 18S rRNA was used as the internal standard. All autoradiographs were quantified using an image analysis system and software purchased commercially (Interactive Technologies International, St. Petersburg, FL). Linearity of signal intensity with increasing quantities of RNA was verified in preliminary tests of the RPA. Myostatin mRNA abundance (arbitrary units of band intensity) is expressed as a percentage of the 18S rRNA signal intensity.

Developmental expression of porcine myostatin mRNA in skeletal muscle. Two dams were euthanized at 21, 35, and 49 days of gestation (normal gestation, ~114 days), and three fetuses (n = 6) were frozen immediately in liquid nitrogen pending total RNA extraction. Additionally, longissimus muscle was collected from fetal pigs at 105 days of gestation (n = 6). Postnatally, longissimus samples were obtained from 12 pigs at each of the following time points: at ~1.5 kg body wt (i.e., within 12 h of birth), at 4 kg body wt (2 wk of age), and at ~23 (9 wk), 55 (15 wk), 107 (21 wk), and 162 kg (34 wk).

Tissue specificity of expression. Myostatin expression was evaluated in multiple fetal piglet tissues (longissimus muscle, brain, kidney, liver, spleen, heart, and lung) at 75 days of gestation and in tissues collected from a male castrate weighing 105 kg. The following tissues were collected from the male castrate: adipose tissue was from the middle layer of the subcutaneous depot over the cervical spinal process and also from the perirenal depot; brain (a composite sample of front, mid, and hindbrain), cardiac and skeletal (longissimus) muscle; and tongue, liver, lung, spleen, kidney, bone marrow, and small intestinal mucosa. Additionally, mammary gland and longissimus muscle samples were obtained from two lactating females 14-19 days after parturition.

Expression of myostatin mRNA in runt piglets and in pigs deprived of food. To determine whether myostatin expression was different in normal and low-birth-weight piglets (littermates), 10 normal (1.37 ± 0.077 kg) and 10 low-birth-weight (0.57 ± 0.032 kg) piglets (5 males and 5 females) were euthanized within 12 h of birth, and longissimus muscle samples were taken for RNA extraction. No attempt was made to verify colostrum intake before sample collection. Also, to determine whether myostatin expression is regulated by food deprivation, piglets at 4 (n = 6) and 7 (n = 5) wk of age were assigned to either ad libitum intake or food deprivation (2 ages × 2 intake levels) for a duration of 3 days. Those assigned to ad libitum intake were fed a commercial-type diet suitable for their stage of growth. All piglets were euthanized after the experimental period, and longissimus muscle samples were collected for RNA extraction.

Effect of growth hormone and polyunsaturated fatty acids on myostatin expression in longissimus and deep and superficial semitendinosus muscles. Myostatin expression was also evaluated in a study designed largely to evaluate the effects of recombinant porcine growth hormone and dietary polyunsaturated fatty acids (PUFA) on markers of insulin sensitivity and postreceptor signaling in adipose and skeletal muscle tissue. Forty-eight pigs weighing ~95 kg were assigned to four treatments (n = 12) arranged as a 2 × 2 factorial in which recombinant growth hormone (0 or 3.0 mg/day) and dietary PUFA (safflower oil, 0 or 10%) were the main effect variables. Commercial-type diets were formulated such that energy and amino acids concentrations were adequate to support the expected improvements in growth rate and muscle mass due to growth hormone. The treatment diets were fed for 16 days, with growth hormone administered (subcutaneously over the omotransversarius muscle) twice daily for the final 9 days of the study. Longissimus and semitendinosus muscle samples were collected immediately when the pigs were euthanized. The semitendinosus muscle was separated into the deep (red) and superficial (white) components.

Myostatin expression was also evaluated in longissimus samples from a longer term growth hormone study conducted to evaluate the impact of growth hormone on -actin and calpain expression (10). In this study, 48 male castrates (65.8 ± 1.6 kg) were assigned to four treatments (n = 12) arranged as a 2 × 2 factorial (± growth hormone, 3 or 6 wk duration) in a randomized complete block design. Twenty-four pigs received a daily injection (subcutaneously over the omotransversarius muscle) of recombinant porcine growth hormone (3 mg). The other 24 pigs received diluent only. At day 21 or 42 after injection, longissimus muscle samples were obtained by surgical biopsy. Total RNA was extracted and used for relative quantification of myostatin mRNA.

Statistical analyses. Data for which quantitative comparisons were made were analyzed using the general linear models procedures of SAS (22). To establish the ontogeny of myostatin expression, the data from whole fetus samples (days 21, 35, and 49 of gestation, n = 6) were analyzed independently of longissimus samples collected at 105 days of gestation (n = 6); within 12 h of birth (n = 12); postnatally at 2 wk of age; and at 23, 55, 107, and 162 kg body wt (n = 12). Within each analysis, means were separated using the least-significant differences (LSD) procedure only when protected by a significant F value (24). For the long-term growth hormone study, the three treatment degrees of freedom were partitioned to establish the main effects and their interaction by orthogonal contrasts. In the growth hormone-safflower oil study, myostatin expression in longissimus muscle and the deep red and superficial white components of semitendinosus muscle was evaluated using a split plot analysis in which pig was the whole plot and muscles were subplots. The effects of growth hormone and safflower (and the interaction) were tested using whole plot as the error term. The effect of muscle and the two-way (muscle by fat and muscle by growth hormone) and three-way (muscle by fat by growth hormone) interactions were tested using the residual error term. Means were separated via the protected LSD procedure.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Northern blot analysis and tissue specificity of expression. Northern blot analysis of poly(A⁺) mRNA prepared from adult longissimus and semitendinosus muscle revealed two distinct transcripts with calculated sizes of ~1.5 and 0.8 kb (Fig. 2). In the porcine fetus (75 days of gestation), nuclease protection assays indicated the presence of myostatin mRNA in skeletal muscle but not in any of the other tissues evaluated (Fig. 3). Furthermore, as shown in Fig. 4, in situ hybridization performed with fetal (day 59) hindlimb muscle indicated that myostatin mRNA is concentrated in the regions of developing muscle fasiculi rather than in the disperse connective tissue surrounding developing muscle bundles. Evaluation of multiple tissues collected from the growing pig substantiated the limited expression of the myostatin gene; myostatin mRNA was again readily detected in skeletal muscle but was not apparent in adipose tissue, brain, tongue, heart, lung, spleen, small intestinal mucosa, kidney, liver, or bone marrow (data not shown). However, myostatin mRNA was detected in lactating mammary gland (Fig. 3). Mammary tissue expression of myostatin was confirmed by RT-PCR using total RNA from lactating and nonlactating females (Fig. 1). The PCR product obtained was identical in size to that produced by RT-PCR of skeletal muscle RNA, and sequence analysis confirmed the identity of this PCR product to be the porcine myostatin cDNA [99% homology with a porcine myostatin sequence now available in GenBank, accession number AF019623 and reported by McPherron and Lee (16)]. Additionally, in situ hybridization (Fig. 4) performed with mammary tissue collected from a female ~48 h after the onset of lactation showed that signal is concentrated in the tubuloalveolar secretory lobules.

View larger version (123K): [in this window] [in a new window]   Fig. 2.   Northern blot analysis of poly(A+) mRNA from porcine superficial semitendinosus (STD; 6.2 µg) and longissimus (LD; 9.4 µg) muscle.

View larger version (56K): [in this window] [in a new window]   Fig. 3.   Tissue specificity of myostatin expression in porcine fetal tissues (75 days of gestation) and adult muscle and mammary gland during lactation as determined by ribonuclease protection assay. Total RNA (20 µg) from multiple tissues was hybridized with the radiolabeled myostatin riboprobe. Protected fragments (MST) were obtained only with fetal skeletal (longissimus) muscle (lane 8) and adult mammary gland (lanes 13 and 14). Other lanes are identified as follows: size markers (100, 200, 300 nt, 1), undigested myostatin riboprobe (2), undigested 18S riboprobe (3), digested myostatin and 18S rRNA riboprobes (4), brain (5), kidney (6), liver (7), spleen (9), heart (10), lung (11), adult longissimus muscle (12).

View larger version (117K): [in this window] [in a new window]   Fig. 4.   In situ hybridization of skeletal muscle and mammary gland. Dark (A and D)- and bright (B and E)-field micrographs of fetal muscle (left) and mammary gland (right) cross sections probed with antisense myostatin cRNA. Abundant silver grains indicate strong myostatin expression in both tissues, and negative (sense) controls confirm probe specificity (C and F). Note that in skeletal muscle mRNA is concentrated in the developing muscle fasiculi (dark regions of B) compared with connective tissue. In mammary gland, mRNA is concentrated in the tubuloalveolar secretory lobules (dark regions of E).

Developmental regulation of myostatin expression. Prenatally, myostatin mRNA abundance was low in whole fetus preparations at 21 and 35 days but was markedly increased (~2-fold, P < 0.05) by 49 days (Fig. 5A). As shown in Fig. 5B, longissimus myostatin mRNA abundance was highest at 105 days of gestation, declined significantly (P < 0.05) at birth, and continued to decrease to its lowest level of expression when piglets were 2 wk of age (P < 0.05). Longissimus myostatin expression increased (P < 0.05) with body weight from 2 wk of age (4 kg) to 55 kg and then plateaued thereafter. Expression remained relatively low throughoutpostnatal growth compared with prenatal (105 days of gestation) level of expression.

View larger version (12K): [in this window] [in a new window]   View larger version (13K): [in this window] [in a new window]   Fig. 5.   Graphic representation of myostatin mRNA expression (ribonuclease protection assay) normalized to 18S rRNA in porcine fetuses during early development (A) and in longissimus muscle at 105 days of development and postnatally (B). Bars represent least-squares means of 6 whole fetus preparations (A) and 6 longissimus preparations at 105 days of development and 12 longissimus samples from birth (B). SEs are shown, and bars without a common letter are different (P < 0.05).

Myostatin expression in normal and low-birth-weight piglets. In the present study, the birth weight of piglets showing the low-birth-weight syndrome was only 42% of that of the normal littermate controls (0.57 ± 0.032 vs. 1.37 ± 0.077 kg). The marked reduction in birth weight was associated with 65% more (P < 0.05) myostatin mRNA in longissimus muscle, irrespective of gender (Fig. 6). The higher expression of myostatin in these piglets was also evident in sections used for in situ hybridization (Fig. 7).

View larger version (11K): [in this window] [in a new window]   Fig. 6.   Myostatin mRNA expression, normalized to 18S rRNA, in longissimus muscle of newborn piglets with normal (Normal BW) and low birth weights (Low BW) determined by ribonuclease protection assay. Longissimus muscle was collected from the low-birth-weight piglets and a normal littermate control (same gender) within 12 h of birth. Myostatin mRNA abundance was higher (P < 0.05) in the muscle of low-birth-weight piglets. Bars represent least-squares treatment means (n = 10). Pooled SE is shown.

View larger version (123K): [in this window] [in a new window]   Fig. 7.   In situ hybridization of skeletal muscle collected from normal and low-birth-weight piglets at birth. Dark (A and D)- and bright (B and E)-field micrographs of longissimus muscle cross sections from low (left)- and normal (right)-birth-weight piglets probed with antisense myostatin cRNA. Consistent with results from the ribonuclease protection assay, hybridization with the radiolabeled antisense myostatin cRNA gave visibly greater signal intensity in muscle from the low-birth-weight piglet. Primary myofiber (arrows in B and E) number is generally greater in the deep semitendinosus muscle of low-birth-weight piglets compared with normal piglets; however, preferential localization of signal to primary or secondary myofibers was not apparent. Specificity was confirmed by hybridization of control sections with a negative (sense) control riboprobe (C and F).

Myostatin expression in relationship to food intake, growth hormone, and dietary safflower oil. Myostatin mRNA abundance was not altered (P > 0.148) by 3-day food deprivation in either 4- or 7-wk-old piglets (Fig. 8). As shown in Fig. 9, myostatin mRNA abundance was not altered by growth hormone (P > 0.504) at 3 or 6 wk and there was no effect of duration of treatment (P > 0.567). Likewise, in the shorter term growth hormone-safflower oil study (Fig. 10), both variables were without effect (P > 0.536) and the interaction was insignificant (P > 0.334). However, the effect of muscle was significant; the deep semitendinosus had lower (P < 0.05) myostatin mRNA abundance than either the longissimus or superficial semitendinosus.

View larger version (13K): [in this window] [in a new window]   Fig. 8.   Myostatin expression (ribonuclease protection assay), normalized to 18S rRNA, in longissimus muscle of pigs at 4 or 7 wk of age having free access to food (AL) or deprived (FD) for 3 days. Neither food deprivation nor piglet age influenced (P > 0.148) myostatin mRNA abundance, and there was no interaction (P > 0.632). Bars represent least-squares treatment means (n = 5 or 6) with the SE shown.

View larger version (12K): [in this window] [in a new window]   Fig. 9.   Myostatin expression (ribonuclease protection assay), normalized to 18S rRNA, in longissimus muscle of growing pigs receiving daily injections of porcine growth hormone [0 (GH) or 3 (+GH) mg] for 3 or 6 wk (least-squares means with SE, n = 10-12). Myostatin expression was not influenced by either growth hormone or duration (P > 0.504), and the interaction was not significant (P > 0.72). Although the abundance of 18S rRNA tended to be higher (12%, P < 0.114) in pigs receiving growth hormone, normalization of the myostatin signal did not change the results.

View larger version (12K): [in this window] [in a new window]   Fig. 10.   Myostatin expression (ribonuclease protection assay) in skeletal muscles of growing pigs injected twice daily for 9 days with vehicle or porcine growth hormone and fed a control diet or the control diet with 10% safflower oil (n = 12). Myostatin expression (normalized to the 18S rRNA signal) was not influenced by growth hormone or safflower oil (P > 0.536), and there was no interaction (P > 0.334). Therefore, the data are presented with respect to muscle only. Bars represent least-squares means with SE shown. Effect of muscle was significant, with longissimus and superficial semitendinosus (SST) having higher (P < 0.05) myostatin mRNA abundance than the deep semitendinosus (DST). 18S rRNA abundance was only slightly greater (10%, P < 0.05) in the DST. Bars without a common letter are signficantly different.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Previous work with mice (15) and cattle (11) indicated a single myostatin transcription product of ~2.9 kb and that myostatin was strongly expressed in skeletal muscle and expressed at low levels in adipose tissue (15). Albeit, with poly(A⁺) mRNA prepared from adult porcine longissimus and semitendinosus muscle, we detected two transcripts of ~1.5 and 0.8 kb. It should be noted that the smaller transcript is not of sufficient size to contain the open reading frame indicated by the available porcine sequence noted above. Characterization of these transcription products will clarify whether they arise via alternative splicing.

Whereas myostatin mRNA was not detected in adipose tissue, expression is certainly not limited to skeletal muscle. Our finding of expression in the tubuloalveolar secretory lobules of lactating mammary gland is intriguing and indicates the possibility that myostatin performs a regulatory role pertaining to gestational or lactational mammary gland growth and development and/or metabolism. It is also an important consideration that milk from numerous species contains a myriad of growth factors, including TGF- and -₂ (7). Accordingly, it seems possible that myostatin is secreted from mammary gland into the milk and serves a regulatory role in the neonatal pig.

In skeletal muscle, in situ hybridization indicated that myostatin mRNA is associated with the developing muscle fasiculi rather than the ancillary cells of the diffuse connective tissue. This finding is consistent with the purported role of myostatin as a negative effector of myogenic events and lends support to the hypothesis that the mechanism of this regulatory factor encompasses an autocrine-paracrine regulatory loop. Whether myostatin regulates myoblast proliferation, differentiation, or fusion events remains to be established.

The developmental pattern of expression obtained in the pig is similar to that reported for cattle (11) in that mRNA was detected at all prenatal and postnatal stages evaluated. Furthermore, as in cattle, there was a sharp increase in mRNA early in development that was sustained until late in gestation. We also documented in porcine longissimus muscle a sharp reduction in myostatin mRNA at birth versus 105 days gestation and a further reduction from birth to 2 wk of age. The developmental pattern of myostatin mRNA abundance (increase in myostatin mRNA in whole fetus preparations at 49 days coupled with the high level of expression in skeletal muscle at 105 days and the reduction at birth) coincides roughly with the progression of primary and secondary muscle fiber formation. Studies of fetal myogenesis in the pig (2) have shown the presence of primary fibers by 45 days of development and that secondary fiber formation occurs between 45 and 75 days. By 90 days, primary and secondary fibers are similar in size, and myofiber number is largely fixed by birth. Thus the reduction in myostatin mRNA abundance at birth and postnatally (vs. 105 days of gestation) may reflect the reduction in myoblast mitogenic activity, differentiation, and fusion events. The biological significance of the increase in myostatin mRNA from 2 wk postnatally (~4 kg body wt) to 55 kg body wt and the plateau thereafter is not clear. Although the rate of muscle growth would be expected to have peaked and be declining well before 162 kg body wt, total muscle mass would be expected to increase far beyond that at 55 kg body wt, the point after which there was no significant increase in myostatin mRNA abundance in this study. The relationship between myostatin and muscle growth must be examined in light of growth rate and total muscle mass to gain a clear understanding of how expression of this growth factor relates to muscle growth.

Mutations or deletions in the myostatin gene, associated hypothetically with lower myostatin bioactivity, have been related to muscular hypertrophy in mice and cattle. Furthermore, the runt piglet data presented provide evidence of a pathological condition accompanied by overexpression of myostatin. Although we did not determine skeletal muscle mass in these piglets, skeletal muscles of the low-birth-weight piglets are smaller than those of normal piglets (26) and runt piglets have fewer myofibers (9, 18). Additionally, Aberle (1) determined that in the red (deep) component of the semitendinosus muscle, runt piglets have lower proportions of type I myofibers, with a higher percentage of those staining as type I fibers appearing to be primary fibers. There is also a lower ratio of secondary to primary fibers. However, these developmental characteristics of the deep semitendinosus muscle are not present in longissimus muscle, perhaps due to its earlier completion of myofiber hyperplasia (1). It may indeed be informative to compare the developmental pattern of myostatin expression in runt and normal piglets to ascertain whether peak myostatin expression is higher and/or sustained longer than in normal piglets. It is conceivable that evaluations of myostatin expression in relationship to intrauterine growth retardation, myofibrillar hypoplasia, and low-birth-weight syndromes may reveal specific therapeutic targets for alleviating these conditions.

Given that low-birth-weight piglets had higher myostatin mRNA abundance than the controls and the heavily muscled phenotype of the myostatin knockout mouse was achieved in part by hypertrophy (15), we thought it possible that myostatin expression would be increased by a metabolic scenario in which muscle growth was suppressed by food deprivation and perhaps decreased in association with the greater muscling caused by exogenous growth hormone. However, myostatin mRNA abundance was not changed by food deprivation. Administration of exogenous growth hormone increases muscle mass in growing animals by as much as 36% (3). The consensus opinion is that this hypertrophy is supported in part by satellite cell proliferation and the incorporation of these nuclei into existing myofibers. Therefore, we hypothesized that myostatin expression would be reduced in conjunction with the muscular hypertrophy achieved by growth hormone. We also thought that dietary PUFA might dampen the efficacy of growth hormone and have direct effects on myostatin via the purported effects of PUFA on insulin sensitivity and glucose transport proteins (5, 17, 23). Growing animals treated with growth hormone show faster growth rates and reductions in feed intake and serum urea nitrogen. These indexes of growth hormone efficacy were evident in both the initial study (10) and the second study (data not shown). Nonetheless, myostatin expression was not altered by growth hormone in either study and safflower oil was without effect in the second study. Collectively, these data suggest that, in muscle, the physiological role of myostatin is largely associated with the prenatal period of muscle growth in which myoblasts are proliferating, differentiating, and fusing to form multinucleated myofibers. However, it is of interest that the abundance of myostatin mRNA in longissimus and superficial semitendinosus was over twice that of the deep semitendinosus muscle in growing pigs. When the difference in fiber type in the white longissimus and superficial semitendinosus muscles versus the red (deep) semitendinosus and the greater potential for white muscle to undergo hypertrophy in response to -adrenergic stimulation (12, 20, 25) and exogenous growth hormone (13, 19) are considered, it is tempting to speculate as to the physiological significance of the difference in myostatin mRNA. However, given that growth hormone did not influence myostatin expression in either red or white muscle, the biological explanation for higher myostatin expression in white muscle is not apparent.

In summary, we have confirmed that myostatin mRNA is localized to developing muscle bundles. We have shown for the first time that this gene is expressed by the tubuloalveolar secretory cells of the mammary gland and also related greater expression of this gene to a physiological condition in which muscle growth is retarded. Finally, we have determined that neither enhanced nor reduced postnatal muscle growth, achieved by growth hormone and food deprivation, respectively, are associated with detectable changes in myostatin expression.

Perspectives