INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

IT IS GENERALLY ASSUMED that fetal growth in most mammalian species, including rats, humans, and pigs, is independent of growth hormone (GH) (2, 8, 11, 19). In contrast, it is well recognized that postnatal growth and tissue development become progressively more GH dependent as an animal matures (18) and that a primary mode of GH action is via the hepatic GH insulin-like growth factor-I (IGF-I) somatotrophic axis (34). This dependence of postnatal growth on GH and IGF-I has major clinical (12) and agricultural (33) significance.

The role of GH in the growth of neonates is unclear. In pigs there is only a low level of GH-specific binding in the liver of newborns (4), and either injections or implants of typical doses of exogenous GH in neonatal pigs have not induced growth responses (1, 5, 27) or improvements in protein utilization (21). However, the expression of GH receptor (GHR) mRNA is relatively high in skeletal muscle of fetal and newborn pigs (15, 25), and GH binding in fetal and neonatal pig muscle and plasma has been reported (35). Furthermore, Gerrard et al. (16) have found that muscle IGF-I mRNA expression was higher at 109 days of gestation and 21 days postnatally than at earlier or later stages of life. This suggests that the somatotrophic axis may be functional in neonates and that skeletal muscle may be more responsive to GH than the liver, which is normally considered to be the primary site of IGF expression and secretion.

Recently, Matteri et al. (27) found that a relatively large dose (0.5 mg/day) of porcine GH (pGH) in 3-day-old pigs increased hepatic IGF-I mRNA and serum IGF-I at 4 and 6 wk of age. Moreover, we (40) have reported that a large dose (1 mg/day) of exogenous pGH in newborn pigs caused a threefold increase in plasma IGF-I and significantly increased weight gain and feed efficiency at 7 days of age. Thus recent experiments have indicated that the hepatic somatotrophic axis is indeed functional or at least inducible in neonatal pigs. However, the extent to which the somatotrophic axis in skeletal muscle can be induced by exogenous GH in neonatal pigs is unknown.

The goals of our research were to 1) establish whether GH induces local somatotrophic function in neonates and 2) identify whether this response is similar in liver and skeletal muscle. To address these questions, we quantified expression of GHR and IGF-I mRNA in the liver and skeletal muscle of neonatal pigs given exogenous GH or saline for 7 days. In contrast to IGF-I, circulating and tissue concentrations of IGF-II are not considered to be directly regulated by GH and thus are usually unresponsive to exogenous GH. We therefore also measured IGF-II mRNA in liver and muscle to assess whether the effect of GH treatment was specific for IGF-I. The present investigation is an extension of our previously published report, which described the metabolic and tissue growth responses to GH treatment (40). In these studies, newborn pigs were used because they are an excellent model for the human infant, especially in relation to endocrine and metabolic function (37).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animals and experimental design. Animal protocols were reviewed and approved by the Animal Protocol Review Committee of Baylor College of Medicine. All animal housing and care complied with United States Department of Agriculture policies and practices. The experimental design has been described in detail previously (40). Briefly, cross-bred pigs from two litters were obtained at birth. They were assigned to one of two treatments: 1) recombinant pGH (Monsanto; St. Louis, MO) at a daily dose of 1 mg/kg body wt (n = 6) or 2) saline (n = 5). The treatments were given as three equal intramuscular injections daily (at 8-h intervals) and were continued for 7 days. Pigs were gavage fed six times daily. During the first 24 h they were given colostrum (195 ml/kg body wt) and then commercial sow-milk replacer (240 ml/kg body wt) for the remainder of the experiment.

Before the start of the experiment, a 20-gauge, Silastic catheter was surgically inserted into the jugular vein of each pig under general isoflurane anesthesia. Blood was sampled from jugular catheters daily for analysis of plasma IGF-I concentrations by radioimmunoassay (40). On day 7, blood was sampled every 20 min for 4 h to determine the plasma GH secretory profile using double-antibody homologous radioimmunoassay (40). At the end of the experiment, pigs were anesthetized with an intravenous dose of pentobarbital (50 mg/kg body wt) and exsanguinated. Samples of liver and longissimus muscle were taken within 6 min of exsanguination, frozen in liquid nitrogen, and stored at 70°C.

RNA isolation and measurement. Total RNA was isolated from 0.5- to 1.0-g portions of frozen tissue by the guanidinium thiocyanate method (6). The amount of RNA isolated was determined by absorbance at 260 nm (1 optical density unit, 40 µg/ml).

Construction of riboprobes. Construction of the GHR probe has been described in detail previously (38). To generate an antisense RNA probe, the template was transcribed with T7 RNA polymerase in the presence of [-³²P]UTP (Amersham International) as described by Saunders et al. (32). The transcript produced from this template is 200 bases, and 140 bases are protected by GHR mRNA in RNase protection assays.

The construction of the IGF-I riboprobe also has been described previously (38). The class 1 probe was used in these experiments. The transcript produced from this template is 270 bases in length and gives rise to the following products in RNase protection assays: 1) a full-length protected band of 200 bases arising from hybridization of mRNA containing exons 1 and 3, 2) an intermediate protected band of 170 bases, which also originates from exons 1 and 3 (39), and 3) a protected band of 147 bases formed by hybridization to exon 3 but not exon 1 (this is assumed to represent the class 2 transcript, which contains exons 2 and 3).

A porcine IGF-II probe specific to the first coding exon was generated based on sequence information from ovine IGF-II (26). Two oligo primers in the exon 8 coding region were designed according to the intronic sequence. The 5' primer was 5'-GCATAT[AAGCTT]CAATGGGGATCACAGCAGGAA-3' and the 3' complementary primer was 5'-ACTCAG[GAATTC]GGTCCCCACAGACAAACTGGA-3'. These primers had Hind III and EcoR I recognition sites, respectively (shown in brackets). A 162-bp fragment was generated by PCR from the two oligo primers. This fragment was then cloned unidirectionally into the Hind III-EcoR I site of Bluescript KS. This recombinant plasmid was used for the generation of the porcine IGF-II specific riboprobe using T7 RNA polymerase. The probe resulted in a 131-base protection product in RNase protection assays.

RNase protection assays. The RNase protection assays were carried out in duplicate on 50 µg of total RNA using methods similar to those described previously (13, 15, 38). In brief, samples were hybridized with a small molar excess of the radiolabeled riboprobe, as determined by titration analysis, to ensure linearity of the assay with respect to RNA. After 16 h of hybridization at 45°C, excess, nonprotected RNA was digested with RNase A (50 µg/ml, ~1 U/sample; Sigma) and RNase T1 (300 U/ml, ~80 U/sample; Sigma). The protected hybridization products were purified by extraction in phenol-chloroform-isoamyl alcohol (25:24:1). Protected fragments were separated on 6% polyacrylamide sequencing gels. For a given tissue (liver or longissimus muscle), samples from all pigs within a litter were separated on the same gel. Dried gels were exposed to X-ray film at 70°C, and the optical densities of protected bands were quantified using image analysis (Seescan; Cambridge, UK). Exposure times were adjusted so that band intensities lay within the linear range of detection of the analyzer.

Major problems associated with the use of inappropriate RNA controls have been highlighted recently (22), and our considerable experience in this field has led us to use a rigorously optimized RNase protection assay (41), combined with a large number of points (n = 5 to 6 animals/group) and duplicate or triplicate analyses of all samples. Any small errors due to differences between samples will be included in the total experimental variation as indicated by our standard errors.

Statistical analysis. Optical densities of the protected bands for GHR, IGF-I, and IGF-II mRNA were analyzed using Statistical Analysis Systems (31). Values for total IGF-I were calculated by adding the values for class 1 and class 2 transcripts before statistical analysis. The model included treatment (control or GH treated) and block (litter/gel). This model corrected for differences among individual assays but did not permit an analysis of litter effects per se. Liver and muscle were not always analyzed on the same gel, and therefore a statistical analysis of differences between tissues was not possible. Because of the unbalanced nature of the data set, least squares means are presented.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

As reported previously (40), injection of GH resulted in large increases in plasma concentrations of GH and IGF-I. During the first hour after GH injection, the plasma GH concentration exceeded 450 ng/ml, compared with <20 ng/ml in control pigs. Concentrations in treated pigs fell steadily until the next GH injection but were greater (P < 0.01) than those of control pigs at all time points. Plasma IGF-I concentrations increased from 37 ng/ml on day 0 to 175 ng/ml on day 6 in pigs treated with GH but remained essentially unchanged in control pigs. Differences in IGF-I between treated and control pigs were significant (P < 0.05) on all days (days 1-7) of treatment.

Representative autoradiographs showing protected bands for all three mRNA probes are illustrated in Fig. 1. The autoradiographs show the results for one litter of pigs, in which two animals served as controls and two were treated with GH. For IGF-I, there are three discrete bands. The first two (200 and 170 bases) correspond to the class 1 transcript and the third (147 bases) to the class 2 transcript. For GHR and IGF-II the protected probes yielded doublet bands at ~140 and ~131 bases, respectively.

View larger version (82K): [in this window] [in a new window]   Fig. 1.   Representative autoradiographs showing effects of growth hormone (GH) administration on GH receptor (GHR), insulin-like growth factors I and II (IGF-I and IGF-II) mRNA in liver and longissimus muscle. Values are for same litter of 4 pigs (2 controls and 2 treated with GH). Duplicate samples are shown in lanes 1 and 2, 3 and 4, etc. There was loss of sample in lane 5 for GHR. Molecular weight markers with number of bases are shown at right.

In liver, treatment with GH increased GHR mRNA by 63% (P < 0.01) and IGF-I mRNA by more than ninefold (P < 0.001; Fig. 2). Furthermore, there was a greater than eightfold increase in both class 1 (P < 0.001) and class 2 (P < 0.01) IGF-I transcripts, with a shift in the class 1 to class 2 proportions from 93:7 to 85:15 (P < 0.001; Fig. 3). The concentration of IGF-II mRNA was unchanged by GH treatment (P = 0.29; Fig. 2).

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Effects of GH administration on GHR (A), IGF-I (B), and IGF-II (C) mRNA in liver. Values are arbitrary density units per 50 µg of total RNA measured from autoradiographs (least squares means ± SE; ** P < 0.01; *** P < 0.001).

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Effects of GH administration on IGF-I mRNA transcripts in liver (A) and longissimus muscle (B). 1, class 1 transcripts derived from exons 1 and 3. 2, class 2 transcripts derived from exons 2 and 3. Values are arbitrary density units per 50 µg of total RNA measured from autoradiographs (least squares means ± SE; * P < 0.05; ** P < 0.01; *** P < 0.001).

In longissimus muscle, GHR mRNA was unaffected (P = 0.31) by GH injection (Fig. 4). However, there was a 72% increase in IGF-I mRNA (P < 0.02) and increases in both class 1 (P < 0.02) and class 2 (P < 0.07) IGF-I transcripts (Fig. 3). In contrast to liver, however, there was no shift in the proportions of class 1 to class 2 transcripts (99:1 for control and 98:2 for GH-treated pigs). Treatment with GH did not affect the concentration of IGF-II mRNA in muscle (P = 0.63; Fig. 4).

View larger version (13K): [in this window] [in a new window]   Fig. 4.   Effects of GH administration on GHR (A), IGF-I (B), and IGF-II (C) mRNA in longissimus muscle. Values are arbitrary density units per 50 µg of total RNA measured from autoradiographs (least squares means ± SE; * P < 0.05).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Our results demonstrate that the somatotrophic axis is functional or can be induced in both the liver and skeletal muscle of neonatal pigs. The changes in hepatic GHR and IGF-I mRNA in response to exogenous GH are consistent with those observed by others in older pigs, whereas the increase in IGF-I mRNA in neonatal muscle in response to GH is a novel finding. Even though both liver and muscle IGF-I mRNA responded to GH, the increase was much greater in liver than muscle.

Previous experiments with pigs weighing 7 and 60 kg (1, 7) have established that treatment with GH increases binding of GH in liver. Furthermore, Combes et al. (10) recently reported that administration of pGH to 75-kg pigs increased both GH binding and GHR mRNA in liver. Brameld et al. (3) also reported increased hepatic GHR mRNA in 42-kg pigs in response to GH treatment. Numerous studies in growing and adult pigs (1, 9, 20, 23, 28, 36) and now studies with very young pigs (21, 40) have shown that administration of exogenous GH increases circulating IGF-I. Most of the circulating IGF-I has been assumed to be derived from the liver because GH injection increases hepatic IGF-I mRNA abundance in both growing (3, 9, 20) and neonatal (27) pigs, and the concentration of circulating IGF-I is more closely related to hepatic IGF-I mRNA than to muscle IGF-I mRNA (38). Our results are the first to demonstrate that all of these effects (increased hepatic GHR and IGF-I mRNA and increased circulating IGF-I) can be induced in newborn pigs, albeit with a much larger dose (on a body wt basis) of pGH than has been used in studies with more mature pigs.

The responsiveness of skeletal muscle to GH is, however, less clearly established, especially in the neonate. Previous studies have indicated both no change (10) and an increase (3) in muscle GHR mRNA of growing pigs after treatment with GH. We observed no change in neonatal pigs. On the other hand, we observed a significant increase in muscle IGF-I mRNA, whereas previous studies with pigs (3, 9, 20) have observed either no change or a decrease after treatment with GH. Our observation that there was an increase in muscle IGF-I mRNA with no significant increase in GHR mRNA implies that there may have been effects on posttranscriptional regulation of GHR or on intracellular signaling. These possibilities should be the subject of future investigation.

A possible explanation for the difference between our findings and those of others is that we used a much higher GH dose than most previous studies using older pigs, although other studies with neonatal pigs also have used relatively high doses (i.e., 0.5 mg/day). We deliberately used a high dose of pGH for two reasons. First, the existing evidence (1, 4, 21) suggested that neonates were less sensitive to GH than older pigs, and thus neonates might require a higher dose. Second, our objective was to establish either the presence or absence of somatotrophic function, thus we used a high dose of pGH to maximize the possibility that we would detect a response, if indeed one existed.

We expected to find a somatotrophic response in muscle based on the relatively high expression of both GHR and IGF-I mRNA reported in neonates. In studies of the ontogeny of the porcine GH-IGF system, several authors (15, 25, 30) have reported that during the perinatal period there was much greater expression of GHR mRNA in muscle than in liver. In addition, recent reports (16, 30) have demonstrated that muscle IGF-I mRNA expression is greatest during late gestation and early postnatal life and that the relative expression is higher in muscle than in liver (30). Findings in our control animals confirm this higher level of IGF-I mRNA in muscle compared with liver. This suggests that muscle may be a significant source of circulating IGF-I under basal physiological conditions in the neonate. Thus one could postulate that somatotrophic function may be more responsive or sensitive in skeletal muscle than in liver during the fetal-neonatal period. However, our evidence indicates that despite differences in basal gene expression in the neonate local somatotrophic function is more responsive to exogenous GH in liver than muscle.

We have described previously the existence of two distinct IGF-I mRNA transcript classes (1 and 2) in pigs (39). Differences in energy status of young pigs that resulted in differences in growth rate and circulating plasma IGF-I concentrations were associated with differences in the relative proportions of the two transcript classes in liver. Specifically, the class 1 to class 2 proportions were 95:5 when plasma IGF-I was 90 ng/ml and 81:19 when plasma IGF-I was 200 ng/ml. Similar effects were observed in a subsequent study (38) in which the hepatic IGF-I mRNA class 1 to class 2 proportions decreased from 100:0 to 77:23 as plasma IGF-I increased. In both studies there was no such relationship between transcripts in muscle. In our experiment, in which circulating IGF-I concentrations were elevated by GH injection, similar results were obtained. In liver, the class 1 to class 2 proportions decreased from 93:7 in control pigs to 85:15 in pigs treated with GH. There was no difference in the proportions in muscle. Differential regulation of hepatic IGF-I transcripts in sheep in response to changed GH and nutritional status also has been reported (29). The significance of these changes in terms of the regulation of growth by the somatotrophic axis has been discussed previously (17, 38). For example, exon 2-derived hepatic IGF-I may have specific properties with respect to its sensitivity to transcription factors or its efficiency of export into blood for action at peripheral tissues.

The relationship between GH and IGF-II is not well understood, especially in neonates. Experiments with growing pigs have shown that exogenous GH may increase, have no effect, or decrease plasma IGF-II concentrations (23, 24, 28, 36). Tissue levels of IGF-II mRNA decline markedly between late gestation and 180 days postnatally (30). This concurs with the major role of IGF-II in skeletal muscle differentiation and the formation of secondary myofibers, which is completed by 90-95 days of gestation in the pig (birth, 115 days of gestation). The well-established finding that primary cultures of skeletal muscle satellite cells, the postnatal myogenic precursor cells, can be induced to proliferate by exposure to physiological levels of IGF-II (14) indicates an important role for IGF-II in the control of postnatal muscle growth. However, our findings clearly indicate that GH does not play a significant role in regulating this response in the young pig. Our finding that GH injection did not affect IGF-II mRNA in liver or muscle is consistent with the recent report that pGH does not affect liver IGF-II mRNA of 4- or 6-wk-old pigs (27).

In summary, it is clear from our results that injection of GH resulted in a specific increase in hepatic GHR and IGF-I gene expression in newborn pigs that was similar to the effects in older pigs. This indicates that the somatotrophic axis in liver is functional, or can be induced, in neonatal pigs and probably in other species, including humans, which are born at a relatively advanced stage of development. The increases in muscle IGF-I gene expression suggest that the somatotrophic axis is also functional in muscle during the perinatal period, but as with older pigs, this response to GH is much less than that observed in the liver. In addition, IGF-II expression in both the liver and skeletal muscle was unresponsive even to a relatively large dose of GH, further suggesting that the somatotrophic response is specific for IGF-I.

Perspectives