Main

During the 3rd wk of postnatal life, dramatic maturational changes occur in small intestinal structure and enzyme activities in the infant rat, enabling the transition from milk to solid food to occur at weaning(1, 2). At this time, the number and turnover of intestinal mucosal epithelial cells are increased due to acceleration of rates of DNA and protein synthesis, cellular proliferation, and cellular migration, resulting in accelerated growth of the small intestine(3). In addition, profound decreases in the catalytic activities of lactase-phlorizin hydrolase (lactase) and the abrupt appearance and increases in sucrase-α-dextrinase (sucrase) and AOP occur at this time(1, 4). These critically important maturational changes in small intestinal structure and function have been shown to require the participation of pituitary hormones for their normal expression(1, 5–8).

Although the roles of corticosteroids and thyroid hormones as regulators of intestinal maturation have been extensively investigated(6–9), the role of GH is far less clear. The few previous studies using rats hypophysectomized in early infancy have produced mixed results in determining whether GH is required for normal postnatal intestinal growth and maturation(5, 8, 10). It is also unclear whether GH acts directly or indirectly on intestinal tissues. Receptors for GH have been found in the gastrointestinal tract of the rat by multiple investigators(11, 12). It is known that many of the growth-promoting effects of GH are mediated by IGF-I, which is a potent polypeptide mitogen and is considered to be the major regulator of somatic growth(13, 14). IGF-I receptors have recently been demonstrated in the intestinal epithelium of young animals(15, 16), and mRNA for IGF-I has been found in the small intestines of both fetal and adult rats(17, 18). It has also been demonstrated that intestinal IGF-I mRNA synthesis in hypophysectomized adult rats increases in response to exogenous GH administration(19). These findings suggest a role for GH and IGF in regulation of small intestinal growth and maturation.

To investigate the role of GH in regulation of postnatal intestinal maturation, we studied a unique strain of Sprague-Dawley rat, the SDR. The dwarfism results from an autosomal recessive point mutation in the GH gene that causes abnormal splicing and a one-base deletion in the GH mRNA, resulting in markedly reduced levels of GH message. As the small amounts of GH mRNA are nontranslatable, there is complete absence of GH protein(20). No GH is detectable in the pituitary of SDR rats when analyzed by PAGE, immunoblot analysis, RIA, or immunocytochemistry. However, the deficiency of GH is isolated, as no effect on prolactin, ACTH, TSH, or LH production has been detected(21). This model offers the opportunity for specific investigation of GH and GH-dependent processes without methodologic complications introduced by use of hypophysectomy or administration of exogenous GH. Using this model, we have examined the role of GH in regulation of intestinal growth and maturation in suckling and postweaned rats.

METHODS

Chemicals and reagents. Sucrose, lactose, glucose oxidase (type X), horseradish peroxidase (type VI), p-hydroxybenzoic acid, aminoantipyrine, leucyl-β-napthylamide, phenylmethyl-sulfonyl flouride, EDTA, and p-hydroxymercuribenzoate were purchased from Sigma Chemical Co. (St. Louis, MO). Hoechst H-33258 was obtained from Calbiochem(San Diego, CA). Bio-Rad reagent for protein assay was purchased from Bio-Rad(Richmond, CA).

Animals. Control animals were the offspring of pregnant dams of the Sprague-Dawley strain purchased from Simonsen Laboratories (Gilroy, CA). The strain of SDR was generously provided by Dr. S. Okuma (Morishita Pharmaceutical Co., Ltd., Shiga, Japan). These rats carry an autosomal recessive mutation in the GH gene, represented by gene symbol dr. Their expression of pituitary GH mRNA is only 3-6% that of control animals, due to an aberrant splice site in the mRNA. Furthermore, this truncated mRNA is not translated into the mature protein due to premature translational termination, such that no GH peptide can be measured in the anterior pituitary gland of SDR(22, 23). Unlike other GH-deficient dwarf rat strains available, these rats are unique, because they lack only GH and are otherwise pituitary-sufficient. The phenotypes of the SDR animals are distinctive with homozygously affected adult males or females (dr/dr) being significantly smaller than unaffected littermates(24). For heterozygotes (DR/dr) or normal control animals (DR/DR), neither somatic growth nor pituitary GH levels differed(24). Thus, DR/dr and DR/DR animals were indistinguishable and subsequently used interchangeably.

Female GH-deficient (dr/dr) rats were mated with GH-deficient (dr/dr) males and female GH-normal (DR/dr and DR/DR) were mated with GH-normal (DR/dr and DR/DR) males. The day of delivery was considered to be postnatal d 1. To control for nutrient intake, each litter was reduced to 10 pups/litter. During the 3rd wk of life, the pups weaned normally from their mothers and began to eat rat chow. Intake of chow was measured in a subset of animals at d 22-23, and no significant differences in grams of intake/body weight were found(0.082 ± 0.002 versus 0.079 ± 0.003). dr/dr animals were easily distinguishable from DR/dr or DR/DR animals by body size. For simplicity, dr/dr animals will be designated simply by dr, and DR/dr as well as DR/DR animals will be designated by DR.

All animals were housed in the facilities of the Department of Laboratory Animal Medicine of the Stanford University Medical Center. The animals were maintained at 21.0 ± 1.0 °C and had a 12-h light, 12-h dark cycle, with lights on at 0700 h. Standard food and water were available ad libitum. The protocols for animal use were approved by the Department of Laboratory Animal Medicine, Stanford University.

Analysis of intestinal tissues. Animals were killed by decapitation on d 7, 14, and 23 of life in the morning between 0900 and 1100 h. The small intestine from the ligament of Treitz to the ileocecal valve was quickly removed and flushed with cold isotonic saline. Intestinal length was measured with application of gentle traction. The intestine was then weighed, divided into equal halves, and stored at -80 °C for biochemical analysis. Before freezing, small sections were placed in formalin for morphometric analysis by light microscopy. At the time of biochemical analysis, the proximal half of the small intestine was homogenized in 10 volumes of homogenization buffer (10 mM sodium phosphate and 5 mM Tris-HCl, pH 6.0) containing the protease inhibitors 1 mM phenylmethylsulfonyl fluoride and 0.1 mM p-hydroxymercuribenzoate. Activities of lactase-phlorizin hydrolase, and sucrase-α-dextrinase were determined by the method of Dahlqvist(25), as modified by Tsuboi et al.(26). The activity of AOP, the major peptidase of the intestinal brush border membrane, was measured using 0.17 mM leucyl-β-naphthylamide as substrate, as previously described(27). Protein was determined by the Bio-Rad assay, using BSA as the standard(28). DNA was determined by the method of Labarca and Paigen(29). Enzyme activity units(U) were defined as micromoles of substrate hydrolyzed per min at 37 °C, and specific activities were expressed as milliunits of enzyme activity per mg of protein.

Measurement of intestinal morphometry. Intestinal sections harvested at the time of sacrifice were fixed in 10% buffered formalin, processed for routine histologic sections, and paraffin embedded. The segments were predominantly embedded perpendicular to the plane of section unless large enough to permit opening so that they could be placed parallel. Sections 5μm in thickness were cut and stained with hematoxylin and eosin. Only one level from each specimen was measured. Crypt and villus measurements were taken with an Olympus microscope equipped with a micrometer. Crypts were defined for purposes of measurement as the glandlike invagination portion of the lamina propria from its base at the muscularis to the surface, or the area where mitotic activity ceases and the brush border becomes more conspicuous. The villi were defined for purposes of measurement as the finger-like evaginations of the lamina propria from the base at the muscularis to the contiguously attached tip(30).

Statistics. All data were analyzed using InStat, a statistical software package provided by GraphPad Software (San Diego, CA). All results are presented as the mean ± SE. Groups were compared using an unpairedt test, with p < 0.05 considered significant.

RESULTS

Somatic growth of GH-deficient and GH-sufficient animals. As indicated in Table 1, postweaned GH-deficient animals are significantly smaller than their GH-sufficient litter mates. Total body weights of the dr animals were 37.6% less than their age-matched GH-sufficient counterparts at d 7 of life, 42.6% less at d 14 of life and 43.9% less at d 23 of life. The increases in weight in GH-deficient animals indicate that, in spite of GH deficiency, they continued to exhibit somatic growth through the period of weaning. Apart from differences in growth, no other phenotypic or behavioral differences were noted between the groups of animals.

Table 1 Somatic growth of GH-deficient and GH-sufficient animals
Full size table

Intestinal growth in GH deficiency. At death, the small intestine from the ligament of Treitz to the cecum was removed, measured, and weighed. The small intestine in GH-deficient (dr) animals was 28.3 ± 0.6 cm in length at d 7, 32.5 ± 0.3 cm in length at d 14, and 52.5± 1.3 cm in length at d 23, whereas in GH-sufficient (DR) animals, small intestinal length was 40.0 ± 0.7 cm at d 7, 45.4 ± 0.7 cm in length at d 14, and 71.9 ± 2.7 cm in length at d 23. At each age, small intestinal length is significantly shorter (p < 0.05) in dr compared with DR. The weight of the small intestine in GH-deficient rats was 0.28 ± 0.03 g at d 7, 0.51±.01 g at d 14, and 1.87 ± 0.08 g at d 23, whereas in GH-sufficient rats, intestinal weight was 0.75 ± 0.03 g at d 7, 1.05 ± 0.03 g at d 14, and 4.51 ± 0.36 g at d 23. Similar to intestinal length, at each age, the small intestine is significantly lighter (p < 0.05) in dr compared with DR. Hence at each age measured, the intestine was significantly shorter and lighter in the GH-deficient animals, but showed evidence of progressive growth through the periods of suckling and weaning.

Table 1 gives the major intestinal parameters measured in this study. When expressed per unit of intestinal length, the weight, DNA content, and protein content of the small intestine is significantly greater in DR at all three ages studied. Additionally, all three parameters are significantly greater in both dr and DR groups at d 14 when compared with d 7, and at d 23 when compared with d 14. A similar pattern of differences is seen in measurement of villus height and crypt depth (Fig. 1). Villus height and crypt depth is significantly greater in GH-sufficient (DR) intestine at all ages. Although there are progressive increases in villus height for both DR and dr groups of animals from d 7 through d 23, crypt depth values are similar in both animal groups at d 7 and 14, whereas there is a significant increase in crypt depth for both groups at d 23.

Figure 1
figure 1

Measurement of villus height and crypt depth in intestines of GH-sufficient and GH-deficient rats. At the time of death, sections from the midpoint of the small intestine were harvested and fixed for measurement of villus heights and crypt depths by techniques given in“Methods.” n = 4 for each treatment group at each age.▪, DR, GH-sufficient; □, dr, GH-deficient.

Full size image

In contrast, intestinal length is significantly greater in GH-deficient(dr) animals at all three ages when expressed per g of body weight, whereas ratios of intestinal length to body weight fell in constant proportion in both animal groups with age (Table 1). Protein/DNA ratios, a measure of intestinal cell size, were not different at any age for either group (Table 1).

Maturational expression of digestive hydrolases in GH deficiency. The catalytic activities of three major digestive hydrolases, lactase, sucrase, and AOP, are shown in Figure 2. As seen in Figure 2A, lactase activities were not different between the rat strains at either d 7 or 23. At d 14, DR level of lactase were significantly greater in dr. d 23 activity levels were significantly lower than d 7 levels in both groups (p < 0.05), demonstrating a normal weaning-associated maturational decline in lactase in both GH deficiency and GH sufficiency. No appreciable sucrase activity was detected in either animal group at d 7 or 14. Sucrase activity was present at d 23, and specific activity levels were not significantly different between the two groups(Fig. 2B). AOP activities are known to be present in readily detectable levels in suckling intestine and increase substantially at the time of weaning. As seen in Figure 2C, AOP levels are not different between dr and DR at d 7 or 14 and rise to similar levels by d 23.

Figure 2
figure 2

Measurement of digestive hydrolases lactase, sucrase, and AOP in Intestines of GH-sufficient and GH-deficient rats. The specific activities of digestive hydrolases were measured in the proximal small intestine of GH-sufficient and GH-deficient rats by techniques given in“Methods.” n = 4 for each treatment group at each age.▪, DR, GH-sufficient; □, dr, GH-deficient.

Full size image

Hydrolytic activity per unit length of intestine for the three hydrolases is given in Table 1. At d 7 and 14, no differences were found for lactase or sucrase, but AOP activities were significantly greater in DR versus dr. At d 23, activities for all three hydrolases were significantly greater in DR.

DISCUSSION

The growth and maintenance of the small intestine appears to be regulated by multiple factors including dietary nutrients(31, 32), luminal secretions(2), systemic hormones(1, 3–8, 30–32), and locally produced growth factors(18). During the 3rd wk of postnatal life in the rat, these factors are thought to interact in complex fashion to result in dramatic alterations in intestinal cell proliferation, kinetics, and enzyme activities that are known to occur in preparation for weaning. Due to the complexity of the nature and interrelationship of these regulatory factors, it has been difficult to isolate and investigate the role of any specific factor.

Although pituitary hormones, such as thyroxine and corticosteroids, have been demonstrated to have major roles in regulation of postnatal intestinal growth and maturation(1, 6, 33, 34), the role of GH remains unclear. In one of the first studies investigating the role of GH and intestinal growth, Yeh and Moog(10) studied Wistar rats hypophysectomized at d 6. At d 24, hypophysectomized animals had body weights and intestinal weights far less than controls(10). Additionally, the proportion of intestinal weight to body weight was also reduced in hypophysectomy, suggesting that pituitary hormones may have a specific trophic effect on intestinal tissues. When GH alone at a dose of 10 μg/d was injected into the hypophysectomized animals from d 15 through d 23, no effect on intestinal weight was found. However, administration of the combination of GH, cortisone, and thyroxine during this same time period resulted in increased intestinal weight to normal proportions of body weight. Using a similar model, we have demonstrated complete restoration of both intestinal growth and somatic growth in infant rats hypophysectomized at d 6 when given both thyroxine and GH, starting on d 7 and extending through d 25(5). Administration of the hormones individually resulted in only partial restoration of intestinal or somatic growth. Both of these studies suggested that, if GH has a role in regulation of intestinal growth, it requires the participation of other systemic hormones.

In this study, we report the investigation of the role of GH in regulation of postnatal intestinal growth and maturation using a unique animal model, the SDR. This strain of animals is selectively GH-deficient due to an autosomal recessive mutation in the GH gene, resulting in low expression of pituitary GH mRNA. In spite of this, apart from their low growth rates, SDR rats are otherwise healthy and develop normally. Because SDR rats are solely deficient in GH, it is possible to investigate the pure effects of GH and GH deficiency without experimental artifact introduced by hypophysectomy or administration of exogenous GH. As expected, the somatic growth of GH-deficient rats (dr) was significantly less than GH-sufficient rats (DR). However, although somatic growth was clearly affected by GH deficiency, longitudinal growth of the small intestine was not. The length of the small intestine in dr was roughly twice that of DR per g of body weight at both d 7 and 23 (Table 1), which confirms earlier data from our group, suggesting that longitudinal growth of the small intestine occurs independently of GH status(5, 35). In contrast to the data relating to longitudinal growth, parameters relating to mucosal epithelial cell growth are all significantly higher in DR. The intestines of dr have lower DNA and protein content per cm as well as diminished villus height and crypt depth compared with DR, and there were no significant differences in Pro/DNA ratios, a commonly used measure of cell size (Table 1). These data indicate that there are fewer epithelial cells in the small intestinal mucosa of GH-deficient animals, suggesting that GH has a direct effect on the growth of the intestinal mucosal epithelium. This concept is supported by work done with transgenic mice that overexpress GH and are a model for chronic GH excess(36). When studied at 60-70 d of age (adulthood), mice with chronic GH excess showed increased growth of the small bowel mucosa that appeared to be specific as increased intestinal growth was present in both ad libitum fed and diet-restricted GH transgenics. However, although these data suggest that GH is required for normal growth of the intestinal mucosa, the parameters of mucosal cell growth are all significantly greater at d 23 than they are at d 7 in GH-deficient animals, indicating that GH is not necessary for the well described increases in growth of the mucosal epithelium known to occur in association with weaning(1–3).

The effect of GH on the intestinal hydrolases is less clear. Yeh and Moog(10) found that GH alone had no effect on the brush border hydrolases alkaline phosphatase or sucrase, and when given in combination with thyroxine or cortisone, GH appeared to inhibit the levels of these enzymes. Similar results were found in our study in which GH administration resulted in lower levels of lactase, sucrase maltase, and AOP(5). In transgenic mice overexpressing GH, jejunal sucrase activities were significantly higher when expressed per cm of intestine, whereas sucrase- and lactase-specific activities were no different from controls(36). The authors speculated that the increased sucrase levels most likely reflected an increase in the number of mature enterocytes. As shown in Table 1, in postweaned animals (d 23), lactase, sucrase, and AOP are all significantly higher per cm of intestine in GH sufficiency compared with GH deficiency. However, no differences in lactase or sucrase levels per cm of bowel are found in the intestines of preweaned animals (d 7 or 14). When taken in combination with our data indicating that GH-deficient animals have fewer mucosal epithelial cells (Table 1, Fig. 1), this finding suggests that, in preweaned animals, epithelial cell turnover may be slowed in GH deficiency, allowing the individual enterocyte more time to accumulate more digestive hydrolase. The linkage between enterocyte cell turnover and digestive hydrolase activities has been investigated by multiple groups and appears to be well accepted(31, 37–40). It is of interest that AOP levels per cm of intestine are all significantly lower in GH deficiency at all ages studied, suggesting GH affects this peptidase differently in comparison with lactase and sucrase.

During the period of weaning in the rat, dramatic maturational alterations are known to occur in the major digestive hydrolases, in addition to the observed marked increases in intestinal growth. Among the well characterized alterations that occur during this time are the fall in intestinal lactase, with concomitant appearance and rise in sucrase and AOP(2, 4, 40–42). We now show that, at both d 7 (suckling) and d 23 (weanling), the specific activities of lactase, sucrase, and AOP were not different between GH-deficient and GH-normal animals (Fig. 2). These findings show that normal maturational expression of these three hydrolases occurred in the absence of GH, thus clearly indicating that GH is not required for normal postnatal maturation of the intestinal hydrolases studied here. These data provide additional support for the possibility that GH may influence digestive hydrolase activities indirectly through effects on epithelial cell number and epithelial cell turnover.

In conclusion, we have demonstrated that pure GH deficiency results in significant alterations in certain aspects of intestinal growth and maturation. During the suckling period, the longitudinal growth of the small bowel was unimpeded, suggesting that this aspect of intestinal growth is independent of GH regulation. GH deficiency at this age resulted in diminished numbers of epithelial mucosal cells but with normal levels of lactase and sucrase per cm of bowel. In postweaned rats, GH deficiency resulted in diminished numbers of epithelial mucosal cells and diminished content of digestive hydrolases. However, the major maturational alterations in intestinal growth and enzyme activities that occur during the weaning period occur normally in the absence of GH. This suggests that GH acts as a mucosal epithelial growth factor that regulates steady-state or homeostatic mucosal growth. The mechanics by which this occurs requires further investigation, but most likely involves effects on regulation of mucosal epithelial cell proliferation and turnover which in turn may affect the regulation of the digestive hydrolases.