Main

Colostrum and milk contain many proteins, peptides, and steroids with biologic activity, including IGFs (1). The highest concentrations of hormones and growth factors occur in colostrum, with many exceeding the levels found in maternal plasma (2). Bioactive substances in milk may function in the regulation of the growth and development of the neonatal gastrointestinal tract, immune system, and several neuroendocrine systems (3).

The adult morphology of the mammalian intestinal epithelium is partially established in utero, with the extent of development varying between species (4). The rat intestine shows morphologic alteration during the final few days of gestation from a simple columnar epithelium without a well developed crypt stem cell population to a well developed epithelium sometime during the postnatal period (5). Most research investigating gut differentiation has used a MF rodent model with emphasis on internal regulators that control cell differentiation, such as cell-cell communication and matrix effects (6). However, milk-borne factors are equally important at influencing the internal regulatory mechanisms in the neonatal rodent gastrointestinal tract (7,8).

The IGF system is complex, involving three polypeptide growth factors (insulin, IGF-I, and IGF-II), two receptors (type I and type II), and six distinct binding proteins (IGFBPs) (9,10). In addition, a seventh IGFBP has been identified in humans (11,12), and candidate proteins have been identified for other IGFBPs (13). IGF receptors occur in the intestines of the human (14), rat (15), pig (16), and calf (17). The mucosa of the small intestine of rats is a principal target tissue for circulating IGF-I (18,19). Resectioning of the distal region of rat intestine stimulates a transitory increase in the concentration of IGF-II receptors. The increase in type II IGF receptors has been speculated to be related to adaptive regenerative response and crypt cell hyperplasia (20). After resectioning treatment with IGF-I improves the function of the remnant rat gut (21). Furthermore, prolonged infusion of specific IGF peptides (long R3 IGF-I) enhances the growth of the gastrointestinal tissue in normal rats (22,23). The in utero esophageal infusion of IGF-I enhances gastrointestinal tract growth in sheep (24), and binding of both IGF-I and IGF-II to cell membranes from suckling rat mucosa is two to six times greater than is binding to adult mucosa membranes (25). However, the oral administration of IGF-I in weaned rats shows no effect on the intestine (26). The role of IGFBPs as transport proteins for the IGFs and as activity modulators of the IGFs is widely recognized. IGF-independent actions are now being recognized for IGFBP-3; IGFBPs 7-10 appear to regulate cell replication although they have low affinity for the IGFs.

Because of the complexity of bioactive milk compounds, we focused upon the specific impact of the IGF ligands on the neonatal rat pup gastrointestinal tract when administered by an oral route. The PIC method, modified in this laboratory (27), involves the cheek cannulation of newborn rat pups with artificial diets delivered into the mouth for subsequent swallowing by the pup. This method allows administration of artificial diets to newborn pups, because most are too fragile to survive gastric cannulation. In these studies, a pharmacologic dose of long R3 IGF-I was chosen to emphasize any changes that might be induced by diet; long R3 IGF-I is an analog of IGF-I that exhibits greater receptor binding because of low IGFBP affinity (28,29). The potential role of the IGFBPs in the modulation of cell activity is becoming increasingly more important. Therefore, we investigated the expression of IGFBP-3 mRNA in the intestines of our artificially reared newborn rat pups. Our hypothesis was that oral long R3 IGF-I provided to newborns enhances intestinal villus growth through the alteration of the IGF receptors or IGFBP-3 expression in intestinal epithelial cells.

METHODS

Animals and PIC protocol. Sprague-Dawley timed pregnant rats were acquired from Harlan (Indianapolis, IN). Pups were randomly assigned to be MF controls (n = 8) or to the PIC protocol (n = 6 per treatment group) with the following two diets: RMR or RMR plus long R3 IGF-I (50 ng/mL; GroPep, Adelaide, Australia). Rat milk substitute-2A was made as described by Auestad et al. (30). For PIC, pups were removed from the dam immediately after birth without suckling. Rat pups were housed in pairs in bedding-lined Styrofoam cups to provide tactile stimuli. A 37-38°C local environment for the pups was maintained by suspending the cups in a 39°C water bath. PIC pups were cheek cannulated and fed on a 2-h feeding and 2-h resting cycle throughout the experimental period. At the end of each feeding cycle the anal-genital area of each pup was stimulated with a moist cotton swab to facilitate elimination. The PIC pups received a diet beginning at 2 mL/d that increased at a rate of 25%/d until completion of the experiment on d 3.

Two separate experiments were performed for the collection of intestinal villi cell and length counts, and BrdU incorporation. IGFBP-3 Northern analysis results are reported from one of the experiments. All experiments were approved by the Institutional Animal Care and Use Committee (no. 92R2042B0).

Tissue for microscopy. Two pups from each treatment group were killed on d 1 through 3 of the experiments. One of the pups from each group received an injection of BrdU (Zymed, South San Francisco, CA) 2 h before sacrifice. Proximal and mid portions of intestinal tissue were collected and fixed for microscopy in buffered formalin, dehydrated, cleared, and embedded in paraffin. Sections for morphologic evaluation and histochemical staining for BrdU were dried onto gelatin-coated microscope slides. BrdU incorporation into DNA was detected using a biotinylated MAb to BrdU with an avidin peroxidase/ 3.3′-diaminobenzidine detection system (Zymed). Fifty villi for each treatment from six (unfed and long R3 IGF-I) or seven (MF and RMR) samples were counted for cell number averages by an associate who was not aware of the treatment. The number of sections representing all animals within treatment and the number of villi measured for each treatment are as follows: unfed, 10 and 87; RMR, 11 and 97; MF, 11 and 107; long R3 IGF-I, 11 and 91.

In situ hybridization. Intestine sections were mounted on 3-aminopropyltriethoxsilane (Sigma Chemical Co., St. Louis, MO), baked for 1 h, deparaffinized, and allowed to air dry. Sections were proteinase K-digested, rinsed in PBS, fixed in formalin, dehydrated, and allowed to air dry. The IGFBP-3 probe was obtained from A. Herington (31). Biotinylated IGFBP-3 cDNA probes were prepared using GIBCO BioPrime (GIBCO-BRL, Gaithersburg, MD). Vanadyl ribonucleoside complex (GIBCO-BRL) was added as an RNase inhibitor. Color was developed using an avidin-biotin-peroxidase conjugate and 3-amino-9-ethylcarbazol substrate which forms a red precipitate. Controls consisted of tissue known to be positive (liver) for the probe as well as a positive probe (ribosomal 18 S) and a negative control (no cDNA in hybridization mix).

Receptor assay. In a separate experiment, rat pups were killed by decapitation at birth or 3 d later after being MF or artificially reared (PIC) on a diet of RMR. Intestines from unfed (n = 18), MF (n = 40), or RMR-fed (n = 18) pups were divided into two equal regions (proximal and distal). The intestinal regions were pooled, and microsomes (32) were prepared (33) as follows. Tissue was homogenized for 15 s with a Polytron in a 300 mM mannitol and 12 mM Tris-HCl (pH 7.4) buffer and centrifuged for 30 min at 10,000 × g. The pellet was discarded and the supernate was centrifuged for 100,000 × g for 1 h. The microsomal pellet was resuspended in a 50 mM Tris-HCl buffer (pH 7.4) containing 2 mM MgCl2 and 100 mM mannitol and held frozen at -80°C until the radioreceptor assay was conducted. Many radioreceptor studies are conducted upon preparations of microsomes (all cell membranes minus most nuclear and mitochondrial membranes) that provide a measure of total cellular membrane receptors. Radioreceptor assays were performed with 125I-IGF-II by a modification of the procedure of Hadsell et al. (34). Nonspecific binding was determined in the presence of 400 ng/mL unlabeled IGF-II obtained from U.S. Biochemical Corp, Cleveland, OH.

Northern analysis. Total cellular RNA was isolated from intestinal sections by a modified method of Chomczynski and Sacchi (35) using solution D (4 M guanidine thiocyanate, 0.5% sarcosyl, 25 mM sodium citrate, and 0.1 M β-mercaptoethanol). Total RNA (5 µg/sample) in loading buffer was pipetted onto a 1.1% agarose-2.2 M formaldehyde gel, electrophoresed in 1× 4-morpholinepropanesulfonic acid running buffer containing ethidium bromide, blotted onto a nylon membrane (GeneScreen, NEN, Boston, MA), UV cross-linked, and baked at 80°C for 2 h. The 18 S RNA bands were visualized under UV and photographed for comparison of RNA loading differences. A cDNA probe for IGFBP-3 (36) was labeled using the method of random priming (Promega, Madison, WI). Hybridization was performed at 42°C in 50% formamide dextran, 5× SSPE, 1% SDS, 5× Denhardt's solution, and 200 µg/mL salmon sperm DNA. The membrane was washed at 65°C for 20 min each in 2× SSC, 2× SSC and 0.1% SDS, and 0.1× SSC and 0.1% SDS. The blot was placed in a cassette and exposed to x-ray film. The autoradiograph and 18 S RNA photograph were scanned by the Eagle Eye II videodensitometry system and analyzed using Scanalytics software (Stratagene, La Jolla, CA).

Statistical analysis. Values for intestinal length, cell number, and receptor binding were analyzed by the SAS (37) general linear model procedure and tested for significance (p < 0.05) by the Duncan-Waller multiple comparison test (intestinal length and cell number) or Tukey's mean separation test (receptor binding).

RESULTS

Dietary effects on rat pup intestine growth and development. Cross-sections of rat pup intestine from different dietary treatments that have been stained for BrdU uptake are shown in Figure 1. The intestinal villi of unfed newborn rat pups (Fig. 1A) are flattened and BrdU-stained, indicating active DNA synthesis is not confined to the crypt-villus axis (38,39). The intestines of pups who were MF for 3 d (Fig. 1B) show BrdU staining that is more localized to the crypt-villus axis and show increased villus height. The proximal intestine morphology of pups fed RMR for 3 d (Fig. 1C) parallel that of the unfed newborn pups, whereas the pups given RMR plus 50 ng/mL long R3 IGF-I (Fig. 1D) show increased villus height with BrdU staining localized to the crypt-villus axis.

Figure 1
figure 1

Morphology of proximal intestine of unfed neonatal rat pups at birth and after 3 d of MF or PIC procedure. Pups were pulsed with BrdU for 2 h before sacrifice. BrdU uptake was detected using a biotinylated MAb to BrdU with peroxidase/3,3′-diaminobenzidine detection (brown) and hematoxylin counterstain (blue). (A) Unfed newborn; (B) MF (3 d); (C) RMR-fed (3 d); (D) RMR + 50 ng/mL long R3 IGF-I-fed (3 d).

Full size image

Figure 2, A and B, is a graphical representation of the diet-induced changes in villus cell number and height. Unfed pup villi contained 29.52 ± 0.8 cells (mean ± SE) and measured at 31.05 ± 0.76 µm (mean ± SE). Pups fed RMR for 3 d showed a slight increase in cell number and increase in villus height, to 31.93 ± 0.98 and 34.59 ± 1.00 µm, respectively, but these measurements were not significantly different (p > 0.05) from unfed animals. Pups who were MF for 3 d showed significantly increased villus cell numbers to 37.14 ± 0.99 (p < 0.05) and villus length to 38.79 ± 1.14 µm that represented a 1.2-fold increase over unfed animals. Three days of feeding RMR plus 50 ng/mL long R3 IGF-I caused an increase (p < 0.05) in both cell number and villus length to 44.48 ± 1.51 and 48.97 ± 1.46 µm when compared with the values from all other treatments (p < 0.05). Long R3 IGF-I treatment increased villi length and cell number 1.5-fold in relation to both unfed and RMR-treated pups, showing that dietary IGF-I stimulates hyperplasia.

Figure 2
figure 2

Comparison of the impact of an oral application of an IGF analog with those MF. (A) Proximal intestine villus cell numbers (crypt to villus tip) at birth and after 3 d of mother-feeding or the PIC procedure. Data are collected from Figure 1 plus other micrographs not shown and are given as the mean ± SE. Different letters represent significant differences at the p < 0.05 level.

Full size image

Age and diet effects on rat pup intestine type II IGF receptors. Rat pups were killed immediately after birth, after being MF for 3 d, or after being fed RMR for 3 d. After sacrifice the intestine was divided into two regions, proximal and distal, and a radioreceptor assay for type II IGF receptors was performed on prepared microsomes. Previous studies with newborn calves have shown the type II IGF receptors are three to four times more prevalent than those of type I IGF receptors (40), making measurement possible to test our hypothesis with the small sample sizes obtained from these studies. Neonatal rat pups are born with an unequal distribution of type II IGF receptors (Fig. 3); the proximal intestinal region exhibits 13.56 ± 1.92% (mean ± SE) specific IGF-II binding, whereas the the distal intestinal region exhibits 9.20 ± 1.16% specific IGF-II binding. The receptor population shifted slightly after RMR feeding for 3 d. The distal intestinal section showed a slight increase in binding to 16.48 ± 0.82% specific binding in contrast to the binding pattern exhibited by unfed newborn pup intestines. MF pup intestines showed a dramatic increase in receptor population for both proximal (19.81 ± 5.37%) and distal (32.5 ± 3.8%) sections after 3 d. In addition, the receptor distribution had shifted so that higher numbers of receptors were now located in the distal portion of the intestine. When receptor-specific binding levels were compared within an intestinal region, the d 3 proximal intestine from MF pups had higher (p < 0.05) IGF-II binding than either the unfed or d 3 RMR-fed pups. In the distal intestine region, all groups were significantly different from each other. Comparing both RMR and MF treatments to newborn unfed pups indicates that the shift in receptor distribution is age-related, whereas aqueous milk factors, including IGF, cause an increase in receptor numbers.

Figure 3
figure 3

Dietary IGF affects binding of IGF to the type II receptor in intestinal regions of neonatal rat pups. Data are expressed as the mean ± SE. Different letters indicate significant differences of p < 0.05.

Full size image

Intestinal expression of IGFBP-3. In situ hybridization was performed on intestine sections of 3-d-old MF pups to demonstrate the cellular location of IGFBP-3 in neonatal rat pup intestine. IGFBP-3 mRNA expression is located in intestine epithelial cells of newborn rat pups (Fig. 4).

Figure 4
figure 4

IGFBP-3 mRNA is expressed in the intestinal epithelial cells of newborn rat pup intestines. In situ hybridization with a cDNA probe for IGFBP-3. (A) MF, d 3, no probe; (B) MF, d 3, cDNA probe for IGFBP-3.

Full size image

Effects of age and diet on IGFBP-3 mRNA expression. Intestinal expression of IGFBP-3 mRNA, as measured by Northern analysis, in unfed newborn rat pups shows 1.9 times greater expression in the proximal intestine region than in the distal intestine region (Fig. 5). IGFBP-3 mRNA expression falls in both proximal and distal regions after 3 d. Reductions in expression were seen irrespective of the diet given, with the proximal region retaining greater expression than the distal region.

Figure 5
figure 5

IGFBP-3 mRNA declines with age in rat pup intestine. Rat pup liver was used as a positive control, and hybrid mix with no cDNA added was used as a negative control (data not shown).

Full size image

DISCUSSION

Our previous reports have shown that rat milk infranatant (milk aqueous phase) contains many peptide growth factors (41). Recently, we demonstrated the effect of milk aqueousborne factor deficiency on newborn rat pup growth and viability. Inclusion of rat milk infranatant in the RMR diet caused morphologic changes in the intestine consistent with differentiation that appeared intermediate between MF and the unfed and RMR-fed controls (15). The study by Baumrucker et al. (27) and the results reported here demonstrate the fact that dietary endocrine factors have the potential to influence intestinal growth and development.

IGF ingested by the neonate survives digestion (42,43) and has been found intact in gastric contents (43). Milk IGF-I has been quantified in several species (44–46); rat milk contains 30 µg/L IGF-I on d 1 of lactation and the level falls to 13 µg/L by d 21 of lactation (47). Pups who were given only RMR, the equivalent of milk without aqueous growth factors, showed very little intestinal development 3 d after birth. Their villi remained flattened with active DNA synthesis occurring outside of the crypt-villus axis. This finding suggests that initial intestinal development in the newborn rat pup requires more than "nutrients" for initial development.

We fed 50 ng/mL long R3 IGF-I, which is 20 ng/mL higher than native wild type IGF-I, found in rat colostrum. This higher dose of IGF-I and the use of the analog with reduced IGFBP binding was expected and intended to cause pronounced results when compared with pups suckled by the mother (MF) to test our hypothesis. The intestinal villi of pups given long R3 IGF-I showed increased cell number and villus height over RMR diet and were surprisingly greater than intestines of the MF pups. Studies using neonatal pigs have shown that oral IGF-I and IGF-II stimulate cell proliferation in the gastrointestinal tract, particularly increasing mucosal thickness (48,49). Our study confirms this finding and demonstrates the efficacy of oral administration of this growth factor in newborn rodents. Oral long R3 IGF-I stimulated enterocyte hyperplasia by increasing mitogenesis in the crypt region. A study by Phillips et al. (50) showed that oral supplementation of IGF-I in rat pups increases the enterocyte migration rate in the duodenum and proximal jejunum, which could be ascribed to the increased mitogenesis of the crypt cells. An increase in enterocyte lifespan by IGF-I exposure would cause macroplasia as well, as suggested by Burrin et al. (49).

Gut responsiveness to IGF-I has been demonstrated under many different conditions. Normal adult rats show increased villus height and crypt depth, with increased cell density throughout, in experiments with long- and short-term IGF-I infusion (22,23). IGF infusion causes increased gut weight and absorptive area in bowels that have been resectioned (21), transplanted (51), or subjected to dexamethasone-induced catabolism (52,53), demonstrating the ability of IGF to stimulate mitogenesis in the adult gastrointestinal tract.

Our study did not measure changes in intestinal enzyme activities, but recent studies have shown that colostrum and oral IGF-I influence intestinal development as measured by changes in alkaline phosphatase and disaccharidase activity. Feeding piglets colostrum increased lactase and alkaline phosphatase activity over that seen when piglets were fed trypsinized colostrum (54). Oral administration of IGF-I to piglets increased lactase and sucrase activities as well as increasing villus height (55).

The type I and II IGF receptor numbers have been characterized by several laboratories over many age ranges in rats and pigs (25,56–58), but the pattern of the type II IGF receptor expression in the newborn rat pup and during the first days of life was unknown. In this study we found that the type II IGF receptor changed its distribution in the intestine during the first 3 d of life and was influenced by diet, thus demonstrating one oral mode of IGF action in the intestine. Newborn unfed pup intestines showed higher receptor numbers in the proximal intestine, whereas 3-d-old MF pups had higher receptor numbers in the distal intestine. Pups given RMR without any additional growth factors for 3 d showed no change in their receptor distribution, indicating that aqueous milk factors can influence the type II IGF receptor distribution. These findings agree with that of Heinz-Erian et al. (56) who found that adult rats had an IGF receptor distribution that was higher in the distal intestine than in the proximal intestine. Young et al. (25) also found that both type I and type II receptor levels were highest proximally during the early suckling period and that receptor distribution shifted distally as the rats approached maturity with overall levels falling with age. It is generally agreed that down-regulation of the receptors occurs as a result of the falling levels of IGF in milk, with receptor levels indicating the level of cellular differentiation (25,56–59). The fact that receptor levels in the RMR-fed pups remain unchanged and our photomicrographs indicate very little organization and differentiation in the crypt-villus axis further supports the concept that milk-borne endocrine factors play a role in newborn intestinal changes.

Although our hypothesis specified oral IGF-I as the agent of intestinal growth and developmental change, we also hypothesized that the change would occur via IGF receptor alterations or changes in intestinal IGFBPs. IGFBP-3 mRNA expression was localized to the epithelial cells of intestinal villi by in situ hybridization. This differs from a previous report by Winesett et al. (60) where they found that IGFBP-3 mRNA was localized to the subepithelial cells of the lamina propria in rat jejunum. The difference in location of mRNA expression is most likely due to the difference in ages of the rats used in the two studies; the study performed by Winesett et al. (60) used adult rats. The results of Northern analysis in our studies show declining IGFBP-3 mRNA with age, which suggests that with time and development of the intestine, IGFBP-3 may not be expressed in epithelial cells. The function of the IGF-binding proteins in the intestine is believed to be growth inhibition by sequestering IGF and preventing it from interacting with the IGF receptors. Knocking out the expressed binding proteins in IEC-6 (61) and HT-29 (62) intestinal cell lines causes increased cell growth. IGFBP-3 mRNA levels have been found to decrease after bowel resectioning (63), and fasting and refeeding (64), further supporting this hypothesis. IGFBP expression has also been found to change with differentiation in Caco-2 cells (65). In our study, IGFBP-3 mRNA levels were found to decrease after birth irrespective of the diet given, suggesting that the changes are solely age-related.

A model of the IGF system in the intestinal tract of a newborn rat pup would be that IGF is ingested in milk during suckling where it passes through the stomach and into the intestines in an active form. When it reaches the intestines it interacts with the IGF receptors and locally produced IGFBPs to increase growth and differentiation of the underdeveloped epithelium. As the IGF content of the ingested milk declines and the intestinal tract becomes more functionally developed, the IGF receptors are down-regulated and local production of IGFBP are reduced. Increasing the level of ingested IGF to the newborn intestine, as demonstrated by the use of long R3 IGF-I, stimulates the growth of the epithelium that is likely to bring it to a more functional state in a shorter period of time. At times when the function of the intestine is compromised, a higher level of IGF is present to stimulate the regeneration of the intestinal epithelium and return it to a normal state.

In summary, it can be concluded from this study that 1) dietary endocrine growth factors may influence rapid intestinal growth and development within short time periods (days) in the newborn, 2) IGF-I analogs retain bioactivity in the newborn's gastrointestinal tract and contribute to neonatal rat pup intestinal growth in a short period of time, 3) IGF system components (type II receptors; IGFBP-3) are expressed in the intestine of neonatal rat pups, 4) type II IGF receptor patterns change with age and diet, and 5) IGFBP-3 mRNA expression occurs in epithelial cells and declines with age, but does not appear to be influenced by oral IGF-I. The study of the effects of oral IGF-I on intestinal development hold promise for enhancing the maturation of infants born prematurely.