Oral Insulin is Biologically Active on Rat Immature Enterocytes

Insulin is a hormone essential for the growth and the control of many cellular metabolic processes, including those occurring in the small intestine (1,2). Several studies have documented that parenteral administration of insulin, even at doses causing no hypoglycemic reaction, to infant mice (3) and rats (4) induces intestinal maturation prematurely and enhances the adult pattern of expression of brush border membrane (BBM hydrolases), lysosomal (N-acetyl-β-glucosaminidase), microsomal (sulfatase C), and cytosolic (lactate dehydrogenase) enzymes (5). Similar effects have been observed on intestinal cells maintained in serum-free organ culture (2) and in adrenalectomized infant rats (6), indicating that enzyme stimulation by parenteral administration of insulin is not mediated by an endogenous release of corticosterone. On the other hand, anti-insulin receptor antibodies given to suckling and weanling rats (6) inhibit mucosal growth and the ontogenic expression of BBM hydrolases. Although significant quantities of insulin have been detected in the colostrum and milk of several species, including humans, with concentrations exceeding 20-30 times those measured in the serum (7,8), it remains uncertain whether milk-borne insulin exerts biological effects on the immature gastrointestinal tract. A recent investigation (9) has shown that the addition of 85 mU/ml porcine insulin to the diet of newborn piglets enhanced lactase activity by day 8 in the ileum but not in the jejunum. Although this corresponds to a daily dose being at least 100-1,000 times higher than the insulin levels measured in pig colostrum and milk (60-625 μU/ml) (7), the authors found no effect of oral insulin on the mucosal levels of high-molecular-weight lactase precursors nor on the expression of lactase mRNA (10), and suggested that the increase in lactase activity could be due to reduced cell proliferation with prolonged enterocyte life span. Using the suckling rat as a model of immature intestinal function, we have analyzed the effects of orogastrically administered insulin on mucosal mass parameters and on the expression of BBM hydrolases. MATERIALS AND METHODS Animals and Insulin Treatments All procedures were approved by the National and University Animal Care Committees (Fonds de la Recherche Scientifique Médicale, FRSM, Belgium). To equalize conditions of feeding and nursing, litters of Wistar rats were reduced to six pups per lactating mother, routinely on day 1 postpartum. Pups were assigned at random to either experimental or control groups. Ten animals per group were treated from days 10 to 14. The first group was treated with human insulin (Actrapid, Novo Industries, Copenhagen, Denmark) given as a mixture incubated for 30 min at 37°C with freshly collected rat milk. The second group was treated with human insulin dissolved in its appropriate buffer, whereas the third group was treated with the vehicle alone (controls). Each dose (5 mU/g body weight) was dissolved in 100 μL of rat milk or vehicle and administered twice daily via a small catheter (1.2 mm inside diameter) that was inserted by mouth to a depth of 3 cm. Assuming that rat milk contains 0.52-0.625 mU insulin/ml (the concentrations measured respectively in human [11] and porcine [7] milk) and that 10-day-old pups consume 2-3 ml of milk per day, each dose expressed per unit of weight would be 10 times higher (pharmacological range) than the estimated daily intake of milk-borne insulin. Preparation of Tissues On the day of the experiment, the animals were killed by rapid decapitation, 2 h after a last dose of insulin. This timing was chosen because the hypoglycemic reaction and the potential stimulation of endogenous corticosterone release occurs 1.5-3 h after administration of the hormone (4,12). Trunk blood was collected into heparinized tubes within 20 seconds after removal of the animals from their mother. The small intestine was removed, trimmed of fat and mesentery, and rinsed with ice-cold saline. After being measured with a constant load (0.1 g), the mucosa was scraped with glass slides and stored at -170°C until use. BBMs were purified according to the calcium chloride precipitation method (13). Preliminary assays revealed that lactase and maltase specific activities were at least 10-15 times higher in the final BBM suspension than in the crude homogenate. Biochemical Determinations Plasma corticosterone (free and bound) concentrations were determined by a specific protein-binding assay on specimens of blood collected from insulin- and control-treated groups (14). Sucrase, lactase, and maltase activities were assayed in BBM samples by standard methods (15). Activities were expressed as micromoles of substrate hydrolyzed per minute and per milligram of BBM protein. Protein contents in mucosa and BBM samples were determined by the method of Lowry et al. (16). Statistics All results are given as mean ± SE. Differences between means were tested for statistical significance (p < 0.05) by the nonparametric Mann-Whitney U test. RESULTS Insulin-treated animals and controls appeared healthy and showed equivalent growth rates during the study period. No mortality was recorded. Changes in intestinal length, intestinal weight, and mucosal mass between insulin-treated rats and controls were not significant. We found virtually no difference in mucosal mass parameters between controls and insulin-treated rats, when insulin was administered in its vehicle, whereas small increases (+11-23%) in intestinal weight and mucosal weight per unit of length were detected in rats treated with insulin incubated in rat milk (not significant). Changes in BBM hydrolases are detailed in Table 1. Compared with controls, sucrase specific activity was enhanced by 6-fold in insulin-treated rats (in vehicle) and by 40-fold in rats treated with insulin mixed with rat milk. Likewise, oral insulin given in its vehicle caused a significant increase in maltase activity (+66%), which was markedly increased when insulin was given as a mixture with rat milk (+2.2-fold). An identical pattern of response to insulin was observed for lactase, whose mean specific activity in insulin-treated rats (in rat milk) was enhanced by 2.5-fold over controls. Protein concentration in BBM remained equivalent in the three groups of rats. At the doses used (5 mU/g body weight), insulin dissolved in its vehicle or mixed with rat milk produced no hypoglycemia. Likewise, we found no difference between insulin-treated rats and controls regarding the mean plasma corticosterone circulating levels. DISCUSSION Although the doses of oral insulin given were in the pharmacological range and ≈10 times higher than the estimated daily intake of milk-borne insulin, the present findings clearly demonstrate that the hormone is able to enhance intestinal BBM enzymes prematurely, when given by the orogastric route in its appropriate vehicle. Because free oral insulin is massively destroyed by endoluminal proteases (17), our findings suggest that a small endoluminal pool of hormone was left intact from the initial dose and remained biologically active on immature intestinal cells. When insulin was given as a mixture with rat milk, BBM enzyme activities were dramatically stimulated. This further supports the concept that milk-borne insulin is biologically active on the gut, probably much more than free oral insulin, because milk-borne hormones such as IGF-I remain stable in the gastrointestinal tract and are protected against endoluminal degradation by several milk antiproteases (17,18). In support of this theory, insulin receptors have been detected in the small intestine of fetuses, newborn, and suckling rats (19), and their concentration peaks at the end of the suckling period (19,20). In addition, we have shown (20) that both α- and β-subunits of the insulin receptors are equally concentrated in BBM and in basolateral membranes of villus and crypt cells (20). After weaning, receptor concentration in intestinal membranes decreases by twofold without change in affinity constants. Thus, the onset of weaning corresponds to a short but critical period during which plasma insulin levels rise actively (21,22), and the concentration of intestinal receptors peaks (20) while milk-borne insulin is still present in the gut, allowing optimal interaction of the hormone with BBM and basolateral membrane receptors. This could explain why immature enterocytes are highly responsive to insulin which is believed to play a key role in the ontogenic changes of intestinal enzymes that occur at the beginning of the weaning period, in conjunction with other maturation effectors such as corticosterone (23) and thyroxine (24). In addition, the lactational stage of the dam and the levels of milk-borne hormones can affect the rate of completion of weaning as well as the terminal phase of sucrase maturation (25). Because oral insulin can pass through the gastrointestinal wall of the infant rat (26) and determine metabolic effects, we measured changes in glycemia and plasma corticosterone levels in controls and in insulin-treated rats. Our data show that in response to oral insulin, no hypoglycemic reaction occurred and that plasma corticosterone levels remained equivalent to that of controls. In addition, the plasma corticosterone levels measured here are identical to the levels measured in preweanling rats and are 15 times lower than active levels measured in weanling rats (6). Obviously, permeability of upper gastrointestinal mucosa to endoluminal insulin is lower in preweanling than in infant rats (26). This suggests that enzyme stimulation is a direct effect of oral insulin, which was not mediated by endogenous stimulation of corticosterone release consecutive to the stress or to hypoglycemia. The direct effect of insulin on the immature enterocyte is further confirmed by three relevant observations. First, after a single low dose of insulin, given intraperitoneally to suckling rats but producing no hypoglycemic reaction, the premature induction of sucrase can be detected very rapidly (within 6 h, corresponding to the half-life time of the enzyme) in both differentiated villus and undifferentiated crypt cells (4). Second, induction of BBM enzymes by insulin does occur in adrenalectomized infant rats with the same magnitude as the enzyme changes noted in normal rats (6). Third, addition of insulin to mouse intestinal epithelium maintained in free-serum organ culture reproduces the effects observed in vivo (3). Taken together, the present observations suggest that oral insulin is biologically active on immature enterocytes. Further studies are warranted to precisely determine the physiological relevance and trophic action of milk-borne insulin in intestinal development. Acknowledgment: These investigations were supported by grant 3-4546-95 from the Fonds de la Recherche Scientifique Médicale (FRSM), Belgium. The authors thank Philippe De Nayer for technical assistance and Dominique Vermeulen for preparation of the manuscript.