Main

Early diet and feeding practices have been linked to the etiology of IDDM(1). Retrospective studies assessing early infant nutrition show an increased risk of disease development with early exposure to foods other than breast milk(1,2). Studies in the BB rat, a valuable model of human IDDM, suggest that semipurified diets may be protective against disease development, whereas standard laboratory chow appears to be more diabetogenic in nature(3,4). The precise mechanism by which semipurified diets confer protection is yet unknown. A "critical window" surrounding weaning appears important for disease incidence(5) in that diabetes incidence is decreased if semipurified diet is introduced before 30 d and continued for at least 100 d(5). Because of the relevance to human infant feeding practices and their possible relationship to the etiology of diabetes mellitus, the diabetes-inducing potential of cow's mil, soy-based products, and early infant feeding practices warrant examination(6).

Cereal-based chow diets contain up to 75% carbohydrate as complex carbohydrate and contain more fermentable dietary fiber(7). The presence of complex carbohydrate slows glucose absorption in the small intestine(8) and can reach the large intestine in significant amounts(9). Fermentation of complex carbohydrate and fiber in the large bowel results in the production of short chain fatty acids, which can increase intestinal cell proliferation and gastrointestinal digestive development(10). Cummins et al.(11) demonstrated that cyclosporin A, a drug shown to prevent autoimmune diabetes in animals(11–13), causes a delay in the maturation of the small intestine during weaning in the rat. Delays in morphologic and enzymologic development can decrease nutrient uptake and hormone production(14,15). Changes in the ontogeny of gut-associated immune function and intestinal morphology and function may be involved in progression of diabetes. Because wheat products containing fermentable carbohydrate are encountered at weaning, the effect of these substances on the development of the intestine should be further investigated.

It is well established that suppressing endogenous insulin secretion can delay the onset of diabetes in the BBdp rat(16,17). Conversely, increasing endogenous insulin secretion by administration of chyclophosphamide accelerates the progression of beta-cell destruction(18,19). Insulin secretion can potentially be increased by enhancing glucose transport and/or increasing the postprandial release of intestinal incretins, or those insulinogenic hormones that are released from the gut in response to oral nutrients. L-cells in the distal ileum and colon contain the proglucagon gene that is postranslationally cleaved into several peptides. The incretin, GLP-1, is a potent insulin secretagogue(20) and delays gastric emptying(21). GLP-2 has been shown to increase jejunal glucose transport(22). Changes in proglucagon expression may therefore influence insulin secretion and glucose absorption from the intestine. We have previously demonstrated that high fiber diets up-regulate intestinal proglucagon expression in normal adult rats(7). Upregulation of proglucagon mRNA is also accompanied by increased plasma GLP-1 and insulin levels. The impact that the presence of complex carbohydrate and fiber in weaning diets has on proglucagon and subsequent insulin secretion release in weanling animals in unknown. Proglucagon may play an important role in increasing the secretory potential of pancreatic islets and ultimate disease pathogenesis.

It remains to be elucidated if differences exist in intestinal incretin and glucose transporter expression as early as 30 d, when clinical symptoms of diabetes are not yet present. The fact that protective diets must be introduced before 30 d to confer protection against diabetes led us to hypothesize that diets differing in diabetic potential may alter intestinal gene expression at this early critical period (30 d). In this study we examined the expression of proglucagon and the brush border glucose and fructose transporters, SGLT-1 and GLUT5, respectively, in BBdp and BBn rats at 30 d of age fed one of three nutritionally complete diets. Chow produces a high incidence (60-80%), casein-based semipurified diet low incidence(20-30%), and soy-based semipurified diet an intermediate incidence (40-60%) of diabetes(3). Disease susceptibility refers to differences between BBn and BBdp rats.

METHODS

Animals and diets. Female and male BBdp and BBn rats were obtained from the University of Alberta, Department of Agricultural, Food and Nutritional Science rat colony. Original breeding pairs were obtained from Health Canada (Animal Resources Division, Health Protection Branch, Ottawa, ON, Canada). Animals were housed in a temperature- and humidity-controlled room with a 12-h light/dark cycle. Weanling rats were housed in groups of two to four animals per cage. The protocol was approved by the University of Alberta Animal Welfare Committee.

Composition of the experimental diets in given in Table 1. The nonpurified chow diet (National Institutes of Health-07 Rodent Diet) was purchased from Ziegler Brothers Inc. (Gardners, PA). The two semipurified diets were made in our laboratory with either soy or casein as the protein source. Dams were fed one of the three nutritionally complete experimental diets beginning 1 wk before parturition. Pups were weaned from the dams onto the same diet as their respective dams at 21 d. Rats were fed until 30 d of age. After an overnight fast, animals were killed by cervical dislocation. The entire small intestine and colon were excised, contents removed, and tissues measured and weighed. The small intestine was divided by length into three equal segments. Approximately 3 cm of distal duodenum, jejunum, and ileum and 3 cm of proximal colon were immersed in liquid nitrogen for later mRNA analysis.

Table 1 Composition of the experimental diets
Full size table

Isolation of total RNA. Total RNA was isolated using Trizolâ„¢ (GIBCO BRL, Burlington, ON, Canada). Isolation was according to the protocol provided with the reagent. RNA was dissolved in diethyl pyrocarbonate-treated water, and quantity and purity were determined by UV spectrophotometry at 260 and 280 nm.

Northern blot analysis. mRNA in all samples was measured using a Northern blot analysis procedure described by Reimer et al.(23). The cDNA proglucagon probe was donated by Dr. P. J. Fuller, Prince Henry's Institute of Medical Research, Melbourne, Australia; the GLUT5 cDNA probe by Dr. G. I. Bell, Howard Hughes Medical Institute, University of Chicago, Chicago, IL, and the SGLT-1 cDNA probe by Dr. N Davidson, University of Chicago. For statistical analysis, signals were quantified using laser densitometry (model GS-670 Imaging Densitometer, Bio-Rad Laboratories (Canada) Ltd., Mississuaga, ONT). The 28 and 18 S ribosomal bands were quantified from negatives of photographs of the membranes. These bands confirm the integrity of the RNA and were used as a denominator for densitometer values to compensate for any loading discrepancies.

Statistical analysis. All data are expressed as mean ± SEM. A significant effect of gender was not found; therefore genders were combined and differences between treatments were determined using a two-way ANOVA model (diet and disease susceptibility) in the general linear model procedure in SAS (Version 6.04, SAS Institute, Cary, NC). Significant differences between groups were determined using least mean squares. Statistical significance is defined as p ≤ 0.05.

RESULTS

Growth parameters. Differences due to diet were largely found in BBn animals with the chow-fed animals having significantly greater body weight, stomach weight, and small intestine length than the BBn animals fed casein or soy diets (p < 0.001) (Table 2). The only diet difference detected in BBdp animals was in small intestine length, with chow-fed animals having the shortest intestine, soy-fed animals the longest, and casein-fed animals were intermediate (p < 0.005).

Table 2 Effect of diet and disease susceptibility on animal characteristics
Full size table

Distinct patterns of growth based on differences in disease susceptibility(BBn versus BBdp) were dependent on the weaning diet. In chow-fed animals, BBn had higher body weight, greater stomach weight, longer small intestine length, lower small intestine weight per body weight, and lower colon weight per body weight (p < 0.005). In soy-fed animals, BBn had higher small intestine weight, higher small intestine weight per body weight, and lower body and colon weights (p < 0.002).

Intestinal RNA content. Diet differences in intestinal RNA content were found with soy-fed BBn animals, having significantly higher total jejunal RNA compared with casein-fed BBn animals (p < 0.003) (Fig. 1). Within BBdp, chow-fed rats had the highest total RNA compared with those fed casein and soy (p < 0.004). When examining differences due to disease susceptibility, only soy-fed BBn had higher total jejunal RNA than soy-fed BBdp (p < 0.004). No significant differences were found for total colonic RNA.

Figure 1
figure 1

Effect of diet and disease susceptibility on total jejunal RNA content. Values were obtained by multiplying RNA concentration (mg/g) by total jejunal weight (g). Values are mean ± SEM (minimum n = 7). Significant effects are expressed as: the asterisk (*) represents a significant difference(p < 0.05) between BBn vs BBdp (disease susceptibility) within a single diet. The dagger († represents a significant difference(p < 0.05) among diets within a disease susceptibility. CAS, casein.

Full size image

Proglucagon mRNA. BBdp animals fed all three diets had significantly lower levels of proglucagon mRNA (p < 0.01) than their BBn counterparts. No differences due to diet were found for any group except chow-fed BBdp animals, which tended (p = 0.09) to have higher proglucagon mRNA than soy-fed BBdp (Fig. 2).

Figure 2
figure 2

Effect of diet and disease susceptibility on colonic proglucagon mRNA expression in BB rats. Values are mean ± SEM (n = 4 rats/treatment). Significant effects are expressed: the asterisk (*) represents a significant (p < 0.05) difference between BBn vs BBdp (disease susceptibility) within a single diet. CAS, casein.

Full size image

When corrected for the total potential colonic mRNA abundance, BBdp fed casein and soy diets had significantly lower total proglucagon than BBn fed casein and soy diets (p < 0.001). A significant difference between the two diseases no longer existed between chow-fed BBdp and chow-fed BBn after correction for total RNA. Within BBn animals, casein-fed had higher total proglucagon than those fed chow (p < 0.05). Within BBdp, animals fed chow diets tended (p = 0.09) to have a higher mRNA abundance than those fed soy diets (Fig. 3).

Figure 3
figure 3

Effect of diet and disease susceptibility on colonic proglucagon mRNA expression adjusted for total colonic RNA content in BB rats. Values were obtained by multiplying total colonic RNA by proglucagon mRNA levels. Values are mean ± SEM(n = 4 rats/treatment). Significant effects are expressed: the asterisk (*) represents a significant (p < 0.05) difference between BBn and BBdp (disease susceptibility) within a single diet. The dagger (†) represents a significant difference (p < 0.05) among diets within a disease susceptibility. CAS, casein.

Full size image

Glucose transporter mRNA. A difference in glucose transporter expression due to the difference in disease susceptibility (BBdp versus BBn) was seen with chow-fed BBn having a greater abundance of SGLT-1 mRNA than BBdp fed chow diets (p < 0.0001)(Table 3). As well, soy-fed BBn rats had a lower abundance of SGLT-1 mRNA than soy-fed BBdp rats (p < 0.05). Within the BBn strain, the ingestion of chow diets resulted in the highest abundance of mRNA with soy-fed animals being the lowest and casein-fed rats were intermediate (p < 0.003).

Table 3 Effect of diet and disease susceptibility on absolute glucose transporter mRNA abundance
Full size table

Adjusting SGLT-1 mRNA for total jejunal RNA content did not alter the pattern seen with the absolute abundance of mRNA with chow-fed BBn rats still having higher total SGLT-1 message than BBdp fed chow-diets (p< 0.01) (Table 4) Due to a greater total jejunal RNA content, chow-fed BBdp rats also had higher total SGLT-1 than casein-fed BBdp rats (p < 0.05). Within BBn, chow-fed animals again had a significantly greater abundance of total SGLT-1 mRNA than did casein- or soy-fed animals (p < 0.001).

Table 4 Effect of diet and disease susceptibility on glucose transporter mRNA abundance corrected for total jejunal RNA content
Full size table

Differences attributable to disease susceptibility were also seen with GLUT5 in which chow-fed BBn rats had a greater abundance than chow-fed BBdp rats (p < 0.0001) (Table 3). As for SGLT-1, BBn rats fed chow diets had the highest abundance of GLUT5 mRNA, whereas soy-fed rats had the lowest, and those fed casein diets were intermediate (p < 0.005).

Adjusting GLUT5 mRNA for total jejunal RNA content resulted in BBn rats fed chow having significantly greater GLUT5 than did BBdp rats fed chow(p < 0.0001) (Table 4). Within BBn, chow-fed animals had a significantly greater total GLUT5 message compared with those fed casein or soy diets (p < 0.0003).

DISCUSSION

Growth and maintenance of the small intestine is regulated by numerous factors including dietary nutrients(14,24), luminal secretions(25), systemic hormones(26,27), and locally produced growth factors(28). In the BB rat, diet is known to be an important environmental factor influencing disease incidence. Because antigens normally expressed in pancreatic islets may mediate the autoaggressive attack(18,19), changes in the response of intestinal incretins and nutrient uptake to different weaning diets in the BB rat are of particular interest. This study examined the effect of different weaning diets and disease susceptibility (BBn versus BBdp) on the growth of the intestine as well as proglucagon and intestinal glucose transporter mRNA abundance. Our data suggest that BBn and BBdp animals respond differently to chow and semipurified nutritionally complete weaning diets.

Perhaps the most striking finding was that early dietary exposure significantly affected growth and intestinal characteristics of BBn rats whereas BBdp animals seemed to be unresponsive. For all growth parameters examined, including body weight, stomach weight, small intestine weight, small intestine weight per body weight, and colon weight per body weight, BBn animals fed chow were different from BBn animals fed casein or soy diets, whereas diet was associated with few significant differences in BBdp rats. This finding is surprising considering the very different disease incidence rates manifested among the three experimental diets. The evidence does, however, suggest that early diet significantly affects the growth rates and intestinal development of normal animals.

A similar response was observed regarding mRNA abundance of glucose transporters and colonic proglucagon. It appears that the physical growth of the intestine and the expression of intestinal gene products reported to determine functional capacity of the intestine were unresponsive to diet in BBdp animals at 30 d of age. Several possible explanations exist regarding this finding. First, the intestinal plasticity observed in normal nondiabetic BBn animals to chow diets containing greater amounts of complex carbohydrate and fiber and less fat may confer protection against disease. The lack of a similar response to diet observed in the BBdp animals might suggest that appropriate mechanisms to adapt to environmental and dietary challenges are lacking or nonfunctional in BBdp animals.

Second, intestinal ontogeny may progress at different rates in BBn versus BBdp animals so that the window of dietary sensitivity is shifted. This concept might explain the observation that the protective effect of semipurified diets seems to be greater when introduced to BBdp rats at less than 30 d of age. Indeed, we have previously demonstrated significant differences in small intestine weight, colon weight, glucose transporter mRNA, and proglucagon mRNA abundance between BBn and BBdp rats fed chow diets as early as 21 d(23). A similar longitudinal study of BBn and BBdp animals fed different diets might provide valuable insight into possible interactions between diet and intestinal ontogeny and the onset and incidence of diabetes mellitus.

Third, rats consuming chow diets usually have faster rates of growth than those fed semipurified diets(29,30). Chow diets tend to be less energy dense than casein- and soy-based diets, suggesting that chow-fed rats must eat more food or absorb nutrients more efficiently. The greater intestinal weight and increased glucose transporter mRNA abundance of BBn rats fed chow implies a greater potential to absorb nutrients. The majority of the energy in the chow diet was provided by carbohydrate whereas the soy- and casein-based diets were much higher in energy derived from fat. In BBdp rats, soy- and casein-weaning diets have been shown to confer more protection against diabetes incidence than chow(3,4). The chow diet was associated with larger small intestinal weight and colonic weights per body weight in BBdp than in BBn animals. Intestinal mass, however, remained similar between BBn and BBdp animals fed casein and soy diets. Differences in intestinal mass between BBn and BBdp animals weaned onto different diets may have important effects on the gut-associated immune system and nutrient transport. Our findings suggest that BBdp animals weaned onto chow diets may have a larger intestine and greater capacity for nutrient absorption.

Proglucagon is produced in the distal ileum and colon and is the precursor for GLP-1, a potent insulin secretagogue(20). Proglucagon mRNA results in the production of enteroglucagon, thought to be trophic to the gut(31); GLP-1, a potent insulin secretagogue(32,33); and GLP-2, which increases glucose transport in the jejunum(22). Because prohormone convertase PCl is solely responsible for the cleavage of proglucagon into its respective posttranslational products(34), one would expect concomitant increased secretion of all intestinal proglucagon-derived peptides theoretically resulting in increased intestinal transport capacity and a greater stimulus for insulin secretion. Feeding a chow diet tended to increase proglucagon expression in the BBdp, whereas casein and soy did not. Adjusting for total colonic RNA removed the effect of disease susceptibility in chow-fed animals whereas soy- and casein-fed BBdp had significantly less total proglucagon mRNA than did their BBn counterparts. The lower proglucagon mRNA found in BBdp animals relative to BBn controls fed soy and casein but not chow suggests that BBdp rats fed semipurified diets may be subjected to less pancreatic challenge immediately after weaning. The greatly reduced level of proglucagon in BBdp animals compared with BBn was unexpected and may signal a defect in the gene machinery of the diabetes-prone animal.

In summary we have shown that weanling diets with differing diabetogenic potential altered intestinal growth and hormone expression in the BB rat. In normal rats, the ingestion of chow diets was associated with greater intestinal growth and increased abundance of glucose transporter mRNA relative to semipurified diets. BBdp animals fed chow responded differently than those fed casein and soy compared with their respective BBn groups. In fact, a lack of intestinal response to diet in the BBdp rat on many parameters, including the relative abundance of SGLT-1, GLUT5, and proglucagon mRNA and the mass of the stomach, small intestine, colon weight per body weight, and overall body weight, may contribute to the diet-related risk of diabetes mellitus in the BBdp rat. This study has implications not only for the understanding of intestinal adaptation in the diabetes prone rat but also suggests that diet significantly influences the development of the normal rat intestine.