The synthesis of pancreatic enzymes and their secretion into the duodenum are essential for digestion of dietary macromolecules and absorption of nutrients. Important digestive changes occur during the early postnatal development. In pigs, the activities of most proteolytic enzymes such as trypsin and chymotrypsin are evident during the fetal period (Weström et al. 1987) despite their low activities at birth. After birth, the activities of the different pancreatic proteases develop divergently, suggesting independent regulation of each pancreatic protein. Moreover, the level of pancreatic hydrolases may be modulated by the composition of the diet. In particular, the activities of proteolytic enzymes are modified either by the protein level and the source of protein in the diet or by feed intake (Efird et al. 1982, Owsley et al. 1986, Makkink et al. 1994a and 1994b, Peiniau et al. 1996). Information about regulation of pancreatic function during postnatal development and during the course of dietary adaptation is scarce. Changes in specific pancreatic mRNA concentrations have been reported in several mammalian species (Giorgi et al. 1985, Lhoste et al. 1993, Le Huërou et al. 1990, Wicker et al. 1984), but these studies are not sufficient to assess the mechanisms involved in the regulation of a given enzyme.

Compared with knowledge of trypsin and chymotrypsin, little information is available on the evolution of expression of pancreatic elastases I and II (Weström et al. 1987). Moreover, many experimental results seem contradictory because of a possible confusion in the determination of elastase II activity. Both enzymes have a unique specificity in hydrolyzing elastin, a fibrous and insoluble protein, but differ greatly in the type of synthetic substrates that they are able to hydrolyse. Elastase I action is specifically limited to Ala-Ala and Ala-Gly bonds (Gertler et al. 1977). Elastase II, which is closely related to the chymotrypsin family (Largman et al. 1976), exhibits a broad specificity for substrates containing medium to large hydrophobic amino acids in the P1 position (DelMar et al. 1980). The specificities of elastases I and II are complementary to those of trypsin and chymotrypsin. Therefore, these enzymes have an obvious role in protein digestion in the intestine. For instance, purified elastase I and particularly elastase II efficiently hydrolyze bovine -lactalbumin (Jakobsson et al. 1983) and -lactoglobulin (Gestin et al. 1997, Jakobsson et al. 1983). In early infancy, trypsin and chymotrypsin are well developed, whereas low amounts of pancreatic elastase II might be present (Borulf and Lindberg 1981). Moreover, the elastase I gene is not expressed in the human pancreas (Tani et al. 1987). Inadequate proteolytic activity in the gut has been proposed as one of several causes of cow's milk protein intolerance in human infants.

For ethical and practical reasons, experiments on human subjects are difficult, and pigs are generally considered a suitable model for studying both human nutrition and gastrointestinal function. Therefore the aim of this study was to determine the influence of age and diet on elastase I and II gene expression (activity and mRNA) in pigs. In particular, the expression of elastases I and II in relation to the source (cow vs. sow) and the nature of protein (animal vs. plant) in the diet was investigated. The changes in specific mRNA levels were analyzed by a reverse transcriptase-polymerase chain reaction (RT-PCR) method.

MATERIALS AND METHODS

Forty-two crossbred Pietrain × Large White piglets from six litters born the same week in a batch of sows of the experimental herd of INRA were used in the experiment. Piglets were assigned to seven groups (Groups 1 to 7) of six animals each, according to their age and dietary treatments. Group 1 consisted of newborn unsuckled piglets. They were slaughtered within 1 h of birth. The piglets of Groups 2 and 3 were suckled by their mothers until slaughter at d 13 and 21, respectively. The piglets of Groups 4 and 7 were suckled by the sows and then removed from the sow on d 14 and fed with free access to pelleted milk substitute until slaughter on d 21 and 56, respectively. The piglets of Group 5 were suckled until d 21 and then fed a dry feed starter corresponding successively to "first age" from d 22 to 42 and then to "second age" from d 43, supplying starch and plant protein and offered ad libitum until slaughter at d 56. The piglets of Group 6 were first suckled by the sows during the 0- to 13-d period and then fed pelleted milk substitute during the 14- to 21-d period and finally given starter feeds until slaughter at d 56. The piglets of Groups 2 to 7 were slaughtered 16-17 h after their last meal. The piglets of Groups 3 and 4 were used to compare the effect of artificial cow's milk to that of sow's milk on the subsequent development of the digestive tract. The piglets of Group 5 were considered a control group because their weaning age and the food offered corresponded to the usual conditions of pig breeding. The effects of age at weaning and of the nature of the diet were analyzed by comparing the piglets of Groups 5, 6 and 7.

The composition and average amount of sow milk suckled by the piglets were estimated using average data of Salmon-Legagneur and Aumaitre (1962) on milk production, whereas the compositions of the dry milk substitute and starter feeds were precisely analyzed (Table 1). Milk substitute was based on spray-dried and refatted cow skim milk. During the first and second weeks of lactation, sow milk contained ~21-27% and 77-81% more protein and fat on a dry matter basis, respectively, and 67-68% less lactose than the cow milk substitute used. The gross energy was slightly higher, i.e., 22.7-24.3 versus 21.8 MJ/kg dry matter, in the sow milk. Compared with cow milk substitute, the protein concentration was slightly lower (9-13%) in the starter feed, and most importantly, the source of protein in this latter was supplied by plant protein from wheat and soybean meal. Energy was supplied mostly by fat and lactose in the milk substitute and by cornstarch in the starter feed.

Table 1. Composition and nutrient content of the experimental diets [View Table]

The RT-PCR method was used to quantify elastase I and II mRNA levels in the pancreas. For reverse transcription, 1 µg of total RNA was heated for 10 min at 70°C in the presence of oligo dT (0.5 µg) and then incubated at 37°C for 1 h in a 20-µL reaction volume consisting of reverse transcriptase buffer (50 mmol/L Tris-HCl, pH 8.3, 75 mmol/L KCl, 3 mmol/L MgCl₂), 10 mmol/L dithiothreitol (DTT), 0.5 mmol/L dNTP and 200 U Superscript RNase H reverse transcriptase (18053-017, Gibco-BRL, Gaithersburg, MD). Before amplification, RT products used for PCR were diluted or not according to the quantity of mRNAs in each sample, which must be in the range of the standard. For elastase I, the sense primer P1 5-CCAGAACGATGGCACCGAG-3 and the antisense primer P2 5-GGGCAGGTAAGCCTGCTG-3 were analogous to nucleotides (nt) 279-297 and the complement to nt 526-543 of the cDNA, and yielded an expected amplicon of 265 bp. For elastase II, the sense primer P3 5-GGCTGGAGC(CT)CTCAG(CT)TG(CT)GGG-3 and the antisense primer P4 5-CACCGAATTGATCCAGTC-3 were analogous to nt 42-62 and the complement to nt 787-804 of the cDNA, and yielded an expected amplicon of 763 bp. Primer pair P1-P2 amplified a fragment that crossed introns 4 and 5 (Swift et al. 1984, Tani et al. 1987) and primer pair P3-P4 amplified a fragment that crossed introns 2 to 6 (Swift et al. 1984) of the rat and human corresponding genes, thereby allowing the detection of contaminating genomic DNA. The PCR reaction was conducted in a total volume of 50 µL, which was composed of PCR buffer (10 mmol/L Tris-HCl, pH 8.3, 1.5 mmol/L MgCl₂, 50 mmol/L KCl, 0.1 g/L Tween 20), 0.2 mmol/L dNTP, 1 mmol/L Primers, 1 U Taq Polymerase (ME-0060, Eurogentec, Seraing, Belgium). The amplification reaction involved denaturation at 94°C for 1 min followed by cycling as follows: 94°C for 1 min, primer annealing at 60° and 54°C for elastases I and II, respectively, for 2 min, and extension at 70°C for 2 min. After cycling, a terminal elongation of 10 min at 70°C was performed. Amplification required 31 and 32 cycles for elastase I and II mRNAs, respectively.

RESULTS

Table 2. Daily nutrient intake and gross energy available during the week before slaughter of piglets that were not fed and slaughtered at birth (Group 1), suckled by their dam up at 13 d (Group 2) or 21 d (Group 3), fed a cow milk substitute from 14 to 21 d (Group 4) or to 56 d (Group 7), suckled up to 21 d and then fed a dry starter feed up to 56 d (Group 5) or fed a cow milk substitute from 14 to 21 d and then fed a dry starter feed up to 56 d (Group 6)1,2 [View Table]

Table 3. Pancreas weight, protein concentration and elastase I and II activities in piglets that were not fed and slaughtered at birth (Group 1), suckled by their dam up to 13 d (Group 2) or 21 d (Group 3), fed a cow milk substitute from 14 to 21 d (Group 4) or to 56 d (Group 7), suckled up to 21 d and then fed a dry starter feed up to 56 d (Group 5) or fed a cow milk substitute from 14 to 21 d and then fed a dry starter feed up to 56 d (Group 6)1,2 [View Table]

Compared with suckled piglets (Group 3), the dry matter and gross energy intakes of piglets receiving milk substitute (Group 4) were 30 and 44% lower, respectively (Table 2). As a consequence, their live weight and weight gain tended to be lower (P = 0.08, Table 3). In contrast, the weight of their pancreas and the level of pancreatic protein did not differ.

At 56 d, dry matter, protein and gross energy intake were not consistently modified when piglets of Groups 5 and 6 were compared (Table 2). Fat and starch intakes were significantly higher in Group 6 than in Group 5. No variations in live weight, pancreas weight or protein content were observed (Table 3). In contrast, the dry matter intake of piglets continuously receiving cow milk substitute (Group 7) tended to be lower (22-31%, P = 0.07), and their fat intake was 100% higher than those receiving starter feed (Groups 5 and 6). Live weight and protein content were significantly greater in Group 7 than in Groups 5 and 6 (Table 3).

The levels of elastase I and II mRNAs were significantly affected by age (Fig. 2). They were minimum at birth, were 3·10³ and 26·10³ times greater at d 13, respectively, and did not change significantly thereafter. The copy numbers of elastase I and II specific mRNAs were sharply higher in 13- and 21-d-old piglets than in newborns (5·10⁶-7·10⁶ vs. 2·10³ copies/µg RNA for elastase I, 3·10⁶-12·10⁶ vs. 0.5·10³ copies/µg RNA for elastase II, respectively). In weaned 56-d-old piglets (Group 5), the elastase II mRNA amount reached 21·10⁶ copies/µg RNA. In newborns and 21-d-old suckled piglets, the copy numbers of elastase II mRNAs were 75 and 32% lower (P < 0.05) than those of elastase I mRNAs, whereas in 13-d-old suckled piglets, it was 89% higher (P < 0.01). In weaned piglets, the elastase II mRNA copy number was 280% greater (significant) than that of elastase I mRNA.

DISCUSSION

The postnatal development of pancreatic proteases, particularly elastases I and II, was dependent on the age of piglets and the source of dietary protein used in the weaning diet. The expression of each gene in the pig pancreas was independently regulated, leading to various gene expression profiles depending on the enzyme, age and weaning. Taken together, our results provide evidence of the early expression of elastase I and II genes in the pig pancreas, expression which could improve protein digestion. The piglet has often been considered as an animal model for human nutrition and digestion in the early stages of life (Dodds 1982, Miller and Ullrey 1987). The present results could be considered as a basis to better understand the digestive changes occuring in the gut of the infant. However, digestive changes must be ascertained with specific assays in human early life.

ACKNOWLEDGMENTS