Main

Dietary lipid is a major nutrient in the developing mammal. Crucial processes in the absorption and assimilation of dietary lipid are intestinal mucosal uptake of the products of lipid digestion, reesterification of these products into complex lipids, packaging of these lipids into lipoprotein particles with apolipoproteins, and secretion of the nascent lipoprotein particles for peripheral metabolism. The predominant lipoproteins produced by the small intestine are chylomicrons, VLDL, and HDL. Understanding the mechanisms of regulation of these processes by dietary and other factors contributes to the development of strategies for the optimal provision of dietary lipid to the developing neonate. Of particular importance is the regulation of intestinal apolipoproteins expression by dietary lipid absorption.

Apolipoproteins are lipid-binding peptides which participate in the assembly, secretion, peripheral metabolism, and receptor-mediated uptake of lipoprotein particles. Apo B is a component of triglyceride-rich lipoproteins and exists as two distinct forms in the human, rat, and swine(1–3). These two forms of apo B have been named on a centile system based upon their relative molecular weights. Apo B-100 is the larger species and is a component of plasma VLDL of hepatic origin and LDL and contains the LDL receptor binding domain(2). Apo B-48 is the smaller form and is found in intestinal chylomicrons and VLDL and does not contain the LDL receptor binding sequence(2). Both forms of apo B are derived from the same gene by a unique posttranscriptional mRNA editing mechanism(4, 5). In the postnatal human and swine, apo B-100 is of hepatic origin, and apo B-48 is synthesized by the intestine(3, 6, 7). However, in fetal intestine, apo B-100 is also produced(8–10). In both liver and intestine, apo B is thought to be essential for the assembly and secretion of triglyceriderich lipoproteins and does not dissociate from the particle of origin after secretion(2).

Apo A-I is the major apolipoprotein of plasma HDL and is produced by both liver and intestine in the human, rat, and swine(3, 6, 11, 12). After secretion from the small intestine apo A-I dissociates from chylomicrons and binds to HDL(13). The major metabolic role of apo A-I is that of an essential cofactor for lecithin:cholesterol acyltransferase, the enzyme responsible for the production of cholesteryl ester in circulating HDL(14).

Our laboratory has focused on study of the regulation of small intestinal apolipoprotein gene expression in the developing mammal using the neonatal swine as a model similar in many ways to the human infant(15). We have demonstrated that absorption of a LCPUT emulsion acutely (over 24 h) up-regulates small intestinal apolipoprotein expression in the newborn piglet when compared with an isocaloric low fat diet(6, 16). However, the patterns of regulation of apo B and A-I expression differ in this model. Apo B expression is regulated at the posttranslational level, and apo A-I expression is regulated at the translational level(6, 17). Furthermore, there appears to be an enhanced regulatory responsiveness to dietary lipid absorption in the immediate newborn period compared with the older suckling animal(3, 6). These previous studies were designed to determine the regulatory effects of the presence or absence of lipid absorption on apolipoprotein expression. However, an important question is whether the specific type of lipid (chain length, degree of saturation) can differentially regulate apolipoprotein expression in the neonatal small intestine. Therefore, in the present study, newborn swine were provided 24-h duodenal infusions of nutritionally complete isocaloric formulas with the same lipid content, but differing in fatty acid composition.

METHODS

Animals. Two-day-old female domestic swine were obtained from Cargill Swine Project, Russellville, AR. From the time of arrival to the time of surgery the next day, the animals were kept in heated isolettes. Animals were fed artificial sow's milk (SPF-LAC, Pet-Ag, Inc., Hampshire, IL) by gavage during this period. This protocol was approved by the University of Arkansas for Medical Sciences Institutional Animal Care and Use Committee.

Preparation of the animal model and experimental design. Anesthesia was induced and maintained by face mask delivering 1 L/min O2 and 0.5-0.8% halothane. A longitudinal midline abdominal incision was made, and the peritoneal cavity was opened. The second portion of the duodenum was cannulated with silicone rubber tubing (inside diameter = 0.030 inch, outside diameter = 0.065 inch) through an incision in the gastric antrum. The tubing was then secured with a purse-string suture, exteriorized through the right flank, tunneled s.c. to the mid-dorsum, and secured through a swivel tether. Postoperatively the animals were allowed to recover in the heated isolettes. An intraduodenal infusion of 5% glucose in 45 mM NaCl and 20 mM KCl at 100 mL/kg/24 h was started during recovery. After a 24-h recovery period, animals were awake, alert; and mobile. Also at this time animals were noted to be tolerating the infusion well without vomiting or abdominal distention and to be passing stool. Experimental infusions were started at the end of this recovery period.

Experimental formulas consisted of nonfat dry milk (Carnation) reconstituted and blended with lipid and infused to provide 120 kcal/kg/24 h with 48% of calories as fat. Three experimental groups were studied:1) MCT: MCT oil (Mead Johnson, Evansville, IN) (predominantly 8:0 and 10:0 fatty acids) (n = 7); 2) ICST: coconut oil (ICN, Aurora, OH) (predominantly 12:0 and 14:0 fatty acids) (n = 8); and3) LCPUT: safflower oil (21st Century Foods, Uniondale, NY)(predominantly 18:2 and 18:1 fatty acids) (n = 8). The fatty acid composition of these infusions is shown in Table 1. Fatty acid composition was determined as previously described(18). Briefly, the fatty acids were extracted, transmethylated, and subjected to capillary gas chromatography(Hewlett-Packard, Avondale, PA) using a 30 m × 0.32-mm Omegawax-320 column (Supelco, Bellefonte, PA) with a flame ionization detector and a splitless injector. The oven temperature was programmed from 80-260°C at 5°C/min. The injector port and detector temperature was 220°C. Fatty acid methyl esters were identified by a comparison of their retention times to those of commercial fatty acid methyl ester standards. Aliquots of the formulas were also extracted and subjected to thin-layer chromatography on silica gel G plates using petroleum etherdiethyl ether-acetic acid 80:20:1(vol/vol). Lipid bands were identified by exposure to iodine vapor and comparison to co-chromatographed standards. The lipid from all three formulas consisted of greater than 99% triglyceride.

Table 1 Fatty acid composition of experimental diets as determined by gas chromatography
Full size table

The stability of the infusates was maintained by constant stirring. The stability was verified by visible appearance and taking aliquots at the beginning and end of the infusion from the top and bottom of the container, as well as from the site of entry of the catheter into the tether. These samples were subjected to triglyceride measurement, and the infusions were considered acceptable only if there was less than 5% variation among measurements from these sites.

All groups received a total fluid volume of 120 mL/kg/24 h during the experimental infusions. During the infusions the animals were kept in the heated isolettes and allowed to move freely within the limits of their tethers. At no time during the recovery phase or during the experimental infusions did the animals receive peroral feedings.

Determination of intestinal apolipoprotein synthesis. At the end of the experimental infusions the animals were anesthetized, and a 10-cm segment of proximal jejunum was isolated 10 cm distal to the ligament of Treitz by two ligatures. Likewise, a 10-cm segment of distal ileum was isolated 10 cm from the ileocecal valve. Radiolabeling was performed by instilling 1.0 mCi of L-[4,5-3H]leucine (>120 Ci/mmol) (Amersham Corp., Arlington Heights, IL) into each segment. Nine minutes later the segments were removed and prepared for immunoprecipitation as described below. This labeling time has previously been shown to be optimal in similar experiments in the adult rat, and we have used it previously in the piglet(3, 6, 12, 16, 19, 20). Adjacent segments of both jejunum and ileum were removed for homogenization and processing for determination of apolipoprotein mass and triglyceride content.

Preparation of mucosal cytosolic supernatants for immunoprecipitation. Radiolabeled intestinal segments were flushed with 50 mL of iced PBS (50 mM phosphate, 100 mM NaCl, pH 7.4)-20 mM leucine, and the mucosa was scraped and homogenized on ice in 1 mL of PBS-1% Triton X-100-2 mM leucine-1 mM phenylmethylsulfonyl fluoride-1 mM benzamidine, pH 7.4, as previously described(3). Aliquots of the homogenate were taken for measurement of total protein concentration, TCA-precipitable radioactivity, and triglyceride content. The remainder was pelleted at 105,000× g for 65 min in a 50.3 Ti rotor (Beckman Instruments, Palo Alto, CA), followed by collection of the cytosolic supernatant. The nonradiolabeled segments were similarly prepared for apolipoprotein mass determination. Although most intracellular apolipoprotein is membrane-associated, this technique has been shown previously to result in extraction and solubilization of 84-94% of total recoverable apolipoprotein mass with no discernible effect of the state of lipid flux(19). All procedures were performed at 0-5°C, and the mucosal supernatant samples were stored at -80°C until analysis.

Apolipoprotein immunoprecipitation. Intestinal cytosolic supernatant fractions were subjected to specific immunoprecipitation of apo B and apo A-I under conditions of antibody excess as described previously(3, 6, 19). Aliquots of cytosolic supernatants were preincubated with washed IgG-SORB (The Enzyme Center, Malden, MA) and subsequently reacted with excess rabbit polyclonal anti-swine apolipoprotein antiserum for 18 h at 4°C. After a second addition of IgG-SORB and extensive washing, the liberated immunocomplex was applied to either 5.6% (apo A-I) or 4% (apo B) SDS polyacrylamide tube gels. After electrophoresis gels were sliced into 2-mm slices and solubilized in Solvable(DuPont, Boston, MA) at 50°C for 3 h followed by the addition of Ultima Gold scintillation fluid (Packard, Meriden, CT) and incubation overnight at room temperature. Liquid scintillation counting was performed in a Packard model 2000 liquid scintillation counter (Packard, Downers Grove, IL). Apolipoprotein species were identified by comparison with stained co-electrophoresed apolipoprotein. Apolipoprotein synthesis rates were expressed as the percentage of specific immunoprecipitated apoprotein counts compared with total protein TCA-precipitable counts. Apolipoprotein synthesis was thereby expressed as a percentage of total protein synthesis.

Apolipoprotein quantitation by ELISA. ELISA assays for swine apo B and apo A-I used polyclonal antibodies produced in New Zealand White rabbits after injection of electrophoretically pure pig apo B-100 from plasma LDL and apo A-I from plasma HDL. These assays were performed as described previously(6). Briefly, 96-well microtiter plates(Falcon, Becton Dickinson, Oxnard, CA) were coated with either purified swine plasma LDL (apo B assay) or HDL (apo A-I assay) at 100 ng of protein/well. Competition assays were performed using serial dilutions of cytosolic supernatants prepared from the tissue homogenates as the competing antigen. After incubation with biotinylated goat anti-rabbit IgG (Vector Laboratories, Burlingame, CA) followed by avidin-biotinylated peroxidase complex(Vectastain, Vector Laboratories), color was developed by addition ofo-phenylenediamine. Color development was stopped by the addition of 2 M sulfuric acid. Standard antigens consisted of swine plasma LDL (apo B) and HDL (apo A-I) applied to the microtiter plate wells in serial dilutions. Assay plates were read at 492 nm by a Bio-Rad ELISA reader (Bio-Rad, Richmond, CA). All samples were run in triplicate in the same assay. Intraassay variability was less than 5%, and interassay variability was less than 10%.

Protein and triglyceride measurement. Protein was measured using the Bradford(21) technique. Total lipid was extracted from aliquots of the mucosal homogenates by the method of Folchet al.(22), followed by triglyceride measurement using an enzymatic assay (Sigma Chemical Co., St. Louis, MO).

Statistical analysis. Data in experimental groups were analyzed by one-way ANOVA followed by the Fisher least significant difference test to compare specific groups. Correlations among variables were analyzed by linear regression. The null hypothesis was rejected at p ≤ 0.05.

RESULTS

There were no significant differences in the total protein specific activities of radiolabeled mucosal homogenates among the three experimental groups as determined by TCA precipitation and protein measurement. Figure 1 shows jejunal and ileal apo B synthesis for the three groups ( Fig. 1, top panel), expressed as percent of total protein synthesis, and apo B mucosal mass( Fig. 1, bottom panel), expressed as nanograms of apo B/μg total protein. In jejunum, only the animals receiving the ICST diet had significantly different apo B synthesis, which was approximately 50% less than the other two groups. However, jejunal apo B mass was significantly different only in the MCT group, which was lower than in the ICST group. In ileum, the pattern of apo B synthesis was similar to that observed in the jejunum with the lowest synthesis in the ICST group. There were no significant differences in ileal apo B mass among the three groups. Ileal synthesis rates were lower than jejunal, indicating a proximal to distal synthetic gradient. As expected, no correlation of jejunal apo B synthesis with mass was found among all groups. Additionally, no significant correlations were found in the ileum (not shown).

Figure 1
figure 1

Newborn piglet jejunal and ileal apo B synthesis(top panel) and mucosal content (bottom panel) in the three experimental groups of animals. Apolipoprotein synthesis is expressed as percentage of total protein synthesis as determined by specific immunoprecipitated apoprotein dpm/TCA-precipitable dpm × 100. Mucosal apoprotein content was measured by ELISA and expressed as ng of apoprotein/μg of total protein. Bars represent mean ± SEM. The p value determined by ANOVA is shown on the horizontal lines placed over each group of bar graphs. Individual bars sharing the same letters on top are not different, whereas those not sharing the same letters are different (p ≤ 0.05). MCT group (n = 7), ICST group(n = 8), and LCPUT group (n = 8).

Full size image

Figure 2 shows jejunal and ileal apo A-I synthesis( Fig. 2, top panel) and mass( Fig. 2, bottom panel) for the three groups. In the jejunum, the group receiving the LCPUT diet had significantly different synthesis, which was approximately twice as high as that of the other two groups. Jejunal apo A-I mass was not different among the three experimental groups. In the ileum there were no differences in synthesis, but apo A-I mass was significantly higher in the LCPUT group. As with apo B, synthesis rates were lower in ileum than in jejunum. There was a significant linear correlation of jejunal apo A-I synthesis with mucosal mass across all groups (Fig. 3). Within individual groups, significant correlations of jejunal apo A-I synthesis with mass were found in the MCT and ICST groups. In the ileum, there was a weak, but statistically significant, correlation of apo A-I synthesis with mass across all groups and within the MCT group (Fig. 4).

Figure 2
figure 2

Newborn piglet jejunal and ileal apo A-I synthesis(top panel) and mucosal content (bottom panel) in the three experimental groups of animals. Apolipoprotein synthesis is expressed as percentage of total protein synthesis as determined by specific immunoprecipitated apoprotein dpm/TCA-precipitable dpm × 100. Mucosal apoprotein content was measured by ELISA and expressed as nanograms of apoprotein/μg total protein. Bars represent mean ± SEM. The p value determined by ANOVA is shown on the horizontal lines placed over each group of bar graphs. Individual bars sharing the same letters on top are not different, whereas those not sharing the same letters are different (p ≤ 0.05). MCT group (n = 7), ICST(n = 8), and LCPUT group (n = 8).

Full size image
Figure 3
figure 3

Correlation of jejunal apo A-I synthesis with mucosal apo A-I mass. Within individual groups the following correlations were found: MCT, r = 0.81, p < 0.05; ICST, r = 0.81,p < 0.02; LCPUT, r = 0.30, p = NS. MCT group(n = 7), ICST group (n = 8), and LCPUT (n = 8).

Full size image
Figure 4
figure 4

Correlation of ileal apo A-I synthesis with mucosal apo A-I mass. Within individual groups the following correlations were found: MCT,r = 0.83, p < 0.02; ICST, r = 0.65,p = NS; LCPUT, r = 0.50, p = NS. MCT(n = 7), ICST group (n = 8), and LCPUT group (n= 8).

Full size image

Jejunal and ileal mucosal triglyceride content in the three experimental groups were as follows (micrograms of triglyceride/mg of total protein, mean± SEM): jejunum, MCT, 94.6 ± 30.9; ICST, 146 ± 45; LCPUT, 71.1 ± 10.7; ileum, MCT, 59.2 ± 31.8; ICST, 15.8 ± 2.4; LCPUT, 57.5 ± 23.3. Although the highest jejunal triglyceride content was found in the ICST group and the lowest in the LCPUT group, these differences did not reach statistical significance. No significant correlations of apolipoprotein synthesis or mass with mucosal triglyceride levels were observed in the jejunum. In the ileum, the lowest triglyceride content was found in the ICST group, although this difference did not reach statistical significance. Apo B synthesis did not correlate with triglyceride levels in the ileum (not shown). However, apo B mass correlated with ileal triglyceride levels across all groups and within the MCT group (Fig. 5). Similarly, apo A-I synthesis did not correlate with ileal triglyceride content (not shown), whereas apo A-I mass did significantly correlate with ileal triglyceride levels across all groups and within the MCT group (Fig. 6). In all groups, jejunal triglyceride content was higher than that in the ileum.

Figure 5
figure 5

Correlation of ileal apo B mass with mucosal triglyceride content. Within individual groups the following correlations were found: MCT, r = 0.90, p < 0.05; ICST, r = 0.04, p = NS; LCPUT, r = 0.73, p = NS. MCT group(n = 5), ICST group (n = 8), and LCPUT group (n= 8).

Full size image
Figure 6
figure 6

Correlation of ileal apo A-I mass with mucosal triglyceride content. Within individual groups the following correlations were found: MCT, r = 0.99, p < 0.002; ICST, r = 0.10, p = NS; LCPUT, r = 0.66, p = NS. MCT group(n = 5), ICST group (n = 8), and LCPUT group (n= 8).

Full size image

DISCUSSION

The biogenesis of intestinal lipoproteins in response to the absorption of dietary lipid is a complex process. This process involves the uptake of the products of lipid digestion, reesterification of these products into complex lipid, packaging with apolipoprotein, and basolateral secretion of nascent lipoproteins by reverse pinocytosis. The regulation of these processes by absorption of dietary lipid in the neonatal period has not been extensively studied, although extensive data are available from adult animal studies. We have developed the neonatal swine as a model for such studies. The developing piglet is similar in many ways to the human infant with regard to the timing and sequence of development of gastrointestinal function and lipoprotein metabolism(11, 15). We have previously described patterns of intestinal apolipoprotein expression with and without the acute(24 h) absorption of long-chain polyunsaturated fatty acids in the newborn piglet(6, 16). In this report, we have described the short-term regulation of apo B and A-I synthesis and mucosal mass by nutritionally complete formulas varying only in the type of fatty acid present in the lipid component.

We hypothesized that, in this study, fatty acid chain length would most likely be the main determinant of the magnitude of regulation of apolipoprotein expression. We had previously found long-chain polyunsaturated lipid absorption to regulate apo A-I at the translational level (synthesis) and apo B at the posttranslational level (mass). Because medium-chain fatty acids bypass incorporation into lipoproteins and are absorbed directly into the portal circulation, the magnitude of apolipoprotein expression was expected to be lowest in the MCT group. The LCPUT group was expected to have the highest level of expression, and the ICST group to be intermediate.

For apo B in jejunum, we found that synthesis was actually lowest in the ICST group, and the MCT and LCPUT groups were comparable. However, jejunal apo B mass was highest in the ICST group. There are several possible explanations for these findings. Because we previously demonstrated posttranslational (apolipoprotein mass) regulation for jejunal apo B by long-chain polyunsaturated lipid absorption with synthesis remaining constant with or without lipid absorption, the present pattern in the MCT and LCPUT groups was as expected with comparable synthesis in the two groups, but lower mucosal mass in the MCT group. However, it is important to note that mucosal mass does not reflect total apo B production, but rather is the net product of production, intracellular degradation, and secretion from the enterocyte. Also, as demonstrated in the adult rat, there appear to be significant intracellular pools of apo B and A-I mainly unassociated with lipoprotein particles(23). With lipid absorption, a small portion of these pools becomes lipoprotein-associated(23).

The lowest jejunal apo B synthesis in the ICST group in our study may be due to the almost exclusive presence of saturated fatty acids (predominantly 12:0) in coconut oil, which, unlike medium-chain fatty acids, do enter the lipoprotein synthetic and secretory pathways. Although the human colon carcinoma cell line, CaCo-2, is a very different system in comparison to ourin vivo swine model, this cell line does have characteristics of immature intestinal epithelial cells, and certain observations in these cells appear to parallel those in the present study. In the CaCo-2 cell line, the saturated long-chain palmitic acid (16:0) induces significantly less triglyceride and apo B secretion than linoleic acid (18:2)(24). Field et al.(25) found that complex lipids containing absorbed 14:0 and 16:0 fatty acids were less efficiently secreted by CaCo-2 cells than those containing absorbed 18:1, 18:2, and 18:3 fatty acids. In CaCo-2 cells exposed to palmitic acid there was more incorporation of fatty acid into phospholipid than triglyceride(24). However, intracellular apo B synthesis and content were not measured in these CaCo-2 cell studies. We speculate that one possible explanation for the relatively high apo B and triglyceride content and low apo B synthesis in the ICST group in our study might be the presence of a relative secretory block with feedback inhibition of apo B synthesis caused by these fatty acids. In vitro studies are underway to test this hypothesis.

Inhibition of microsomal degradation of apo B is also likely to be a major posttranslational regulatory mechanism. In hepatocytes, oleic acid decreases apo B degradation with a concomitant increase in secretion. Also, Murthyet al.(26) demonstrated a similar mechanism in CaCo-2 cells after exposure to oleic acid. That there is a significant intracellular pool of apo B with a slow turnover is supported by the observation of enhanced mass secretion of apo B in oleic acid-treated CaCo-2 cells in the absence of change in secretion of radiolabeled apo B(26). Therefore, the dissociation of apo B jejunal mucosal mass from synthesis in our study may be due to independent regulation of apo B synthesis, degradation, and secretion. Interestingly, jejunal triglyceride content was highest in the ICST group (although not reaching statistical significance), suggesting retention of triglyceride in the mucosa. It is possible that some or all of these suggested mechanisms for the observations in the ICST group may be secondary to changes in microsomal fluidity with incorporation of saturated fatty acids into membrane phospholipids.

The pattern of regulation of jejunal apo A-I in this study differs from that of apo B. Apo A-I synthesis was highest in the LCPUT group and approximately twice that of the MCT and ICST groups. This pattern would generally fit our hypothesis, predicting the greatest magnitude of regulation with long-chain fatty acids. No significant differences in jejunal apo A-I content were found among all three groups. However, as mentioned before, this parameter reflects the net effect of both production and secretion and does not quantitate total apolipoprotein production.

Apo B synthesis in the ileum among all three groups of animals in our study generally followed the same pattern as in jejunum. However, as we had observed previously(6), ileal synthesis was lower than jejunal in all three groups. Ileal apo B mass was not different among groups and was comparable to that found in the jejunum. Interestingly, ileal apo B mass in the MCT group highly correlated with triglyceride content. However, because presumably medium-chain fatty acids would not be reesterified even if taken up by the ileum, the reason for this correlation is not apparent and may be related to endogenous ileal triglyceride metabolism. As was the case for apo B, ileal apo A-I synthesis was lower than jejunal with no significant differences among groups. However, ileal apo A-I content was highest in the LCPUT group, which would be indicative of posttranslational regulation. Among all groups and within the MCT group, ileal apo A-I synthesis significantly correlated with apo A-I mass, as was observed in the jejunum. Similar to the observation for apo B, ileal apo A-I mass correlated significantly with triglyceride content in the MCT group. It is likely that ileal apolipoprotein expression is regulated by factors other than dietary fatty acid flux, because the majority of dietary lipid would be expected to be absorbed proximal to the terminal ileum. Other regulatory factors might include bile acids and gastrointestinal hormones released proximally. We previously found that duodenal infusion of long-chain polyunsaturated lipid induced posttranslational regulation of both apo B and A-I in the terminal ileum of newborn piglets compared with a low fat isocaloric infusion(6).

In summary, this study defines the patterns of small intestinal apo B and A-I synthesis in the newborn piglet, a unique model for the newborn human, receiving a nutritionally complete formula with varying fatty acid composition by intraduodenal infusion. Intestinal apo B and A-I expression is acutely and differentially regulated by dietary lipid varying in fatty acid chain length and saturation. Patterns of regulation are complex and vary among specific apolipoprotein, as well as location (proximal versus distal intestine), and these patterns include translational and posttranslational mechanisms. Studies are underway in vitro to define these regulatory mechanisms at the cellular level.