Chylomicrons promote intestinal absorption and systemic dissemination of dietary antigen (ovalbumin) in mice

Abstract

Background: A small fraction of dietary protein survives enzymatic degradation and is absorbed in potentially antigenic form. This can trigger inflammatory responses in patients with celiac disease or food allergies, but typically induces systemic immunological tolerance (oral tolerance). At present it is not clear how dietary antigens are absorbed. Most food staples, including those with common antigens such as peanuts, eggs, and milk, contain long-chain triglycerides (LCT), which stimulate mesenteric lymph flux and postprandial transport of chylomicrons through mesenteric lymph nodes (MLN) and blood. Most dietary antigens, like ovalbumin (OVA), are emulsifiers, predicting affinity for chylomicrons. We hypothesized that chylomicron formation promotes intestinal absorption and systemic dissemination of dietary antigens.

Methodology/principal findings: Absorption of OVA into MLN and blood was significantly enhanced when OVA was gavaged into fasted mice together with LCT compared with medium-chain triglycerides (MCT), which do not stimulate chylomicron formation. The effect of LCT was blocked by the addition of an inhibitor of chylomicron secretion, Pluronic L-81. Adoptively transferred OVA-specific DO11.10 T-cells proliferated more extensively in peripheral lymph nodes when OVA was gavaged with LCT than with MCT or LCT plus Pluronic L-81, suggesting that dietary OVA is systemically disseminated. Most dietary OVA in plasma was associated with chylomicrons, suggesting that these particles mediate systemic antigen dissemination. Intestinal-epithelial CaCo-2 cells secreted more cell-associated, exogenous OVA when stimulated with oleic-acid than with butyric acid, and the secreted OVA appeared to be associated with chylomicrons.

Conclusions/significance: Postprandial chylomicron formation profoundly affects absorption and systemic dissemination of dietary antigens. The fat content of a meal may affect immune responses to dietary antigens by modulating antigen absorption and transport.

Figure 1. Chylomicron formation promotes intestinal OVA absorption.

Fasted mice were gavaged with a dispersion of 0.05 ml ¹²⁵I-labeled OVA (black bars) in PBS plus 0.15 ml of either LCT, MCT, or LCT plus 6 µl of Pluronic L-81 (Pl-81). Radioactivity in the entire plasma per mouse (top panel) and in pooled MLN per mouse (bottom panel) was measured 90 minutes later. Another group of mice was gavaged with identical solutions, except that ¹²⁵I-OVA was replaced with [³H]-retinol (white bars). Shown are averages±S.D. of 4 mice per experimental group; * indicates statistically significant differences between feeding groups (P<0.05; ANOVA, Bonferroni's posthoc analysis). The figure shows a representative outcome of two repeats.

Figure 2. Chylomicron formation promotes intestinal absorption of full-length, antigenic OVA.

Fasted mice were gavaged with 0.2 ml emulsions as described in Figure 1, except that ¹²⁵I-OVA was replaced with 25 mg OVA. Blood samples were obtained from the submandibular vein at indicated time points and analyzed for OVA by Western blotting.

Figure 3. Chylomicron formation promotes systemic dissemination and antigen presentation of dietary antigen.

Naïve BALB/C mice were injected with 2.5×10⁶ CFSE labeled T cells from DO11.10 TCR transgenic mice. After 24 h, the mice were fasted (4 h) and gavaged with OVA (25 mg) in 0.2 ml PBS or 25 mg OVA in 0.05 ml PBS+0.15 ml of either MCT, LCT, or LCT plus Pl-81. Mice were then fasted for an additional 6 h. Three days later, inguinal LN cells were isolated, stained with anti-CD4 and KJ1-26 (TCR clonotypic antibody), and analyzed by flow cytometry. Histograms show representative CFSE dilution profiles of gated CD4+, KJ1-26+ T cells as a measure of cell division. The % of cells under markers M1 and M2 represent cells which have not or have undergone cell division respectively. Each panel represents a typical result of three experimental repeats.

Figure 4. Plasma chylomicrons transport dietary OVA.

Fasted mice were gavaged with 0.2 ml LCT-containing emulsions also containing 25 mg OVA. Plasma was isolated 1 h later, and 55 µl were fractionated via FPLC. The grey line of the chromatogram shows the elution profile of a mouse injected i.p. with Poloxamer P-407 1 h prior to gavage to inhibit chylomicron clearance, which caused a milky plasma appearance (inset) and a greatly increased first peak. The solid line shows the elution profile of a mouse not previously injected with Poloxamer P-407. The fractions of this mouse, indicated by the vertical separators, were subjected to immunoprecipitation for detection of OVA (lower panel).The experiment was repeated three times with similar outcomes.

Figure 5. Uptake and secretion of OVA in association with chylomicrons by intestinal epithelial cells.

(A) CaCo-2 cells were incubated with 20 µg/ml Alexa-red OVA for 1 h at 37°C. Nuclei were stained with DAPI (blue). (B) CaCo-2 cells on Transwell filters were incubated overnight at the apical side with 0.1 mg/ml OVA, washed from both sides, then incubated apically with 1.6 mM oleic acid (OA), butyric acid(BA), or oleic acid plus Pl-81 (2 µl/ml). Basolateral medium was collected 16 h later, and OVA was detected by immunoblotting. OVA was immunoprecipitated from the basolateral medium with anti-OVA coupled to protein A-Sepharose (or protein A-Sepharose only), followed by Western blotting of unwashed precipitate for detection of Apo-B. As shown in (C), ApoB-48 co-precipitated with OVA.