A metabolomics approach used to profile plasma from portal-arterial pigs revealed differences between breads incurred by dietary fibre and protein contents

Abstract

A liquid chromatography-MS (LC-MS) metabolomics analysis of plasma from portal-arterial catheterised pigs fed breads prepared with whole-grain rye or wheat flour with added concentrated arabinoxylan (AX) or β-glucan (BG) was conducted. Comparison of the effects of concentrated fibres with whole grains has received little attention. Six female catheterised pigs were given two white wheat breads with wheat AX or oat BG, two rye breads with ground rye (GR) or intact rye kernels (RK), and a control white wheat bread (WF) on separate occasions in a randomised cross-over design. The amount of available carbohydrate was similar for the five breads but varied in the content of protein. Plasma was collected continuously for 4 h after feeding. Glucose levels in the portal vein were reduced postprandially in response to the AX, GR and RK breads that had high contents of AX compared with WF bread (P < 0·03). AX and RK breads further tended to decrease plasma levels of some lysophosphatidylcholine species (P ≤ 0·10). The abundance of amino acids in plasma correlated with the protein contents in the breads and leucine uptake significantly affected insulin secretion in the mesenteric artery. In conclusion, the present study revealed that concentrated AX in wheat bread had similar positive effects as whole-grain rye bread on glucose and lipid metabolism.

Keywords: AX, arabinoxylan; Arabinoxylan; BCAA, branched-chain amino acid; BG, β-glucan; Catheterised pigs; DF, dietary fibre; GR, dark ground rye; LC-MS, liquid chromatography–MS; Metabolomics; PCA, principal components analysis; RK, rye kernels; Rye; WF, white wheat; lysoPC, lysophosphatidylcholine; ΔAV, portal–arterial difference; β-Glucan.

Fig. 1.

Principal components analysis scores plot of plasma from the mesenteric artery (a) and the portal vein (b) of pigs taken at 0, 30, 60, 90, 120, 180 and 240 min after consumption of five different test breads. ■, Fasting values; Δ, white wheat bread; ●, arabinoxylan bread; ○, β-glucan bread; +, dark ground rye bread; □, rye bread with kernels. The amount of total variation accounted for by the principal components PC1 and PC2 is shown in parentheses.

Fig. 2.

Principal components analysis scores plot of portal–arterial differences in pigs fed five different test breads. Plasma was sampled at 0, 30, 60, 90, 120, 180 and 240 min postprandial. ■, Fasting values; Δ, white wheat bread; ●, arabinoxylan bread; ○, β-glucan bread; +, dark ground rye bread; □, rye bread with kernels. The amount of total variation accounted for by the principal components PC1 and PC2 is shown in parentheses.

Fig. 3.

Time profiles of identified plasma metabolites in pigs fed five different test breads: (a) glucose (pooled sem = 8·0 × 10⁴); (b) leucine (pooled sem = 1·7 × 10⁴); (c) phenylalanine (pooled sem = 4·1 × 10⁴); (d) lysophosphatidylcholine 18 : 2 (pooled sem = 1·1 × 10⁵). ▴, White wheat bread (WF); ●, arabinoxylan bread (AX); ○, β-glucan bread (BG); +, dark ground rye bread (GR); □, rye bread with kernels (RK). Means at the same time point differ: * RK < AX, BG (P < 0·05); † RK < BG, WF (P < 0·003); ‡ RK < AX, BG, GR, WF (P < 0·05); § GR < AX and RK < AX, BG, WF (P < 0·03); ‖ GR, RK < AX (P < 0·03). ΔAV, portal–arterial difference.