Comparison of simulated gastrointestinal digestion platforms with varying complexity and sample mass requirements, using common bean ( Phaseolus vulgaris L.) as a model

BACKGROUND Measuring bioaccessible nutrient levels during digestion in a physiologically relevant (dynamic) manner is a bottleneck in plant breeding and food product development. The first aim of this study was to develop and compare four simulated gastric digestion models paired with a static small‐intestinal phase for use with small sample masses. The four gastric models tested included three dynamic models, which emulate peristaltic contractions with or without additional emulation of enzymatic secretion and flow of food, and one static model. The second aim was to examine bioaccessible nutrient levels in common bean ( Phaseolus vulgaris L.) samples representing four genotype by growing environment combinations with differing seed coat coloration, patterning, and phenolic profiles. RESULTS While the highest bioaccessible nutrient levels (starch and protein hydrolysis, total phenolics, and antioxidant power) were observed from more complex dynamic digestion models, even simple dynamic models showed higher values for these traits than a static digestion model. The use of in vitro models varying in complexity provided insight into nutrient bioaccessibility; for example, differences were observed during digestion between genotypes for protein hydrolysis and between environments for both total phenolics and protein hydrolysis, even in a higher‐throughput digestion model. CONCLUSION Smaller‐scale dynamic models show promise as higher‐throughput dynamic digestion platforms. These results inform the application of simulated digestion models to assay bioaccessible nutrient levels in plant‐based foods.