Western diet-based NutriCol medium: A high-pectin, low-inulin culture medium promoted gut microbiota stability and diversity in PolyFermS and M-ARCOL continuous in vitro models

Abstract

Optimizing fermentation media to accurately reflect the colonic environment remains a challenge in developing in vitro models that simulate the human colon. This study aimed to develop a fermentation medium, Nutritive Colonic (NutriCol), which mimics colonic chyme with fiber content reflective of a typical Western diet and compared to the widely used MacFarlane medium. MacFarlane/NutriCol media contained the following fiber (g/L): potato starch (5/0.1), pectin (2/5.6), xylan (2/4.4), arabinogalactan (2/1.8), guar gum (1/0.4), glucomannan (0/0.8), and inulin (1/0.2). The performance of NutriCol was evaluated using two in vitro models: PolyFermS, which simulates the human proximal colon, and M-ARCOL, which mimics both the lumen and mucosa of the human colon.

In the PolyFermS model, findings revealed that NutriCol maintained microbiota α-diversity closer to the donor fecal samples and significantly higher than MacFarlane (Shannon's p ≤ 0.01; Simpson's p ≤ 0.001). In contrast, no significant differences in α-diversity were observed between NutriCol and MacFarlane in the M-ARCOL model, likely due to differences in model design and donor microbiome composition. Microbial community structure, assessed by Bray-Curtis distance and A Permutational multivariate analysis of variance (PERMANOVA), revealed significant variations between the two media in both models (PolyFermS: p = 0.02; M-ARCOL: p = 0.01). Additionally, NutriCol demonstrated a higher capacity to cultivate gut microbes, with increased ASV numbers compared to MacFarlane across PolyFermS and M-ARCOL.

SCFAs production was influenced by media composition, individual microbiome structure, and the colonic model used. In the M-ARCOL, NutriCol significantly increased acetate (p = 0.0006) and butyrate (p = 0.02) levels compared to MacFarlane. While a similar trend was observed with the PolyFermS, the differences were not statistically significant (p > 0.05). This increase is attributed to the enrichment of SCFA-producing bacteria, such as Butyricicoccus, Lachnospira, Oscillospiraceae UCG-003, Clostridium butyricum, and Lachnospiraceae NK4A136-group. Additionally, NutriCol generated lower levels of intestinal gases (H₂, O₂, CO₂, and CH₄) than MacFarlane in the M-ARCOL model.

In conclusion, our study demonstrates that NutriCol, a growth medium specifically designed to replicate the typical fiber content of a Western diet, supports gut microbiota diversity and structure better than the established MacFarlane medium. NutriCol's impact was model- and donor-dependent, enhancing microbiota diversity in PolyFermS, while promoting SCFA production and reducing gas levels in M-ARCOL.