Use of two in vitro approaches to simulate rumen fermentation to investigate the effect of gallic and ellagic acids on methane production and microbial communities

Ruminants upgrade human-inedible biomass into proteins and energy but they contribute to greenhouse gas emissions, such as methane (CH4), and ammonia (NH3) [1]. Tannin-rich plants can alter rumen fermentation and mitigate CH4 emissions, but the effect of individual tannin subunits in rumen has not been fully elucidated. We investigated the effect of gallic acid (GA) and ellagic acid (EA) on rumen fermentation, gas and NH3 production and microbiota. Two in vitro simulation approaches were applied: Hohenheim gas test (HGT) [2] and Rusitec [3]. Rumen fluid was collected from 4 (HGT) and 3 (Rusitec) fistulated lactating Original Brown Swiss cows (1 cow/run). For HGT, a control diet (200 mg hay) was supplemented with (% of DM): i) EA7.5, ii) EA15, iii) GA7.5, iv) GA15, v) EA+GA (both 7.5). After 24-h at 39°C, NH3, volatile fatty acids (VFA), and nutrient degradability were measured. Gas production was evaluated by gas chromatography. EA15 and EA+GA significantly reduced CH4 (-22%) and NH3 (-17%), but reduced VFA (-13%) and nutrient degradability (-9%). GA and EA at 7.5 did not reduce CH4 and NH3 and similarly neither VFA and nutrient degradability. The HGT outcomes were used to set up Rusitec, for which a control diet (hay and barley, 10 g DM/d) was supplemented with (% DM): i) EA7.5, ii) GA7.5, iii) EA+GA (both at 7.5). The 10-d incubation was set at 39°C. All HGT parameters were determined, including urolithins A and B and bacterial and archaeal microbiome. Data were analysed by ANOVA with RStudio, using linear mixed-effects regression (Lme4) models. If data were not normally distributed, the non-parametric Kruskal–Wallis test was applied. Multiple comparisons were performed with pairwise Wilcoxon post-hoc test. Differences were considered significant if P<0.05. EA and EA+GA but not GA alone significantly reduced CH4 (-53%) and NH3 (-51%), but also VFA (-20%) and nutrient degradability. EA and EA+GA increased urolithins A and B, and reduced bacterial richness more than archaeal richness. Considering taxonomy, bacterial communities were dominated by Megasphaera elsdenii following EA and EA+GA treatments, while archaeal communities were dominated by Methanomethylophilaceae in all treatments. M. elsdenii is able to redirect H2 from methanogenesis to other H2 acceptors (pyruvate or propionate). Here, M. elsdenii showed the highest negative correlation with CH4 (r = -0.78), suggesting that a redirection of H2 flow, together with reduced feed fermentation and bacterial richness indirectly contributed to lower CH4. Overall, EA addition altered bacterial diversity and taxonomy composition, influencing CH4, VFA, NH3, and nutrient degradation outcomes, whereas GA addition resulted in minor detrimental effects on VFA production and nutrient degradation. HGT was useful to better set up the Rusitec trial for robust data on fermentation and microbiome. Although EA showed promising results, balancing the positive and negative effects of EA in rumen is imperative before potential in vivo applications.