Colon-Specific Gut Microbiome Remodelling Is Associated with Host Metabolism in Paracetamol-Induced Liver Injury in Rats

Background/Objectives: The gut and liver are functionally coupled along the gut–liver axis, yet most studies treat the intestine as a single compartment and sample feces alone. We tested whether a plant-derived beetroot–tarragon microgreen beverage engages this axis in a segment-selective, colon-anchored manner in a rat model of paracetamol-induced liver injury with the intestinal dysbiosis characteristic of this model. Methods: In a rat model of paracetamol-induced liver injury with the intestinal dysbiosis characteristic of this model, three paracetamol-exposed groups—a paracetamol-only model (APAP) and two treatment arms receiving either the beetroot–tarragon microgreen beverage (APAP+MG) or a licorice-root decoction (APAP+LIC)—were compared against an intact control serving as a reference. The gut–liver axis was interrogated end-to-end by combining liver histology, serum 1H-NMR metabolomics and vitamin profiling, and 16S rRNA gene sequencing across five intestinal sites (duodenum, jejunum, ileum, colon, and faeces). Group differences were resolved with consensus differential-abundance testing, and colonic microbiota were linked to circulating metabolites through paired microbiome–metabolome integration. Results: Paracetamol produced a reproducible sub-massive injury. Microgreens were associated with less histological injury, less weight loss, and a serum metabolome closest to the intact state among the treated arms, although no individual metabolite contrast survived FDR correction. Community separation localised to the colon as a genuine centroid shift (Aitchison PERMANOVA R2 = 0.22; PERMDISP non-significant), whereas small-intestinal separation partly reflected dispersion. The microgreen arm showed higher abundance of SCFA-associated Lachnospiraceae and lower Escherichia–Shigella, with a broader colonic response than licorice (27 vs. 4 differentially abundant genera). The microbiome–metabolome association was directional but exploratory; no genus–metabolite pair survived FDR correction. Conclusions: The beverage was associated with attenuation of paracetamol-induced hepatic injury (assessed histologically only; serum ALT/AST were not measured) through a colon-localised, SCFA-oriented remodelling with a directional but non-significant shift of the serum metabolome toward the intact profile, extending the microbiome-modulating benefits of microgreens to an acute drug-injury setting and supporting their further development as functional beverages, subject to a dedicated safety evaluation. As a pilot study with modest sample sizes and non-time-matched group endpoints, these findings require confirmation in adequately powered, time-matched cohorts.