Evaluating the prebiotic effect of oligosaccharides on gut microbiome wellness using *in vitro* fecal fermentation

Diet modulates the gut microbiome, which in turn affects host health through metabolic and immune pathways. Traditional definitions of prebiotics focused on promoting presumed beneficial taxa (e.g., Lactobacillus, Bifidobacterium) and prebiotic intake is known to influence host biomarkers and short-chain fatty acid production. However, a quantitative, evidence-based index linking prebiotic-driven microbiome changes to health status is lacking. The original Prebiotic Index (PI) and its modification relied on a limited set of taxa with assumptions (e.g., Bacteroides as harmful) that are outdated given current microbiome knowledge and taxonomy revisions. The Gut Microbiome Wellness Index (GMWI) was previously developed as a species-level, data-driven metric that predicts likelihood of disease based on the collective abundances of health-prevalent versus health-scarce species. Research question: Can GMWI serve as a robust, quantitative tool to assess the prebiotic effect of dietary oligosaccharides on gut microbiome wellness? Purpose and importance: To demonstrate a proof-of-concept application of GMWI for evaluating the impact of common prebiotic oligosaccharides (FOS, GOS, XOS, IN, 2FL) on fecal microbiome communities in vitro, and to compare GMWI with traditional indices and alpha-diversity metrics for food science and personalized nutrition applications.

Prebiotics were first defined as non-digestible ingredients that stimulate beneficial gut bacteria and later broadened to substrates selectively utilized by host microorganisms conferring health benefits. The original PI (Palframan et al., 2003) incorporated changes in Bifidobacteria, Lactobacilli (assumed beneficial) and Bacteroides, Clostridia (assumed harmful), with a modified version (2007) adding growth-rate weighting and more taxa. Subsequent research has shown many Bacteroides spp. are commensal/beneficial, and the genus Lactobacillus has been reclassified into 25 genera, challenging the assumptions underlying PI/PIm. With next-generation sequencing enabling species-level resolution, GMWI was introduced to classify health status using 50 species (7 health-prevalent, 43 health-scarce) identified across 4347 human stool metagenomes, outperforming alpha-diversity metrics for distinguishing healthy from disease states. This background motivates evaluating GMWI as an updated, evidence-based index for prebiotic assessment.

Study design: In vitro anaerobic batch fecal fermentations (24 h) were conducted to assess the effects of five commercial prebiotics—fructooligosaccharides (FOS), galactooligosaccharides (GOS), xylooligosaccharides (XOS), inulin (IN), and 2'-fucosyllactose (2FL)—on gut microbial community composition and GMWI. Two control conditions were included: no substrate at 0 h (NS0) and no substrate for 24 h (NS24). Each of the seven groups was run in triplicate. Raw materials: Food-grade prebiotics were sourced at high purity: FOS (95.38%), GOS (≥98.99%), XOS (≥95.20%), IN (≥90%), and 2FL (≥94%). Basal medium components included peptone water, yeast extract, salts, bile salts, L-cysteine, hemin, vitamin K1, and Tween 80. In vitro fecal fermentation: Mixed fecal slurry (10% w/v) was prepared from 19 healthy adults (14 males, 5 females; 25–30 years; no antibiotics or pre/probiotics; no recent GI disorders) under anaerobic conditions with informed consent and IRB approval (CBNU-201905-BR-839-01). Each 300 mL water-jacketed fermenter contained 135 mL basal medium inoculated with 15 mL slurry. Prebiotics were added to 1% (w/v). Fermentations were maintained at pH 6.8 and 37 °C with magnetic stirring; anaerobiosis was maintained by nitrogen sparging (15 mL/min). Sampling: 5 mL samples collected at 0 h (NS0) and after 24 h for all groups (NS24, FS24, IN24, GS24, XS24, FL24) for metagenomic analysis. Shotgun metagenome sequencing and taxonomic profiling: DNA was QC-checked (Nanodrop OD260/OD280) and libraries prepared with NEBNext Ultra II for Illumina. Sequencing was 2×150 bp on Illumina NovaSeq 6000 (Macrogen). Reads were processed by KneadData v0.5.1, using Trimmomatic v0.39 (SLIDINGWINDOW:4:30; Q<30 trimming, adapters removed; reads <36 bp discarded) and Bowtie2 to remove human reads. Taxonomic profiling used MetaPhlAn2 v2.7.0 (default parameters). GMWI computation: Using the GMWI-webtool, species-level relative abundances were mapped to the predefined sets of 7 health-prevalent and 43 health-scarce species. For each sample, collective abundances Ψ of each set were computed and compared via log10 ratio H=log10(ΨMH/ΨMN); positive values indicate dominance of health-prevalent species. Statistical analysis: Relative abundances are mean ± SD (n=3). One-way ANOVA with Tukey’s HSD was used for group comparisons (P<0.05). Principal component analysis (PCA) assessed community-level differences. Alpha-diversity metrics (Shannon index, species richness, evenness, inverse Simpson) were calculated and compared.

- Community composition: Across all five prebiotics, Actinobacteria, Bacteroidetes, Fusobacteria, and Proteobacteria phyla increased on average over 24 h, while Firmicutes decreased. PCA showed prebiotic-treated groups (FS24, IN24, GS24, XS24, FL24) clustered together and were distinct from controls (NS0, NS24). - Species-level shifts: Of 236 detected species, 33 overlapped with the 50 GMWI species (5 health-prevalent; 28 health-scarce). Health-prevalent species increased with prebiotics. Notably, Bifidobacterium adolescentis rose from 1.74% (NS0) to 4.73% (FOS), 3.99% (IN), 5.59% (XOS), and 3.61% (2FL) (P<0.05). Bifidobacterium catenulatum and Sutterella wadsworthensis also increased with all prebiotics (P<0.05). Many health-scarce species decreased or remained low after prebiotic fermentation. Examples include reductions relative to NS24 in Clostridium bolteae (e.g., 0.12% in NS24 vs 0.00–0.02% in prebiotics) and Clostridium hathewayi (0.66% NS24 vs 0.01–0.21% with prebiotics). - GMWI outcomes (mean ± SD, n=3): IN 0.48 ± 0.06; FOS 0.47 ± 0.03; XOS 0.33 ± 0.02; GOS 0.26 ± 0.02; 2FL 0.16 ± 0.06. Controls: NS0 0.21 ± 0.06; NS24 -0.60 ± 0.05. All five prebiotics had significantly higher GMWI than NS24 (P<0.05). FOS and IN were significantly higher than NS0; others trended higher than NS0 but not significant in all cases. - Alpha-diversity: In contrast to GMWI, alpha-diversity metrics (Shannon, richness, evenness, inverse Simpson) were significantly lower in all prebiotic groups versus NS0 (P<0.05), yet higher than NS24. - Comparison with traditional PI (literature values): Signs and relative magnitudes generally aligned between GMWI and original PI across treatments (higher with prebiotics vs NS24), supporting GMWI as a consistent, data-driven alternative for assessing prebiotic effects.

This study demonstrates that five representative oligosaccharide prebiotics shift fecal microbiome communities toward higher abundances of species associated with a healthy state, as quantified by positive GMWI values. The findings address the central hypothesis that GMWI can serve as a quantitative, evidence-based index for prebiotic evaluation, overcoming limitations of legacy indices (PI/PIm) that rely on limited taxa and outdated assumptions. The clear increase in health-prevalent species (e.g., Bifidobacterium adolescentis, B. catenulatum, Sutterella wadsworthensis) alongside decreases in multiple health-scarce species explains the positive GMWI responses. The divergence between alpha-diversity decreases and GMWI increases underscores that diversity metrics alone may not reflect health-relevant compositional shifts, highlighting the value of a targeted species-based wellness index. Concordance in directionality between GMWI and literature-reported PI further validates GMWI’s utility, while offering improved alignment with contemporary microbiome knowledge and species-level resolution. These results support the application of GMWI in food science to quantify prebiotic effects and inform personalized dietary interventions aimed at improving gut microbiome wellness.

GMWI effectively quantified the prebiotic impact of FOS, GOS, XOS, IN, and 2FL in an in vitro fecal fermentation model, with all prebiotics producing positive GMWI values and outperforming the 24 h no-substrate control. The approach provides a modern, data-driven alternative to traditional prebiotic indices and reveals that alpha-diversity metrics may not capture health-relevant compositional changes. This work positions GMWI as a practical tool for dietary intervention studies and the design of personalized nutrition strategies based on gut microbiome wellness. Future work should evaluate strain-level and functional (metagenomic/metabolomic) dynamics, extend GMWI applicability beyond adult cohorts, and validate findings in in vivo human trials to link GMWI shifts to clinical outcomes and host metabolites (e.g., SCFAs).

- Taxonomic resolution: Analyses focused on species-level profiles; strain-level differences, which may have clinical relevance, were not assessed. - Functional profiling: Metagenomic functional potential and metabolomic outputs were not incorporated into GMWI in this study. - Population scope: GMWI was derived excluding individuals under 10 years; the index is currently limited to adult microbiomes. - Model system: In vitro fermentation does not fully replicate in vivo physiology, host interactions, and long-term dynamics, potentially limiting generalizability. - Substrate-specific response: 2FL showed comparatively lower GMWI increases, possibly reflecting age-related applicability or substrate-specific utilization patterns.