Integrated microbiome and metabolome analysis reveals microbial and metabolic dynamics associated with goose foie gras spoilage during refrigerated storage.

Abstract

Goose foie gras spoils rapidly under refrigeration, yet its microbial and metabolite dynamics during storage are poorly described. Samples from a single Landes production batch were stored at 4 °C in air-sealed polyethylene pouches and analysed on days 0, 2, 4, 6, 8, 10 and 12. Three biological replicates per time point were processed for total viable count (TVC), total volatile basic nitrogen (TVB-N), thiobarbituric acid reactive substances (TBARS), 16S ribosomal RNA (rRNA) amplicon sequencing, and ultra-high-performance liquid chromatography-mass spectrometry (UHPLC-MS)-based untargeted metabolomics. TVC increased from 3.85 to 6.16 log CFU/g, TVB-N from 8.99 to 27.77 mg/100 g and TBARS from 0.18 to 0.72 mg MDA/kg between day 0 and day 12, with the steepest changes appearing after day 6. The maximum TVC reached 6.16 log CFU g-1 on day 12, below the 7 log CFU g-1 level cited for some meat products; no sensory evaluation was performed. The late storage period is therefore described as a phase of concurrent rises in TVC, TVB-N and TBARS rather than as a confirmed spoilage endpoint. The 14 most abundant genera were retained for multivariate analysis. Seven genera, including Brochothrix, Pseudomonas, Lactobacillus and Lactococcus, met the operational definition of candidate spoilage-associated taxa. Forty candidate metabolites were screened by random forest. Five (tyramine, 4-hydroxyphenylacetaldehyde, adenosine monophosphate, oxidized glutathione and γ-glutamylcysteine) were prioritized on the joint basis of random forest importance, Benjamini-Hochberg-adjusted Kruskal-Wallis p < 0.05, ROC AUC ≥ 0.85 and KEGG pathway interpretability. Microbe-metabolite co-variation was quantified through Benjamini-Hochberg-adjusted Spearman correlation, yielding 108 microbe-metabolite pairs at BH-adjusted P < 0.05 (|ρ| ≥ 0.6), and canonical correlation analysis explained 73.39% of the joint cross-covariance in the first two variates. This study establishes a quantitative microbial-metabolic deterioration signature for refrigerated goose foie gras under air-sealed storage, provides a set of biomarker candidates with defined ROC performance, and identifies shifts in specific tyrosine, glutathione, purine and phospholipid pathways. The concordance between microbial succession and these pathway changes is hypothesis-generating and requires metagenomic confirmation, and these findings are presented as preliminary biochemical anchors for future targeted validation and preservation research.