The authors ground their argument in an ecological reframing of why fungi make these toxins in the first place. Mycotoxins emerge from specialised biosynthetic machinery, including polyketide synthases, non-ribosomal peptide synthetases and terpenoid pathways, and recent surveys have reported more than 900 new fungal metabolites, predominantly polyketides, terpenoids and peptides. Rather than evolving to poison mammals, aflatoxin production increases the fitness of Aspergillus flavus by up to 26-fold against insect herbivores while suppressing competing soil microorganisms, a trade-off that persists in natural populations. The same biosynthetic continuum yields antibiotics, antifungals and immunosuppressants, suggesting that the line between mycotoxin and pharmaceutical reflects the producer organism and the exposure context rather than any fundamental chemical mechanism. When dietary mycotoxins reach the gut, they therefore encounter a microbial community that treats them much as it treats any xenobiotic, transforming them enzymatically, binding them physically and, in some cases, releasing them back into active forms.

Microbial metabolism governs mycotoxin fate through conserved mechanisms whose relative contributions depend on the chemistry of the toxin and the composition of the microbiome. Detoxification is exemplified by the de-epoxidation of deoxynivalenol, the most prevalent Fusarium trichothecene, into the far less toxic metabolite DOM-1. Bioactivation runs in the opposite direction: zearalenone can be converted by gut bacteria into alpha-zearalenol, a compound with greater estrogenic activity than its precursor. Microbial cell walls, particularly the peptidoglycan of Lactobacillus and Bifidobacterium strains, adsorb aflatoxin B1 and reduce its free absorption in the small intestine. Meanwhile, microbial deconjugation during enterohepatic cycling can prolong systemic exposure, and masked plant-derived mycotoxins, such as deoxynivalenol-glucosides, can be hydrolysed by colonic beta-glucosidases back into their toxic aglycones in species- and site-dependent ways. The authors emphasise that mycotoxin fate is therefore shaped by microbiota-specific metabolic capacity rather than being intrinsically fixed.

Aflatoxin B1, a Group 1 human carcinogen that contaminates maize, peanuts and animal feed in warm climates, illustrates how epithelial injury intersects with microbiome-dependent immune dysfunction. Multiple microbial taxa degrade aflatoxin B1 through reductive and oxidative ring-modifying pathways that diverge fundamentally from hepatic cytochrome P450 bioactivation, including laccase-mediated oxidation from Bacillus licheniformis and Streptomyces species, F420H2-dependent reductases from mycobacteria and peroxidase-mediated degradation from Pleurotus ostreatus. These activities avoid epoxidation into the DNA-reactive aflatoxin B1-8,9-epoxide, the metabolite responsible for hepatic carcinogenicity. Yet the microbiome also amplifies aflatoxin toxicity. In mice, antibiotic-mediated microbiota depletion eliminated aflatoxin-induced splenic pyroptosis, while faecal microbiota transplantation and even sterile faecal filtrate transfer recapitulated hepatic pyroptosis, implicating gut-derived pipecolic acid and norepinephrine as causal microbial mediators that activate TLR4 and NLRP3 inflammasome signalling, driving caspase-1 activation, gasdermin D cleavage and the release of interleukin-1beta and interleukin-18.

Ochratoxin A, classified as a Group 2B possible human carcinogen and found in cereals, dried fruits, coffee, wine and cocoa, poses sustained dietary risk owing to its chemical stability, long biological half-life and renal accumulation. Microbial detoxification proceeds primarily through carboxypeptidase- and peptidase-mediated hydrolysis into the less toxic ochratoxin alpha, together with cell-wall-mediated binding. Fermented food consortia, particularly Tibetan kefir communities enriched in Lactobacillus kefiranofaciens and Kazachstania species, bind and degrade ochratoxin A through multivalent interactions involving cell wall beta-glucans, polysaccharides and peptidoglycans. In vivo, kefir intervention lowered ochratoxin burden, restored tryptophan and NAD+ homeostasis, upregulated tight junction proteins including claudin-1, occludin and ZO-1, and reduced liver enzyme levels, with metabolomics confirming glycerophospholipid rebalancing. Without such protection, ochratoxin A downregulates tight junction proteins through reactive oxygen species, calcium and myosin light chain kinase-dependent signalling, induces mitochondrial ROS and caspase-dependent apoptosis, and depletes tryptophan-metabolising taxa, weakening aryl hydrocarbon receptor signalling and regulatory T cell function while LPS translocation activates hepatic TLR4 and exacerbates inflammation.

Fumonisin B1, prevalent in maize due to Fusarium contamination, competitively inhibits ceramide synthase, disrupting de novo sphingolipid biosynthesis and driving the characteristic elevation in sphinganine-to-sphingosine ratios used as a biomarker of exposure. Gut-associated lactobacilli contribute to detoxification through esterase- and amidase-mediated hydrolysis that removes tricarballylic acid side chains to form hydrolysed fumonisin B1. Among the best-characterised bacteria, Lactobacillus pontis, L. amylovorus and L. ultunensis isolated from porcine caecum converted fumonisin B1 via surface adsorption followed by enzymatic hydrolysis, achieving up to approximately 50 percent conversion within 48 hours in vitro. However, hydrolysed fumonisin B1 retains biological activity and can sustain intestinal and hepatic injury through MAPK- and NF-kappaB-dependent pathways largely independent of sphingolipid disruption. Comparative animal studies demonstrate that both compounds compromise epithelial barrier integrity, alter gut microbial composition, reduce short-chain fatty acid production by 40 to 60 percent and disrupt hepatic metabolism, although effects are generally milder following exposure to the hydrolysed form. Antibiotic treatment eliminated 75 to 80 percent of fumonisin hepatotoxicity, while faecal microbiota transplantation reproduced hepatic injury, confirming the microbiome as a causal determinant of systemic toxicity.

The trichothecenes deoxynivalenol and T-2 toxin exert toxicity primarily through inhibition of translation and activation of ribotoxic stress pathways that phosphorylate eIF2alpha and engage MAPK and NF-kappaB cascades. Deoxynivalenol reshapes intestinal microbial communities and compromises barrier function, and these effects were partially mitigated by selected strains of Bacillus velezensis, which restored tight junction protein expression, suppressed TLR4/NF-kappaB signalling and normalised microbial composition and short-chain fatty acid profiles. Lactobacillus rhamnosus GG binds deoxynivalenol via S-layer proteins and rescues food intake in anorexia models, increasing butyrate and lowering TNF-alpha and interleukin-6. For T-2 toxin, microbiota-dependent activation of interleukin-6/Janus kinase/signal transducer and activator of transcription 1 signalling mediates splenic injury, as the phenotype was abolished by antibiotic treatment and recapitulated by faecal microbiota transplantation. Elemental nano-selenium attenuated this inflammatory response by reducing reactive oxygen species, suggesting that targeting microbiome-immune interactions may represent a viable strategy to reduce trichothecene toxicity. In contrast to aflatoxin and ochratoxin, evidence for binding-mediated sequestration of trichothecenes remains limited, indicating that biotransformation rather than adsorption is the dominant microbial mechanism influencing their toxicity.

Perhaps the most provocative implication concerns antimicrobial resistance. The authors highlight emerging evidence that deoxynivalenol exposure selectively enriches Gram-positive taxa carrying antibiotic resistance genes, particularly those associated with last-resort antibiotics such as daptomycin. Enterococcus faecalis isolated from deoxynivalenol-exposed microbiota showed extensive drug resistance, with resistance exceeding 128 micrograms per millilitre to daptomycin, vancomycin and linezolid. Deoxynivalenol-induced stress promotes genomic plasticity, enhances the co-occurrence and dissemination of resistance genes via mobile genetic elements, and increases reactive oxygen species production and membrane permeability, amplifying resistance phenotypes. Bacterial defence responses to the toxin converge with classical antibiotic resistance mechanisms, including activation of ATP-binding cassette transporters, toxin-detoxifying enzymes and stress-response proteins such as porins, glutathione S-transferases and phosphotransferases. The authors caution that deoxynivalenol is not an antimicrobial in the conventional sense but may act as a selective ecological pressure that favours resistant populations, and they note that evidence that aflatoxin, ochratoxin or zearalenone directly promote horizontal resistance gene transfer remains limited.

The authors are equally candid about the limitations of the evidence base. Most reports of mycotoxin-induced microbiota perturbation remain descriptive, with limited causal or functional resolution, and many experiments use high doses and non-human models whose relevance to real-world conditions, where exposure is usually chronic and low-dose, is uncertain. Reported microbiota responses are inconsistent, with deoxynivalenol, zearalenone and aflatoxin producing opposing shifts in Proteobacteria, Firmicutes and Bacteroidetes across experimental systems, likely reflecting differences in host biology, exposure intensity, duration and matrix composition. Results for Lactobacillus are particularly inconsistent despite its relevance in detoxification, and aflatoxin exposure can even increase Lactobacillus populations, raising mechanistic questions about whether microbial enrichment reflects adaptive detoxification, ecological resilience or selection for intrinsically resistant taxa with altered functional potential. High in vitro degradation or binding rates by probiotics do not necessarily persist in vivo because of gastric pH, transit time, competing food matrices, host metabolism and enterohepatic recirculation.

Looking ahead, Garcia-Cela and Marcon Gasperini call for a shift beyond descriptive microbiota profiling toward integrative frameworks that link microbial ecology, host physiology and metabolomic outputs. Priority areas include in vivo validation of microbial detoxification pathways, systematic comparison of in vitro and dietary exposure models, and strain-level functional characterisation to resolve inter- and intra-species variability in detoxification capacity. Multi-strain consortia, including kefir-like communities, are particularly attractive because they can couple adsorption, biodegradation and ecological competition within a single system. Engineered microbial systems could be designed to express defined detoxification enzymes or high-affinity binding modules, but their deployment will require careful evaluation of containment, regulatory approval and ecological impact. Future studies must also incorporate co-exposure scenarios, enterohepatic cycling and host-specific variables to enhance translational relevance, and mechanistic resolution of feedback loops between epithelial injury and microbial metabolic reprogramming will be essential to understand whether host-microbiome interactions amplify or constrain toxicity over time. The next generation of interventions, the authors suggest, will likely be modular, mechanism-based and tailored to the toxin, matrix and host context rather than dependent on any single probiotic solution.

Subject of Research: Microbiology / Toxicology

Article Title: Mechanistic evidence and assumptions in mycotoxin-gut interaction

Article References: Garcia-Cela, E., & Marcon Gasperini, A. (2026). Mechanistic evidence and assumptions in mycotoxin-gut interaction. Frontiers in Fungal Biology, 7, Article 1855881. https://doi.org/10.3389/ffunb.2026.1855881

Image Credits: AI Generated

DOI: 10.3389/ffunb.2026.1855881

Keywords: mycotoxins, gut microbiota, aflatoxin B1, ochratoxin A, fumonisin B1, deoxynivalenol, T-2 toxin, antimicrobial resistance, intestinal barrier, microbial detoxification, Mechanistic, evidence