William Vevers



Deoxynivalenol (DON) contamination of grain destined for animal feeds is a major toxicological risk to monogastrics and is suspected of restricting productivity in ruminants. Whereas bacterial additives have been developed that can detoxify DON in the rumen and lower intestine, there are currently no commercial inoculants able to perform this task in crimped grain (CG) silage, a regionally important method of moist grain preservation based on homo- and heterofermentative lactic acid bacteria or chemical additives. Determining whether this ensiling process alongside the action of detoxifying bacteria has the potential to remove DON in CG prior to ingestion, was explored in mini-silo ensiling experiments. CG was heat treated (100 °C, 60 min) or ensiled fresh in triplicate 50 g silos, spiked with 5 mg/kg DON and inoculated with lactic acid bacteria derived from wild birds, natural epiphytic inoculants and commercially sourced silage additives (21 d). DON recovery was only significantly reduced (31.2 ± 14.4% recovery, p<0.001, n= 30) by heat treatment, as determined by IAC-RP-HPLC-UV. Bacterial assemblage analysis by 16S rRNA PCR-DGGE-SEQ identified Weissella cibaria, Pantoea agglomerans, Bacillus subtilis, B. licheniformis and Hafnia alvei as candidate detoxification agents, of which W. cibaria and H. alvei decreased DON recovery in vitro (11.3 and 6.2% recovery respectively, p<0.05, n = 18), which translated to inoculated W. cibaria yielding a decrease in DON recovery (67.2± 14.4%, 28 d) in naturally contaminated crimped wheat (13.5 ± 1.0 mg/kg, 35-40% moisture, p<0.05, n =15). As W. cibaria is a lactic acid bacteria already associated with fermented CG by default it has promise as a novel DON detoxification agent in CG silage. DON is however just one of many hepatotoxic co-contaminants. Retrorsine, a DNA-crosslinking pyrrolizidine alkaloid derived from Ragwort (Senecio sp.) was investigated for interactive toxicity with DON in an in vitro co-exposure experiment. HepG2 cells were exposed to Log10 multifactorial binary exposures for 48 h followed by a suite of assays to elucidate mechanisms of interactive cytotoxicity, genotoxicity and modulation of the proteome. Retrorsine was tentatively confirmed to form DNA/protein crosslinks in the comet, micronucleus and crosslinking assays, whilst DON was found to potently induce cytotoxicity and apoptosis. Co-exposure yielded a complex toxicity response, with low doses yielding antagonistic effects and high doses trending towards additive effects, although DON dose was generally the principle component. The difficulties associated with undertaking an interactive toxicity study where both toxins have multiple metabolic and cellular targets are highlighted.

Document Type


Publication Date