Key points are not available for this paper at this time.
In hematopoietic cell transplant (HCT) patients, intestinal Candida parapsilosis expansion and translocation can cause life-threatening candidemia, yet whether commensal intestinal bacteria prevent Candida expansion remains incompletely defined. Here, we trained a machine learning model on supernatant metabolomic profiles of Lachnospiraceae to identify bacteria-derived inhibitors of fungal growth, identifying valeric and butyric acids as top hits. Fecal samples from HCT patients supported this association, with valeric and butyric acid levels inversely correlating with C. parapsilosis growth. In cell culture and mice, valeric acid inhibited C. parapsilosis growth by increasing intracellular acidification. Administration of glycerol valerate, or free or microencapsulated valeric acid, to release valeric acid along the entire intestinal tract blunted C. parapsilosis growth at murine intestinal sites where valeric acid was detected. Thus, machine learning identified a mechanistic driver of trans-kingdom ecology limiting C. parapsilosis intestinal expansion and may inform strategies to reduce patient risk of candidiasis.
Yasuma-Mitobe et al. (Mon,) studied this question.