Intestinal Microbiota: From Inflammatory Bowel Disease to Bone Marrow Transplantation

Abstract SCI-50

A dysregulated interaction of the intestinal microbiome with the patient's innate and adaptive immune response seems to contribute to both inflammatory bowel disease (IBD) and intestinal graft-versus-host disease (GVHD). In IBD, polymorphisms within genes involved in antibacterial defense have been identified as genetic risk factors, such as in NOD2, a gene coding for an intracytoplasmatic receptor for muramyldipeptide, a bacterial cell wall compound, or in ATG16L1, a gene involved in autophagy of bacteria. Disruption of these genes results in dysfunction of Paneth cells, which are the major producers of antimicrobial peptides such as defensins. They thereby protect epithelial stem cells from invasion and destruction by intestinal bacteria and contribute to homeostasis of the intestinal microbiota. Based on van Bekkum's finding of the absence of intestinal GVHD in germ-free mice and the central role of TNF-α release in intestinal GVHD, our group focused on the microbiome/host interactions in GVHD. In prospective studies on genetic risk factors of GVHD, single-nucleotide polymorphisms (SNPs) of NOD2 and also ATG16L1 turned out to be predictive for severe intestinal GVHD and IBD. In addition, however, pulmonary complications revealed an altered production of antibacterial peptides in the presence of NOD2 SNPs. We therefore speculated that human intestinal GVHD similar to IBD might be associated with disruption of the bacterial diversity. We applied metabolomic analyses of metabolites processed in the presence of intestinal bacteria as well as 16s rRNA sequencing to serial urine and stool samples from patients receiving allogeneic stem cell transplantation. Urinary indoxylsulfate (IS) levels dropped during the period of decontamination and use of antibiotics during the neutropenic period but recovered to pretransplant levels in patients with uneventful courses. In contrast, patients developing intestinal GVHD had significantly lower IS levels, suggesting suppression of bacterial diversity in intestinal GVHD. Analysis of 16s rRNA confirmed a major shift from an almost normal distribution pretransplant toward a loss of Firmicutes and an increase in enterococci in the neutropenic period. Although this shift may be partially explained by antibiotic decontamination or treatment during this period that was given to all patients, those patients with subsequent development of intestinal GVHD showed a significantly stronger shift toward enterococci in this period (p=0.002). Whereas patients without intestinal GVHD returned to pretransplant diversity thereafter, predominance of enteroccal flora persisted in patients with intestinal GVHD. These data indicate early microbiome changes in patients with intestinal GVHD. We are currently addressing potential Paneth cell damage and loss of antimicrobial peptides as an underlying mechanism. In summary, our data confirm the relevance of the close interaction of microbiome and host defense in GVHD patients, similar to what has been described in IBD, and raise new options for immune system modulation by restoration of intestinal tolerance.