In this issue of Blood, Lauder et al1 characterize the microbiome and metabolite profiles of gut contents from different intestinal regions. The relationship between the gut microbiota, the metabolome, and graft-versus-host disease (GVHD) has been extensively studied in recent years. However, in experimental models, previous studies have only considered the microbiome and metabolome of excreted stool separately and neither provided insight into their variability throughout the gastrointestinal (GI) tract, nor the temporal dynamics associated with different gut locations. Here, the authors describe microbial and metabolite changes between syngeneic and allogeneic transplant recipients and how they varied by GI location and time after transplantation, using integrated multiomic analyses in well-established murine models.
Gut dysbiosis has been associated with many diseases such as inflammatory bowel diseases, obesity, diabetes, cancer, and GVHD. In their seminal paper, Jenq et al 2 first demonstrated in murine and human recipients of allogeneic hematopoietic stem cell transplantation (HSCT) that intestinal inflammation secondary to GVHD is associated with major shifts in the composition of the intestinal microbiota with loss of overall diversity and expansion of some species including Lactobacillales and Clostridiales. This microbial chaos early after allogeneic bone marrow transplantation was identified as a potential risk factor for subsequent GVHD. Subsequently, in a large multicentric cohort of patients profiled by means of 16S ribosomal RNA gene sequencing, the same group provided strong evidence that the microbiota disruption characterized by loss of diversity and domination by single taxa were associated with risk of death after allogeneic HSCT.3
Further study of the role of intestinal microbiota on microbial metabolites on GVHD identified alterations in GI-derived short-chain fatty acids (SCFAs) after HSCT. Changes in the amount of 1 SCFA, butyrate, were observed in intestinal epithelial cells and resulted in decreased histone acetylation in tissues that mitigated disease severity.4 Dysbiosis and loss of microbe-derived bile acids were also shown to be an important mechanism to amplify T-cell mediated damage in experimental models of GVHD.5
However, these previous studies have only considered the microbiome and metabolome separately and were performed on excreted stool. In their study, Lauder et al provide insight into the variability of microbiome metabolite composition throughout the GI tract and of the time-dependent dynamic of the process. They provide the first multiomic analysis of the microbiome and metabolite profiles from different intestinal regions in mouse models of GVHD. Their analyses provide new insights on how changes vary by GI location and time after transplantation. Their integrated analysis confirmed the role of SCFA synthesis and bile acid metabolism and identified additional pathways and metabolites, such as amino acids, fatty acids, and sphingolipids, linked to GI GVHD and validated the biological relevance of a newly identified microbial metabolite, phenyl lactate, which had not been previously linked to GI GVHD.
As reviewed by Lauder et al, previous studies have already demonstrated that the microbial species recovered from different intestine regions vary from stool-associated ones and are referred to as geographical variation. However, whether the microbial metabolites also vary with the geographical variation of GI microbiome and whether this pattern and disease severity changes in GI GVHD was unknown. Significant alterations in both the microbiome composition and the metabolome were found in the ileum and the cecum. These locations have long been recognized as a major initiating site of the allogeneic reaction in GVHD.6 The current study identified several novel pathways and metabolites associated with GVHD, but more importantly, to our knowledge, it is one of the most comprehensive studies as it integrates not only microbiological and metabolomic data but also the location (geography) and changes over time after the transplant (see figure). Such time- and location-dependent multiomic data require careful bioinformatic analyses. The authors used MEFISTO, an algorithm that allows considering several time points during multiomic analysis. Although such time-dependent multiomic approaches have been used in humans using only fecal samples in GVHD and in posttransplant relapse,7,8 this is the only study that used an animal model. The use of an experimental model, as was done in the study for obvious practical and ethical reasons, is of profound advantage because human data can disclose location/geography.
Finally, after integrating metabolomics data with microbiome data, the authors found that an increase in phenyl lactic acid (PLA) correlated with a bloom of Lactobacillus on day +21 in allogeneic transplant recipients compared to syngeneic controls. Addition of PLA to in vitro cultures resulted in decreased budding of organoids and administration of PLA to mice posttransplant decreased survival and increased interleukin 1β (IL1β) and IL17α, suggesting that PLA has direct effects on the host and GVHD pathology. However, as stated by the authors, the mechanism by which it affects the immune subsets and intestinal epithelial cells will require further investigation.
In summary, this very elegant and thoroughly analyzed set of data integrating multiomic analyses in the setting of experimental transplantation provides critical insights and data that will further help in deciphering the complexity of the allogeneic response and foster future clinical developments.9
Conflict-of-interest disclosure: G.S. declares no competing financial interests.
