Microbiota can exert immunomodulatory effects by short-chain fatty acids (SCFA) in experimental models of graft-versus-host disease (GvHD) after allogeneic hematopoietic stem cell transplantation (allo-SCT). Therefore we aimed to analyze the expression of SCFAs sensing G-protein coupled receptor GPR109A and GPR43 by quantitative PCR in 338 gastrointestinal (GI) biopsies obtained from 199 adult patients undergoing allo-SCT and assessed the interaction of GPR with FOXP3 expression and regulatory T cell infiltrates. GPR expression was strongly upregulated in patients with stage II-IV GvHD (p=0.000 for GPR109A, p=0.01 for GPR43) and at the onset of GvHD (p 0.000 for GPR109A, p=0.006 for GPR43) and correlated strongly with FOXP3 and NLRP3 expression. The use of broad-spectrum antibiotics (Abx) drastically suppressed GPR expression as well as FOXP3 expression in patients’ gut biopsies (p=0.000 for GPRs, FOXP3 mRNA and FOXP3+ cellular infiltrates). Logistic regression analysis revealed treatment with Abx as an independent factor associated with GPR and FOXP3 loss. The upregulation of GPRs was evident only in the absence of Abx (p=0.001 for GPR109A, p=0.014 for GPR43) at GvHD onset. Thus, GPR expression seems to be upregulated in the presence of commensal bacteria and associates with infiltration of FOXP3+ T regs, suggesting a protective, regenerative immunomodulatory response. However, Abx, which has been shown to induce dysbiosis, interferes with this protective response.

The role of IL-22 in adult patients undergoing allogeneic stem cell transplantation (SCT) is of major interest since animal studies showed a protective and regenerative effect of IL-22 in graft versus host disease (GvHD). However, no clinical data exist on the tissue expression. Here we demonstrate that patients not suffering from transplant-related mortality (TRM) show significantly upregulated IL22 expression during histological and clinical GI-GvHD (p = 0.048 and p = 0.022, respectively). In contrast, in GvHD patients suffering from TRM, IL22 was significantly lower (p = 0.007). Accordingly, lower IL22 was associated with a higher probability of TRM in survival analysis (p = 0.005). In a multivariable competing risk Cox regression analysis, low IL22 was identified as an independent risk factor for TRM (p = 0.007, hazard ratio 2.72, 95% CI 1.32 to 5.61). The expression of IL22 seemed to be microbiota dependent as broad-spectrum antibiotics significantly diminished IL22 expression (p = 0.019). Furthermore, IL22 expression significantly correlated with G-protein coupled receptor (GPR)43 (r = 0.263, p = 0.015) and GPR41 expression (r = 0.284, p = 0.009). In conclusion, our findings reveal an essential role of IL22 for the prognosis of patients undergoing allogeneic SCT.

The vitamin D receptor (VDR) is critical in regulating intestinal homeostasis and emerging evidence demonstrates that VDR deficiency is a critical factor in inflammatory bowel disease pathology. However, no clinical data exist regarding the intestinal expression of VDR in patients after allogeneic haematopoietic stem cell transplantation (HSCT). Analyzing intestinal biopsies from 90 patients undergoing HSCT with mortality follow-up, we demonstrated that patients with severe acute gastrointestinal graft versus host disease (GI-GvHD) showed significant downregulation of VDR gene expression compared to mild or no acute GI-GvHD patients (p = 0.007). Reduced VDR expression was already detectable at acute GI-GvHD onset compared to GvHD-free patients (p = 0.01). These results were confirmed by immunohistochemistry (IHC) where patients with severe acute GI-GvHD showed fewer VDR+ cells (p = 0.03) and a reduced VDR staining score (p = 0.02) as compared to mild or no acute GI-GvHD patients. Accordingly, low VDR gene expression was associated with a higher cumulative incidence of treatment-related mortality (TRM) (p = 1.6x10-6) but not with relapse-related mortality (RRM). A multivariate Cox regression analysis identified low VDR as an independent risk factor for TRM (p = 0.001, hazard ratio 4.14, 95% CI 1.78-9.63). Furthermore, VDR gene expression significantly correlated with anti-microbial peptides (AMPs) gene expression (DEFA5: r = 0.637, p = 7x10-5, DEFA6: r 0 0.546, p = 0.001). In conclusion, our findings suggest an essential role of the VDR in the pathogenesis of gut GvHD and the prognosis of patients undergoing HSCT.

Calprotectin, a calcium- and zinc-binding protein that comprises the subunits S100A8 and S100A9, has been extensively studied as a biomarker of gastrointestinal (GI) inflammation through fecal and serum analyses. However, its role in intestinal tissue remains poorly understood because of the limited availability of biopsy specimens. In this study, we analyzed S100A8 and S100A9 messenger RNA (mRNA) expression in 579 intestinal biopsy specimens from allogeneic stem cell transplantation recipients and observed a strong association with acute GI graft-versus-host disease (aGI-GVHD; P< .001). Neutrophil infiltration correlated with the severity of aGI-GVHD (P< .001), and calprotectin expression was strongly linked to Toll-like receptor 4 (TLR4; P< .001) and TLR2 (P< .001) expression. Both TLR4 and aGI-GVHD were associated with elevated calprotectin mRNA levels (P< .001). When patients received broad-spectrum antibiotics at disease onset, calprotectin expression was suppressed (S100A8, P = .001; S100A9, P = .01). GI site–specific differences in calprotectin expression were identified: during severe aGI-GVHD, levels increased up to 30-fold in the small intestine and up to fivefold in the large intestine with respect to mild or no aGI-GVHD, whereas under homeostasis, the large intestine exhibited higher baseline calprotectin (P = .001). The high clinical relevance of this finding is evident from the observation that calprotectin expression was prognostic for transplant-related mortality. Our study suggests that (1) calprotectin is a potential biopsy biomarker in aGI-GvHD and (2) calprotectin expression and neutrophil infiltration possibly indicate translocation of microbiota, which (3) may be modulated by antibiotics.

We previously described that elevated levels of the active vitamin D3 metabolite 1,25-dihydroxyvitamin D3 (1,25(OH)2D3) during the early phase of allogeneic hematopoietic stem-cell transplantation (HSCT) can predict one-year transplant-related mortality (1y-TRM). Given that the liver and kidneys are the primary organs responsible for the effective conversion of vitamin D3, we investigated whether liver and/or kidney function, inflammation, or patient sex might influence vitamin D3 metabolism and, consequently, patient outcomes during transplantation. We found that female patients exhibited higher levels of 1,25(OH)2D3 at the time of transplantation compared with male patients. However, 1,25(OH)2D3 levels were associated with 1y-TRM in both sexes. No correlation was found between liver-associated markers, such as bilirubin, or the inflammation marker C-reactive protein (CRP) and serum levels of vitamin D3 metabolites in either female or male patients. However, serum levels of 1,25(OH)2D3, but not 25(OH)D3 correlated with the creatinine-based estimated glomerular filtration rate (eGFR), indicating that 1,25(OH)2D3 levels are associated with kidney function in HSCT patients. However, a Cox regression analysis, adjusted for baseline risk factors, demonstrated that high peri-transplant levels of 1,25(OH)2D3 (measured from days −2 to 7) remained a significant predictor of patient survival, even when eGFR was taken into account (hazard ratio = 0.99; p = 0.004). These findings suggest that optimal serum levels of 1,25(OH)2D3 may not be achievable in some HSCT patients and that kidney function alone cannot explain why some patients fail to reach the optimal 1,25(OH)2D3 threshold. These data support the potential use of 1,25(OH)2D3 as a prophylactic agent, particularly in patients with pre-existing kidney disease.