Graft-versus-host disease (GVHD) mediated by alloreactive donor T cells is the most dreaded complication after allogeneic bone marrow transplantation (BMT). Conditioning therapy in the context of BMT creates a proinflammatory milieu, which is thought to be central to the development of GVHD. Interfering with the conditioning-induced inflammatory response could be an approach to prevent GVHD without compromising the graft-versus-malignancy reaction. Histone deacetylase (HDAC) inhibitors belong to a new family of anti-cancer drugs with potent anti-inflammatory properties and have recently been shown to reduce the development of GVHD. The aim of this study was understand the mechanisms underlying the downregulation of GVHD after treatment with the HDAC inhibitor suberonylanilide hydroxamic acid (SAHA). Using the fully MHC-mismatched strain combination B6 to BALB/c, treatment with SAHA resulted in a significantly reduced GVHD mortality. Thus, at days +10 or +37 post-BMT survival for vehicle-treated or SAHA-treated mice was 33 % versus 86 % and 8 % versus 57 % respectively (Chi2 test, p = 0,027 and p = 0,02, respectively). This was associated with a significant reduction in IFN-g and IL-5 serum levels of SAHA-treated animals. As we could not detect any effect of SAHA treatment on T cell activation or T cell expansion in vitro and in vivo, we hypothesized that the inhibitory effect of SAHA treatment on the development of GVHD might be primarily due to an interference in the early events of the inflammatory cascade occurring after conditioning and initial alloactivation. Therefore, we performed gene expression profiling studies in classical GVHD target organs of animals treated with SAHA or vehicle to further understand the mechanisms underlying this effect. SAHA treated animals revealed a significant upregulation of the mRNA expression of the Protein inhibitor of activated stat 1 (PIAS1) gene in the liver compared to vehicle-treated animals. To further strengthen the hypothesis that SAHA might exert its action by interfering with inflammatory reaction and subsequent signaling through the JAK/STAT pathway, we analyzed the effects of SAHA on STAT-1, 3, and 5 activation and expression of SOCS-1 and SOCS-3 in vitro and in vivo. Thus, BALB/c responder splenocytes were incubated with or without irradiated B6 stimulators in the presence or absence of LPS in order to allow for the separate analysis of LPS and alloactivation-induced JAK/STAT activation. Treatment for 24 hours with SAHA completely inhibited phosphorylation of STAT-1 and STAT-3 in response to LPS and alloactivation using western blot analysis. Furthermore, analysis of liver tissue from GVHD animals showed a sustained expression of SOCS-1 protein in SAHA treated animals whereas SOCS-1 was downregulated in the absence of SAHA.

In conclusion our data suggest that the inhibitory effect of SAHA on the development of GVHD is associated with an inhibition of the JAK/STAT signaling pathway. Further studies are warranted to understand the precise mechanisms how SAHA interferes with JAK/STAT signaling and how this leads to inhibition of GVHD.

However, it is conceivable that interfering with inflammatory signaling pathways using pharmacological inhibitors of the JAK/STAT pathway might provide a highly attractive treatment strategy for the prevention of GVHD.

Author notes

Corresponding author