Recipient Signaling through the Cannabinoid Type 2 Receptor Regulates IL-6 Independent Neuroinflammation during Graft Versus Host Disease

Neurotoxicity is an increasingly recognized complication of cancer immunotherapy occurring after allogeneic hematopoietic stem cell transplantation (HSCT), immune checkpoint inhibitor administration, and chimeric antigen receptor T cell therapy. While inflammatory cytokines are thought to play a pivotal role, the immunological and biochemical pathways by which neuroinflammation occurs in these settings are still incompletely understood. Using well-defined murine models of graft versus host disease (GVHD) [C57BL/6 (H-2^b)→Balb/c (H-2^d) and B10.BR(H-2^k)→B6(H-2^b)] to examine T cell-mediated inflammation in the central nervous system (CNS) after allogeneic HSCT, we previously identified host IL-6 production as a pivotal cytokine responsible for inducing neuroinflammation. Notably, whereas blockade of this signaling pathway reduced donor T cell infiltration, inflammatory cytokine production, and microglial cell expansion in the brain, it did not reverse dysregulation in the tryptophan metabolic pathway and the subsequent production of neurotoxic metabolites, indicating the presence of an IL-6 independent pathway that promoted CNS inflammation. To understand the pathophysiology of this process, we examined the role of endocannabinoid signaling through the type 2 cannabinoid receptor (CB2R) since this receptor is expressed on all immune cells and the natural ligands for this receptor modulate cognitive and behavioral function. Using genetic approaches to define the role of CB2R signaling, we observed that absence of the CB2R on donor cells had no effect on T cell infiltration or expression of inflammatory cytokines (IFN-γ, GM-CSF, IL-6 or TNF-α) when compared to animals transplanted with wild type (WT) grafts. There was also no difference in the production of neurotoxic tryptophan metabolites when assessed by mass spectrometry. Conversely, absence of CB2R expression on host tissues significantly reduced donor T cell infiltration and inflammatory cytokine gene expression, indicating that host CB2R expression potentiated CNS inflammation. In addition, there was a significant decrease in gene expression of indole 2,3 dioxygenase and kynureninase which are key enzymes involved in the pathway by which tryptophan is metabolized to neurotoxic metabolites under inflammatory conditions, as well as a decreased ratio of neurotoxic (i.e. 3-hydroxykynurenine) to neuroprotective (i.e. kynurenic acid) tryptophan metabolites. Pharmacological inhibition of CB2R signaling with a peripherally restricted CB2R-selective antagonist had no effect on reducing GVHD-induced inflammation indicating that peripheral regulation of CB2R did not alter CNS inflammation and that a CB2R-expressing population resident in the brain was critical for regulating disease. To identify CB2R⁺ cells in the brain, we employed a novel CB2R-EGFP reporter mouse and examined regions with diverse functional roles (i.e. amygdala, brainstem, and cerebellum). These studies revealed that CB2R expression was detected only on host microglial cells. To determine whether microglial cells promoted CNS inflammation, we employed IL-34^-/- mice, which specifically lack microglial cells, as recipients and demonstrated a significant reduction in donor T cells and inflammatory cytokines in the brain compared to WT controls. Thus, CB2R expression on microglial cells appeared to be critical for the regulation of inflammation in the CNS. Since we had shown that blockade of IL-6 failed to prevent tryptophan metabolite dysregulation, we examined whether this treatment had any effect on microglial CB2R expression. Treatment with an anti-IL-6 receptor antibody did not alter CB2R expression, indicating that the proinflammatory effects mediated by host CB2R⁺ microglial cells were IL-6 independent. Collectively, these results indicate that signaling through the CB2R expressed on microglial cells appears to constitute an IL-6 independent pathway by which neuroinflammation and concurrent tryptophan metabolite dysregulation is propagated. These data also provide an explanation for the failure of anti-IL-6R antibody administration with tocilizumab to prevent and/or treat neurotoxicity occurring after T cell-directed immunotherapy, and suggests that strategies targeting key enzymes in the tryptophan metabolic pathway or the CB2R directly may be novel therapeutic approaches to ameliorate this complication.