PRMT5 Inhibition Reduces Autoinflammation in a Murine Model of Secondary Hemophagocytic Lymphohistiocytosis

Background: Hemophagocytic lymphohistiocytosis (HLH) is a rare but frequently fatal autoinflammatory syndrome characterized by dysregulated immune activation, excessive production of proinflammatory cytokines and hyperactivation of T cells and macrophages leading to tissue damage. Primary HLH occurs due to predisposing genetic mutations in the perforin/granzyme pathway, while secondary HLH can be triggered by infections, rheumatologic disorders, or malignancies. HLH has a poor overall prognosis despite appropriate therapy, and new avenues targeting pathologic disease mechanisms are urgently needed. The protein arginine methyltransferase 5 (PRMT5) enzyme has been identified as an important mediator in inflammatory T cell subsets in autoimmune disease and graft versus host disease. We hypothesized that PRMT5 activity is relevant to inflammation loops that drive autoinflammation observed in HLH.

Methods: We explored inhibition of type II protein arginine methyltransferase enzyme, PRMT5, in a murine model of secondary HLH. HLH was triggered in C57BL/6-J mice (Jackson Laboratories) by repeated injections of CpG 1826 (Integrated DNA Technologies) and an IL-10 receptor–blocking antibody (clone 1B1.3A; Bio X Cell) on days 0, 2, 4, and 7. Injected mice were left untreated or were treated with ruxolitinib 90 mg/kg (validated treatment of HLH) orally twice daily, starting on Day 4, or increasing doses of oral inhibitor of PRMT5, PRT382 (0.25, 1 or 2 mg/kg daily), starting on Day 0. Mice were weighed daily. On day 9, mice were euthanized and analyzed for spleen and liver weight, serum IFNγ levels and immune profiling of lymphoid and myeloid splenocyte subsets. Splenocytes were left unstimulated, or stimulated with lipopolysaccharide (LPS, myeloid cell stimulation) or CD3/CD28-directed antibodies (T cell stimulation) and analyzed by flow cytometry. Data analysis was performed by unsupervised ViSNE clustering in Cytobank software. Statistical analysis was performed by unpaired t test.

Results: We observed that 1 mg/kg PRT382 lessened the hallmarks of secondary HLH, compared to the untreated HLH cohort, including weight loss (21% reduction, p-value 0.2746), the development of splenomegaly (71% reduction, p-value <0.001) and hepatomegaly (61% reduction, p-value <0.05), similarly to ruxolitinib (p-values <0.01, <0.001, and 0.0795, respectively). Furthermore, mice with untreated HLH showed a 17X increase in serum IFNγ levels compared to controls (p-value <0.01), while mice treated with ruxolitinib showed a 10X increase (p-value <0.05 compared to controls, 0.2577 compared to untreated HLH), and mice treated with 1 mg/kg PRT382 showed undetectable serum IFNγ (p-value 0.2104 compared to controls, <0.01 compared to untreated HLH). Flow cytometric analysis of mouse splenocytes showed that HLH stimulated the expansion of neutrophils (CD11b+/Ly6Clo/Ly6G+) and inflammatory monocytes (CD11b+/Ly6C+/Ly6G-). Ruxolitinib partially counteracted this expansion. PRT382 inhibited expansion of these myeloid cell subsets in a dose-dependent manner and also limited production of IFNγ and TNFα by myeloid cells in response to stimulation with LPS in presence of Golgi blockade with Brefeldin A. Furthermore, PRT382 limited the upregulation of CD25 on the surface of T cells in response to T cell stimulation, although no change in cytokine production by T cells was observed.

Conclusions: PRMT5 inhibition counteracts pro-inflammatory features of pathologic hyper-inflammation driven by myeloid immune cell subsets in a model of secondary HLH. This may represent a novel avenue for urgently needed immune modulatory targeted treatment of HLH. Studies of PRMT5 expression and targeting in models of primary HLH are underway.