HDAC3 Is Required for Pathognomonic Features of Langerhans Cell Histiocytes

Langerhans cell histiocytosis (LCH) is a pediatric inflammatory myeloid neoplasm that develops due to dysregulated myeloid cell development. BRAFV600E is the most common disease-causing mutation and constitutively activates the mitogen-activated protein kinase (MAPK) pathway in myeloid lineage precursors leading to the key pathognomonic features of LCH cells. Enhanced myelopoiesis and reduced CCR7 expression promote accumulation of LCH cells in tissues by simultaneously increasing the production of pathological DCs and preventing tissue egress. Furthermore, LCH cells acquire an oncogene induced senescence-associated secretory phenotype (SASP) that depends on mammalian target of rapamycin (mTOR), which is hallmarked by increased expression of anti-apoptotic proteins, inflammatory cytokines, and matrix metalloproteinases enhancing survival of LCH cells and promoting recruitment of inflammatory immune cells forming characteristic granulomatous lesions. These pathognomonic features result in the accumulation of LCH cells in any organ causing a wide range of clinical symptoms. Frontline therapy for LCH involves combination chemotherapy and steroidal anti-inflammatories, or MAPK inhibitors, which have significant toxicity and fail to eliminate disease causing precursors. New therapeutic approaches are urgently needed.

Here, using multiple genetic mouse models, we show that normal epidermal Langerhans cells (LCs) depend on HDAC3 for their development, differentiation, and survival. Integrative RNA and chromatin-immunoprecipitation-sequencing show loss of HDAC3 abrogates the expression of master regulators of myeloid development and function including Csf1r, Spi1, Id2, and Runx3. LCH cells are also known to rely on Csf1r and Pu.1 for their development and homeostasis, thus we hypothesized that LCH cells similarly rely on HDAC3 for their development and survival.

CD11c^Cre LSL-BRAFV600E (BRAFV600E^CD11c) mice develop severe multifocal LCH with pronounced lesion development in their livers and lungs, hepatosplenomegaly, and a reduced lifespan due to the accumulation of pathological dendritic cells (DCs). We generated BRAFV600E^CD11c HDAC3^fl/fl (BRAFV600E^HD3KO) mice, which produce pathological DCs that simultaneously express BRAFV600E and harbor a conditional deletion in the deacetylase domain of HDAC3. Compared to BRAFV600E^CD11c mice, BRAFV600E^HD3KO mice exhibited significantly less hepatosplenomegaly, reduced lesional burden (Panel A), attenuated disease progression, and improved survival indicating reduced LCH disease burden. Compared to BRAFV600E^CD11c, flow cytometry showed BRAFV600E^HD3KO had reduced numbers of LCH cells in lungs and livers linking improved disease outcomes to abrogation of pathological DCs. Flow cytometric analysis of circulating myeloid cells further found reduced frequency of circulating DCs and DC progenitors, indicating that a lack of HDAC3 activity prevents the development of pathological DCs (Panel B).

LCH-like cells can be generated in vitro by culturing BRAFV600E^CD11c bone marrow with granulocyte-monocyte colony-stimulating factor (GM-CSF), providing a valuable drug screening tool. Treating LCH-like cells with RGFP966, an HDAC3-specific inhibitor, increased apoptosis indicated by annexin-V and DAPI staining, reduced expression of Bcl-2, increased CCR7 expression, and decreased S6 phosphorylation (an indication of decreased mTOR activity), showing that pharmacological inhibition of HDAC3 may prove therapeutically efficacious by abrogating pathognomonic features of LCH cells.

Together, our findings identify HDAC3 as a critical epigenetic regulator for both healthy and pathological LCs. We further show, HDAC3 is required for multiple pathognomonic features of LCH cells and could be a promising drug target. Furthermore, if HDAC3 is required for the development of pathological DC and DC progenitors, HDAC3 blockade would address a great need in treatment of patients with LCH.