Epigenetic Regulation to Enhance Graft-versus-Leukemia Activity

With over 55% of deaths attributed to primary disease, relapse represents the leading cause of mortality after allogeneic hematopoietic cell transplantation (allo-HCT; ref. 1). Therapeutic approaches for acute myeloid leukemia (AML) relapse after allo-HCT include a second allo-HCT, donor lymphocyte infusion, kinase inhibitors, or demethylating therapies alone or in combination with venetoclax and others (reviewed in ref. 2). It is unclear why the alloreactive T cells, which initially controlled the leukemia cells, at some point fail to maintain this control, which leads to relapse. Lactic acid release (3), oncogenic signaling, and reduced proinflammatory cytokine production in leukemia cells (4) have been shown to contribute to immune paralysis post–allo-HCT. Reduced leukemia cell recognition via the T-cell receptor may lead to immune escape. In agreement with the central role of HLA expression by AML cells, a genetic loss of mismatched HLA molecules in AML relapse was observed (5) as well as downregulation of MHC class II genes as a nongenomic loss of HLA (6, 7). The role of MHC II in AML relapse supports the concept that not only CD8 T cells but also CD4 T cells contribute to graft-versus-leukemia (GVL) effects. Additionally, the upregulation of inhibitory ligands on leukemia cells has been recognized as an important driver of leukemia immune escape (8). It was shown in two independent patient cohorts that expression of multiple inhibitory molecules, including PD-L1, B7-H3, and PVRL2, was increased on AML blasts at posttransplantation relapse (7). Loss of HLA class II expression and upregulation of inhibitory checkpoint molecules were responsible for the reduced recognition of AML cells by donor-derived T cells because antileukemic T-cell responses could be partially restored in vitro by treatment with an anti–PD-L1 antibody (7). Consistent with these studies, it was reported that patients experiencing AML relapse after allo-HCT had more exhausted T cells (PD-1⁺CTLA4⁺TIM-3⁺) in the bone marrow at relapse with a restricted T-cell receptor repertoire and impaired effector functions (9). These different studies indicate that the loss of HLA class II molecule expression is a major cause of immune escape post–allo-HCT—involved in 21% to 50% of posttransplantation relapses (6, 7). However, the biological processes driving nongenomic HLA loss were unclear.

In this issue of Cancer Discovery, the groups of Luca Vago and Raffaella Di Micco identified a new molecular mechanism that mediates HLA downregulation in leukemia blasts at relapse, linking epigenetics and leukemia immune escape. Gambacorta and colleagues describe the epigenetic regulator polycomb repressive complex 2 (PRC2) as an epigenetic driver of leukemia immune escape in AML at relapse after allo-HCT (Fig. 1) by using a state-of-the-art multiomics approach combining RNA sequencing (RNA-seq), reduced representation bisulfite sequencing (RRBS) for DNA methylation, and the assay for transposase-accessible chromatin with sequencing (ATAC-seq) for chromatin accessibility (10). The group first established patient-derived xenograft (PDX) mouse models to expand paired patient-derived AML cells from five different patients at primary diagnosis and post–allo-HCT relapse in vivo. ATAC-seq of both primary and PDX samples revealed that relapse samples displayed reduced chromatin accessibility and that HLA class II genes and the CIITA gene acquired a closed chromatin conformation status at relapse, while RRBS did not detect any significant differences at the level of DNA methylation between diagnosis and relapse samples (10). Data integration by multiomics factor analysis and gene set enrichment analysis revealed that the main differences between diagnosis and relapse samples were in the expression and accessibility of genes targeted by the PRC2 multiprotein complex. PRC2-catalyzed histone H3 lysine 27 trimethylation (H3K27me3) induces transcriptional repression at its target loci, leading to the hypothesis that PRC2 might be involved in the silencing of HLA class II genes at relapse after allo-HCT (Fig. 1). Importantly, the authors were able to propose a pharmacologic approach to counteract reduced HLA expression using different PRC2 inhibitors including EPZ-6438 (tazemetostat), which targets the catalytic PRC2 subunit enhancer of zeste homolog 2 (EZH2). Indeed, ex vivo pharmacologic inhibition of PRC2 subunits rescued HLA class II gene expression and improved leukemia recognition by CD4 T cells. In contrast, expression of HLA class I and PD-L1 was not affected by PRC2 inhibition. Additionally, immunodeficient mice that were engrafted with relapsed leukemia cells and allogeneic T cells showed a trend toward reduced leukemia burden upon treatment with EPZ-6438 as compared with vehicle-treated mice.

A question that remains is whether the PRC2-mediated mechanism is the main instrument of AML relapse or one of many instruments given the mutational heterogeneity of AML. Different mutations might contribute to different immune escape strategies, and the specific mechanisms or mutations driving PRC2 expression in AML cells at relapse remain to be elucidated. To complement the insights from bulk sequencing, single-cell RNA-seq and single-cell ATAC-seq could help to answer these questions, taking account of intertumoral and intratumoral heterogeneity. In addition, in vitro coculture experiments by Gambacorta and colleagues pointed toward an inhibitory effect of pharmacologic PRC2 inhibition on T-cell proliferation, which might counteract its effect on reestablishing HLA class II expression (10). Previous studies have shown that demethylating agents expand regulatory T cells, which inhibit antileukemic immune responses, and it is unclear if PRC2 inhibition leads to expansion of immunosuppressive cells, which could be tested in future studies.

In aggregate, the authors not only identify a new immune escape mechanism of AML post–allo-HCT via the epigenetic regulator protein PRC2 that leads to HLA downregulation but also offer a new therapeutic strategy using PRC2 inhibition combined with T-cell transfer. Relapse of the underlying disease represents the main limitation for the success of allo-HCT for patients with AML, and loss of HLA class II expression is a frequent cause of immune escape after allo-HCT. Here, Gambacorta and colleagues describe a mechanism for how AML blasts escape from the allogeneic T-cell response that occurs via PRC2-mediated downregulation of HLA class II molecules. This link between an epigenetic regulator and leukemia immune escape not only is highly relevant and novel but also reveals a targetable vulnerability of AML cells by using PRC2 inhibitors. Several inhibitors of PRC2 subunits, including EPZ-6438 and GSK126, are already in clinical development. The data presented in this study pave the way for epigenetic modification as a therapeutic modality to rescue HLA class II expression on leukemia blasts at relapse after allo-HCT.