KMT2D deficiency confers a therapeutic vulnerability to glycolytic and IGFR inhibitors in melanoma.

We reported that histone H3 lysine (K) 4 methyltransferase, KMT2D, serves as a potent tumor-suppressor in melanoma, which was identified via in vivo epigenome-focused RNA interference (RNAi) screen. KMT2D-deficient tumors show substantial reprogramming of key metabolic pathways including glycolysis via reduction of H3K4me1 (Histone H3K4 mono-methylation)-marked active enhancers, conferring sensitivity to inhibitors of glycolysis and IGFR (Insulin Growth Factor Receptor) pathway.

Metastatic melanoma is an aggressive skin cancer with a 5-year survival of less than 25%, and in the past decade, the number of people affected by this disease has increased alarmingly.[1] Melanoma is also notorious for its resistance mechanisms to current therapies and is characterized by genetic and epigenetic alterations.[2,3]

Melanoma pathogenesis, commonly known as melanomagenesis, involves the acquisition of sequential alterations in specific genes and molecular pathways that control vital cellular processes. The epigenome is an important player in cancer progression;[4] however, we have a limited understanding of how specific epigenetic modifiers aberration impact melanomagenesis. In this context, our main aim is to further learn about the disease onset. Thus, systematic functional approaches are needed to elucidate how aberrations in epigenetic regulators impact tumorigenesis through reprogramming of chromatin states and downstream gene expression changes.

Although the landscape of current treatment options for metastatic melanoma has expanded, it remains insufficient due to poor treatment outcomes which continue to cause several thousand patients deaths annually.[1] A detailed mechanistic understanding of role of epigenetic regulators in melanomagenesis will pave the path for new therapeutic strategies, which will guide patients to receiving appropriate treatments.

To identify epigenetic regulators that function as a tumor suppressor in melanoma, we isolated KMT2D (Lysine methyltransferase 2D) in an in vivo RNAi (RNA interference) screen, a cell-based system for discovering tumor-promoting events. Utilizing the unbiased epigenome-focused RNAi screen in vivo, we identified and validated eight epigenetic modifiers (Lysine Methyltransferase – KMT2D and KMT2F, Lysine Demethylase – KDM1A and KDM5B, Lysine Acetyltransferase – KAT4, Histone Deacetylase – HDAC6, histone methyltransferase – SET domain-containing protein 4 (SETD4) and cytidine deaminase – Apolipoprotein B mRNA Editing enzyme Complex 2 (APOBEC2)) whose loss can significantly accelerate tumor growth.[5] As, the strongest phenotypes were seen for KMT2D, we focused on this gene for deep mechanistic studies.

To investigate the molecular phenotypes conferred by KMT2D loss, RNA sequencing-based transcriptome profiling was performed in the wild-type (WT) and mutant murine melanoma lines. We found that genes overexpressed in the KMT2D mutant cells were associated with immune response, cell adhesion, and several metabolic pathways (including glycolysis as the top pathway). This observation was also made across multiple different cancer types where KMT2D mutations occur in more than 5% abundance. Inhibition of the glycolysis pathway using three different inhibitors – 2-Deoxy-D-glucose (2-DG, a glucose competitor), pomhex (an Enolase 1 inhibitor[6]), and lonidamine (Hexokinase inhibitor) particularly reduced the proliferation of KMT2D mutant melanoma cells compared with that of KMT2D WT melanoma cells in both murine as well as human systems. In vivo treatment with 2-DG also showed specific sensitivity in KMT2D mutant mouse models,[5] suggesting upregulated glycolysis to be a vital contributor to enhanced tumorigenesis in KMT2D mutant melanomas. Importantly, such a sensitivity to glycolysis inhibitors was also observed in accompanying paper from Alam et al.,[7] suggesting that this could be a potential treatment strategy in this specific genetic context.

Unbiased epigenome profiling of KMT2D-deficient versus WT cells showed specific chromatin switches in H3K4me1 (Histone H3 lysine K4 mono-methylation)-enriched active enhancer states. To understand how enhancer deregulation can lead to metabolic reprogramming, we integrated gene expression and active enhancer differences between KMT2D WT and mutant-murine tumor lines. This suggested a significant association between the loss of expression and loss of H3K4me1 patterns in nearby loci; the genes were involved in various cell signaling processes and are tumor suppressors.[5] With a focus on IGF (Insulin Growth Factor) signaling pathway, we noted higher levels of phosphorylated-Serine-AKT (pAKT) and phosphorylated-Insulin Growth Factor Receptor 1 (pIGF1R) in KMT2D mutant murine and human lines, which suggests aberrant activation of the IGF-AKT-glycolysis pathway (Figure 1).[8]

Figure 1.

Impact of KMT2D function in melanoma. Our data suggest that KMT2D (Lysine methyltransferase 2D) loss leads to enhancer reprogramming on tumor suppressor genes including IGFBP5 (Insulin Like Growth Factor Binding Protein 5) which control various pathways such as IGF1R (Insulin Growth Factor Receptor 1) signaling that leads to activation of AKT and rewires metabolic pathways. Glycolytic and IGFR inhibitors serve as a novel therapeutic strategy in the patients with melanoma harboring KMT2D mutations. 2DG (2-Deoxy-D-glucose); TS1/TS2/TS3 (Tumor suppressor 1/2/3)

To gain a direct mechanistic link between KMT2D and IGF1R pathway deregulation, we searched for presumed regulators of IGF signaling and subsequent metabolic reprogramming that lose active enhancers and gene expression in KMT2D mutants. IGFBP5 (Insulin Like Growth Factor Binding Protein 5) expression was consistently lost in KMT2D mutant murine and human cell lines and IGFBP5 expression was markedly reduced in KMT2D mutant human and murine melanoma tumors.[5] Follow-up mechanistic data suggested that KMT2D plays the role of a tumor suppressor through eliciting enhancer reprogramming on tumor suppressor genes, like IGFBP5,[9] that regulate key pathways such as IGF1R signaling which ultimately leads to metabolic rewiring (Figure 1).

Over 6% of all cancers harbor loss-of-function mutations in KMT2D, but there is little information about why these mutations are selective over the duration of tumor evolution.[10] Enhancer reprogramming via KMT2D loss can rewire metabolic pathways for increased energy needs of cancer cells as suggested by drastic deregulation of several metabolic pathways in KMT2D mutant melanomas and lung cancers.[5,7] KMT2D mutant cancers are dependent on glycolysis as this pathway is a central node for catering to needs of proliferating cells by contributing to several different biomass needs.

Further research may be needed to better stratify the functional driver mutations in KMT2D because it is likely to harbor high mutations due to its excessive length and some of the observed mutations may be passenger events, particularly in cancers such as melanoma and lung cancers, which possess a high mutation burden. Additionally, while we show importance in focusing on the role of glycolysis, many other metabolic pathways are highly upregulated in KMT2D mutant cancers which require further exploration. In fact, metabolic reprogramming functions of KMT2D is only one of a multitude of factors contributing to the growth of KMT2D-deficient cells. Future studies will bring attention to other key aspects of KMT2D biology such as investigating the tumor microenvironment.

Overall, our study serves as evidence for the dependency of KMT2D mutant melanomas on the glycolysis pathway and the IGF pathway through enhancer reprogramming. These results implicate glycolysis inhibition as a potential therapeutic strategy in patients with melanoma and other cancers harboring mutations in this epigenetic regulator, thus providing a novel biomarker-driven precision oncology approach that can be tested in clinic.