Proton export takes charge of proliferation

In this issue of Blood, Man et al¹ report that cellular proton balance regulates the proliferation of hematopoietic progenitors and that inhibition of proton export in acute myeloid leukemia (AML) suppresses malignancy in vivo.

Increased intracellular pH (pHi) has emerged as a metabolic adaptation that is linked to increased cancer anabolism, but evidence supporting a direct role for pHi changes in cancer growth has been lacking.² The monocarboxylate transporter (MCT) family, including MCT1 and MCT4, consists of enzymes that mediate proton-linked cellular transport of monocarboxylates, such as lactate and pyruvate, across the plasma membrane; they are upregulated in a number of cancers.^3,4 Such enhanced expression has been considered a compensatory adaptation to Warburg metabolism to mitigate cellular toxicity by increasing lactate export.² The investigators uncovered that simply increasing pHi via an MCT4-dependent mechanism reprograms carbon metabolism and drives growth in normal and malignant myeloid cells; however, it is dispensable for normal hematopoietic stem cells (HSCs), pointing to MCT4 inhibition as a promising metabolic intervention to limit AML proliferation (see figure).

It was shown >20 years ago that pHi in cells from hematological malignancies, including AML, is alkaline compared with normal hematopoietic cells and is dependent on the sodium/proton exchange protein NHE1; however, the in vivo relevance of these observations was unclear.⁵ Man et al undertook a deeper analysis of pHi across the murine and human hematopoietic hierarchies, as well as in murine AML models and human AML cell lines and patient samples. pHi in normal human hematopoietic stem and progenitor cells (HSPCs), including hematopoietic stem cells (HSCs), was ∼7.2 and was significantly higher in AML patient samples (≥7.6) and in transformed murine models compared with their normal counterparts. They found that pH^high cells from primary AML samples were enriched for MCT4 expression and displayed increased proliferation over pH^low cells. Moreover, increased MCT4 expression was associated with adverse outcomes in AML patient cohorts and, thus, is clinically relevant. Remarkably, Man et al found that decreasing proton alkalization through knockdown (KD) or CRISPR-CAS9 knockout (KO) of MCT4 restricted proliferation in mouse and human AML models. Crucially, increasing pHi through MCT4 or NHE1 overexpression in normal hematopoietic cells was sufficient to confer a competitive growth advantage, but MCT4 KD did not disrupt growth in normal HSPCs. Instead, they were dependent on MCT1. The distinct dependency on MCT1 and MCT4 for proton export in normal HSPCs and AML opens a therapeutic window to target MCT4 in AML.

Energy metabolism has emerged as an intriguing avenue that distinguishes HSCs from leukemia stem cells (LSCs).⁶ HSCs have a distinct metabolic program that is reliant on glycolysis and is reprogrammed upon differentiation to committed progenitors to center on oxidative phosphorylation. In contrast, LSCs are highly reliant on oxidative phosphorylation to produce adenosine triphosphate (ATP), whereas AML blasts use glycolysis.⁶ The carbon cycle is also responsible for producing the biomolecules that are necessary for growth. Thus, Man et al undertook carbon flux analysis and Seahorse assays, which functionally measure glcolysis and oxidative phosphorylation, of wild-type and MCT4-KO MLL-AF9 murine AML cells to determine the mechanism for MCT-dependent alkaline pHi in promoting growth. They found that alkaline pHi activated metabolic enzymatic activity in glycolysis and pentose phosphate pathway proteins without changes in ATP production. Metabolic assays following pharmacologic inhibition of NHE1 demonstrated a general role for pHi in regulating AML proliferation through carbon flux that was not solely dependent on lactate/proton transport. The investigators proposed that subtle changes in pHi directly impact the enzymatic activity of various metabolic enzymes to reprogram carbon metabolism to drive anabolism.

To understand how MCT4 was upregulated in AML, the investigators assessed the effect of AML mutations on MCT4 expression. Intriguingly, AML patients with mutations in DNMT3A and EZH2 have significantly increased expression of MCT4 compared with patients with wild-type AML, and this was linked to epigenetic remodeling. Although AML is associated with a global decrease in a number of histone marks, including H3K27 acetylation (ac),^7,8 the investigators found that the MCT4 promoter in AML showed increased H3K27ac. They further connected H3K27ac enrichment on the MCT4 promoter in various AML cell lines and models to increased pHi and showed that inhibition of the H3K27ac reader BRD4 with JQ-1 was sufficient to decrease MCT4 expression, decrease pHi, and suppress proliferation in AML cells. Crucially, JQ-1–mediated suppression of the MLL-AF9 AML model in vivo was rescued by MCT4 overexpression. Thus, targeting the promoter availability of the MCT4 locus and directly inhibiting MCT4 protein function represent potential therapeutic approaches in AML.

The study found a subtle difference in pHi between HSCs and downstream committed progenitors that is unexplained and warrants future investigation. Granulocyte monocyte progenitors (GMPs) displayed slightly elevated pHi (∼7.3) over HSCs, and MCT4 and NHE1 overexpression increased pHi and enhanced proliferation of GMPs but not HSC. Aging leads to dysregulation of HSC function, including a skewing to myeloid differentiation, and HSPCs from aged bone marrow also manifest epigenetic alterations, including a global decrease in H3K27ac.⁷ It is important to determine how aging impacts promoter availability and expression of MCT4 in HSPC subpopulations and whether pHi regulatory mechanisms are altered upon aging to promote malignancy.

Targeting metabolic vulnerabilities in AML is emerging as a feasible, but challenging, therapeutic approach.^6,9 Patients with AML display intra- and interpatient heterogeneity that is connected to stemness, which complicates the effectiveness of any targeted treatment modality.^8-10 This study raises a number of questions, including whether proton export is distinct in LSCs and non-LSCs from patients with AML. pHi regulatory mechanisms may be subverted in different ways across the landscape of AML subtypes, at diagnosis and upon relapse. Patient-derived xenograft studies with serial transplantation will help to clarify the role of pHi in driving AML heterogeneity and LSC function. This study provides an important preclinical basis for measuring pHi across patients with AML and identifies alkaline pHi as a promising biomarker and therapeutic target for clinical investigation.