ALKBH5 modulates hematopoietic stem and progenitor cell energy metabolism through m6A modification-mediated RNA stability control

Abstract

N⁶-methyladenosine (m⁶A) RNA modification controls numerous cellular processes. To what extent these post-transcriptional regulatory mechanisms play a role in hematopoiesis has not been fully elucidated. We here show that the m⁶A demethylase alkB homolog 5 (ALKBH5) controls mitochondrial ATP production and modulates hematopoietic stem and progenitor cell (HSPC) fitness in an m⁶A-dependent manner. Loss of ALKBH5 results in increased RNA methylation and instability of oxoglutarate-dehydrogenase (Ogdh) messenger RNA and reduction of OGDH protein levels. Limited OGDH availability slows the tricarboxylic acid (TCA) cycle with accumulation of α-ketoglutarate (α-KG) and conversion of α-KG into L-2-hydroxyglutarate (L-2-HG). L-2-HG inhibits energy production in both murine and human hematopoietic cells in vitro. Impaired mitochondrial energy production confers competitive disadvantage to HSPCs and limits clonogenicity of Mll-AF9-induced leukemia. Our study uncovers a mechanism whereby the RNA m⁶A demethylase ALKBH5 regulates the stability of metabolic enzyme transcripts, thereby controlling energy metabolism in hematopoiesis and leukemia.

Keywords: ALKBH5; ATP production; CP: Molecular biology; CP: Stem cell research; OXPHOS; RNA stability; energy metabolism; hematopoietic stem and progenitor cells; leukemia; m(6)A modification; oxidative phosphorylation; stress hematopoiesis.

Graphical abstract

Figure 1

ALKBH5 is dispensable for steady-state hematopoiesis but required for competitive repopulation (A) Measurement of Alkbh5 mRNA expression levels in mouse bone marrow (BM) by qRT-PCR (n = 5 of each group). (B) Measurement of ALKBH5 protein levels in mouse BM by immunoblot. (C) Quantification of RNA m⁶A modification in mouse BM by ELISA (n = 5 of each group). (D) Gating strategy and quantification of long-term hematopoietic stem cell (LT-HSC) and multipotent progenitor (MPP) frequencies by flow cytometry (n = 3 of each group). (E) Gating strategy and quantification of granulocyte-monocyte progenitors (GMPs), common myeloid progenitors (CMPs), and megakaryocyte-erythroid progenitors (MEPs) in BM (n = 3 of each group). (F) Competitive transplantation assay measuring CD45.2⁺ donor-derived cells in the peripheral blood (PB) of recipient mice 4, 8, 12, and 16 weeks after BM transplantation (n = 5 recipients of each group). (G) Gating strategy and quantification of PB engraftment of CD45.2⁺ donor cells, highlighted by red quadrants, and WT CD45.1⁺ competitor cells by flow cytometry, at 16 weeks post-transplantation (n = 5 recipients of each group). (H) Contribution of donor cells to myeloid and lymphoid lineages in PB as determined by flow cytometry (n = 5 recipients of each group). Myeloid, CD11b⁺; B cells, B220⁺; T cells, CD3⁺. (I) Contribution of CD45.2⁺ donor cells to the HSC and progenitor compartments in the BM as determined by flow cytometry (n = 5 recipients of each group). LT-HSC, Lin^–Sca-1⁺c-Kit⁺CD150⁺CD48^–; ST-HSC (short-term HSC), Lin^–Sca-1⁺c-Kit⁺CD150^–CD48^–; MPP, Lin^–Sca-1⁺c-Kit⁺CD150^–CD48⁺; HPC (hematopoietic progenitor cell), Lin^–Sca-1⁺c-Kit⁺CD150⁺CD48⁺; CMP, Lin^–Sca-1^–c-Kit⁺CD34⁺CD16/32^–; GMP, Lin^–Sca-1^–c-Kit⁺CD34⁺CD16/32⁺. Data are represented as mean ± SEM and are representative of at least three independent experiments; p values were calculated using two-tailed Student’s t test. n.s., not significant, ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001, ^∗∗∗∗p < 0.0001.

Figure 2

Loss of ALKBH5 attenuates hematopoietic stem and progenitor cell proliferation without causing apoptosis in transplanted mice (A) Schematic diagram of assessment of cell proliferation by BrdU administration in competitively transplanted mice. (B–D) Quantification of CD45.2⁺ singlet (B), MPP (C), and HSC (D) proliferation via BrdU uptake and 7-AAD staining of DNA content by flow cytometry (n = 3 of each group). (E–G) Determination and quantification of apoptotic rate of CD45.2⁺ singlets (E), MPPs (F), and HSCs (G) via Annexin V staining (n = 3 of each group). Data are represented as mean ± SEM and are representative of at least three independent experiments; p values were calculated using two-tailed Student’s t test. n.s., not significant, ^∗p < 0.05.

Figure 3

Loss of ALKBH5 results in reduced OGDH mRNA and protein level, and restoration of ODGH expression rescues competitive reconstitution (A) Scatterplot of gene expression fold change and the relative contribution of RNA degradation of vcAlkbh5^−/− vs. WT lineage-depleted BM cells (n = 3 of each group). (B) TimeLapse-seq tracks depicting the read coverage over Ogdh for WT and vcAlkbh5^−/− groups. Bar graphs of normalized read counts aligned to the mature Ogdh transcript are shown at the right. Reads are colored according to their U-to-C mutational content. (C) Measurement of OGDH protein level in WT and vcAlkbh5^−/− BM cells. (D) Measurement of m⁶A enrichment of Ogdh mRNA in BM cells by m⁶A-RIP-qPCR (n = 3 of each group). Results are presented relative to input. Gapdh serves as negative control, and Myc serve as positive control. (E) Ogdh mRNA levels in different cell types of WT and vcAlkbh5^−/− BM cells, normalized to Actb (n = 3 of each group). (F) Determination of gene expression levels of TCA cycle enzymes by qRT-PCR, normalized to Actb (n = 3 of each group). (G) Ogdh expression level upon knockdown of Ythdf2 as measured by qPCR. (H) Engraftment rate of vcAlkbh5^−/− lineage-depleted BM cells transduced with empty or Alkbh5-, Ogdh-, or Alkbh5-H205A-expressing retroviral vector, 16 weeks after transplantation (n = 3 of each group). (I) Characterization of multilineage population of rescued cells in the recipient mice (n = 3 of each group). Data are represented as mean ± SEM and are representative of at least three independent experiments; the p values were calculated using two-tailed Student’s t test. ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001.

Figure 4

vcAlkbh5^−/− hematopoietic cells are defective in mitochondrial energy production (A) Diagram depicting expected consequences of reduced OGDH levels on energy production. (B) Real-time analysis of oxygen consumption rates (OCRs) of WT and vcAlkbh5^−/− lineage-depleted BM cells via the Seahorse XF ATP Rate Assay (n = 6 of each group). (C) Quantification of ATP production rate via glycolysis versus mitochondrial respiration, determined via Seahorse ATP Rate Assay (n = 6 of each group). (D) Determination of mitochondrial respiration function via measurement of the OCR using the Cell Mito Stress Assay in WT and vcAlkbh5^−/− lineage-depleted BM cells. (E and F) Quantification of the maximal respiration rate (E) and spare respiratory capacity (F) determined by the Seahorse XF Cell Mito Stress Assay (n = 6 of each group). (G) Determination of NAD⁺ and NADH levels in the WT and vcAlkbh5^−/− lineage-depleted BM cells (n = 3 of each group). (H) Quantification of metabolite levels of TCA cycle in murine plasma via ELISA (n = 4 of each group). (I) Ultrastructure of mitochondria in lineage-depleted BM cells of WT and vcAlkbh5^−/− mice imaged via electron microscopy. Scale bar, 2 μm (left) and 1 μm (right). Data are represented as mean ± SEM and are representative of at least two independent experiments; p values were calculated using two-tailed Student’s t test. n.s., not significant, ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001, ^∗∗∗∗p < 0.0001.

Figure 5

Reduced OGDH results in accumulation of L-2-HG that compromises mitochondrial respiration (A) Diagram depicting expected consequences of reduced OGDH levels on TCA cycle metabolites. (B) Chiral liquid chromatography-mass spectrometry (LC-MS) analysis resolving enantiomers of L-2-HG and D-2-HG in plasma of WT and vcAlkbh5^−/− mice. Shaded areas serve as reference marking L- (gray) and D-enantiomer (mint) retention times. (C) Quantification of L-2-HG levels in plasma of WT and vcAlkbh5^−/− mice (n = 13 biological independent samples of each group). (D) Real-time analysis of the OCR of murine WT lineage-depleted BM cells treated with vehicle control or L-2-HG. (E) Quantification of the ATP production rate via glycolysis versus mitochondrial respiration in murine lineage-depleted BM cells treated with vehicle control or L-2-HG (n = 6 of each group). (F) Real-time analysis of the OCR of MOLM13 cells treated with vehicle control and increasing concentrations of L-2-HG. (G) Quantification of the ATP production rate via glycolysis versus mitochondrial respiration in MOLM13 cells treated with vehicle control or increasing concentrations of L-2-HG as determined by the Seahorse ATP Rate Assay (n = 6 of each group). Oligo, oligomycin; Rot/AA, rotenone and antimycin A. Data are represented as mean ± SEM and are representative of at least two independent experiments; the p values were calculated using two-tailed Student’s t test. n.s., not significant, ^∗p < 0.05, ^∗∗∗p < 0.001, ^∗∗∗∗p < 0.0001.

Figure 6

Loss of ALKBH5 limits Mll-AF9-induced leukemogenicity via diminished energy metabolism (A) Diagram depicting the establishment of Mll-AF9 acute myeloid leukemia (AML). (B) Determination of deletion of Alkbh5 mRNA level in Mll-AF9 BM cells by qRT-PCR after treatment of Dox for 4 days (n = 3 of each group). (C) Measurement of OGDH and ALKBH5 protein levels in the Mll-AF9 AML cell lines. (D) Assessment of colony-forming unit (CFU) potential of MA9-WT and MA9-Alkbh5^−/− AML cells (n = 3 of each group). (E) Kaplan-Meier survival curves for recipient mice of MA9-WT and MA9-Alkbh5^−/− AML cells (n = 7 recipients of each group). (F) Determination of mitochondrial respiration function via measurement of the OCR using the Cell Mito Stress Assay in Mll-AF9 cells (n = 6 of each group). (G) Quantification of the ATP production, maximal respiration, and spare respiratory capacity by the Seahorse XF Cell Mito Stress Assay (n = 6 of each group). (H) Schematic depicting the role of ALKBH5 and its downstream effect on hematopoietic cell function. Oligo, oligomycin; Rot/AA, rotenone and antimycin A. Data are represented as mean ± SEM and are representative of at least two independent experiments; the p values of (B), (D), and (G) were calculated using two-tailed Student’s t test. The p value of (E) was calculated using log-rank test. ^∗∗p < 0.01, ^∗∗∗p < 0.001, ^∗∗∗∗p < 0.0001.