8-Cl-Ado and 8-NH2-Ado synergize with venetoclax to target the methionine-MAT2A-SAM axis in acute myeloid leukemia

Abstract

Targeting the metabolic dependencies of acute myeloid leukemia (AML) cells is a promising therapeutical strategy. In particular, the cysteine and methionine metabolism pathway (C/M) is significantly altered in AML cells compared to healthy blood cells. Moreover, methionine has been identified as one of the dominant amino acid dependencies of AML cells. Through RNA-seq, we found that the two nucleoside analogs 8-chloro-adenosine (8CA) and 8-amino-adenosine (8AA) significantly suppress the C/M pathway in AML cells, and methionine-adenosyltransferase-2A (MAT2A) is one of most significantly downregulated genes. Additionally, mass spectrometry analysis revealed that Venetoclax (VEN), a BCL-2 inhibitor recently approved by the FDA for AML treatment, significantly decreases the intracellular level of methionine in AML cells. Based on these findings, we hypothesized that combining 8CA or 8AA with VEN can efficiently target the Methionine-MAT2A-S-adenosyl-methionine (SAM) axis in AML. Our results demonstrate that VEN and 8CA/8AA synergistically decrease the SAM biosynthesis and effectively target AML cells both in vivo and in vitro. These findings suggest the promising potential of combining 8CA/8AA and VEN for AML treatment by inhibiting Methionine-MAT2A-SAM axis and provide a strong rationale for our recently activated clinical trial.

Fig. 1. 8CA/8AA downregulate MAT2A expression in the cysteine and methionine metabolism pathway.

A AML cell lines MV4-11 and KG-1a were treated with 1 μM 8CA/8AA for 24 h, and the differentially expressed genes compared to nontreated group were analyzed by RNA sequencing and enriched using KEGG database and GSEA software (n = 2). The ranking metric scores of genes in the cysteine and methionine metabolism pathway are shown. B, C NCBI GEO and GenomicScape analysis of the expression level of MAT2A in AML patients. Complete remission (n = 27); relapse (n = 25); MAT2A-Low (n = 21, 26.6%), MAT2A-High (n = 58, 73.4%). D, E Protein levels of MAT2A in MV4-11 and KG-1a cells were measured through western blotting after 48 h treatment with vehicle (NT), 0.4 μM or 1 μM 8CA /8CA (n = 3). F, G Protein levels of MAT2A in MV4-11 and KG-1a cells were measured after 24 h pre-treatment with complete media or methionine-depleted media and followed by 24 h treatment with vehicle (NT) or 1 μM 8CA /8CA (n = 3). **p ≤ 0.01.

Fig. 2. 8CA/8AA increase intron retention in MAT2A RNA by decreasing METTL16 occupancy.

A Diagram of primers covering different regions of MAT2A RNA. B, C The ratios of intron retention in MV4-11 and KG-1a cells were measured after 24 h treatment with vehicle (NT) or 1 μM 8CA/8AA (n = 3). D–G The ratios of intron retention in MV4-11 and KG-1a cells were measured after 24 h pre-treatment with complete media or methionine-depleted media and followed by 24 h treatment with vehicle (NT) or 1 μM 8CA/8AA (n = 3). The respective protein levels of MAT2A were measured as well. H, I The binding of hp1 and hp2–6 clusters of MAT2A RNA and GAPDH RNA to METTL16 protein in MV4-11 and KG-1a cells were measured after 48 h treatment with vehicle (NT) or 1 μM 8CA/8AA (n = 3). J–M Protein and RNA levels of METTL16 in MV4-11 and KG-1a cells were measured after 48 h treatment with vehicle (NT) or 1 μM 8CA/8AA (n = 3). N, O MV4-11 and KG-1a cells were pretreated with vehicle (NT) or 1 μM 8CA/8AA for 24 h, and then incubated with 10 μg/mL puromycin for 10 min. Puromycin incorporation during nascent protein synthesis was measured through western blotting. P The components of METTL16 RNA in MV4-11 cells after 24 h treatment with vehicle (NT) or 1 μM 8CA/8AA in different fractions extracted from the polysome profiling assay were measured by qRT-PCR and normalized to inputs (n = 3). ^nsp > 0.05, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001.

Fig. 3. 8CA/8AA and VEN synergistically target the viability and methylations of AML cell lines.

A, B MV4-11 and KG-1a cells were treated with vehicle (NT), 10 nM or 20 nM VEN for 48 h, then the cellular methionine levels were measured through mass spectrometry (n = 3). C–E MV4-11 and KG-1a cells were treated with vehicle (NT), 1 μM 8CA/8AA, 0.02 μM VEN or their combinations for 48 h, then the intracellular SAM were measured through mass spectrometry (n = 3) and the histone methylations H3K4me3, H3k9me3, H3K27me3, and H3K36me3 were measured through western blotting. F MV4-11 and KG-1a cells were treated with 8AA, VEN or 8AA plus VEN at a series of concentrations for 48 h, and the cell viability was measured (n = 4). The combination index (CI) is calculated through CompuSyn software to define drug synergism (CI < 1.0), addition (CI = 1.0) and antagonism (CI > 1.0). *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001.

Fig. 4. 8CA/8AA and VEN synergistically target primary R/R AML blasts and Lin⁻CD45^dimCD34⁺CD38⁻ LSCs.

A Protein levels of MAT2A in primary R/R AML blasts isolated from three patients (#046, #288, #434) were measured after 48 h treatment with vehicle (NT) or 1 μM 8CA/8AA. B, C Primary R/R AML blasts were treated with 8CA/8AA, VEN or their combinations at a series of concentrations for 48 h, and the cell proliferations were measured (n = 4). The combination index (CI) is calculated through CompuSyn software to define drug synergism (CI < 1.0), addition (CI = 1.0) and antagonism (CI > 1.0). D Primary R/R AML blasts were treated with vehicle (NT), 1 μM 8CA/8AA, 0.02 μM VEN or their combinations for 48 h, then the histone methylations H3K4me3 and H3k9me3 were measured by western blotting (n = 3). E Cell apoptosis of primary R/R AML blasts was measured after 24 h pre-treatment with complete media or media containing 500 μM methionine ( + MET) or 500 μM SAM (+SAM), followed by 24 h treatment with vehicle (NT), 1 μM 8CA/8AA, 0.02 μM VEN or their combinations (n = 2). F, G RNA and protein levels of METTL16 in Lin⁻CD45^dimCD34⁺CD38^- LSCs were measured after 48 h treatment with vehicle (NT) or 1 μM 8CA/8AA (n = 3). Median Fluorescence Intensity (MFI) of anti-METTL16 was measured through flow cytometry to represent the protein level of METTL16. H The ratios of intron retention in LSCs were measured after 24 h treatment with vehicle (NT) or 1 μM 8CA/8AA (n = 2). I Protein levels of MAT2A in LSCs (represented by MFI of anti-MAT2A) were measured after 24 h pre treatment with complete media or methionine-depleted media and followed by 24 h treatment with vehicle (NT) or 1 μM 8CA/8AA (n = 2). J Cell viability of LSCs was measured through flow cytometry after 24 h treatment with vehicle (NT), 1 μM 8CA/8AA, 0.02 μM VEN or their combinations (n = 2). K LSCs were pretreated with vehicle (NT), 1 μM 8CA/8AA, 0.02 μM VEN or their combinations for 24 h, and then cultured in 3D media at the same density to form colonies. Colonies were counted using microscopy and quantified (n = 3). ^nsp > 0.05, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001.

Fig. 5. 8CA/8AA and VEN synergistically inhibit the progression of R/R AML in vivo.

A NSG mice were engrafted with 0.8 M R/R AML blasts, and after 5 days, they were treated with vehicle (n = 8), 12.5 mg/kg/day 8CA/5 mg/kg/day 8AA (n = 8), 20 mg/kg/day VEN (n = 7), or 8CA/8AA + VEN (n = 10) for 7 weeks. B The survival of mice receiving different treatments was monitored. *p ≤ 0.05, ***p ≤ 0.001.