Mitochondrial suppression in Acute Myeloid Leukemia and Myelodysplastic Syndromes: A two-cohort transcriptomic analysis using GEO Free

Mitochondrial Suppression in Acute Myeloid Leukemia and Myelodysplastic Syndrome: A Two Cohort Transcriptomic Analysis Using GEO Database Background Mitochondria play a central role in the biology of myeloid malignancies, regulating not only energy production but also redox balance, apoptotic signaling, and epigenetic control. In acute myeloid leukemia (AML), venetoclax-based regimens that target BCL2 have demonstrated substantial efficacy, in part due to their ability to impair mitochondrial integrity and trigger apoptosis. However, the broader transcriptomic effects of venetoclax on mitochondrial energetics and redox pathways remain incompletely defined in clinical samples. Similarly, in myelodysplastic syndromes (MDS), mitochondrial dysfunction has been implicated in clonal evolution and hematopoietic stem cell failure, but transcriptomic evidence across disease risk strata is limited. We hypothesized that mitochondrial pathway suppression would be evident both as a treatment effect in AML and as an intrinsic feature of high-risk MDS.

Methods We conducted a two-cohort transcriptomic study using publicly available datasets from the Gene Expression Omnibus (GEO). In the AML cohort (GSE138144), we analyzed RNA expression profiles from 14 patients with paired pre- and post-treatment bone marrow aspirates following venetoclax and azacitidine therapy. In the MDS cohort (GSE114922), we evaluated CD34+ sorted cells from 55 patients spanning the full range of IPSS risk categories. Differential gene expression was performed using GEO2R. Gene set enrichment analysis (GSEA) was conducted using Enrichr and MSigDB. A mitochondrial suppression score was defined as the mean log2 fold-change of a curated gene panel derived from the MitoCarta 3.0 database, covering pathways such as oxidative phosphorylation (OXPHOS), electron transport, ROS detoxification, and mitochondrial biogenesis.

Results In the AML cohort, venetoclax-based therapy induced significant downregulation of mitochondrial and redox-related genes. Notably, key components of mitochondrial complex I and ATP synthase—including NDUFS1 and ATP5F1A—were significantly repressed (FDR < 0.01). Redox regulators such as PRDX3, SOD2, and GPX4, as well as mitochondrial biogenesis markers TFAM and SIRT3, also showed decreased expression. GSEA revealed strong negative enrichment of oxidative phosphorylation (normalized enrichment score [NES] -2.4, FDR q = 0.001), electron transport chain (NES -2.1, q = 0.003), and ROS detoxification pathways (NES -1.9, q = 0.004) following treatment, supporting a global suppression of mitochondrial energetics.

In the MDS cohort, we observed a progressive decline in mitochondrial suppression scores across IPSS risk categories. Median scores were -0.55 in low-risk, -1.07 in intermediate-risk, -2.25 in high-risk, and -3.00 in very high-risk patients. High-risk and very high-risk groups demonstrated coordinated downregulation of mitochondrial respiratory chain genes, including COX7A2, UQCRC1, and NDUFA9. These findings suggest that mitochondrial dysfunction is not only a therapy-induced consequence but also a potential intrinsic marker of advanced disease biology.

Conclusion Our cross-cohort analysis reveals that mitochondrial suppression is a recurrent feature in myeloid malignancies. Venetoclax-based therapy in AML leads to significant downregulation of mitochondrial energy metabolism and redox regulation genes, reinforcing the therapeutic relevance of mitochondrial priming. Simultaneously, patients with high-risk MDS exhibit intrinsic suppression of mitochondrial programs, possibly contributing to ineffective hematopoiesis and clonal progression. These findings highlight mitochondrial energetics as a shared vulnerability across the AML-MDS spectrum and support ongoing efforts to combine BCL2 inhibitors with metabolic or redox-modulating agents such as NAD⁺ precursors, CoQ10, or mitochondrial-targeted antioxidants. Further investigation in prospective cohorts is warranted to validate mitochondrial suppression as a predictive biomarker and therapeutic target.