Author response:
The following is the authors’ response to the original reviews
Reviewer #1:
(1) The data are generated using ATP read-out (CTG assay). For any inhibitor of mitochondrial function, ATP assays are highly sensitive reflecting metabolic stress, yet these do not necessarily translate into cell growth inhibition using standard Trypan blue assays and tend to overestimate the effects. Please show orthogonal more robust assays of cell growth or proliferation.
We acknowledge the sensitivity of the ATP read-out assay in reflecting metabolic stress. While additional cell growth assays such as Trypan blue exclusion could provide further insights, we believe that the current ATP assay data robustly demonstrate the effect of the IMT and venetoclax combination on cellular metabolism, which is a critical aspect of our study. The scope of our current work focused on metabolic inhibition, and we suggest that future studies could further explore cell proliferation assays to complement these findings.
(2) It is concluded that AML cells do not utilize glucose for ATP production. Please provide formal measurements of glycolysis/lactate upon combinatorial treatment.
We appreciate the reviewer’s suggestion to include glycolysis and lactate measurements, which could indeed add further granularity to our metabolic analysis. However, the primary focus of our study is on mitochondrial function and oxidative phosphorylation (OXPHOS) in AML cells treated with IMT and venetoclax. We believe the data presented in Figure 3 provide strong support for the conclusion that glycolysis is not a major energy source in these cells.
Specifically, in Figure 3C, we demonstrate that AML cells maintain ATP levels and viability when cultured in galactose, a condition that restricts ATP production through glycolysis and forces cells to rely on OXPHOS. This result strongly suggests that these AML cells are not dependent on glycolysis for ATP production. Furthermore, in Supplementary Figure S3B, we show that oxygen consumption rate (OCR) measurements remain stable in the presence of excess glucose, further supporting our conclusion that the cells do not switch to glycolysis when OXPHOS is inhibited.
These findings collectively indicate a primary reliance on OXPHOS for energy generation in AML cells, consistent with our study’s objectives to explore mitochondrial dependency and the therapeutic potential of targeting mitochondrial transcription in AML. Future studies could certainly expand on these insights by incorporating a more detailed analysis of glycolytic flux and lactate production under combinatorial treatment, but we believe the current data are sufficient to support our main conclusions.
(3) The transcriptome data are shown without any analysis of pathways. The conclusion from this data beyond the higher number of genes impacted in the combination arm is unclear. Please provide analysis for example GO pathways and interpret in the context of the drugs' mechanism of action.
In response to the reviewer’s question, we have added gene ontology (GO) pathway analysis to clarify the transcriptomic impact of our combination treatment with IMT and venetoclax. Functional annotation identified significant enrichment in pathways relevant to innate immune response, mitochondrial function, and cellular signaling processes. Specifically, pathways associated with immune defense, mitochondrial signaling, and intracellular signaling were notably affected. These findings suggest that the combination treatment not only disrupts cellular energy metabolism but also potentially primes immune signaling mechanisms. This aligns with the proposed mechanism, where IMT targets mitochondrial transcription and venetoclax induces apoptosis, together enhancing sensitivity in AML cells. The enriched pathways, therefore, support the mechanism of action of both drugs, showing how the combined inhibition of BCL-2 and mitochondrial transcription creates a compounded cellular disruption that enhances the therapeutic effect.
(4) Please demonstrate (could be in supplement) matrix of combination to support the statement that the combination is synergistic using Bliss index. The actual Bliss values are missing.
For the revision, we have now included a matrix of combination treatment effects with the corresponding Bliss synergy index values to substantiate our claim of synergy between IMT and venetoclax. This analysis, provided in the supplement, demonstrates that the observed effects exceed the expected additive impact of each drug alone, as calculated by the Bliss independence model. Specifically, the Bliss values confirm a synergistic interaction in venetoclax-sensitive AML cell lines, highlighting that the combined treatment significantly enhances inhibition of cell viability and apoptosis induction compared to single treatments. This data supports our interpretation of synergy and strengthens the mechanistic conclusions drawn from our findings on the combination therapy’s efficacy.
(5) Please show KG1 data (OCR), here or in Supplement.
In response to the reviewer’s request to include OCR data for the KG-1 cell line, we would like to clarify that OCR measurements were attempted; however, they did not yield conclusive results. This is noted in the revised manuscript (Results section), where we explain that the KG-1 cell line did not provide usable OCR data, likely due to limitations in detecting reliable mitochondrial respiration in this particular line under our experimental conditions. Therefore, we were unable to include KG-1 OCR data in the main figures or the supplement.
Reviewer #2:
(1) It's important that the authors show that the drug's effects in AML are due to on-target inhibition. It's critical that they show that IMT actually inhibits the mito polymerase in the AML cells in the dose range employed.
We appreciate the importance of demonstrating on-target inhibition of mitochondrial RNA polymerase by IMT1, especially in light of the detailed characterization of IMT1b, a closely related compound, as presented in Bonekamp et al., Nature 2020. The work by Bonekamp et al. established the specificity and efficacy of IMT1b in targeting mitochondrial RNA polymerase across various tumor models. Building on these findings, we designed our study to primarily evaluate the combinatorial efficacy of IMT1 with venetoclax in AML models, assuming a similar mechanism of action as described for IMT1b. While direct confirmation of on-target inhibition in AML cells by IMT1 would undoubtedly provide additional mechanistic insight, we focused on translational aspects in this study. We believe that the foundational work provided by Bonekamp et al. supports the assumption of on-target effects by IMT1, and we suggest that future studies could explicitly verify this in the context of AML.
(2) For Fig 1, the stated synergism between Venetoclax (Vex) and IMT in p53 mutant THP1 cells is really not evident, despite what the statistical analysis says. In some ways, the more interesting conclusion is that inhibiting mitochondrial transcription does NOT potentiate the efficacy of Bcl2 inhibition in TP53 mutant AML.
We appreciate the reviewer’s observation regarding the lack of evident synergy between IMT and venetoclax in TP53 mutant THP-1 cells. In line with this comment, we have now expanded the discussion to emphasize that, while statistical analysis suggested a potential interaction, the biological response in TP53 mutant cells was minimal. This contrasts with the strong synergy observed in TP53 wild-type cell lines, such as MV4-11 and MOLM-13. We have now highlighted that TP53 mutation status may limit the effectiveness of mitochondrial transcription inhibition in potentiating BCL-2 inhibition. This addition underscores the importance of mutation profiles, such as TP53 status, in predicting response to combination therapies in AML and is now clearly addressed in the revised discussion.
(3) They combine IMT with Vex, but Vex plus azacytidine or decitabine is the approved therapy for AML. Any clinical trial would likely start with this backbone (like Vex+Aza). They should test combinations of IMT with Vex/Aza or Vex/Dec.
While we recognize the importance of testing IMT in combination with clinically approved therapies like Vex+Aza, our current study was designed to explore the potential of IMT in combination with venetoclax alone. Expanding to other combinations would be an excellent direction for future research but is beyond the scope of our current investigation.
(4) It's interesting that AML cell lines do not show any reliance on ATP generation from glycolysis, but would this still be the case when OxPhos is inhibited with IMT? Such a simple experiment would be much more interesting and could help them better understand the mechanism of IMT efficacy.
We thank the reviewer for highlighting this point regarding the reliance of AML cell lines on glycolysis under OxPhos inhibition. In our study, we observed that AML cells predominantly rely on OxPhos, and we did test for ATP production in conditions that favored glycolysis by growing AML cells with galactose instead of glucose in the medium. As described in the manuscript, we did not observe significant ATP production or cell viability from glycolysis, even under these conditions. This finding suggests that AML cells have a low capacity to adapt to glycolytic ATP generation when OxPhos is disrupted by IMT, reinforcing the view that they are highly dependent on mitochondrial function for energy production. We agree that this adaptation—or lack thereof—is an intriguing aspect of IMT efficacy in targeting energy metabolism in AML cells, and we have clarified this point in the discussion.
(5) OxPhos measurements need statistical analyses.
We appreciate the reviewer’s suggestion to include statistical analyses for the OXPHOS measurements. We would like to clarify that statistical analyses were included in the initial submission. These are detailed in Figure 3 and its legend, as well as in the Statistical Analysis section, where we specify methods such as the calculation of standard error across replicates. This approach was implemented to ensure the rigor of our OCR data and its conclusions on OXPHOS inhibition in AML cells.
(6) Given that the combo-treated mice do not exhibit much leukemia in the blood through ~180 days, and yet start dying after 100 days, the authors should comment on this, given that the bone marrow has been shown to be a refuge that protects leukemia cells from various therapies.
We thank the reviewer for highlighting the observed discrepancy between peripheral blood leukemia levels and survival in combo-treated mice. While leukemic cells were minimally detected in the blood up to approximately 180 days, treated mice began to show signs of disease progression and reduced survival around 100 days. This may suggest that residual leukemic cells persist within the bone marrow, which has been established as a sanctuary site for leukemic cells, providing protection from various therapies. The bone marrow environment likely supports a survival niche, enabling these residual cells to evade treatment effects and potentially initiate disease relapse. We have added this interpretation to the discussion to acknowledge the possibility of bone marrow as a protective refuge, which may limit the full eradication of leukemia in these models despite apparent peripheral blood clearance.
(7) For Fig 5C, the authors should statistically compare the Combo with Vex alone.
We have now included statistical comparisons between the combination treatment and venetoclax alone in Fig 5C to provide a clearer interpretation of the data.
(8) The analyses of gene expression using RNAseq of harvested leukemia cells from the PDX model (Table S2), some more discussion of these results would be helpful, particularly given that neither drug is directly targeting nuclear gene expression.
We thank the reviewer for their suggestion to discuss the RNAseq findings in more detail. In the revised manuscript, we have expanded on the functional annotation of the gene expression changes observed in leukemia cells from the PDX model following combination treatment (Table S2). The enriched pathways include innate immune involvement, mitochondrial function and immune signaling, and intracellular signaling. This suggests that while neither IMT nor venetoclax directly targets nuclear gene expression, the combined treatment induces secondary effects that alter these pathways, potentially contributing to the treatment’s efficacy in AML. This expanded discussion provides greater insight into how the drug combination impacts gene expression and cellular pathways.
(9) We need more information on the PDX models, in terms of the classification (M1 to M6) of the patient AMLs and genetics (specific mutations, not just the genes mutated, and chromosomal alterations).
Additional details regarding the classification and genetic background of the PDX models have been included in the manuscript to better contextualize our findings.
(10) The authors should discuss whether or not IMT represents an improvement over other therapies intended to target Oxphos in AML (clearly, the low toxicity of IMT is a plus, at least in mice).
We appreciate the reviewer’s suggestion to discuss IMT in comparison with other OXPHOS-targeting therapies for AML. In the revised discussion, we highlight IMT’s unique properties, particularly its low toxicity profile, which may offer advantages over other OXPHOS inhibitors. This low toxicity, demonstrated in preclinical studies, suggests that IMT might improve patient tolerability compared to existing therapies that target mitochondrial function.
(11) The authors examined toxicity by weighing the mice and performing CBCs. Measurements of liver and kidney toxicity will be necessary for further clinical development.
We thank the reviewer for the suggestion to further investigate liver and kidney toxicity. In our study, we assessed toxicity through regular weight monitoring and complete blood counts (CBCs) to evaluate overall health status. While additional liver and kidney toxicity measurements will indeed be important in future studies, resource limitations currently prevent us from performing these additional analyses in this model. We agree that these assessments will be essential as we progress towards clinical development, and we plan to address them in upcoming preclinical studies.
