The SF3B1 inhibitor pladienolide B massively inhibits DNA damage signaling and repair and counteracts resistance to platinum salts in Non-Small Cell Lung Cancer

Article activity feed

1. 

   Review Commons

   Jul 27, 2026

   Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Reply to the reviewers

   We thank the two reviewers for their constructive comments regarding our manuscript. Below is our point-by-point response (un-bold) to each reviewer’s comments together with the experiments we propose to carry-out in order to strengthen our main conclusions.

   Reviewer #1

   In this study, the authors investigate the effects of pharmacological inhibition of the spliceosome using the SF3B1 inhibitor pladienolide B in models of platinum-resistant non-small cell lung cancer (NSCLC). Using a combination of cell lines, platinum-resistant derivatives, and patient-derived xenograft (PDX) models, the authors show that spliceosome inhibition sensitizes …

   Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Reply to the reviewers

   We thank the two reviewers for their constructive comments regarding our manuscript. Below is our point-by-point response (un-bold) to each reviewer’s comments together with the experiments we propose to carry-out in order to strengthen our main conclusions.

   Reviewer #1

   In this study, the authors investigate the effects of pharmacological inhibition of the spliceosome using the SF3B1 inhibitor pladienolide B in models of platinum-resistant non-small cell lung cancer (NSCLC). Using a combination of cell lines, platinum-resistant derivatives, and patient-derived xenograft (PDX) models, the authors show that spliceosome inhibition sensitizes platinum-resistant tumors to treatment and leads to increased DNA damage accumulation and impaired DNA damage response signaling. Transcriptomic analyses indicate that transcripts encoding DNA damage regulators are particularly sensitive to alternative splicing perturbations, and selected mechanistic experiments suggest involvement of specific regulators such as MLH3. The study further explores links between splicing inhibition, transcriptional activity, and cell cycle progression. Overall, the manuscript presents extensive datasets across multiple experimental systems and provides strong evidence that spliceosome inhibition can sensitize platinum-resistant tumors to DNA damage. However, several aspects of the mechanistic interpretation, data consistency, and presentation require clarification or strengthening to fully support the central claims.

   We thank the reviewer for his/her positive comments on our work and we agree that the manuscript requires clarification to support our main claims.

   Major comments* *

   Conceptual clarity and synthesis of mechanistic model

   The manuscript presents multiple mechanistic observations-including altered splicing of DNA repair genes, increased DNA damage accumulation, transcriptional perturbation, and cell cycle changes-but these are not integrated into a coherent conceptual framework. While it is reasonable that not all mechanistic details are fully resolved, the current presentation leaves the reader uncertain about the relative contributions of these processes. A clearer synthesis of the proposed mechanism, possibly including a summary model figure, would substantially improve the conceptual clarity of the study.

   __We thank the reviewer for his/her comment. We acknowledge that our study provides multiple mechanistic observations. However, all these observations converge towards a more global mechanism by which pladienolide B induces cell death in NSCLC. Hence, we demonstrate that pladienolide B, which targets the SF3B1 protein, a core component of the spliceosome machinery, induces a massive shutdown of the DNA Damage Response signaling pathway, both at the transcriptional and splicing levels, which leads to enhanced genomic instability and cell death notably in NSCLC cells with acquired resistance to platinum salts. More specifically, we identified ATR, DNA-PKcs and MLH3 as novel targets of SF3B1 in NSCLC cell lines, as well as more importantly, in NSCLC Patient-Derived Xenografts. To our knowledge, and as also highlighted by the reviewer, this is the first evidence that pladienolide B slows-down tumor growth by negatively impacting DNA Damage Response in patient-derived xenografts. We agree with the reviewer that providing a summary model figure would improve the clarity of the study. We will provide such a graphical abstract in our revised manuscript. __

   Biological specificity of platinum-resistant cell sensitivity

   A central premise of the study is that platinum-resistant cells exhibit enhanced sensitivity to spliceosome inhibition. However, in several experiments (e.g., cell cycle analysis in Figure 2A), similar responses to pladienolide B appear to occur in both platinum-sensitive and resistant cells. This observation complicates the interpretation that resistant cells exhibit uniquely distinct vulnerability. The authors should clarify how these findings align with the proposed model and more explicitly distinguish shared versus resistance-specific responses.

   __We thank the reviewer for this remark. In this study, we did not want to claim that platinum-salts resistant cells exhibit unique vulnerability to pladienolide B as we fully agree with the reviewer that pladienolide B also exhibits cytotoxic effects in parental sensitive cells, although with a delayed kinetic. We acknowledge that the way we introduced our results section in the former manuscript could have contributed to such a misunderstanding. Rather, we propose a model in which pladienolide B induces cell death in NSCLC cells by massively impairing the expression of key components of the DNA damage and repair signaling pathways, at the transcriptional and/or splicing level. As NSCLC cells with acquired resistance to platinum salts are likely more dependent to these pathways for their survival than parental cells, this explains enhanced susceptibility of these resistant cells to pladienolide B-induced cell death. We think that these results highlight spliceosome targeting compounds as an alternative therapeutic strategy in NSCLC patients who escape chemotherapy. According to the remark of the reviewer, we will modify the way we introduce our results and we will clarify all these points in the discussion section of the revised manuscript. __

   Heterogeneity in PDX responses and lack of platinum-sensitive controls

   The PDX experiments represent a major strength of the study. However, resistant tumors display heterogeneous responses to pladienolide treatment, suggesting the presence of additional determinants of sensitivity.

   __We thank the rewiever for this remark. In this study, we used seven distinct NSCLC PDXs We initially selected these PDXs based on their low responsive rate to cisplatin rather than their mutational status. As discussed in the discussion section, we did not find any common mutation(s) that could predict the differential response of these PDXs to pladienolide B. We only noticed that the LCIM10 PDX, which is the most responsive to pladienolide B, exhibits *ATRX *mutation. ATRX has been shown to protect stalled replication forks from collapsing. As we found that pladienolide B induces early replicative stress in NSCLC cells, it is tempting to speculate that ATRX mutation might interfere with the replicative stress response and potentiates pladienolide B’s cytotoxic effects. Noteworthy, LCIM10 PDX also displayed higher basal levels of both P-DNA-PKcs(Ser2056) and P-ATR(Thr1989) proteins as compared to LCF26, ML1 and LCIM1 PDXs that were less responsive to pladienolide B (data to be added in the revised version of the manuscript as supplementary data). Therefore, and although this remains to be further clarified, this suggests that NSCLC patients with higher basal level of replicative stress, such as those who escape chemotherapy, might be more susceptible to SF3B1 inhibition. __

   Including platinum-sensitive PDX tumors, if available, would provide valuable baseline comparison and strengthen interpretation of resistance-specific effects. If not feasible, the limitations should be acknowledged and discussed.

   To our knowledge, and as also mentioned by the reviewer, our study provides the first demonstration of the effects of pladienolide B on the growth of NSCLC PDXs. We think that these results pave the way for further investigations in additional NSCLC PDXs and we agree with the reviewer that adding platinum-sensitive PDX tumors would be interesting. However, due to cost limitations and time constraints, we will be unable to repeat them for this specific study. Based on the results we already obtained in 7 platinum salts-resistant PDXs as regard to their response heterogeneity, one might also speculate that the comparisons between numerous sensitive and resistant PDXs should be complicated as sensitive PDXs might also display distinct mutational status. According to the remark of the reviewer, we will discuss these aspects in the discussion section of the revised manuscript.

   Consistency between pharmacological inhibition and genetic depletion

   In Figure 4, the authors compare pladienolide treatment with SF3B1 knockdown to demonstrate target specificity. However, the effects observed with the two perturbations are not entirely consistent-for example, pladienolide affects phosphorylation of DNA-PKcs, while SF3B1 knockdown appears to produce broader effects at both protein and mRNA levels. Additionally, differences are observed between resistant cell lines in the response to SF3B1 knockdown. These discrepancies should be addressed and discussed, as they may reflect mechanistic differences between acute pharmacological inhibition and genetic depletion.

   __We thank the reviewer for this remark and we agree that acute pharmacological inhibition using pladienolide B and genetic depletion using SF3B1 siRNA might produce distinct effects as they do not exhibit the same mechanism of action. Hence, pharmacological inhibitors target the protein while siRNA targets mRNA with effects depending on the basal mRNA level/stability. This could explain why we did not exactly observe the same effects on ATR/DNA-PKcs mRNA and protein levels using both approachs in H460/A549 parental and resistant cells (Fig 3-4). For pladienolide B treatment, the effects were analyzed at “early” timepoint [i.e. after 4-6 hours treatment (Fig 3a-d)] or at a “later timepoint” [i.e. after 24-48 hours treatment (Fig 3e-f)]. At early timepoint, pladienolide B induced replicative stress that correlated with DNA-PKcs phosphorylation, while at later timepoints it decreased ATR and DNA-PKcs mRNA/protein levels. The biological consequences of SF3B1 knock-down were analyzed after 72 hours of transfection in resistant cells. This could explain why SF3B1 knock-down appears to produce broader effects at both protein and mRNA levels. However, we acknowledge the existence of differences between SF3B1 knocked-down-H460 and -A549 cells as regard to the downregulation of ATR and DNA-PKcs that occurs at the mRNA and/or protein level depending on the cell line. Again, this could depend on the time as well as the efficiency of SF3B1 knock-down which was more prominent in H460 resistant cells compared to A549 cells. Nevertheless, and despite these mechanistic differences in cell lines and between pladienolide B and SF3B1 siRNA, our results identify ATR and DNA-PKcs as novel targets of SF3B1 in NSCLC cell lines, including cells with acquired resistance to cisplatin, as well as more importantly in NSCLC PDXs (Fig 9c). To our knowledge, this is the first evidence that pladienolide B or SF3B1 knock-down negatively targets ATR or DNA-PKcs in solid tumors. Owing to the crucial role played by both kinases in the maintenance of genomic stability in cancer cells, we think that this result is of importance. Nevertheless, as suggested by the reviewer, we propose to acknowledge and discuss more in details the discrepancies between cellular models and pladienolide B / SF3B1 knock-down in the revised version of the manuscript. __

   Selection and interpretation of splicing-sensitive transcripts

   Transcriptomic analyses in Figure 5 identify both shared and differential splicing changes between sensitive and resistant cells. However, much of the analysis focuses on transcripts that are commonly affected in both conditions, rather than those uniquely altered in resistant cells. Given that the central phenotype is resistance-specific sensitivity, transcripts uniquely mis-spliced in resistant cells may represent more informative candidates. The authors should clarify the rationale behind focusing on shared events and discuss the implications of resistance-specific versus common splicing changes.

   We thank the reviewer for this remark. Indeed, after obtaining RNA-Seq data, we initially looked for genes which differential expression and/or splicing upon pladienolide B treatment could only be observed in resistant cells but not parental ones. We focused first on the 121 genes belonging to the DNA repair pathways full network (WikiPathway WP4946) since Gene-Ontology analyses based on RNA-Seq data demonstrated enrichment of genes involved in DNA metabolic process, which includes DNA repair, among genes down-regulated after pladienolide B treatment (Fig S3). The focus on DNA repair was also justified by our observation showing accumulation of DNA double strand breaks upon pladienolide B treatment in H460 resistant cells (Fig 2d-f). Doing this comparison, we showed that pladienolide B regulates the expression of 41 (33%) and 47 (39%) genes of this network in H460S and H460R cells respectively, ____which were mostly down-regulated in both H460S (33/41) and H460R (33/47) cells (Table 3). Fifteen genes involved in all DNA repair processes were found to be specifically down-regulated in H460R cells upon pladienolide B treatment, including PARP-1. As a whole, these results demonstrated that pladienolide B down-regulates the expression of numerous DNA repair genes in both NSCLC parental and resistant cells. As discussed above, we propose that the enhanced sensitivity of resistant cells to pladienolide B is related to their increased dependency for survival to functional DNA repair pathways.

   When differentially spliced genes were considered, and focusing on exon skipping events, as they were the more prominent (Fig 5e), we found that pladienolide B regulates the splicing of 107 and 87 genes of the WikiPathway WP4946 in H460 parental and resistant cells, respectively (Table 5). Forty five genes were predicted to be regulated in both cell lines. Only 3 genes, namely DCLRE1C, *POLD3 *and PNKP, were predicted to be differentially spliced upon pladienolide B treatment in H460R cells only. Trying to increase the number of genes to study, we extended our analysis to genes belonging to another DNA repair database (Human DNA Repair Genes, Resources from Wood laboratory, UT MD Anderson), and we found six additional genes, namely MLH3, MSH5, RAD54L, EME1, SETMAR and SMC6, that were also predicted to be differentially spliced upon pladienolide B treatment in H460R cells only. However, four of these exon skipping events (i.e. PNKP-Ex9, DCLRE1C-Ex11, SETMAR-Ex2, SMC6-Ex6) were not validated and we did not observe clear difference between H460 resistant and parental cells for the others (Fig 5g and Fig S6). Skipping of MLH3-Ex8 was the sole event displaying a slight difference between both cell lines, mainly in term of kinetic of recovery. This is why we decided to further analyze this specific splicing event. Noteworthy, we focused only on genes involved in DNA damage and repair signaling pathways. Therefore, we cannot exclude that genes involved in other biological processes might be differentially transcribed or spliced in response to pladienolide B in H460 parental and resistant cells.

   Transient nature of splicing effects

   The authors report transient alternative splicing effects upon prolonged pladienolide treatment. This observation is counterintuitive, as continued spliceosome inhibition might be expected to produce cumulative splicing defects. While the authors reference studies showing that transient inhibition can produce lasting effects, the current observations involve continuous exposure. This apparent discrepancy should be clarified and discussed.

   We thank the reviewer for this remark. We agree with him/her that the transient effect of pladienolide B on most of the splicing events we studied was unexpected as pladienolide B treatment was prolonged. We do not have a clear explanation for that. One possibility is that pladienolide B is not stable in the cell culture supernatant and is degraded rapidly. Another, not exclusive, possibility relies on the structure of studied transcripts. As discussed in the discussion section, it is possible that the nature of the transcripts involved in DNA damage response/DNA repair which have a long length, a large number of small exons per transcript, and an elevated number of introns could explain why they recover very rapidly from pladienolide B inhibition. Alternatively, upon SF3B1 inhibition, compensatory regulations by other splicing factors might occur. We will discuss these aspects in the revised version of the manuscript.

   Use of unrelated cell lines in reporter assays

   The DNA damage reporter assays (Figure 6A-D) appear to be performed in cell lines not directly linked to platinum sensitivity or resistance. Given the central importance of resistance-specific responses, repeating key reporter assays in both sensitive and resistant paired models would strengthen the conclusions.

   __We initially engineered these cellular models to assess the role of SRSF2, a splicing factor, in DNA repair ____(Khalife M. et al., NAR Cancer, 2025). In these models derived of either H1299 or A549 NSCLC cell lines, DNA double strand breaks (DSBs) are produced after cleavage by the SceI enzyme and the efficiency of repair is assessed based on the expression of either GFP (for homologous recombination) or CD4 (for c-NHEJ). These cellular models were difficult to engineer and to work with as they require a first stable transfection with either PBL174 pDR-GFP (for HR analysis) or PBL230 (for c-NHEJ) plasmid followed by a second transient transfection with the PLBL133 plasmid that encodes SceI enzyme. As an example, we tried to generate stable A549 cells with PBL174 plasmid but we never succeeded. The reverse was true for H1299 cells and transfection with PBL230. In addition, during the time course of this project, we also tried to transiently transfect plasmid encoding MLH3 or MLH3 protein devoid of exon 8-encoding amino acids in H460 or A549 resistant cells but we never obtained good efficiency of transfection, although we tested several transfection reagents. So, it looks like that resistant cells are hardly transfectable. This is why repeating these experiments in resistant models will not be possible. However, and although we agree with the reviewer that the cell lines we used were not directly linked to platinum sensitivity or resistance, the idea behind these experiments was to test whether pladienolide B could have a general impact on DNA repair by homologous recombination or c-NHEJ using these SceI-induced DSBs systems that are already widely used in the DNA repair field. As shown in Fig 6b and 6d, pladienolide B prevented DNA repair in these cellular models, confirming that it widely negatively impacts DNA repair pathways. __

   Interpretation of MLH3 splicing results

   In Figure 7, differences between pharmacological inhibition and SF3B1 knockdown in MLH3 exon 8 regulation are not entirely consistent.

   We do not strictly agree with this comment. __As also discussed above, the differences seen between pharmacological inhibition of SF3B1 using pladienolide B and SF3B1 knock-down in term of MLH3-exon 8 regulation could be related to differences in term of mechanism of action and/or the fact that the effects of SF3B1 knockdown were analyzed after 72 hours treatment (as we obtained the best knock-down efficiency at this time point) while those of pladienolide B were studied between 6 to 48 hours treatment. However, as illustrated in Fig 7a and 7c (right panel), we showed that both pladienolide B and SF3B1 knock-down promote MLH3-exon 8 exclusion in H460 and A549 resistant cells. The effects of SF3B1 knock-down were less pronounced in A549R cells which could be consistent with the decreased efficiency of SF3B1 knockdown as depicted in Fig 7c (left panel). Pladienolide B also promoted MLH3-Ex8 exclusion in H460 and A549 parental cells but the recovery was faster in parental cells as compared to resistant cells (Fig 7a). __

   Furthermore, inclusion levels of regulated and non-regulated exons appear similarly correlated with SF3B1 expression, potentially weakening the argument for exon-specific regulation. These observations should be clarified.

   __As regard to MLH3-exon 5 exclusion, we observed its exclusion in A549 parental and resistant cells upon pladienolide B treatment, while this was not observed in H460 parental and resistant cells (Fig 7a-b). In SF3B1 knocked-down H460R and A549R cells, the exclusion of MLH3-exon 5 was seen in A549R cells. When analyzing MLH3 exon 8 or exon 5 usage in lung adenocarcinoma patients (Fig 7e), we agree with the reviewer that there was a significant correlation between SF3B1 mRNA level and MLH3 exons 5 and 8 usage. Therefore, these and our data indicate that both MLH3 exons 5 and 8 could be regulated by SF3B1 in NSCLC although differences might occur depending on the cell line. To make this point clearer, we will clarify the text of the results for Figure 7 and our conclusion. __

   Combination treatment logic

   In Figure 8, co-treatment experiments with pladienolide B and cisplatin are performed primarily in resistant cells. Performing similar experiments in platinum-sensitive cells would provide an important reference point to distinguish additive versus resistance-specific effects.

   __We thank the reviewer for this important remark. The objective of figure 8 was to investigate whether pladienolide B that induces a shutdown of numerous DNA damage response-related genes could resensitize NSCLC cells with acquired resistance to cisplatin-induced apoptosis. The results presented in Figures 8a and 8b show that this is indeed the case. However, and according to the remark of the reviewer, we propose to illustrate in the revised version of the manuscript the results of the co-treatment experiments in sensitive parental cells also. Indeed, we already had the results for H460 parental cells. They did not show any additive or synergistic effects of the combination in these cells. Rather adding pladienolide B to cisplatin tended to decrease apoptosis as compared to cisplatin alone. We will reiterate these experiments in A549 cells. If confirmed, and to a translational point of view, these results would support the idea that treating NSCLC patients who relapse from chemotherapy with a combination of platinum salts and pladienolide B could provide therapeutic benefits, whereas NSCLC patients that primary respond to platinum salts could less benefit from this combination. __

   Minor comments

   • In Figure 1, pladienolide treatment in platinum-sensitive cells appears to plateau at approximately 50% cell killing. Extending the concentration range may help clarify whether maximal efficacy was reached.

   __ We agree with this remark. We will reiterate our MTS experiments increasing the dose of pladienolide B. __

   In Figure 1H, the difference between 2.5 mg/kg and 5 mg/kg pladienolide in PDX models appears disproportionately large relative to the dose change. This should be discussed or experimentally clarified.

   We agree with the reviewer but these are the results we obtained. In Figure 1h, we illustrated the probability of progression based on Relative Tumor Volume (RTV) = 2. The difference between the two doses was less, although remaining significant, when considering RTV = 4 as a marker of progression (Fig S2d).

   In Figure 5, differential expression and splicing analyses are presented using multiple cutoffs (e.g., log₂FC > 0.4 and >1; ΔPSI thresholds). This introduces redundancy and may obscure key findings. A single well-justified cutoff would improve clarity.

   We agree with the reviewer that in initial Figure 5 we provided graphs illustrating different analysis thresholds based on our transcriptomic analyses. We will select one cut-off for Differentially Expressed Genes [absolute Log2Fold Change ≥ 0.4 and p ≤ 0.05 (Fig 5a)] and Differentially Spliced Genes [absolute percent splice in (PSI) ≥ 0.2 and p ≤ 0.05 (Fig 5e)]. We will remove Fig 5b and 5d for more clarity.

   Several isoform-specific RT-PCR gels are difficult to interpret due to low image clarity. Improving gel presentation or focusing on key timepoints would strengthen data readability.

   We will improve gel presentation.

   Some figure panels appear redundant, showing similar datasets under different analysis thresholds.

   We agree with the reviewer that in Figure 5 we provided graphs illustrating different analysis thresholds based on our transcriptomic analyses. We will select one cut-off for Differentially Expressed Genes [absolute Log2Fold Change ≥ 0.4 and p ≤ 0.05] and for Differentially Spliced Genes [absolute percent splice in (PSI) ≥ 0.2 and p ≤ 0.05]. We will remove Fig 5b and 5d for more clarity.

   A graphical summary model illustrating the proposed mechanism would improve reader comprehension.

   We agree with the reviewer. We will provide such a graphical abstract in the revised version of our manuscript.

   In Figure 3D, representative images of γH2AX foci appear visually similar across conditions, whereas quantification shows large differences. The authors should ensure that representative images accurately reflect quantified trends and clarify selection criteria for displayed images.

   We agree with the reviewer. We will select additional images illustrating more the differences we highlighted after quantification of more than 500 nuclei (Figure 3D, right panel).

   Reviewer #1 (Significance (Required)):

   This study represents a comprehensive investigation of spliceosome inhibition as a therapeutic strategy to overcome platinum resistance in NSCLC. The use of resistant cell lines, PDX models, and functional reporters provides strong experimental depth. The most compelling aspects of the study include the demonstration that spliceosome inhibition enhances DNA damage accumulation and sensitizes resistant tumors to platinum-based therapies. However, several mechanistic interpretations require clarification, and data presentation could be streamlined to improve logical coherence.

   We thank the reviewer for his/her constructive remarks. We hope that our answers and the additional works we now propose to carry-out in order to revise the manuscript will get agreement to him/her and will strengthen our main claims

   Advance

   The work provides evidence that targeting spliceosome function-specifically via SF3B1 inhibition-can sensitize platinum-resistant tumors to DNA-damaging agents. To my knowledge, this represents one of the first comprehensive demonstrations that spliceosome-targeting compounds can effectively overcome acquired platinum resistance in solid tumor models. The study also contributes to the emerging understanding that DNA damage response transcripts may represent particularly sensitive targets of splicing perturbation.

   Audience

   The study will be of interest to researchers in RNA biology, cancer therapeutics, DNA damage response, and translational oncology. It is particularly relevant to scientists investigating therapeutic vulnerabilities in drug-resistant cancers and those exploring RNA processing as a therapeutic target.

   Expertise I have expertise in RNA biology, alternative splicing and cancer models. My expertise is more limited in pharmacological dosing strategies and some aspects of in vivo xenograft modeling.

   Keywords: RNA biology, alternative splicing, spliceosome function, cancer biology, DNA damage response, transcriptomics

   Reviewer #2 (Evidence, reproducibility and clarity (Required)):

   Summary

   In the manuscript by Jamal-El-Hussein et al., the authors demonstrated that pladienolide B, an inhibitor of splicing factor 3B subunit 1 (SF3B1), inhibits cell viability in NSCLC cells and PDXs with resistance to platinum-based chemotherapy. Mechanistically, they identified that pladienolide B regulates splicing events of genes associated with DNA damage signaling and repair, specifically through exon skipping of MLH3, thus increasing vulnerability to chemotherapy. This study provides therapeutic insights into combining pladienolide B with chemotherapy for overcoming therapy resistance.

   __Major comments __

   The authors showed that both H460R and A549R cell lines are sensitive to pladienolide B; however, they exhibit distinct molecular responses. For example, SF3B1 knockdown downregulates the mRNA expression of ATR and PRKDC in H460R cells, while the expression of these genes remains unchanged in A549R cells (Fig. 4b). Furthermore, exon 8 skipping of MLH3 is not observed in A549R cells with pladienolide B treatment (Fig. 7a). These discrepancies suggest that the two cell lines respond to pladienolide B through different mechanisms. The authors should also perform a bulk RNA-seq on A549 cell line to better understand the different behavior of the two cell lines.

   We disagree with the remark regarding Fig 7a as this figure demonstrated MLH3 exon 8 skipping in A549R cells also after 6 and 24 hours pladienolide B treatment. __Although we agree with the reviewer that some differences exist in term of molecular mechanisms regulated by pladienolide B or SF3B1 knock-down in H460R and A549R cell lines, we identified in this study ATR, DNA-PKcs and MLH3 as novel targets of SF3B1 in both cell lines as well as more importantly in NSCLC PDXs. To our knowledge, this is the first evidence that SF3B1 inhibition negatively impacts ATR or DNA-PKcs and regulates MLH3 alternative splicing in solid tumors. In addition, we found a significant correlation between SF3B1 and DNA-PKCs or ATR protein levels in 77 NSCLC cell lines (Fig 4c) which supports a close relationship between these proteins in lung cancer. As also discussed in the response to reviewer 1, the differences between pladienolide B and SF3B1 knock-down could be related to the distinct timepoints at which we analyzed their effects as well as to the different mechanisms of action between pharmacological and siRNA inhibition. Nevertheless, we propose to acknowledge and discuss more in details the discrepancies between cellular models and pladienolide B or SF3B1 knock-down in the revised version of the manuscript. __

   The authors' central claim is that platinum-based chemotherapy-resistant cells are more sensitive to pladienolide B treatment.

   __We thank the reviewer for this remark. In this study, we did not want to claim that platinum-salts resistant cells exhibit unique vulnerability to pladienolide B as we fully agree with the reviewer that pladienolide B also exhibits cytotoxic effects in parental sensitive cells, although with a delayed kinetic. We acknowledge that the way we introduced our results section in the former manuscript could have contributed to such a misunderstanding. Rather, we propose a model in which pladienolide B induces cell death in NSCLC cells by massively impairing the expression of key components of the DNA damage and repair signaling pathways, at the transcriptional and/or splicing level. As NSCLC cells with acquired resistance to platinum salts are likely more “addict” to these pathways for their survival than parental cells, this might explain enhanced susceptibility of these resistant cells to pladienolide B-induced cell death. We think that these results highlight spliceosome targeting compounds as an alternative therapeutic strategy in NSCLC patients who escape chemotherapy. According to the remark of the reviewer, we will modify the way we introduce our results and we will clarify all these points in the discussion section of the revised manuscript. __

   In their bulk RNA-seq analysis, the authors selected genes commonly regulated by pladienolide B in both parental and resistant cells for further validation. This approach raises a critical concern: the observed sensitivity to pladienolide B may already be present in parental cells rather than representing a mechanism uniquely acquired by the resistant cells. To substantiate their central claim, the authors should analyze the differentially expressed genes between parental and resistant cells to identify resistance-specific molecular alterations that may confer enhanced sensitivity to pladienolide B, thereby providing a more mechanistically rigorous basis for their conclusions.

   We thank the reviewer for this remark and we agree with him/her. Indeed, after obtaining RNA-Seq data, we initially looked for genes which differential expression and/or splicing upon pladienolide B treatment could be only observed in resistant cells but not parental ones. We initially focused on the 121 genes belonging to the DNA repair pathways full network (WikiPathway WP4946) since Gene-Ontology analyses demonstrated enrichment of genes involved in DNA metabolic process, which includes DNA repair, among genes down-regulated after pladienolide B treatment (Fig S3). The focus on DNA repair was also justified by our observation showing accumulation of DNA double strand breaks upon pladienolide B treatment in H460 resistant cells (Fig 2d-f). Doing this comparison, we showed that pladienolide B regulates the expression of 41 (33%) and 47 (39%) genes of this network in H460S and H460R cells respectively, ____which were mostly down-regulated in both H460S (33/41) and H460R (33/47) cells (Table 3). Fifteen genes involved in all DNA repair processes were found to be specifically down-regulated in H460R cells upon pladienolide B treatment, including PARP-1. As a whole, these results demonstrated that pladienolide B down-regulates the expression of numerous DNA repair genes in both NSCLC parental and resistant cells. However, and as discussed above, we propose that the enhanced sensitivity of resistant cells to pladienolide B is related to their increased dependency for their survival to functional DNA repair pathways.

   __When differentially spliced genes were considered, and focusing on exon skipping events, as they were the more prominent (Fig 5e), we found that pladienolide B regulates the splicing of 107 and 87 genes of the WikiPathway WP4946 in H460 parental and resistant cells, respectively (Table 5). Forty five genes were predicted to be regulated in both cell lines. Only 3 genes, namely DCLRE1C, *POLD3 *and PNKP, were predicted to be differentially spliced upon pladienolide B treatment in H460R cells only. Trying to increase the number of genes to study, we extended our analysis to genes belonging to another DNA repair database (Human DNA Repair Genes, Resources from Wood laboratory, UT MD Anderson), and we found six additional genes, namely MLH3, MSH5, RAD54L, EME1, SETMAR and SMC6, that were also predicted to be differentially spliced in H460R cells only. However, four of these exon skipping events (i.e. PNKP-Ex9, DCLRE1C-Ex11, SETMAR-Ex2, SMC6-Ex6) were not validated and we did not observe clear difference between H460 resistant and parental cells for the others (Fig 5g and Fig S6). Skipping of MLH3-Ex8 was the sole event displaying a slight difference between both cell lines, mainly in term of kinetic of recovery. This is why we decided to further analyze this specific splicing event. Noteworthy, we focused only on genes involved in DNA damage and repair signaling pathways. Therefore, we cannot exclude that genes involved in other biological processes might be differentially transcribed or spliced in response to pladienolide B in H460 parental and resistant cells. __

   The authors conclude that pladienolide B treatment correlates with activation of DNA-PKcs signaling followed by a shutdown of ATR and DNA-PKcs pathways. However, the data presented do not fully support this interpretation. P-ATR levels are already elevated in resistant cells and remain unchanged following pladienolide B treatment (Fig. 3a). However, prolonged pladienolide B treatment leads to decreased total ATR protein and mRNA expression (Fig. 3e-f), suggesting that pladienolide B maintains an initial constitutive ATR activation followed by transcriptional downregulation. Since pladienolide B is a splicing inhibitor, the authors should determine whether ATR and DNA-PKcs mRNA downregulation is a direct consequence of aberrant splicing of their transcripts, or a non-specific effect of prolonged cellular toxicity. To strengthen their mechanistic conclusions, the authors should perform time-course experiments to establish the temporal relationship between these signaling events, analyze splicing changes specifically in ATR and DNA-PKcs transcripts (are they in the differential genes from the bulk RNA-seq analysis?), and also check the downstream targets of the ATR and DNA-PKcs signaling.

   __We thank the reviewer for his/her comment. In our RNA-Seq analyses, we did not recover PRKDC among the differential genes expressed or spliced upon pladienolide B treatment whatever the cell line. However, PRKDC was also not in the full RNA-Seq data list of not significant genes. Therefore, it remains unclear whether PRKDC splicing could account for the decrease of PRKDC mRNA level upon pladienolide B treatment. Concerning ATR, it was not in the RNA-Seq data list of the genes significantly up- or down-regulated upon pladienolide B treatment in either H460 parental or resistant cells. However, transcriptomic analyses were performed after 8 hours pladienolide B treatment while the decrease of ATR mRNA was observed after 24 hours (Fig 3f). Regarding splicing, *ATR *was predicted to be spliced, skipping of exon 30, upon pladienolide B treatment in both H460 parental and resistant cells. We validated this splicing event after 8 hours treatment with pladienolide B in both H460 cellular models but we did not analyze this splicing event at later timepoints, nor in the A549 parental or resistant cells. Therefore, and according to the remarks of the reviewer, we propose to deepen the temporal relationships between all these signaling events by performing time-course experiments for analysis of ATR exon 30 splicing by RT-PCR, ATR and PRKDC mRNA levels by RT-qPCR, and expression of downstream targets of ATR and DNA-PKcs, such as P-CHK1(Ser345) or P-RPA32(Ser4/8) by immunoblotting. __

   The authors report that skipping of exon 8 of MLH3 leads to the complete absence of the protein (Fig. 7d). However, this observation needs further clarification, as at least two alternative explanations exist. First, the antibody used to detect MLH3 may specifically recognize an epitope encoded by exon 8 or downstream exons, in which case the loss of signal would reflect antibody incompatibility rather than true protein absence.

   __We thank the reviewer for this remark. The anti-MLH3 antibody recognizes the C-terminal part of the MLH3 full-length protein (between amino acids 1228-1453). MLH3 exon 8 is 72 base pair and does not encode for the amino acids recognized by the anti-MLH3 antibody. In ENSEMBL, the MLH3-201 transcript encodes for the full-length protein (1453 amino acids) and the MLH3-202 transcript encodes for a MLH3 protein (1429 amino acids) devoid of the amino acids encoded by exon 8. Nevertheless, the two products have the same C-terminus recognized by the anti-MLH3 antibody used in this study. So, the loss of the signal depicted in Figure 7d is not due to antibody incompatibility. __

   Second, skipping of exon 8 may introduce a premature stop codon, triggering nonsense-mediated mRNA decay (NMD) and consequent loss of the transcript. To distinguish between these possibilities, the authors should perform qPCR using primers targeting sequences both upstream and downstream of the skipped exon, as well as consider NMD inhibition experiments, to clarify whether the observed protein loss occurs at the transcriptional or translational level.

   __As shown in Figure 8f, we demonstrated by RT-qPCR that pladienolide B alone or the combination of pladienolide B with cisplatin does not negatively impact* MLH3* mRNA level in both H460R and A549R cells. These results were confirmed in a time-course experiment of pladienolide B treatment performed in H460 and A549 parental and resistant cells, as well as in NSCLC PDXs. Similar results were obtained in H460R or A549R cells deprived of SF3B1. These new data will be added in the revised version of the manuscript. The couple of primers we used for MLH3 amplification was located downstream of exon 8, respectively on constitutive MLH3-exon 9 (forward primer) and MLH3-exon 10 (reverse primer). These results indicate that pladienolide B regulates MLH3 splicing but not MLH3 total mRNA level. Considering NMD, in the FASTER DB database, none of the MLH3 transcripts devoid of exon 8 are predicted to be degraded by NMD. So we do not think that pladienolide B-induced MLH3 exon 8 skipping promotes the synthesis of transcripts recognized by the NMD machinery. As discussed above, the anti-MLH3 antibody does not allow to distinguish between the full length MLH3 protein and the MLH3 product encoded by transcript devoid of exon 8. In addition, only 24 amino acids (around 2-3KDa) differentiate both products which could render difficult their specific detection in SDS-PAGE. So, we speculate that the decrease of MLH3 signal detected by immunoblotting in pladienolide B-treated and SF3B1 knocked-down cells is mostly related to the decrease of MLH3 full-length protein due to the decreased level of MLH3 transcript retaining exon 8 and encoding MLH3 full-length protein. Alternatively, and not exclusively, MLH3 product devoid of exon 8 might also be less stable. __

   The difference shown in Fig. 6b after pladienolide B treatment decreases from 2.2% to 1%. This raises concern about whether the observed difference reflects a true biological effect or is confounded by technical limitations such as low transfection efficiency. The authors should consider optimizing their transfection conditions to achieve a more robust and convincing result.

   We agree with the reviewer’s comment. However, using this SceI-inducible system to analyze DNA double strand breaks repair by homologous recombination, it is very frequent to have only a very low percentage of cells able to perform homologous recombination thereby expressing the GFP protein ____(as examples: Yoshino Y et al., Sci Reports, 2019; Brustel et al., Sci Rep., 2018; Croglio et al., Oncotarget, 2016; Mamouni et al., Mol Cell Biol., 2014). This is why the difference is low between each condition but it is significant. Indeed, these engineered cellular models are not easy to manipulate as we first need to obtain stable clones having incorporated the PBL174 pDR-GFP-plasmid and then to transiently transfect them using a second plasmid encoding the SceI enzyme which creates DNA Double Strand Breaks. The efficiency of the second round of transfection might therefore be decreased as the cells already experienced a first round of transfection.

   __Minor comments __

   The abbreviation "S" in H460S and A549S cells is not defined in the manuscript. As this designation is used throughout the text, the authors should clarify what "S" denotes upon its first appearance.

   __We thank the reviewer for this remark. We will correct the text. __

   The current presentation of Figure S1a does not clearly demonstrate that different NSCLC cell lines exhibit differential sensitivity to pladienolide B. The authors should calculate and report IC50 values for each cell line to enable a more rigorous and quantitative comparison of their respective dose-response relationships.

   We thank the reviewer for this remark and agree with it. We will calculate and report in the revised version of Fig S1a the IC50 for pladienolide B for each cell line.

   In Figure 4c, two dashed black lines are present in the plot but are not described or explained in the figure legend or the main text.

   We thank the reviewer for this remark. The two dashed black lines represent the upper and lower boundaries of the 95% confidence interval for the fitted linear regression line. We have now clarified their meaning in the revised figure legend and indicated section of the main text also.

   In Figure 5a, the authors combine the downregulated and upregulated genes in a single Venn diagram. This approach may obscure biologically meaningful differences, as overlapping genes between conditions could reflect opposing directions of regulation. The authors should separate upregulated and downregulated genes into distinct Venn diagrams to provide a more accurate and interpretable comparison.

   We thank the reviewer for this remark and agree with it. Hence, in Fig 5a-b and Fig S3, we already highlighted the number of genes down-regulated or up-regulated upon pladienolide B treatment in either H460 parental and resistant cells using bar graphs. We will provide new Venn diagrams separating up-regulated and down-regulated genes for both cell lines.

   In the figure legend of Fig. 6a, the panel is incorrectly described as a "quantification." As the panel depicts a schematic representation of the experimental construct rather than numerical data, the term "illustration" or "schematic" would be more accurate and should be used instead.

   We thank the reviewer for this remark and we agree with it. We will modify the legend of Figure 6a accordingly.

   Reviewer #2 (Significance (Required)):

   This manuscript provides evidence that pladienolide B can overcome chemotherapy resistance in NSCLC by modulating splicing events of genes associated with DNA damage signaling and repair. Although the underlying mechanism requires further elucidation, this study offers valuable mechanistic insights into how aberrant splicing regulates therapy resistance, with potential implications for the development of novel therapeutic strategies targeting splicing factors in chemotherapy-resistant cancers.

   My research field is in tumor heterogeneity and tumor microenvironment.

   Read the original source
2. 

   Review Commons

   Jul 27, 2026

   Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Referee #2

   Evidence, reproducibility and clarity

   Summary

   In the manuscript by Jamal-El-Hussein et al., the authors demonstrated that pladienolide B, an inhibitor of splicing factor 3B subunit 1 (SF3B1), inhibits cell viability in NSCLC cells and PDXs with resistance to platinum-based chemotherapy. Mechanistically, they identified that pladienolide B regulates splicing events of genes associated with DNA damage signaling and repair, specifically through exon skipping of MLH3, thus increasing vulnerability to chemotherapy. This study provides therapeutic insights into combining pladienolide B with chemotherapy for overcoming therapy resistance.

   Major comments

   1. The authors showed that …

   Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Referee #2

   Evidence, reproducibility and clarity

   Summary

   In the manuscript by Jamal-El-Hussein et al., the authors demonstrated that pladienolide B, an inhibitor of splicing factor 3B subunit 1 (SF3B1), inhibits cell viability in NSCLC cells and PDXs with resistance to platinum-based chemotherapy. Mechanistically, they identified that pladienolide B regulates splicing events of genes associated with DNA damage signaling and repair, specifically through exon skipping of MLH3, thus increasing vulnerability to chemotherapy. This study provides therapeutic insights into combining pladienolide B with chemotherapy for overcoming therapy resistance.

   Major comments

   1. The authors showed that both H460R and A549R cell lines are sensitive to pladienolide B; however, they exhibit distinct molecular responses. For example, SF3B1 knockdown downregulates the mRNA expression of ATR and PRKDC in H460R cells, while the expression of these genes remains unchanged in A549R cells (Fig. 4b). Furthermore, exon 8 skipping of MLH3 is not observed in A549R cells with pladienolide B treatment (Fig. 7a). These discrepancies suggest that the two cell lines respond to pladienolide B through different mechanisms. The authors should also perform a bulk RNA-seq on A549 cell line to better understand the different behavior of the two cell lines.
   2. The authors' central claim is that platinum-based chemotherapy-resistant cells are more sensitive to pladienolide B treatment. However, in their bulk RNA-seq analysis, the authors selected genes commonly regulated by pladienolide B in both parental and resistant cells for further validation. This approach raises a critical concern: the observed sensitivity to pladienolide B may already be present in parental cells rather than representing a mechanism uniquely acquired by the resistant cells. To substantiate their central claim, the authors should analyze the differentially expressed genes between parental and resistant cells to identify resistance-specific molecular alterations that may confer enhanced sensitivity to pladienolide B, thereby providing a more mechanistically rigorous basis for their conclusions.
   3. The authors conclude that pladienolide B treatment correlates with activation of DNA-PKcs signaling followed by a shutdown of ATR and DNA-PKcs pathways. However, the data presented do not fully support this interpretation. P-ATR levels are already elevated in resistant cells and remain unchanged following pladienolide B treatment (Fig. 3a). However, prolonged pladienolide B treatment leads to decreased total ATR protein and mRNA expression (Fig. 3e-f), suggesting that pladienolide B maintains an initial constitutive ATR activation followed by transcriptional downregulation. Since pladienolide B is a splicing inhibitor, the authors should determine whether ATR and DNA-PKcs mRNA downregulation is a direct consequence of aberrant splicing of their transcripts, or a non-specific effect of prolonged cellular toxicity. To strengthen their mechanistic conclusions, the authors should perform time-course experiments to establish the temporal relationship between these signaling events, analyze splicing changes specifically in ATR and DNA-PKcs transcripts (are they in the differential genes from the bulk RNA-seq analysis?), and also check the downstream targets of the ATR and DNA-PKcs signaling.
   4. The authors report that skipping of exon 8 of MLH3 leads to the complete absence of the protein (Fig. 7d). However, this observation needs further clarification, as at least two alternative explanations exist. First, the antibody used to detect MLH3 may specifically recognize an epitope encoded by exon 8 or downstream exons, in which case the loss of signal would reflect antibody incompatibility rather than true protein absence. Second, skipping of exon 8 may introduce a premature stop codon, triggering nonsense-mediated mRNA decay (NMD) and consequent loss of the transcript. To distinguish between these possibilities, the authors should perform qPCR using primers targeting sequences both upstream and downstream of the skipped exon, as well as consider NMD inhibition experiments, to clarify whether the observed protein loss occurs at the transcriptional or translational level.
   5. The difference shown in Fig. 6b after pladienolide B treatment decreases from 2.2% to 1%. This raises concern about whether the observed difference reflects a true biological effect or is confounded by technical limitations such as low transfection efficiency. The authors should consider optimizing their transfection conditions to achieve a more robust and convincing result.

   Minor comments

   1. The abbreviation "S" in H460S and A549S cells is not defined in the manuscript. As this designation is used throughout the text, the authors should clarify what "S" denotes upon its first appearance.
   2. The current presentation of Figure S1a does not clearly demonstrate that different NSCLC cell lines exhibit differential sensitivity to pladienolide B. The authors should calculate and report IC50 values for each cell line to enable a more rigorous and quantitative comparison of their respective dose-response relationships.
   3. In Figure 4c, two dashed black lines are present in the plot but are not described or explained in the figure legend or the main text.
   4. In Figure 5a, the authors combine the downregulated and upregulated genes in a single Venn diagram. This approach may obscure biologically meaningful differences, as overlapping genes between conditions could reflect opposing directions of regulation. The authors should separate upregulated and downregulated genes into distinct Venn diagrams to provide a more accurate and interpretable comparison.
   5. In the figure legend of Fig. 6a, the panel is incorrectly described as a "quantification." As the panel depicts a schematic representation of the experimental construct rather than numerical data, the term "illustration" or "schematic" would be more accurate and should be used instead.

   Significance

   This manuscript provides evidence that pladienolide B can overcome chemotherapy resistance in NSCLC by modulating splicing events of genes associated with DNA damage signaling and repair. Although the underlying mechanism requires further elucidation, this study offers valuable mechanistic insights into how aberrant splicing regulates therapy resistance, with potential implications for the development of novel therapeutic strategies targeting splicing factors in chemotherapy-resistant cancers.

   My research field is in tumor heterogeneity and tumor microenvironment.

   Read the original source
3. 

   Review Commons

   Jul 27, 2026

   Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Referee #1

   Evidence, reproducibility and clarity

   In this study, the authors investigate the effects of pharmacological inhibition of the spliceosome using the SF3B1 inhibitor pladienolide B in models of platinum-resistant non-small cell lung cancer (NSCLC). Using a combination of cell lines, platinum-resistant derivatives, and patient-derived xenograft (PDX) models, the authors show that spliceosome inhibition sensitizes platinum-resistant tumors to treatment and leads to increased DNA damage accumulation and impaired DNA damage response signaling. Transcriptomic analyses indicate that transcripts encoding DNA damage regulators are particularly sensitive to alternative splicing …

   Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Referee #1

   Evidence, reproducibility and clarity

   In this study, the authors investigate the effects of pharmacological inhibition of the spliceosome using the SF3B1 inhibitor pladienolide B in models of platinum-resistant non-small cell lung cancer (NSCLC). Using a combination of cell lines, platinum-resistant derivatives, and patient-derived xenograft (PDX) models, the authors show that spliceosome inhibition sensitizes platinum-resistant tumors to treatment and leads to increased DNA damage accumulation and impaired DNA damage response signaling. Transcriptomic analyses indicate that transcripts encoding DNA damage regulators are particularly sensitive to alternative splicing perturbations, and selected mechanistic experiments suggest involvement of specific regulators such as MLH3. The study further explores links between splicing inhibition, transcriptional activity, and cell cycle progression.

   Overall, the manuscript presents extensive datasets across multiple experimental systems and provides strong evidence that spliceosome inhibition can sensitize platinum-resistant tumors to DNA damage. However, several aspects of the mechanistic interpretation, data consistency, and presentation require clarification or strengthening to fully support the central claims.

   Major comments

   Conceptual clarity and synthesis of mechanistic model

   The manuscript presents multiple mechanistic observations-including altered splicing of DNA repair genes, increased DNA damage accumulation, transcriptional perturbation, and cell cycle changes-but these are not integrated into a coherent conceptual framework. While it is reasonable that not all mechanistic details are fully resolved, the current presentation leaves the reader uncertain about the relative contributions of these processes. A clearer synthesis of the proposed mechanism, possibly including a summary model figure, would substantially improve the conceptual clarity of the study.

   Biological specificity of platinum-resistant cell sensitivity

   A central premise of the study is that platinum-resistant cells exhibit enhanced sensitivity to spliceosome inhibition. However, in several experiments (e.g., cell cycle analysis in Figure 2A), similar responses to pladienolide B appear to occur in both platinum-sensitive and resistant cells. This observation complicates the interpretation that resistant cells exhibit uniquely distinct vulnerability. The authors should clarify how these findings align with the proposed model and more explicitly distinguish shared versus resistance-specific responses.

   Heterogeneity in PDX responses and lack of platinum-sensitive controls

   The PDX experiments represent a major strength of the study. However, resistant tumors display heterogeneous responses to pladienolide treatment, suggesting the presence of additional determinants of sensitivity. Including platinum-sensitive PDX tumors, if available, would provide valuable baseline comparison and strengthen interpretation of resistance-specific effects. If not feasible, the limitations should be acknowledged and discussed.

   (OPTIONAL - would strengthen study but may require substantial additional work.)

   Consistency between pharmacological inhibition and genetic depletion

   In Figure 4, the authors compare pladienolide treatment with SF3B1 knockdown to demonstrate target specificity. However, the effects observed with the two perturbations are not entirely consistent-for example, pladienolide affects phosphorylation of DNA-PKcs, while SF3B1 knockdown appears to produce broader effects at both protein and mRNA levels. Additionally, differences are observed between resistant cell lines in the response to SF3B1 knockdown. These discrepancies should be addressed and discussed, as they may reflect mechanistic differences between acute pharmacological inhibition and genetic depletion.

   Selection and interpretation of splicing-sensitive transcripts

   Transcriptomic analyses in Figure 5 identify both shared and differential splicing changes between sensitive and resistant cells. However, much of the analysis focuses on transcripts that are commonly affected in both conditions, rather than those uniquely altered in resistant cells. Given that the central phenotype is resistance-specific sensitivity, transcripts uniquely mis-spliced in resistant cells may represent more informative candidates. The authors should clarify the rationale behind focusing on shared events and discuss the implications of resistance-specific versus common splicing changes.

   Transient nature of splicing effects

   The authors report transient alternative splicing effects upon prolonged pladienolide treatment. This observation is counterintuitive, as continued spliceosome inhibition might be expected to produce cumulative splicing defects. While the authors reference studies showing that transient inhibition can produce lasting effects, the current observations involve continuous exposure. This apparent discrepancy should be clarified and discussed.

   Use of unrelated cell lines in reporter assays

   The DNA damage reporter assays (Figure 6A-D) appear to be performed in cell lines not directly linked to platinum sensitivity or resistance. Given the central importance of resistance-specific responses, repeating key reporter assays in both sensitive and resistant paired models would strengthen the conclusions.

   (OPTIONAL - likely moderate experimental effort.)

   Interpretation of MLH3 splicing results

   In Figure 7, differences between pharmacological inhibition and SF3B1 knockdown in MLH3 exon 8 regulation are not entirely consistent. Furthermore, inclusion levels of regulated and non-regulated exons appear similarly correlated with SF3B1 expression, potentially weakening the argument for exon-specific regulation. These observations should be clarified.

   Combination treatment logic

   In Figure 8, co-treatment experiments with pladienolide B and cisplatin are performed primarily in resistant cells. Performing similar experiments in platinum-sensitive cells would provide an important reference point to distinguish additive versus resistance-specific effects.

   (OPTIONAL - moderate experimental effort.)

   Minor comments

   • In Figure 1, pladienolide treatment in platinum-sensitive cells appears to plateau at approximately 50% cell killing. Extending the concentration range may help clarify whether maximal efficacy was reached.
   • In Figure 1H, the difference between 2.5 mg/kg and 5 mg/kg pladienolide in PDX models appears disproportionately large relative to the dose change. This should be discussed or experimentally clarified.
   • In Figure 5, differential expression and splicing analyses are presented using multiple cutoffs (e.g., log₂FC > 0.4 and >1; ΔPSI thresholds). This introduces redundancy and may obscure key findings. A single well-justified cutoff would improve clarity.
   • Several isoform-specific RT-PCR gels are difficult to interpret due to low image clarity. Improving gel presentation or focusing on key timepoints would strengthen data readability.
   • Some figure panels appear redundant, showing similar datasets under different analysis thresholds.
   • A graphical summary model illustrating the proposed mechanism would improve reader comprehension.
   • In Figure 3D, representative images of γH2AX foci appear visually similar across conditions, whereas quantification shows large differences. The authors should ensure that representative images accurately reflect quantified trends and clarify selection criteria for displayed images.

   Significance

   This study represents a comprehensive investigation of spliceosome inhibition as a therapeutic strategy to overcome platinum resistance in NSCLC. The use of resistant cell lines, PDX models, and functional reporters provides strong experimental depth. The most compelling aspects of the study include the demonstration that spliceosome inhibition enhances DNA damage accumulation and sensitizes resistant tumors to platinum-based therapies. However, several mechanistic interpretations require clarification, and data presentation could be streamlined to improve logical coherence.

   Advance

   The work provides evidence that targeting spliceosome function-specifically via SF3B1 inhibition-can sensitize platinum-resistant tumors to DNA-damaging agents. To my knowledge, this represents one of the first comprehensive demonstrations that spliceosome-targeting compounds can effectively overcome acquired platinum resistance in solid tumor models. The study also contributes to the emerging understanding that DNA damage response transcripts may represent particularly sensitive targets of splicing perturbation.

   Audience

   The study will be of interest to researchers in RNA biology, cancer therapeutics, DNA damage response, and translational oncology. It is particularly relevant to scientists investigating therapeutic vulnerabilities in drug-resistant cancers and those exploring RNA processing as a therapeutic target.

   Expertise

   I have expertise in RNA biology, alternative splicing and cancer models. My expertise is more limited in pharmacological dosing strategies and some aspects of in vivo xenograft modeling.

   Keywords:

   RNA biology, alternative splicing, spliceosome function, cancer biology, DNA damage response, transcriptomics

   Read the original source
4. 

   Version published to 10.64898/2026.02.17.706284 on bioRxiv

   Feb 19, 2026