Review Reports

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript is well written, and the data are clearly presented. However, a limitation is the use of only a single cell line (KMS-28-BM) in the validation experiments. Using additional multiple myeloma cell lines would strengthen the findings and support the results. In addition, several aspects of the manuscript would benefit from further clarification or improvement:

1. In the introduction part related to mechanisms underlying drug resistance, authors should aim to include original works, for example: Second, MM cells can upregulate proteasome activity through the overexpression of proteasome subunits in response to PI treatment – PMID: 19225532. Moreover, since authors identify in CRISPR screen genes involved in oxidative phosphorylation, already described changes in cellular metabolism related to response to PI treatment should be included, such as: PMID: 27118406, 30792209, 28855983.
2. The authors report using a p-value threshold of < 0.01 for statistical significance in the CRISPR screen. This approach is relatively uncommon, as standard practice typically involves correction for multiple hypothesis testing, such as using a false discovery rate (FDR) < 0.05. The authors should provide justification for this lower stringency.
3. While the data reveal candidate genes associated with sensitivity and resistance to bortezomib and carfilzomib, validation in additional MM cell lines beyond KMS-28-BM is critical. This is particularly important to determine whether the observed role of NUDCD2 in PI resistance is specific to this cell line or not.
4. Figure revisions:

Figure 1E and F: Please indicate the treatment conditions directly in the figure.

Figure 2B: Clarify in the figure legend that the cells were treated with bortezomib.

Figure 3: Clearly specify whether cells were treated with carfilzomib or ixazomib.

Figure 5B: The volcano plots are difficult to interpret. Please consider alternative visualizations such as a heatmap to represent upregulated and downregulated genes.

Figure 5C–E: Indicate treatment conditions explicitly—i.e., untreated KMS-28-BM cells (C), carfilzomib-treated (D), and bortezomib-treated (E).

5. The discussion should include a part addressing the observed decrease in drug response following PSMC4 gene knockout. This finding aligns with a body of literature indicating that subunits of the 19S proteasome contribute to drug resistance in multiple myeloma and other cancer types (PMID: 26327694, 26327695, 28028240).

6. The concluding statement that these results could contribute to the development of novel specific biomarkers appears to overstate the findings. Since all experiments were conducted in a single MM cell line and no patient-derived data are provided, this claim should be changed. The authors should either validate the role of NUDCD2 knockout in additional models or acknowledge the limitations of their study, noting that further validation, especially in patient-derived MM cells,is necessary for the use of biomarkers.

Author Response

Comment 1: In the introduction part related to mechanisms underlying drug resistance, authors should aim to include original works, for example: Second, MM cells can upregulate proteasome activity through the overexpression of proteasome subunits in response to PI treatment – PMID: 19225532. Moreover, since authors identify in CRISPR screen genes involved in oxidative phosphorylation, already described changes in cellular metabolism related to response to PI treatment should be included, such as: PMID: 27118406, 30792209, 28855983.

Answer 1: We have added original works to the introduction (line numbers: 81-91), including the reference PMID 19225532 (line number: 84). We have added a part regarding alterations in cellular metabolism as a resistance mechanism and included the above-mentioned references (line numbers: 90-91).

 

Comment 2: The authors report using a p-value threshold of < 0.01 for statistical significance in the CRISPR screen. This approach is relatively uncommon, as standard practice typically involves correction for multiple hypothesis testing, such as using a false discovery rate (FDR) < 0.05. The authors should provide justification for this lower stringency.

Answer 2: Correction for multiple testing was indeed performed. Unfortunately, an (typographical) error was made reporting the stringency. We corrected the information about the stringency (line number: 155).

 

Comment 3: While the data reveal candidate genes associated with sensitivity and resistance to bortezomib and carfilzomib, validation in additional MM cell lines beyond KMS-28-BM is critical. This is particularly important to determine whether the observed role of NUDCD2 in PI resistance is specific to this cell line or not.

Answer 3: We fully agree that a more extensive validation is required in multiple myeloma cell lines. Unfortunately, this is not feasible anymore for us. However, we formulated this limitation more clearly in the discussion (line numbers: 508-510).

 

Comment 4: Figure revisions:

Figure 1E and F: Please indicate the treatment conditions directly in the figure.

Answer: We have indicated the treatment conditions directly in figures 1E and 1F.

Figure 2B: Clarify in the figure legend that the cells were treated with bortezomib.

Answer: The figure legend of Figure 2B already indicates that the cells were treated with bortezomib. However, we have now also indicated this in the figure itself to clarify the treatment condition.

Figure 3: Clearly specify whether cells were treated with carfilzomib or ixazomib.

Answer: The figure legend of Figure 3 already indicates whether the cells were treated with carfilzomib or ixazomib. We have now also indicated this in the figure itself to clarify the treatment conditions.

Figure 5B: The volcano plots are difficult to interpret. Please consider alternative visualizations such as a heatmap to represent upregulated and downregulated genes.

Answer: We have removed the volcano plots and have created a heatmap representing the differentially expressed genes (DEGs) that have both a significant log2FC and adjusted P-value.

Figure 5C–E: Indicate treatment conditions explicitly—i.e., untreated KMS-28-BM cells (C), carfilzomib-treated (D), and bortezomib-treated (E).

Answer: We have indicated the treatment conditions in the figure itself (Figure 6A, 6B and 6C).

 

Comment 5: The discussion should include a part addressing the observed decrease in drug response following PSMC4 gene knockout. This finding aligns with a body of literature indicating that subunits of the 19S proteasome contribute to drug resistance in multiple myeloma and other cancer types (PMID: 26327694, 26327695, 28028240).

Answer 5: We have included a part in the discussion addressing that the observed decrease in drug response following PSMC4 KO is in line with published literature (line numbers: 454-457).

 

Comment 6: The concluding statement that these results could contribute to the development of novel specific biomarkers appears to overstate the findings. Since all experiments were conducted in a single MM cell line and no patient-derived data are provided, this claim should be changed. The authors should either validate the role of NUDCD2 knockout in additional models or acknowledge the limitations of their study, noting that further validation, especially in patient-derived MM cells,is necessary for the use of biomarkers.

Answer 6: Unfortunately, it is not feasible for us to perform further validation experiments in additional models. Therefore, as suggested, we now better indicate the limitations of our study (line numbers: 508-510). In the concluding statement, we acknowledge that the obtained results need to be validated in patient-derived MM cells (line numbers: 523-525).

Reviewer 2 Report

Comments and Suggestions for Authors

In this study entitled "Genome-wide CRISPR-Cas9 knockout screening identifies NUDCD2 depletion as sensitizer for bortezomib, carfilzomib and ixazomib in multiple myeloma" the authors used a genome-wide CRISPR-Cas9 knockout (KO) screening approach in KMS-28-BM MM cells to identify several genes that modulate sensitivity or resistance to PIs. And they also identified that KO of NUDCD2 sensitized myeloma cells to PIs. However, several issues need to be addressed:

1. The authors employed the IC50 concentration of proteasome inhibitors to identify both resistance and sensitizer genes in their CRISPR-Cas9 screen. However, previous studies(PMID: 35166240) have demonstrated that using low-dose drug exposure is more appropriate for identifying resistance genes via negative selection, while higher doses are more effective for detecting sensitizer genes via positive selection， Therefore, relying on a single concentration may limit the ability to distinguish between these two categories, and the authors should clarify and justify their dosing strategy accordingly.
2. It is acknowledged that the selection of KMS-28-BM cells for the CRISPR-Cas9 screen is substantiated by the assertion that they exhibit optimal transduction efficiency. Nevertheless, in consideration of the genetic heterogeneity inherent to MM, it is recommended to include several MM cell lines in the validation experiments.
3. It is evident that both bortezomib and carfilzomib are proteasome inhibitors with similar mechanisms of action; however, it is unclear why the overlap between the resistance and sensitization genes identified under treatment with these two drugs is limited.
4. The mechanistic link between NUDCD2 depletion and these transcriptional changes remains unclear; validation of key pathway alterations at the protein level is recommended.
5. Have the authors considered or explored the potential synergistic effects of combining NUDCD2 inhibition (pharmacological or genetic) with PIs in MM?
6. Please include statistical analysis and indicate significance in Figures 2 and 3.

 

Author Response

Comment 1: The authors employed the IC50 concentration of proteasome inhibitors to identify both resistance and sensitizer genes in their CRISPR-Cas9 screen. However, previous studies (PMID: 35166240) have demonstrated that using low-dose drug exposure is more appropriate for identifying resistance genes via negative selection, while higher doses are more effective for detecting sensitizer genes via positive selection. Therefore, relying on a single concentration may limit the ability to distinguish between these two categories, and the authors should clarify and justify their dosing strategy accordingly.

Answer 1: We used an IC50 concentration to identify both resistance and sensitizing genes in one single KO screen. Instead of performing two separate KO screens, one with a very high IC (e.g. 90%) to identify resistance genes and one with a very low IC (e.g. 20%) to identify sensitization genes, as this is more time consuming. In addition, a high IC can be too toxic for cells and a low IC can be too mild. We thoroughly optimized the concentration of the PIs to be used in the KO screens, as this is an important parameter, which influences the success of a CRISPR KO screen. The design of our CRISPR KO screens also took into account the duration of the screen, as a too brief treatment duration might miss resistant clones. Myeloma cells were cultured with a proteasome inhibitor for at least 20 cell doublings. Additionally, as the use of a proper untreated (vehicle) control is important to distinguish drug-specific effects, sgRNAs in the PI-treated cell population were compared to DMSO-treated cells to identify genes that are specifically involved in PI treatment. Importantly, the candidate genes of the CRISPR KO screens were validated individually afterwards.

 

Comment 2: It is acknowledged that the selection of KMS-28-BM cells for the CRISPR-Cas9 screen is substantiated by the assertion that they exhibit optimal transduction efficiency. Nevertheless, in consideration of the genetic heterogeneity inherent to MM, it is recommended to include several MM cell lines in the validation experiments.

Answer 2: We fully agree that a more extensive validation in myeloma cell lines is required to represent the heterogeneity of different MM subtypes. Unfortunately, this is not feasible anymore for us. Therefore, we indicated this limitation of our study in the discussion (line numbers: 508-510). 

 

Comment 3: It is evident that both bortezomib and carfilzomib are proteasome inhibitors with similar mechanisms of action; however, it is unclear why the overlap between the resistance and sensitization genes identified under treatment with these two drugs is limited.

Answer 3: We expected that there would be some overlap between the two screens, as they are both proteasome inhibitors with a similar mechanism of action. However, as expected, we identified also resistance and sensitization genes unique for bortezomib and carfilzomib, as bortezomib is a reversible boronate, while carfilzomib is an irreversible epoxy ketone. Both proteasome inhibitors inhibit the chymotrypsin-like activity of the 20S proteasome. However, bortezomib can also inhibit other protease activities at a higher concentration, while carfilzomib is more selective for the chymotrypsin-like site.

 

Comment 4: The mechanistic link between NUDCD2 depletion and these transcriptional changes remains unclear; validation of key pathway alterations at the protein level is recommended.

Answer 4: We agree that a key mechanistic link between NUDCD2 KO and the discovered transcriptional changes is missing and that further validation of the mechanism of action of NUDCD2 depletion is critical. Unfortunately, it is not feasible for us to perform additional validation experiments. We indicated this limitation more clearly in the discussion (line numbers: 505-507).

 

Comment 5: Have the authors considered or explored the potential synergistic effects of combining NUDCD2 inhibition (pharmacological or genetic) with PIs in MM?

Answer 5: Combination of NUDCD2 KO and a PI sensitizes myeloma cells to the PI, resulting in a higher loss of cell viability. It would be interesting to further explore potential synergistic effects, however, we have not looked into this.

 

Comment 6: Please include statistical analysis and indicate significance in Figures 2 and 3.

Answer 6: The figures shown depict N=1, therefore, it is not possible to perform statistical analysis to estimate significance. However, we have performed this experiment multiple times and show here a representative experiment. We have confirmed our results with multiple sgRNAs per gene as shown in Figures 2 and 3. We added to the figure legends of Figures 2 and 3 that a representative experiment is shown.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have corrected all of my suggestions. I recommend acceptance of the manuscript.

Author Response

The authors have corrected all of my suggestions. I recommend acceptance of the manuscript.

Answer: We are delighted that Reviewer 1 is pleased with our corrections made to the manuscript. We would like to thank Reviewer 1 for his recommendation of accepting the manuscript.