Leveraging a disulfidptosis-based signature to characterize heterogeneity and optimize treatment in multiple myeloma.

BACKGROUND

Disulfidptosis is an emerging type of programmed cell death related to ROS accumulation and aberrant disulfide bond formation. Multiple myeloma (MM) is the second most prevalent hematologic malignancy characterized by a high synthesis rate of disulfide bond-rich proteins and chronic oxidative stress. However, the relationship between disulfidptosis and MM is still unclear.

METHODS

Using the non-negative matrix factorization and lasso algorithm, we constructed the disulfidptosis-associated subtypes and the prognostic model on the GEO dataset. We further explored genetic mutation mapping, protein-protein interactions, functional enrichment, drug sensitivity, drug prediction, and immune infiltration analysis among subtypes and risk subgroups. To improve the clinical benefits, we combined risk scores and clinical metrics to build a nomogram. Finally, experiments examined the expression patterns of disulfidptosis-related genes (DRGs) in MM.

RESULTS

By cluster analysis, we obtained three subtypes with C2 having a worse prognosis than C3. Consistently, C2 exhibited significantly lower sensitivity to doxorubicin and lenalidomide, as well as a higher propensity for T-cell depletion and a non-responsive state to immunotherapy. Similarly, in the subsequent prognostic model, the high-scoring group had a worse prognosis and a higher probability of T-cell dysfunction, immunotherapy resistance, and cancer cell self-renewal. DRGs and risk genes were widely mutated in cancers. Subtypes and risk subgroups differed in ROS metabolism and the p53 signaling pathway. We further identified eight genes differentially expressed in risk subgroups as drug targets against MM. Then 27 drugs targeting the high-risk group were predicted. Based on the DRGs and risk genes, we constructed the miRNA and TF regulatory networks. The nomogram of combined ISS, age, and risk score showed good predictive performance. qRT-PCR of cell lines and clinical specimens provided further support for prognostic modeling.

CONCLUSION

Our research reveals the prognostic value of disulfidptosis in MM and provides new perspectives for identifying heterogeneity and therapeutic targets.