Proteomics analysis of histone deacetylase inhibitor-resistant solid tumors reveals resistant signatures and potential drug combinations

Abstract

Histone deacetylase inhibitors (HDACis) are important drugs for cancer therapy, but the indistinct resistant mechanisms of solid tumor therapy greatly limit their clinical application. In this study we conducted HDACi-perturbated proteomics and phosphoproteomics analyses in HDACi-sensitive and -resistant cell lines using a tandem mass tag (TMT)-based quantitative proteomic strategy. We found that the ribosome biogenesis proteins MRTO4, PES1, WDR74 and NOP16 vital to tumorigenesis might regulate the tumor sensitivity to HDACi. By integrating HDACi-perturbated protein signature with previously reported proteomics and drug sensitivity data, we predicted and validated a series of drug combination pairs potentially to enhance the sensitivity of HDACi in diverse solid tumor. Functional phosphoproteomic analysis further identified the kinase PDK1 and ROCK as potential HDACi-resistant signatures. Overall, this study reveals the potential HDACi-resistant signatures and may provide promising drug combination strategies to attenuate the resistance of solid tumor to HDACi.

Keywords: drug resistance; histone deacetylase inhibitor; proteomics; solid tumor; vorinostat (SAHA).

Fig. 1. TMT-based proteomics and phosphoproteomics analysis of the HDACi-sensitive and resistant cell lines treated with SAHA.

a The experimental design and data analysis. b The IC₅₀ value of the SAHA-sensitive and -resistant cells. c The number of phosphosites, phosphoproteins and protein groups identified in four TMT groups. d The number and proportion of identified phosphorylation sites at the amino acid of S, Y and T. e Pearson correlation value of common reference in the different TMT groups in the proteome and phosphoproteome data.

Fig. 2. GSEA analysis of proteome in the HDACi-sensitive and resistant cell lines treated with SAHA.

a Biological process analysis of proteome in the HDACi-sensitive and resistant cell lines treated with SAHA. Nodes indicate the enrichment of biological processes from all proteins. The biological processes with NES > 0 indicated the enriched processes in the sensitive group, and NES < 0 indicated the enriched processes in the resistant group. The network was visualized by the Cytoscape app. b KEGG analysis of proteome in the HDACi-sensitive and resistant cell lines treated with SAHA. c Hallmark analysis of proteome in the HDACi-sensitive and resistant cell lines treated with SAHA.

Fig. 3. Proteomics analysis of ribosome biogenesis and MYC target in the sensitive and resistant cell lines treated with SAHA.

GSEA enrichment plots of biological process ribosome biogenesis (a) and hallmark MYC_TARGET_V2 (b). c The overlap of proteins enriched in ribosome biogenesis and hallmark MYC_TARGET_V2. d The changes of 15 shared proteins in ribosome biogenesis and hallmark MYC_TARGET_V2 between HDACi sensitive and resistant cells treated with SAHA. e The protein expression of four proteins MRTO4, NOP16, PES1 and WDR74 in NAT (normal tissue adjacent to the tumor) and tumor tissue of NSCLC. ***P < 0.001. f Survival analysis of NSCLC patients with high and low expression of PES1 protein.

Fig. 4. The prediction and validation of potential combined drugs with HDACi based on proteomics data.

a The differential proteins between sensitive and resistant group after SAHA treatment. b Workflow of predicting potential combined drugs of HDACi. c The proportion of top 30 drugs predicted by PDCP methods. d Correlation of proteomics data from CCLE database and drug sensitivity data from GDSC database. e Validation of potential drug combination effect on the NCI-H226, A549 and MDA-MB-468 cell lines. *P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 5. Functional phosphoproteomics analysis of differential phosphorylation sites in the sensitive and resistant cells after SAHA treatment.

a Volcano map of differential regulated phosphorylation site in the sensitive and resistant cell lines after SAHA treatment. The up/down-regulated modification sites (foldchange >1.5 and P < 0.05) at the phosphorylation level were marked with red/blue. The up/down-regulated proteins at the proteome level (fold change >1.2 and P < 0.05) were marked with light red/light blue. Solid circles indicate the phosphosites with functional scores >0.5, and open circles indicate the phosphosites with functional scores <0.5. Phosphosites with functional scores >0.5 and only changed in the phosphorylation level were annotated by GO database. b Biological process analysis of phosphoproteins with differential phosphosites by DAVID software against GO database. c Differential phosphoproteins enriched in the GOBP RNA splicing and KEGG pathway spliceosome. d The change level of phosphosites in the significant enrichment pathway DNA repair, RNA splicing and Rho signaling transduction after SAHA treatment. Red asterisk indicated the phosphosites with functional score >0.5.

Fig. 6. Kinase analysis in the sensitive and resistant cells after SAHA treatment.

a Heatmap of the differential changed phosphosites on protein kinase in the protein and phsosphoprotein level between sensitive and resistant cell lines after HDACi treatment. b The analysis of kinase activity by KSEA software in the sensitive and resistant cells after SAHA treatment. Each dot represented a kinase. Red dots represented kinases with significantly increased activity (enrichment score >0 and P < 0.01), and the blue dots represented kinases with significantly decreased activity (enrichment score <0 and P < 0.01). c The validation of the synergistic effect of ROCKi GSK429286A and HDACi SAHA on the NCI-H226, A549 and MDA-MB-468 cell lines. *P < 0.05, **P < 0.01, ***P < 0.001.