Cisplatin enriches polyploidy in ovarian cancer cell lines
Considering that polyploidy is proposed as a leading player in chemo-drug resistance [10], the effect of cisplatin was examined on two ovarian cancer cell lines. A2789 and SCOV-3 cells were treated with cisplatin for 72 hours and then recovered for 7 days. As expected, cisplatin resulted in massive cell death. Notably, the surviving sub-population was enriched for polyploid cells as revealed by fluorescent microscopy following DNA-staining. Indeed, PGCCs as large as 100µm were evident in the culture of cisplatin-treated A2789 and SCOV-3 lines. For further validation and quantification, DNA content analysis was performed which demonstrated 2.0- and 4.7-fold enhancement in the ratio of PGCCs in A2789 and SCOV-3 cells, respectively (Fig. 1). This finding suggests that polyploid cells can be a source of cisplatin-resistance in ovarian cancer.
Transcriptomics profile of cisplatin-resistant ovarian cancer cells highlights the role of nuclear processes
To investigate molecular mechanisms involved in polyploid-resistant cells, the GSE58470 microarray dataset originally generated by Cossa et al. [21, 22] was reanalyzed. They established a cisplatin-resistant ovarian cancer line and performed mRNA microarray profiling concluding that the ERK1/2 pathway has an important role in cisplatin resistance.
Re-analysis of this dataset was started with a quality control assessment. As previously shown, about half of the publicly available transcriptomics profiles suffer from the lack of sufficient quality, underscoring the importance of pre-assessments of such data [23]. PCA and hierarchical clustering denoted the acceptable quality of the selected dataset as the samples were separated according to experimental groups in an unsupervised manner (Fig. 2a, b). A comparison of cisplatin-resistant and control group determined 1930 DEGs with at least two-fold overexpression or down-regulation (|log2FC| ≥1) and adjusted p-value < 0.05 (Fig. 2c).
To assess the functional role of the identified DEGs, pathway enrichment analysis was performed determining the 21 signaling pathways with p-value < 0.05 (Fig. 3a). Among them, the protein ubiquitination pathway, FAT10 signaling pathway, and phagosome maturation have been previously reported to be involved in chemotherapy resistance [24–26]. Moreover, mTOR, 14-3-3, HIPPO and Hypoxia are known as polyploidy-related pathways [12, 13, 15, 27]. Additionally, enrichment for biological process and molecular function GO terms highlighted RNA polymerase regulation and protein metabolism. Noteworthy, nucleoplasm was an enriched cellular component term with the maximum number of child terms (Fig. 3b). Overall, these findings indicate that gene regulation mechanisms and nucleus processes are critical for drug resistance and potentially cancer polyploidy.
The interaction network was constructed to identify the map of communications between the identified DEGs. To identify the most influential genes in the network, centrality analysis was performed and top 10 genes with the highest degree and betweenness centrality measures were detected (Fig. 4a). A considerable fraction of these central genes is involved in nuclear processes; POLR2C, RANBP2, UPF3B, SP1 and ELOB contributed to RNA metabolism and gene expression mechanisms. Also, a group of central genes including UBC, UBE2D2 and SUMO1 are involved in protein ubiquitination. Previous studies have also suggested the role of PLK1 in the cell cycle regulation of PGCCs [28].
Densely connected regions of the constructed network, known as structural modules were determined based on clustering coefficient centrality feature. These modules were mainly related to RNA metabolism, protein localization, chromatid segregation, microtubule polymerization and membrane budding (Fig. 4b).