Differentiated UCPs mRNA expressions in OV patients
We compared the mRNA levels of UCPs between ovarian cancer and control tissues through the GEPIA (Gene Expression Profiling Interactive Analysis) dataset (http://gepia.cancer-pku.cn/). The results showed that ovarian cancer tissues have a higher expression level of UCP2 than normal tissues (Figure S1A); meanwhile, the expression levels of UCP4 and UCP5 were lower in ovarian cancer tissues (Figure S1I, K). And no differences in UCP1(Fig. 1A) and UCP3 (Figure S1E) gene expression were observed in tumors when compared to normal tissues. We also analyzed the mRNA levels of UCPs among OV patients with different tumor stages. UCP3 (Figure S1F) and UCP4 (Figure S1J) groups significantly differed, whereas UCP1 (Fig. 1B), UCP2, and UCP5 (Figure S1B, L) groups did not significantly vary.
Prognostic values of UCPs mRNAs in OV patients
To determine the prognostic value of UCPs, the correlation between the mRNA levels of UCPs and the survival of patients with ovarian cancer was explored by using Kaplan- Meier Plotter tools. Log rank test analyses showed a significant association between the elevated UCP1 (Fig. 1C-E) mRNA levels and improved OS, post-progression survival (PPS), and progression-free survival (PFS) (p < 0.05) of all ovarian cancer patients. Ovarian cancer patients with high mRNA levels of UCP2 and UCP3 were predicted to have improved OS and PFS (p < 0.05) (Figure S1C-D, G-H). In addition, ovarian cancer patients with higher mRNA levels of the UCP5 (SLC25A14) were predicted to have improved OS and PPS (p < 0.05) (Figure S1M-N). Data of UCP4 were not available for ovarian cancer patients.
Association of UCP mRNA levels with tumor stages and grades of OV patients
We further explored the correlations between UCPs mRNA levels and the survival of OV patients with different tumor stages and grades using Kaplan-Meier Plotter tools. The results revealed that the increased UCP1 mRNA was significantly associated with improved overall survival of Stage 3 and Grade 3 ovarian cancer patients (p < 0.05) (Fig. 2A-B). The elevated UCP2 mRNA expression was associated with improved OS of Stage 1–2 and 3 and Grade 3 ovarian cancer patients (Table S1). In addition, ovarian cancer patients at Stage 3 and Grade 2, 3 were predicted to have longer OS when they have higher mRNA levels of the UCP3 (Table S1). Meanwhile, ovarian cancer patients at Stage 1, 2, 4 and Grade 2, 3 were predicted to have longer OS when they have higher mRNA levels of the UCP5(SLC25A14) (Table S1). The data of UCP4 were not available for ovarian cancer patients.
Differentiated UCP mRNA expressions in OV patients with different treatments
Surgery and adjuvant chemotherapy are the most common treatments for ovarian cancer. Effectively predicting the prognosis of patients can guide the clinical treatment of cancer patients and avoid waste of medical resources. Therefore, we further analyzed the role of UCPs in the prognosis of ovarian cancer patients after surgery or chemotherapy. Patients with lower mRNA levels of UCP1, 2, 3, and 5 have inferior prognosis after optimal and suboptimal debulk (Fig. 2C-D, Table S2). The most significant prognosis values were shown in UCP3 (HR = 0.69, 95%CI: 0.55–0.86, p = 0.00074) for optimal debulk patients and UCP2 (HR = 0.7, 95%CI: 0.57–0.86, p = 0.00053) in suboptimal debulk patients (Table S2). Patients with higher mRNA levels of UCP1, 2, 3, and 5 have better prognosis after docetaxel and platinum chemotherapy, with UCP3 showing the largest potential for predicting prognosis of patients with both docetaxel chemotherapy (HR = 0.56, 95%CI: 0.33–0.96, p = 0.029) and platinum chemotherapy (HR = 0.76, 95%CI: 0.66–0.89, p = 0.00033) (Fig. 2E-F, Table S2).
Correlations between UCP expressions and immune marker genes
The correlation between UCPs and certain marker genes of immune cells, involving tumor-associated macrophage (TAM), M2 macrophages, T-helper 1 (Th1), regulatory T cells (Tregs) and exhausted T-cells, was analyzed by TIMER. Correlation coefficients greater than 0.2 and less than − 0.2 were considered as significant correlation at p < 0.05[15]. Specifically, UCP1 showed significant negative correlation with T cell marker genes CD3E, monocyte marker genes CD86, TAM marker genes CCL2 and CD68; M2 marker genes CD163 VSIG4 (V-set and immunoglobulin domain containing 4) and MS4A4A and Th2 marker genes GATA3, as well as positive correlation with B cell marker genes CD19 (Table 1). However, no significant correlation between the other UCPs and immune cell marker genes.
Relationship between UCP mRNA levels and tumor functional states
Additionally, we analyzed the relevance of UCPs mRNA expression levels to 14 functional states in ovarian cancers, including angiogenesis, apoptosis, cell cycle, differentiation, DNA damage, DNA repair, EMT, hypoxia, inflammation, invasion, metastasis, proliferation, quiescence and stemness, using the CancerSEA database. The results showed significant negative correlations between UCP1 and invasion (r =-0.49, p = 0.00), EMT (r =-0.39, p = 0.00), metastasis (r =-0.39, p = 0.00), DNA repair (r =-0.31, p = 0.00) (Fig. 3A-D). At the same time, negative correlations between UCP3 and invasion (r =-0.35, p = 0.00) (Figure S2A), UCP5 and angiogenesis (r =-0.38, p = 0.00) (Figure S2B), were identified. However, UCP2 was correlated with none of 14 functional states.
Relationship between UCP mRNA levels and tumor marker genes
Several genes have been confirmed to be closely related to the occurrence and progression of ovarian cancer, including SNAI2[16], VCAM1[17], BRCA1[18], MMP2[19], MECOM[20], MXS1[21], PARP1[22] and so on. We further analyzed the association between UCP1 and these gene expression. Results showed that UCP1 has negative correlation with SNAI2 (r = -0.262, p = 3.76E-06), VCAM1 (r = -0.291, p = 2.4E-07) and MMP2 (r = -0.3, p = 9.87E-08), while positive correlation with BRCA1 (r = 0.281, p = 6.67E-07), MECOM (r = 0.309, p = 4.2E-08), MSX1 (r = 0.31, p = 3.44E-08) and PARP1 (r = 0.275, p = 1.22E-06) (Fig. 3E-K). The above evidence indicates that UCP1 is negatively correlated with proto-oncogenes and positively correlated with tumor suppressor genes.