CORO6 upregulation in ccRCC patients
To explore the clinical significance of CORO6, we first examined its mRNA level in ccRCC patients and normal kidney tissues from the TCGA KIRC dataset. Our analysis revealed that the CORO6 mRNA level was clearly increased in ccRCC tissues compared to normal kidney tissues (Fig. 1A). However, the CORO6 mRNA level was indiscriminate in 72 pairs of ccRCC samples and their corresponding noncancerous tissues (Fig. 1B). The CORO6 expression level was also not associated with sex hormones (Fig. 1C). Interestingly, upregulation of CORO6 was observed in tumor (T) stages III and IV compared to T stages I and II (Fig. 1D, top). Moreover, the expression level of CORO6 was also highly associated with grade (G), stage, and pathological tumor/node/metastasis (TNM Classification of Malignant Tumors; Union for International Cancer Control) stage in ccRCC patients (Fig. 1E–F, top). Of note, a higher CORO6 level was observed in T stage III, TNM stage III/IV, and G4 compared to the corresponding controls (Fig. 1D-F, bottom). Taken together, all of these data suggest that the CORO6 level is significantly upregulated in ccRCC patients and is further increased as the tumor progresses to the lethal stage.
Association between CORO6 level and poor prognosis in ccRCC patients
We then sought to investigate whether the expression level of CORO6 was associated with ccRCC progression. Data revealed that ccRCC patients with distant metastasis tended to express a high level of CORO6 compared to non-metastatic patients (Fig. 2A). Of note, the CORO6 level was significantly elevated in lymph node metastatic ccRCC patients (Fig. 2B). Importantly, the CORO6 mRNA level clearly classified the overall survival (OS) of ccRCC patients (Fig. 2C), in which patients with poor OS tended to express a high level of CORO6. Our analysis also showed that CORO6 was highly expressed in dead ccRCC patients compared to living ones (Fig. 2D). Together, all of these data support the notion that a high CORO6 level serves as an independent predictor of poor prognosis in ccRCC patients.
Correlation between CORO6 level and OS in ccRCC patients
The correlation between CORO6 level and the OS of ccRCC patients was further analyzed. We first divided the 533 patients from the TCGA-KIRC dataset into two groups using the median expression level of CORO6 as a cutoff. Figure 3A demonstrates that patients with a high CORO6 level had a shorter OS time (P < 0.0001). Furthermore, OS analyses among various subgroups of ccRCC patients were performed. The results demonstrated that CORO6 expression level may serve as a prognostic factor for ccRCC patients with different classifications, such as female (Fig. 3B; P = 0.0052), male (Fig. 3C; P < 0.0001), age > 65 years (Fig. 3D; P = 0.0058), age ≤ 65 years (Fig. 3E; P < 0.0001), T stage I + II (Fig. 3F; P = 0.0094), T stage III + IV (Fig. 3G; P < 0.0001), N0 (Fig. 3H; P < 0.0001), M0 (Fig. 3I; P = 0.0036), M1 (Fig. 3J; p < 0.0001), TNM stage III + V (Fig. 3K; P < 0.0001), and G3 + G4 (Fig. 3L; P < 0.0001). However, we failed to observe a significant association between CORO6 expression and the OS of ccRCC patients with TNM stage I + II and G1 + G2 (data not shown). Collectively, all of these analyses indicate that CORO6 expression level is tightly correlated with the OS of ccRCC patients.
Experimental validation of CORO6 expression in ccRCC samples and RCC cell lines
To verify the above online analyses, we next collected ccRCC samples for the detection of CORO6 at both the protein and mRNA levels. The CORO6 mRNA level was significantly increased in ccRCC patients (n = 40) compared to adjacent normal kidney tissues (Fig. 4A). Western blotting detection also confirmed that the CORO6 protein level was considerably upregulated in ccRCC samples (Fig. 4B & 4D). We also applied immunohistochemistry (IHC) to stain CORO6 in tissue microarrays with ccRCC tumors (n = 75) and paired adjacent tissues (n = 75). The IHC results consistently supported the potential tumor-promoting role of CORO6 in ccRCC development (Fig. 4C & 4E). Moreover, RCC cell lines, including Caki-1, A-498, SN12-PM6, and 786-O, tended to highly express CORO6 in contrast to normal kidney cell line HK2 using both quantitative polymerase chain reaction (qPCR; Fig. 4F) and western blotting (Fig. 4G). All of these data suggest that CORO6 is highly expressed in RCC cell lines and ccRCC patients and may serve as a tumor-promoting factor in determining ccRCC progression.
CORO6 enhancement of ccRCC cell growth
To test whether CORO6 contributed to ccRCC development, we first knocked down CORO6 by two independent short hairpin RNAs (shRNAs) in Caki-1 and SN12-PM6 cells. Our results showed that these two shRNAs against CORO6 successfully depleted CORO6 at both the mRNA and protein levels (Fig. 5A & 5B). As expected, the abrogation of CORO6 significantly reduced Caki-1 and SN12-PM6 cell growth, which was monitored by an MTT assay (Fig. 5C). Cell growth is at least somewhat determined by cell proliferation and cell apoptosis. To further identify the roles of CORO6, we performed BrdU staining to examine ccRCC cell proliferation with or without CORO6 manipulation. The results illustrated that the numbers of BrdU-positive Caki-1 and SN12-PM6 cells were considerably decreased when CORO6 was depleted (Fig. 5D-E), indicating that CORO6 attenuation may suppress Caki-1 and SN12-PM6 cell proliferation. In addition, we also found that the loss of CORO6 in Caki-1 and SN12-PM6 cells led to cell cycle arrest at the G0/G1 phase (Fig. 5F-G) and annexin V/PI-monitored cell apoptosis (Fig. 5H-I). Together, all of these data suggest that CORO6 promotes ccRCC cell growth by increasing cell proliferation and suppressing cell apoptosis.
CORO6 promotion of ccRCC cell migration and invasion
Cell migration and cell invasion are two essential hallmarks of cancer cells [22]. Given this notion, we sought to test whether CORO6 contributes to cell migration and cell invasion of ccRCC cells. The wound healing assay illustrated that a strong inhibition of Caki-1 and SN12-PM6 cell migration was clearly observed when CORO6 expression was efficiently attenuated (Fig. 6A). To confirm this, we performed a transwell migration assay, which revealed that CORO6 knockdown was indeed able to suppress ccRCC cell migration (Fig. 6B & 6C). In addition, the results from the Matrigel invasion assay also showed that loss of CORO6 in Caki-1 and SN12-PM6 cells dramatically inhibited their invasion ability (Fig. 6D & 6E). In summary, all of these findings indicate that CORO6 promotes ccRCC cell migration and invasion.
Mechanistic dissection of CORO6 promotion of ccRCC progression
To identify the underlying mechanisms responsible for CORO6-mediated ccRCC cell growth and cell migration/invasion, we first grouped patients from the TCGA KIRC dataset into high and low CORO6 groups using the median expression level of CORO6 as a cutoff and performed KEGG pathway analyses using the GSEA tool. The data revealed that Gene Ontology (GO) receptor agonist activity was tightly associated with the CORO6 expression level (Fig. 7A). Subsequent analysis pinpointed that WNT1, WNT3, and WNT10B were highly enriched in the high CORO6 group (Fig. 7B), indicating that CORO6 may regulate the WNT pathway to affect ccRCC cell growth and invasion. Indeed, knockdown of CORO6 reduced the expression levels of WNT1, WNT3, and WNT10B in Caki-1 and SN12-PM6 cells (Fig. 7C & 7D). In addition, TOP flash luciferase activity, which is used to monitor the WNT signaling pathway, was clearly stimulated by CORO6 in Caki-1 and SN12-PM6 cells. However, CORO6 failed to activate negative FOPflash activity (Fig. 7E & 7F). Taken together, these results prove that WNT signaling is one of the mechanisms responsible for CORO6-mediated ccRCC development.
CORO6-induced ccRCC cell growth and cell invasion/migration associations with WNT activation
To validate the involvement of WNT signaling in CORO6-induced ccRCC cell growth and cell invasion/migration, we sought to explore whether WNT signaling inhibitor IWP-O1 could attenuate the CORO6 effect on ccRCC cells. Of note, our data illustrated that IWP-O1 had no ability to affect the expression level of CORO6 at both the mRNA and protein levels (Fig. 8A & 8B), suggesting that the anti-cancer effect of IWP-O1 was due to WNT inhibition but not the indirect effect from CORO6 reduction. As expected, CORO6-induced Caki-1 cell growth was reversed in the presence of IWP-O1 (Fig. 8C). Inhibition of WNT signaling also prevented CORO6 from increasing the cell proliferation rate (Fig. 8D), turning cell cycle entry into the S phase (Fig. 8E) and decreasing Caki-1 cell apoptosis (Fig. 8F). All of these data suggest that CORO6-induced cell growth of ccRCC cells was at least partially caused by WNT activation.
In addition, we also found that CORO6 lost its ability to increase Caki-1 cell migration (Fig. 8G) and cell invasion (Fig. 8H) when there cells were supplemented with IWP-O1, suggesting that the activation of WNT signaling contributes to CORO6-induced cell invasion/migration.
Taken together, the data from Fig. 8A-H suggest that the phenotypic alterations of ccRCC cells caused by CORO6 were partially attributable to WNT activation.
Xenograft mouse model to support the oncogenic role of CORO6 in ccRCC development
To translate our findings into a pre-clinical animal model, we subcutaneously implanted shCtrl- or shCORO6-bearing Caki-1 cells (1 × 106) into nude mice and monitored ccRCC tumor growth. Consistent with our in vitro findings, CORO6 depletion by two independent shRNAs led to a significant suppression of ccRCC tumor growth monitored by tumor size (Fig. 9A), growth curve (Fig. 9B), and tumor weight (Fig. 9C), suggesting that CORO6 indeed acted as a tumor-promoting factor in the development of ccRCC. Importantly, our data also showed that CORO6-depleted ccRCC tumors expressed low mRNA expression levels of WNT1, WNT3, and WNT10B (Fig. 9D). Moreover, protein detections of the well-known downstream targets of WNT signaling, including c-Myc, AXIN2, and CCND1, showed that they were dramatically reduced in CORO6-depleted tumors (Fig. 9E). Furthermore, a second xenograft mouse model validated that CORO6 significantly promoted ccRCC tumor growth (Fig. 9F-H) attenuated by WNT signaling inhibitor IWP-O1 (Fig. 9F-H), suggesting the involvement of WNT signaling in CORO6-mediated ccRCC growth in vivo.