TIP-B1 promotes kidney clear cell carcinoma growth and metastasis via EGFR/AKT signaling

TIP-B1 is overexpressed in KIRC and has aggressive clinicopathological prediction trait

To identify potential molecules which may play key role in KIRC metastasis, we download three KIRC microarray and RNA sequencing data (GSE781, GSE15641, GSE73121) from GEO database. GSE781 contained 5 adjacent non-tumor tissue and 12 KIRC tumors. GSE15641 contained 23 adjacent non-tumor tissue and 32 KIRC tumors. GSE73121 contained 47 primary KIRC and 37 metastatic KIRC patient-derived xenografts (PDX) tissues. The detailed gene expression data were listed Supplementary Table 6. After merging those 3 datasets, 10 significantly upregulated genes both in tumor and metastasis tissues were identified (Figure 1A, Supplementary Table 6). We further validated those 10 genes in TCGA-KIRC cohort and found only TIP-B1 expression was significantly correlative in receiver operating characteristics (ROC) curves associating with KIRC genesis (Figure 1F, Supplementary Figure 1), which suggested TIP-B1 might be a good indicator for KIRC. As shown in Figure 1B, TIP-B1 was dramatically increased both in tumor tissues and metastatic PDX models. Then we analyzed the expression of TIP-B1 in TCGA-KIRC cohort and found that TIP-B1 expression was also significantly upregulated in tumor tissues when compared with adjacent non-tumor kidney tissues (Figure 1C). Besides, upregulation of KIRC mRNA level were also confirmed by the Oncomine database (https://www.oncomine.org/) (Supplementary Figure 2). In addition, GEPIA webtools (http://gepia.cancer-pku.cn/), Kaplan–Meier overall survival (OS) analysis and disease free survival (DFS) analysis indicated that KIRC patients with the higher level of TIP-B1 had dramatically shorter OS (Figure 1D) and DFS (Figure 1E) than those with the lower level of TIP-B1, which indicated TIP-B1 might paly central role in KIRC progression. Moreover, tumor (red) samples had significantly higher TIP-B1 expression when compared to paired normal (black) samples across 26 human tumor types (Figure 1G) from GEPIA webtools, and poorer prognosis in many human cancers, including esophageal carcinoma, head-neck carcinoma, liver carcinoma, pancreatic ductal adenocarcinoma, and rectum adenocarcinoma (Supplementary Figure 3), which further suggested that TIP-B1 might play an oncogenic role in various human cancer types.

To future identify changes of TIP-B1 expression during KIRC development, we analyzed the TCGA-KIRC dataset subgroups and found TIP-B1 expression was significantly correlated with several clinicopathological characteristics. For example, with the progression of pathological stage (Figure 1H), tumor grade (Figure 1I) and tumor stage (Figure 1K), the TIP-B1 expression significantly increased step by step. The Oncomine database also confirmed those phenomenon (Supplementary Figure 4).

Furthermore, we also analyzed the 5-year overall survival rate of each stage and grade, and results indicated high expression of TIP-B1 in each stage and grade also have worse OS rate (Supplementary Figure 5). Besides, higher TIP-B1 expression was also strongly correlated with easier tumor recurrence (Figure 1J), lymphatic invasion (Figure 1L) and metastasis (Figure 1M). Furthermore, we used ROC analysis to calculate the optimal cut-off value (11.767) of the expression of TIP-B1 with KIRC patients survival state, and revealed that TIP-B1 might serve as a good indicator for predicting KIRC progression (Figure 1N, Supplementary Figure 6 and Supplementary Table 1). Univariate and multivariable logistic regression models were performed to analyze the correlation of TIP-B1 levels with overall survival of KIRC patients. Patient characteristics were provided in Supplementary Table 2. Both univariate analysis and multivariate analysis (Figure 1O, Supplementary Table 3) indicated that high TIP-B1 expression level was an independent risk factor for worse overall survival, together with tumor recurrence, grade and metastasis.

In conclusion, the above data from GEO and TCGA human clinical samples demonstrated that higher expression of TIP-B1 was associated with the advanced tumor grade and stage. More importantly, higher expression of TIP-B1 was associated with tumor recurrence and metastasis and could be used as an independent prognostic marker for KIRC patients.

TIP-B1 is frequently upregulated in KIRC patients and correlates with poor prognosis

Because TCGA dataset results were RNA-seq data, we wanted to demonstrate TIP-B1 expression both in RNA level and protein level. Therefore, we firstly collected 18 pairs of KIRC tissues and adjacent normal tissues, and detected the expression of TIP-B1 by RT-PCR and WB assay. The RT-PCR results demonstrated that TIP-B1 expression was significantly upregulated in KIRC tissues compared with adjacent normal tissues (Figure 2A) and WB assays also confirmed the RT-PCR findings (Figure 2B). Next, we focused on the relationship between TIP-B1 and KIRC prognosis. 8 recurrent patients samples were collected. WB assay showed that the protein level of TIP-B1 in paracancerous tissues, primary tumors and recurrent samples increased gradually (Figure 2C), which confirmed the TCGA cohort. Moreover, IHC staining from 112 KIRC patients tissue microarray (TMA) samples showed that KIRC tissues expressed higher protein level of TIP-B1 than adjacent kidney tissues (Figure 2D).

Then we calculated the expression of TIP-B1 and selected the median expression point as the cutoff criterion and divided those patients into high TIP-B1 group and low TIP-B1 group. As shown in Table 1, TIP-B1 expression significantly correlated with tumor size (P=0.033), Fuhrman grade (P=0.037), tumor stage (P=0.031) and metastasis (P=0.019) in KIRC. Overall survival analysis indicated that KIRC patients with high expression of TIP-B1 had significantly lower 5-year overall survival rates (Figure 2E), the same results were also confirmed by the Human Protein Atlas website (https://www.proteinatlas.org/) (Supplementary Figure 7). what’s more, higher expression of TIP-B1 significantly correlated with worse OS rate both in early Fuhrman stage (Figure 2F) and advanced Fuhrman stage (Figure 2G). To further study whether the dysregulation of TIP-B1 correlated with the OS rate of KIRC patients, we then performed Kaplan–Meier method and Cox proportional hazard model. As shown in Table 2, the up-regulation of TIP-B1 significantly correlated with worse overall survival in TMA patients, suggesting TIP-B1 might be a useful prognosis predictor for KIRC patients.

Collectively, from public database to our TMA cohort, from RNA level to protein level, our data indicated that high level of TIP-B1 might play a critical role in KIRC progression and metastasis.

TIP-B1 promotes proliferation, migration and invasion of KIRC cells

In order to investigate the function of TIP-B1 in KIRC cells, we firstly detected TIP-B1 expression in human immortalized proximal tubule epithelial cell line HK-2 and KIRC cell lines 786-O, ACHN, OS-RC-2 and A498. The expression of TIP-B1 was significantly increased in KIRC cell lines when compared with immortalized adult human kidney cell HK2 both in RNA level (Figure 3A) and protein level (Figure 3B). Then we knocked down TIP-B1 by shRNA (sh-TIP-B1) in OS-RC-2 (Figure 3C) and 786-O (Figure 3D) cells which have a higher endogenous TIP-B1 expression, and chose the higher efficiency shRNA1 for further investigation. Functionally, cells with sh-TIP-B1 exhibited significantly decreased proliferation potential both in OS-RC-2 (Figure 3E) and 786-O (Figure 3F) compared with controls according to CCK-8 assay. Besides, cells with sh-TIP-B1 performed less migratory capability in OS-RC-2 (Figure 3G) and 786-O (Figure 3H) cells using wound-healing assay. Similarly, using the transwell migration and matrigel-coated invasion assay, we found that knocking down TIP-B1 in OS-RC-2 and 786-O cells decreased cell migration (Figure 3I) and invasion (Figure 3J) abilities compared to the vector control (pLKO) group.

In conclusion, results from in vitro assays demonstrated that TIP-B1 plays key role in regulating KIRC cell proliferation and a high level of TIP-B1 could increase KIRC cell migration and invasion abilities.

TIP-B1 knockdown inhibits KIRC tumor growth and metastasis

To further confirm TIP-B1 function in vivo, 786-O cell line transfected with sh-TIP-B1 were subcutaneously implanted into nude mice. As expected, the tumor volume of sh-TIP-B1 group was much smaller than that of control group at 5 weeks (Figure 4A and 4B). Besides, the linear curve also recorded that knockdown TIP-B1 dramatically inhibited the growth (Figure 4C) and average weight (Figure 4D) of tumors in nude mice.

Next, we injected 786-O sh-TIP-B1 cells into tail vein of nude mice to simulate tumor metastasis. IVIS image showed that luciferase signal strength and area of sh-TIP-B1 group was significantly lower than control group (Figure 4E). Besides, the volume of micro-metastatic nodules markedly decreased in sh-TIP-B1 group (Figure 4F). HE analysis indicated the number and volume of pulmonary metastatic nodules were significantly decreased in sh-TIP-B1 group compared with control group (Figure 4G and 4H). In addition, the average lung weight in sh-TIP-B1 group was also dramatically lower than in control group (Figure 4I). More importantly, the mice injected with 786-O sh-TIP-B1 cells had significantly higher survival rates than control group (Figure 4J).

Taken together, our data demonstrated that inhibiting TIP-B1 could modulate the aggressive and metastatic abilities of KIRC in vivo.

TIP-B1 triggers epithelial-mesenchymal transition by activating the AKT pathway

To explore the underlying molecular mechanisms of TIP-B1 in KIRC, we chose the TCGA-KIRC dataset and used median expression of TIP-B1 as the cutoff criterion and divided TCGA-KIRC dataset into high TIP-B1 group and low TIP-B1 group. Then we used the Gene set enrichment (GSEA) analysis to compare the different biological processes between the TIP-B1 high group and low group. Interestingly, GSEA analysis revealed that the tight junction function, an important part of the epithelial-mesenchymal transition (EMT) process, was significantly decreased in TIP-B1 high group (Figure 5A). It is well known that KIRC is mostly derived from renal tubular epithelial cells, and EMT is a key factor in the invasion and metastasis of epithelial-derived tumors [12, 13]. So firstly we tested EMT related markers. RT-PCR (Figure 5B) and WB (Figure 5C) results found that down regulated TIP-B1 had significantly increased the expression of epithelial related genes ZO-1, E-cadherin, and dramatically decreased mesenchymal related markers N-cadherin, vimentin, twist, snail and slug.

To elucidate the signal pathway that TIP-B1 plays role in, we detected AKT, p38(MAPK) and ERK, which were thought to be associated with KIRC deterioration and metastasis [14–16], and confirmed the phosphorylation of AKT molecules was clearly inactivated in both OS-RC-2 and 786-O sh-TIP-B1 cell lines, but ERK and p38 exhibited no changes (Figure 5D). Besides, the GSEA indicated phosphatidylinositol 3 phosphate binding genes enriched in TIP-B1 high group (Figure 5E), and TCGA correlation analysis also showed that TIP-B1 expression was positively correlated with AKT1 (Figure 5F), AKT2 (Figure 5G), and negatively with PTEN (Figure 5H), further demonstrating that TIP-B1 played a key role in the AKT pathway. In order to examine the effects of TIP-B1 on KIRC cell activities through AKT pathway, we treated KIRC cell with MK-2206 (a highly selective inhibitor of pan-AKT) or SC-79 (AKT phosphorylation activator) and divided cells into four groups (DMSO, MK-2206, sh-TIP-B1, sh-TIP-B1+SC-79). The CCK-8 assay showed both AKT inhibition and TIP-B1 knockdown could hindered KIRC cells proliferation, whereas AKT activation could attenuate the inhibitory effects of TIP-B1 knockdown on OS-RC-2 (Figure 5I) and 786-O (Figure 5J) cells. Besides, the increase of epithelial gene ZO-1, E-cadherin, and the decrease of mesenchymal gene N-cadherin, vimentin, twist, snail and slug caused by TIP-B1 knockdown could be rescued by AKT activation through WB assays in KIRC cells (Figure 5K). The rescue experiments of transwell assay also confirmed AKT inhibition could suppressed KIRC cells migration and invasion capacities, whereas AKT activation could release the inhibitory effects of TIP-B1 knockdown on KIRC cells (Figure 5L–5O, Supplementary Figure 8).

Together, our data suggested that TIP-B1 could trigger KIRC cells epithelial-mesenchymal transition by activating the AKT signaling pathway.

TIP-B1/EGFR axis modulates AKT signaling in KIRC

To investigate how TIP-B1 induces AKT signaling, we used GSEA analysis and found GO_RECEPTOR_ AGONIST_ACTIVITY was significantly enriched in TIP-B1 high group (Figure 6A), which means TIP-B1 interacts with some receptors to effect a change in the activity of those receptors. Previous studies have confirmed that activation of AKT signaling pathway is primarily through the interaction of receptor tyrosine kinases (RTKs) with their specific ligands, such as EGFR, which play important role in urological related tumors [17, 18]. So we tested whether TIP-B1 affected AKT signaling by interacting with EGFR. Interesting, knockdown TIP-B1 significantly inhibited AKT activation triggered by EGF stimulation compared to the control group (Figure 6B). More importantly, the phosphorylation levels of the three residues in EGFR (Tyr992, Tyr1068 and Tyr1086) were significantly reduced after downregulation of TIP-B1 (Figure 6C). Moreover, under the stimulation of EGFR-specific tyrosine kinase inhibitor (AG1478), the inhibitory effect of knockdown TIP-B1 on EGFR and AKT phosphorylation was furtherly enhanced, demonstrating that TIP-B1 might play a key role in EGFR mediated activation of AKT (Figure 6D).

To further investigate whether TIP-B1 knockdown could affect EGFR protein degradation, we detected the protein level of EGFR under EGF stimulation. The stability of EGFR protein was evaluated by cycloheximide (CHX, 100μg/ml), an inhibitor of protein synthesis, treatment for indicated times. Our data of WB analysis revealed that TIP-B1 knockdown could dramatically enhanced the degradation of EGFR (Figure 6E). Moreover, we sought to detect the physical interactions between TIP-B1 and EGFR by co-immunoprecipitation assays, and results revealed that the EGFR successfully co-immunoprecipitated TIP-B1 (Figure 6F).

Together, these results strongly suggested TIP-B1 might promote KIRC cells progression and metastasis by binding EGFR and suppressing EGFR degradation, then promoting EGF-induced AKT signaling.

Human clinical samples

KIRC tumor samples and adjacent normal tissues were collected from the Department of Urology, Shanghai Tenth People’s Hospital, Tongji University. There were three groups were enrolled in this study. Briefly speaking, in group 1, fresh tumor tissues and adjacent non-tumor tissues from 18 KIRC patients were collected for RT-PCR and Western blot (WB) assays. In group 2, eight recurrent KIRC adjacent non-tumor, primary tumor and recurrent tumor tissues were collected for further WB assays. In group 3, 112 KIRC para-tumor and tumor specimens from 2008 to 2012 were collected for tissue microarrays (TMA) establishment.

GEO database, TCGA database, ROC and GSEA analysis

Three public KIRC gene microarray profiling data (GSE781, GSE15641, GSE73121) were obtained from GEO database. GEO2R and R language were used to preprocess those three RNA sequencing and microarray data. The TIP-B1 RNA sequence data of KIRC were obtained from TCGA database. The ROC analyses were used to MedCalc software to calculate the sensitivities, specificities and accuracies at best cut-off point. As for single gene Gene set enrichment (GSEA) analysis, we used GSEA v3.0 to perform, and the median expression level of TIP-B1 was used as the cutoff criterion.

Cell culture and lentiviral vector construction

The human immortalized proximal tubule epithelial cell line HK-2 and KIRC cell lines 786-O, ACHN, OS-RC-2 and A498 were obtained from Cell Bank of the Chinese Academy of Sciences. The HK-2 cells were cultured in Keratinocyte Serum Free Medium (K-SFM, Gibco, USA). The KIRC cell lines were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM, Gibco, USA). All cells were supplement with 10% Fetal Bovine Serum (FBS, Gibco, USA). As for lentiviral vector construction, the pLKO/pLKO.1-shTIP-B1 plasmids, pMD2G envelope plasmid, psPAX2 packaging plasmid were transfected into HEK-293 cells using the standard calcium chloride transfection method. After incubating for 48 or 72 h, the supernatant medium was filtered and frozen in −80 °C for later use.

CCK8, wound-healing assay and transwell assay

The CCK8 assay were conducted to calculate cell proliferation. Briefly speaking, cells were seeded in 200μL of fresh medium into 96-well plates (1000 cells/well) for various lengths of time at 37 °C in 5% CO2. For the wound-healing migration assay, a wound was made by scraping a 200-uL plastic yellow pipette tip along the bottom of the 6-well plate and observed and photographed at 0 h and 24 h after wounding. As for transwell migration and invasion (using matrigel) assay, 5×10⁴ cells/well in 200ul serum-free medium were added into the upper chambers, 600ul contained 10% FBS medium was added to the lower chambers. After 24 hours of incubation, the KIRC cells on the upper surface were clean up with cotton swab, and the lower surface were fixed with 4% paraformaldehyde, then stained by 0.1% crystal violet.

Quantitative real-time PCR assay

Total RNAs were extracted using Trizol reagent (Invitrogen, CA). The cDNA were synthesized with Reverse Transcriptase Kit (Invitrogen, CA). The real-time PCR reactions were as follows: 5 min at 95°C, 40 cycles of 95°C for 10 s, then 60°C for 60 s. The relative expression were calculated using the ΔΔCt method. The primers sequence were listed in Supplementary Table 4.

Western blot assay

Cells were lysed in RIPA lysis buffer (Beyotime, China) containing proteinase and phosphatase inhibitors on ice, and the concentration of protein was calculated through BCA method. Protein extraction (50μg) were seeded on SDS-polyacrylamide gel, then transferred to polyvinylidene fluoride (PVDF) membrane (Millipore, Billerica, MA). The PVDF membranes were incubated with primary antibodies overnight at 4 °C on a rocking table, and then incubated with the secondary antibodies for 1 hour at room temperature. The antibodies and reagents were listed in Supplementary Table 5.

Tissue microarray and immunohistochemistry staining

The tissue microarray (TMA) construction has been previously described in detail [40]. Briefly speaking, 0.6 mm diameter tissue cylinders were punched from representative tissue area of each donor tissue block then brought into one paraffin block. Patients’ demographic and clinic-pathological were retrieved from our clinical database. Immunohistochemistry (IHC) experiments were carried out as described previously [41]. In general, the clinical human samples were incubated with the primary antibodies in 3% BSA in PBS overnight at 4 °C, followed by appropriate secondary antibodies. A semiquantitative scoring criterion which the staining index (values 0–10) were independently evaluated by three experienced pathologists.

Animal studies

The animal protocols were approved by the Institutional Animal Use and Care Committee of Tongji University. All BALB/c nude mice (6 weeks) were purchased from the SLAC Laboratory Animal Company (Shanghai, China). As for xenografts group, 1×10⁶/100ul 786-O cells were inoculated subcutaneously into nude mice. The tumor volume of each mouse was measured every 7 days, and tumor weight were calculated after 5 weeks when mice were killed. In the pulmonary metastatic group, to evaluate the potential of the cells to metastasize to the lungs, ten male nude mice were randomized into 2 groups (n=5 each), and 1×10⁶/100ul cells were injected into the tail vein. The metastases were detected using the Lumina II IVIS system (Perkinelmer, USA). Those mice were also sacrificed after 5 weeks and weighed lung weight, then using H&E staining to analyze the metastatic lung foci. As for metastatic survival group, the same amount of cells injected into the tail vein, and the mouse survival was monitored in a survival analysis.

Statistical analysis

Statistical analyses were using SPSS 23.0 software (IBM, USA). Continuous variables were presented as mean± SD. Two groups were analyzed by student t-test and multiple groups were used one way analysis of variance (ANOVA). The optimal cut-off value of the TIP-B1 expression was calculated by a ROC curve analysis through MedCalc software (MedCalc, Korea). Survival analysis were calculated using the Kaplan–Meier method, and evaluated by a log-rank test. The Cox proportional hazard model was used to determine the influencing factors based on the variables selected by univariate analysis. p < 0.05 was considered statistically significant. The detailed results were *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001.