Overexpression of HNF1A regulated the invasion power of ESCC cells.
In our previous study, we successfully constructed ESCC cell lines TE1 and KYSE150 with overexpression protein of HNF1A. The expression levels of mRNA and protein of HNF1A were verified by RT-qPCR and WB assay. All results showed that HNF1A was still highly expressed in lentivirus transfected cell lines, and its levels of mRNA and protein were significantly higher than those in NC group (Fig. 1. a, b and c). In TE1 and KYSE150 cells overexpressing HNF1A, mRNA level was more than 2000 times (P = 0.032, P = 0.009) and protein level was more than 2 times (P = 0.000, P = 0.001). Our primary experiment found that HNF1A was involved in the invasion of ESCC cells. In this study, the invasion ability of TE1 and KYSE150 cells were clearly increased after ectopic expression of HNF1A (P = 0.025, P = 0.023) (Fig. 1. d and e). When TE1 and KYSE150 cells were irradiated, the invasion ability of NC group and HNF1A group observably enhanced (P = 0.011, P = 0.006, P = 0.014, P = 0.006, respectively), and the increasing level was much significant in HNF1A group than in NC group (P = 0.01, P = 0.006).
HNF1A interacted with HSPD1 in ESCC cells.
In this paper, we paid more attention to explore mechanism of HNF1A affecting the invasion of ESCC cells. We used IP and liquid chromatography-tandem mass spectrometry (LC-MS/MS) assay to analyze HNF1A correlated proteins in TE1 cells. The differential protein bands were examined on the silver-stained SDS-PAGE compared with IgG control (Fig. 2, a). Next, we adopted mass spectrometry (MS) to identify differential protein bands (Report Number: R201901097). We had identified several bands proteins that possibly interact with HNF1A, such as HSPD1, hnRPA1/A3, and TCP1. HSPD1 protein was selected for subsequent research based on its score and protein banding localization.
Through GEPIA online database analysis, we obtained the association between HNF1A and HSPD1 in TCGA tumor tissues and normal tissues (P < 0.001) (Fig. 2. c). In order to clarify relationship between HNF1A and HSPD1, CO-IP technology was used to verify its function. Results showed that strongly confirmed interaction between HNF1A and HSPD1 (Fig. 2. b). The IF assay was used to detect the subcellular localization of HNF1A and HSPD1. Before IR, we discovered HNF1A protein expressed in nucleus, while HSPD1 protein ubiquitously existed in nucleus and cytoplasm. Moreover, HNF1A protein expression was enhanced and translocated from nucleus to cytoplasm after IR, which clearly demonstrated that HNF1A and HSPD1 co-localized in nucleus and cytoplasm (Fig. 2. d). This results again proved reciprocity between HNF1A and HSPD1.
Radiation-induced cells reduced HSPD1 expression, which was not regulated by HNF1A.
In order to illustrate how HNF1A and HSPD1 regulated each other, we employed WB experiment to verify. The HSPD1 protein expression was decreased in TE1 and KYSE150 cells after ectopic expression of HNF1A (P = 0.001, P = 0.009). From this, we may consider that HNF1A negatively regulated HSPD1 protein expression. However, after IR 2h, HSPD1 protein in NC group reduced (P = 0.011, P = 0.003), and the change trend in HNF1A group was not significant, and even a little bit higher. Notably, expression of HSPD1 protein in HNF1A group was higher than that in NC group after IR (P = 0.009, P = 0.004) (Fig. 2. e). According to this results which HNF1A negatively regulated HSPD1 protein expression, the result was obviously contrary to it. Therefore, we had a novel discover that HNF1A was not involved in regulating HSPD1 expression, but their combination might play a pivotal role in tumor progression. In addition, our foregone study uncovered that the HNF1A expression reached its highest point after IR 2 h, and the HSPD1 expression cut down the lowest point at 4 h after IR in our current research (Fig. 2. f). Based on the above, it was explicit that HSPD1 protein expression was decreased after IR, and it was not subject to HNF1A.
The plasmid of HSPD1 knockdown was successfully screened and HSPD1 relieved radio-resistance from ectopic expression HNF1A.
Based on these results, we hypothesized that HNF1A was not involved in regulating HSPD1 expression, but HNF1A needed HSPD1 as a molecular chaperone to perform its biological functions. Next, we adopted transwell experiment to verify whether absence of HSPD1 affect the phenomenon of HNF1A overexpression and promoted ESCC cells invasion ability. First, we learned from the GEPIA database that HSPD1 was widely high expressed in multifarious cancer types (Fig. 3. a), and HSPD1 expression was higher in ESCC tissues than in paracancerous tissues form TCGA database (P < 0.05) (Fig. 3. b). Using the GEPIA database to analyze the function of HSPD1 on survival of patients with ESCC, it revealed that patients with high expression of HSPD1 had poor OS (HR = 2.1, P = 0.025). Therefore, we had reasons to speculate that HSPD1 might play an oncogene role in ESCC. Next, we transfected the knockdown plasmids of Sh-HSPD1 or Sh-NC into TE1 and KYSE150 cells by using lipofectamine 2000. These results disclosed that all knockdown plasmids inhibited HSPD1 expression at mRNA level and protein level (Fig. 3. d, e and f). ShRNA1 and ShRNA2 were more effective than ShRNA3 in HSPD1 protein knockdown by WB (Fig. 3. d, e). Meanwhile, the knockdown change of ShRNA1 and ShRNA2 were more significant at mRNA level (Fig. 3. f). Therefore, ShRNA1 and ShRNA2 were selected for subsequent experiments. Colony formation assay revealed that deletion of HSPD1 could enhance radiosensitivity of TE1 and KYSE150 cells (SER = 1.32, SER = 1.33), and overexpression of HNF1A could induce radio-resistance of TE1 and KYSE150 cells (SER = 0.64, SER = 0.66). Notably, knockout HSPD1 reversed that HNF1A promoted radio-resistance in TE1 and KYSE150 cells (SER = 0.80, SER = 0.89, SER increasing compared with alone overexpression of HNF1A) (Fig. 3. g).
Knockdown HSPD1 gene protein resulted to inhibit EMT process and affect the invasion ability in ESCC cells.
After plasmid knockdown screening, HSPD1 effect on the invasion of ESCC cells was detected by transwell assay. These data showed that the invasion ability was significantly reduced before IR in HSPD1 knock-out TE1 and KYSE150 cells (shRNA1, P = 0.002, P = 0.007; ShRNA2, P = 0.004, P = 0.044, respectively) (Fig. 4. a, b). At the same time, invasion ability of all groups were increased after IR (NC, P = 0.000, P = 0.015; shRNA1, P = 0.004, P = 0.002; ShRNA2, P = 0.000, P = 0.007, respectively) (Fig. 4. a, b). However, the invasion ability of ShRNA groups were decreased compared with NC group in TE1 and KYSE150 cells after IR (shRNA1, P = 0.000, P = 0.014; ShRNA2, P = 0.000, P = 0.02, respectively) (Fig. 4. a, b). Furthermore, the expression of EMT-related proteins was detected by WB. These results suggested that with or without IR, E-cadherin proteins were intensive in ShRNA group compared to NC group, but N-cadherin/Vimentin proteins were inverse (Fig. 5. a). In the meantime, E-cadherin proteins were depressed and N-cadherin/Vimentin proteins were enhanced in all groups after IR (Fig. 5. a). Consequences indicated that the knockdown of HSPD1 gene protein reduced invasion ability via inhibiting the EMT process in ESCC cells.
HSPD1 reversed increasing cell invasion ability of HNF1A overexpression by inhibited EMT process.
Next, the HSPD1 knock-out effect on the invasion ability was observed in HNF1A overexpressing cells. We discovered that knockdown HSPD1 reversed the tendency of HNF1A to enhance cell invasion power (Fig. 4. c). Regardless of IR, HSPD1 knockout in HNF1A ectopic expressing cells resulted in cut down invasion ability (all P < 0.05) (Fig. 4. d). In addition, HSPD1 was knocked down in HNF1Aoverexpressing cells, and total proteins of cells were extracted. These results revealed that E-cadherin proteins were improved in ShRNA group compared to control check (CK) and NC groups, while N-cadherin/Vimentin proteins were decreased with or without IR (Fig. 5. b).
The exosomal secretion inhibitor GW4869 kept HSPD1 proteins inside cell after IR.
Based on the results of above experimental, we concluded that HNF1A could not regulate HSPD1 expression, but HNF1A required the assistance of HSPD1 in regulating EMT and influenced cell invasion. On the other hand, radiation-induced decreased protein expression of HSPD1, but ectopic express HNF1A could keep from its decreasing. The regulating mechanism remained unclear still now. Recently, it had been reported that HSPD1 was distributed on exosome membrane and released into the tumor microenvironment, then affecting tumor cell production and progression(11). We used GW4869, an inhibitor of exosome secretion, to observe that irradiated may cause changes of HSPD1 protein in cells total protein or not. These results revealed that content of HSPD1 in GW4869 group was significantly higher than that in DMSO group before IR (P = 0.000, P = 0.009) (Fig. 5. e, f). Moreover, intracellular HSPD1 protein in DMSO group was significantly reduced after IR (P = 0.002, P = 0.006) (Fig. 5e, f). At the same time, HSPD1 protein in GW4869 group was higher than it in DMSO group (P = 0.013, P = 0.032) (Fig. 5. e, f). The results indirectly demonstrated that HSPD1 protein was secreted via exosomes. According to instructions of the exosome extraction kit, exosomes of TE1 cells were extracted and verified by TEM and WB assay. TEM showed that exosomes were successfully presented (Fig. 5. c). Positive markers Tsg101/Alix/CD9 proteins were checked by WB assay, while negative markers Calnexin protein was not detected (Fig. 5. d). It had been confirmed that exosomes were successfully extracted from supernatant of TE1 cells culture again.
Ectopic expression HNF1A reduced HSPD1 secretion by exosomes outside cells after IR.
Notably, all results also revealed that the presence of HSPD1 in exosomes, which directly proved HSPD1 protein could be secreted into extracellular environment through exosomes (Fig. 5. d). After IR, trendency HSPD1 protein expression in GW4869 group was consistent with those in HNF1A group. We hypothesized that ectopic expression HNF1A reduced releasing of HSPD1 protein in exosomes, resulting in HSPD1 stop in cells and assisting HNF1A to play biological functions. Next, we employed WB verifying our conjecture. The content of HSPD1 in exosomes significantly decreased with or without IR after overexpression of HNF1A (P = 0.102, P = 0.006) (Fig. 5g). After IR, HSPD1 protein content of exosomes in NC group was significantly increased (P = 0.005), but there were no statistically significant difference in HNF1A group (P = 0.197) (Fig. 5. g). Our results affirmed that radiation-induced HSPD1 from exosomes pathway were largely released outside cells and that ectopic expression HNF1A could weaken this phenomenon.