CRT silencing inhibited ERS-induced autophagy in PC cells
We first detected the expression of critical markers involving ERS and autophagy signaling in CRT silencing PC cells combing with ERS inducer TG and ERS inhibitor 4-PBA treatment. It is well known that IRE1/XBP1 and PERK/eIF2α axis of UPR signaling play a significant role in the initiation and progression of autophagy [20, 21]. Our previous study had verified that CRT regulated TG-induced ERS in PC cells independent of IRE1/XBP1 signaling [16]. Thus, we focus on the potential and specific role of CRT in regulating PERK/eIF2α axis targeted autophagy in current study. WB first showed that TG significantly induced CRT, PERK/eIF2α and autophagy signaling in human Capan-2 and Miapaca-2 cells, and mouse Panc02 PC cells, following the upregulation of CRT, p-PERK, p-eIF2α, ATG5, ATG12 and LC3II protein expression (Fig. 1A-C). However, the early autophagy markers, ULK1 and Beculin1 were unchanged, which indicated that TG-activated ERS mainly contributed to the late progression of autophagy. Moreover, both CRT silencing and 4-PBA treatment could reverse TG-induced the activation of above targeted proteins (Fig. 1A-C). Therefore, CRT performed the same function of 4-PBA, specifically inhibited ERS-induced late autophagy in vitro. We next found that autophagy inhibitor CQ inhibited TG-induce autophagy by the accumulation of LC3II expression. This trend was much more significant in CRT silencing PC cells combing TG treatment (Fig. 1D-F), indicating a close mutual interaction of CRT in regulating ERS-induced autophagy signaling in PC cells. The corresponding quantified data of WB was shown in supplemental Fig. 1.
Similarly, TG increased the number of autophagosome in Capan-2 and Panc02 cells along with the disorder of Endoplasmic reticulum (ER) by TEM (Fig. 2A and B), indicating a close relationship of ERS and autophagy in vitro. However, CRT silencing and 4-PBA treatment inhibited the functional role of TG, following the decrease of the number of autophagosome and the alleviation of ER disorder (Fig. 2A and B). mRFP-GFP-LC3 assays showed that CRT silencing and 4-PBA treatment abolished TG-induced the increase of autophagic flux in the same cell lines (Fig. 2C and D), which further verified that CRT mediated ERS-induced the late progression of autophagy.
Taken together, a close interaction between CRT mediated ERS-induced autophagy phenotype in PC cells drives us to deeply investigate their coordinate function in PC malignant behavior in vitro and vivo.
CRT silencing inhibited ERS-induced cell mobility and chemoresistance via promoting autophagy in vitro
ERS, as a series of adaptive changes mediated by UPR, takes part in cancer development via mediated tumor malignant behavior (proliferation, migration, apoptosis and survival) and autophagy [22, 23]. In current study, transwell assays first showed that TG significantly induced cell invasion and migration in Capan-2, Miapaca-2 and Panc02 cells (Fig. 3A-F). However, both CRT silencing and 4-PBA reversed TG induced cell mobility accordingly (Fig. 3A-F). In normal condition, CRT silencing had a certain suppression in cell invasion and migration. However, this suppressive trend was much significant in TG treated group. For example, the change of cell invasion and migration in scramble and sgCRT1 groups of Capan-2 cells with TG treatment was much significant than that in corresponding groups without TG treatment (Fig. 3A and B). The same results were also observed in Miapaca-2 and Panc02 cells (Fig. 3C-F). Similarly, autophagy inhibitor CQ also significantly reversed TG induced cell mobility in the same cell lines, which indicated that ERS induced autophagy enhanced cell invasion and migration in vitro. Therefore, CRT performed the same function of 4-PBA, specifically inhibited ERS-induced cell mobility via promoting autophagy in vitro.
It is well known that ERS induced autophagy plays a significant role in tumor chemoresistance [10, 24]. MTT first showed that TG enhanced Gemcitabine (GEM) chemoresistance in above three cell lines. For example, with various concentration (0 to 40ug/ml) of GEM treatment, cell growth (viability) in TG treated Capan-2 cells was significantly increased than that in Ctrl group. However, both CRT silencing and 4-PBA treatment significantly reversed the increase of cell growth induced by TG (Fig. 3F). The same results were also observed in Miapaca-2 and Panc02 cells (Fig. 3G and H). Similarly, CQ also significantly reversed TG induced cell mobility in the same cell lines, which indicated that ERS induced autophagy enhanced GEM chemoresistance in vitro. Therefore, CRT performed the same function of 4-PBA, specifically inhibited ERS-induced GEM chemoresistance via promoting autophagy in vitro.
Taken together, CRT silencing inhibited ERS-induced cell mobility and chemoresistance via promoting autophagy in vitro.
CRT silencing inhibited TG induced pancreatic tumor size in situ and distant liver metastasis in vivo
We next investigated the mutual interaction of CRT, ERS induced autophagy and malignant behavior of PC in vivo. The pancreatic tumor size in situ in TG treated group was much larger than that in Ctrl group (Fig. 4A and C). However, both CRT silencing and CQ treatment reversed TG-induced the increase of pancreatic tumor size in situ. Briefly, pancreatic tumor size in situ in TG treated group was much larger than that in sgCRT1 and CQ groups plus TG treatment (Fig. 4A and C). The pancreatic tumor was verified by HE staining (Fig. 4B). IHC further showed that CRT, p-eIF2a, ATG5 and LC3 was significantly increased in pancreatic tumor tissues of TG group in contrast with Ctrl group. However, TG induced the expression of above proteins was downregulated in sgCRT1 group plus TG treatment (Fig. 4D-G). TEM further showed that TG induced the increase of the number of autophagosome along with the disorder of ER in pancreatic tumor tissues in contrast with Ctrl group, indicating a close relationship of ERS and autophagy in the development of primary PC. However, CRT silencing reversed the function of TG, following the decrease of the number of autophagosome and the alleviation of ER disorder in vivo (Fig. 4H).
Similarly, the number of distant liver metastases in TG treated group was significantly increased compared with that in Ctrl group (Fig. 5A and C). However, both CRT silencing and CQ treatment reversed TG-induced the increase of the number of distant liver metastases. Briefly, the number of distant liver metastases in TG treated group was much higher than that in sgCRT1 and CQ groups plus TG treatment (Fig. 5A and C). The liver metastases were verified by HE staining (Fig. 5B). IHC further showed that CRT, p-eIF2a, ATG5 and LC3 were profoundly increased in liver metastasis tissues of TG group in contrast with Ctrl group. However, TG induced the expression of CRT, p-eIF2a, ATG5 and LC3 was downregulated in sgCRT1 combing with TG group (Fig. 5D-G). TEM further showed that TG induced the increase of the number of autophagosome along with the disorder of ER in liver metastasis tissues in contrast with Ctrl group, indicating a close relationship of ERS and autophagy in the development of distant metastasis. However, CRT silencing reversed the function of TG, following the decrease of the number of autophagosome and the alleviation of ER disorder in vivo (Fig. 5H).
CRT interacts with LC3 through the LIR motif to enhance the augmentation of ERS-induced autophagy and EMT in vitro.
To investigate deep insight into the link between CRT and autophagic pathway, we next studied the interaction between CRT and autophagic proteins. IF showed that CRT and LC3 were colocalized in Capan-2 cells under TG and CQ treatment (Fig. 6A). Moreover, CRT and LC3 could form co-immunoprecipitation from endogenous protein in vitro under the same condition (Fig. 6B). It is well known that LC3-interacting region (LIR) motif is crucial for autophagy, which ensures the targeting of autophagy receptors to LC3 (or other ATG8 family proteins) anchored in the phagophore membrane [25]. The LIR motif, containing the motif W/Y/FXXL/I/V (X represents any residue) and existing in most LC3-interacting proteins [26], including CRT protein from amino acids 200 to 204 (Fig. 6C). To identify the binding site of CRT with LC3, we truncated the LIR motif (WDFL) of CRT tagged with GST (GST-▲LIR). GST tagged wildtype CRT (GST-CRT) and His tagged LC3 (His-LC3) were also successfully constructed. HEK293T cells were transfected with His-LC3, GST-CRT and GST-▲LIR, respective following the subsequent GST pulldown assays. As we predicted, GST-CRT (wildtype CRT), but not GST-▲LIR (lack of LIR motif from CRT) could bind His-tagged LC3 (Fig. 6D), which indicated that LIR motif is essential for the interaction between CRT and LC3. We next investigated if the binding capacity of CRT with LC3 might be enhanced under ERS. To avoid the influence of endogenous CRT protein expression under ERS, GST-CRT and His-LC3 were overexpressed in HEK293T cells following the subsequent GST pulldown assays. The binding capacity between CRT and LC3 was profoundly increased in dose-dependent manner to TG treatment (from 50nM and 200nM) (Fig. 6E). We next transfected GST-CRT and GST-▲LIR in Capan-2 cells along with TG treatment, in contrast to wildtype CRT overexpressing cells, CRT-induced LC3II was largely abrogated in CRT-▲LIR (truncation)– expressing cells (Fig. 6F). Thus, LIR motif is required for CRT in augmentation of autophagic activity.
Further study showed that the augmentation of ERS-induced autophagy profoundly promoted EMT in vitro, following the obvious EMT-like cell morphology and the activation of EMT signaling. TG treated Capan-2 cells (85–90%) exhibited a spindle-shaped/fibroblast-like morphology (Fig. 6G). However, CRT silencing or CQ treatment reversed TG-stimulated EMT-like phenotype in vitro. Only 20–35% of spindle-shaped/fibroblast-like cellular morphology was observed in both TG plus sgCRT and TG plus CQ groups (Fig. 6G). Meanwhile, TG induced the decrease of E-cad and the increase of Slug independent of ZEB1 and N-cad in Capan-2 cells, which was also inhibited by CRT silencing or CQ treatment. The similar experiment was also repeated in Panc02 cells (Fig. 6G). Thus, CRT interacts with LC3 through the LIR motif to enhance the augmentation of ERS-induced autophagy and EMT in vitro.
A close relationship of CRT with UPR and autophagy signaling in human PC specimens
Finally, we investigated the relationship of CRT expression with the critical markers of ERS (p-eIF2a) and autophagy (ATG5 and LC3) signaling in 74 cases of human PC tissues. All above proteins showed mainly cytoplasmic expression in PC tissue. In 74 PC specimens, proteins were overexpressed in 67.6% (50/74) of CRT, 54.0% (34/74) of p-eIF2a, 58.1% (43/74) of ATG5, and 62.1% (46/74) of LC3 by IHC assays, respectively (Table 1).
Table 1
The relationship of CRT with the critical markers of ERS and autophagy signaling.
Parameters | CRT | Spearman ranks | P |
Low (24) | High (50) |
p-eIF2a | Negative (34) | 16 | 18 | 0.288 | 0.013 |
Positive (40) | 8 | 32 | | |
ATG5 | Negative (31) | 14 | 17 | 0.231 | 0.047 |
Positive (43) | 10 | 33 | | |
LC3B | Negative (28) | 15 | 13 | 0.352 | 0.002 |
| Positive (46) | 9 | 37 | | |
CRT overexpression was positively associated with p-eIF2a (P = 0.013), ATG5 (P = 0.047) and LC3 expression (P = 0.002) (Table 1). In serial sections (#3), PC tissues with CRT overexpression were also accompanied with positive p-eIF2a, ATG5 and LC3 expression (Fig. 7A), and vice versa in #11 sample (Fig. 7B). These results were consistent with our observation in vitro and vivo. Moreover, patients with high CRT or p-eIF2a expression showed the bad prognosis (P = 0.005 and P = 0.017) (Fig. 7C and D). Though ATG5 and LC3 had no relationship with the survival (Fig. 7E and F), patients combining with high CRT and ATG5 expression or high CRT and LC3 expression were associated with the worse prognosis (P = 0.008 and P = 0.006) (Fig. 7G and H). Taken together, a close relationship between CRT with UPR and autophagy signaling was also observed in clinical PC samples, which coordinately promoted poor prognosis of PC patients.