Polycystin‐1 affects cancer cell behaviour and interacts with mTOR and Jak signalling pathways in cancer cell lines

Abstract Polycystic Kidney Disease (PKD), which is attributable to mutations in the PKD1 and PKD2 genes encoding polycystin‐1 (PC1) and polycystin‐2 (PC2) respectively, shares common cellular defects with cancer, such as uncontrolled cell proliferation, abnormal differentiation and increased apoptosis. Interestingly, PC1 regulates many signalling pathways including Jak/STAT, mTOR, Wnt, AP‐1 and calcineurin‐NFAT which are also used by cancer cells for sending signals that will allow them to acquire and maintain malignant phenotypes. Nevertheless, the molecular relationship between polycystins and cancer is unknown. In this study, we investigated the role of PC1 in cancer biology using glioblastoma (GOS3), prostate (PC3), breast (MCF7), lung (A549) and colorectal (HT29) cancer cell lines. Our in vitro results propose that PC1 promotes cell migration in GOS3 cells and suppresses cell migration in A549 cells. In addition, PC1 enhances cell proliferation in GOS3 cells but inhibits it in MCF7, A549 and HT29 cells. We also found that PC1 up‐regulates mTOR signalling and down‐regulates Jak signalling in GOS3 cells, while it up‐regulates mTOR signalling in PC3 and HT29 cells. Together, our study suggests that PC1 modulates cell proliferation and migration and interacts with mTOR and Jak signalling pathways in different cancer cell lines. Understanding the molecular details of how polycystins are associated with cancer may lead to the identification of new players in this devastating disease.

modulates cell adhesion 16,17 and to cell-matrix contacts. 18 PC1 has also been located at the primary cilium of kidney cells, where it is thought to act as a mechanosensitive receptor that transduces mechanical stimuli (fluid flow) into intracellular biochemical signals. [19][20][21] PC2 is a smaller transmembrane protein that contains six transmembrane domains, with intracellular C-and N-termini. 3,22 PC2 belongs to the transient receptor potential family of calcium channels that regulate intracellular calcium and affects various cellular features such as cell proliferation, differentiation and planar cell polarity. [23][24][25] Accumulating evidence suggests that both polycystins act as conductors to tune the overall mechanosensitivity of cells. 26 The function of polycystins has mainly been explored in the context of PKD where mutations in the polycystins PC1 and PC2 give rise to a complex cell phenotype, characterized by increased cell proliferation and apoptosis, de-differentiation, disturbed planar cell polarity, extracellular matrix alterations and abnormal fluid secretion. 27 In cancer, however, the function of polycystins is unknown. A comparison between cancer and PKD reveals that both diseases exhibit a deregulation in many important cellular features, such as proliferation, differentiation and apoptosis. 27,28 Surprisingly, ADPKD cells activate some of the same signalling pathways that are utilized by cancer cells in order to promote their malignant cell behaviour. For example, the mTOR pathway is a critical pathway that is deregulated in both cancer and PKD. mTOR signalling is up-regulated in a wide variety of cancers and is regarded as one of the most frequently altered cascades in this heterogeneous disease. [29][30][31] mTOR signalling is increased in mouse models of PKD and human ADPKD, while mTOR inhibitors, such as sirolimus and everolimus, slow disease progression in PKD animal models. 12,[32][33][34] The Jak/STAT pathway is also deregulated in both cancer and PKD. Jak/STAT signalling is activated in haematological malignancies, particularly in myeloproliferative neoplasms and solid tumours. [35][36][37] In PKD, Jak/STAT signalling activity is abnormally activated and promotes cystic growth. [38][39][40][41][42] Despite these similarities between cancer and PKD, up to date, there is only one study on the function of polycystins in cancer.
Analysing colorectal cancer (CRC) cell lines (HCT116, HT29 and SW480), HT29 tumour xenografts and cancer tissue samples from CRC patients, Gargalionis et al provided evidence of a role for polycystins in CRC aggressiveness. 43 In the present study, our goal was to examine the in vitro role of PC1 in cancer using cancer cell lines derived from five different types of human cancer (brain-GOS3, lung-A549, prostate-PC3, colon-HT29, breast-MCF7). We found that PC1 modulates the proliferation and migration of cancer cells. We also found that PC1 interacts with mTOR and Jak signalling and affects their activity in cancer cells. Importantly, our study represents the first steps in understanding the function of polycystins in cancer biology.

| Antibodies
The following primary antibodies were used for Western blot analysis:

| Semi-quantitative PCR and quantitative realtime PCR
Total RNA was extracted from cultured cells using RNeasy Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. PrimeScript RT reagent kit-Perfect Real Time (Takara Bio, Japan) for RT-PCR was used for cDNA synthesis according to the manufacturer's protocol.
For semi-quantitative PCR, the produced cDNA was am-

| Cell Proliferation Assay
Cells were seeded in a 96-well plate at a density of 10 3 -10 5 cells/well in 100 μL of culture medium with IgPKD1 (1:50 and 1:100 dilutions) or non-immune rabbit serum. Cells were cultured in a CO 2 incubator at 37°C for 24 and 48 hours. Ten microlitres of the prepared XTT Mixture (XTT Cell Proliferation Assay Kit, 10010200; Cayman Chemical, USA) was added to each well and mixed gently. The cells were incubated for 4 hours at 37°C in a CO 2 incubator. The absorbance of each sample was measured using a microplate reader at a wavelength of 450 nm.
Cells were synchronized with serum starvation for 6 hours.

| Cell migration assay
HT29, MCF7, PC3, A549 and GOS3 cells were cultured in 12-well cell plates until confluent and synchronized by serum starvation for 6 hours. The cellular layer was etched with a 200 μL sterile pipette tip.
Cells were incubated with IgPKD1 or non-immune rabbit serum. Each location was photographed in a computer-connected microscope (×10 magnification) at zero hour and after 24 hour incubation. Images were analysed using TScratch software. The results were expressed as percentages of the incised and the cell-coated region.

| Statistical and image analysis
All experiments were performed at least three times. Data are presented as mean ± SD and were analysed by one-way ANOVA.
GraphPad Prism 6 software was employed for these statistical analyses. All statistical tests were two-sided. P < 0.05 was considered statistically significant.

| Endogenous mRNA and protein expression of PC1 and PC2 in cell lines
PC1 and PC2 proteins have only been detected in SW480 CRC cells, 43 therefore we firstly sought to determine the endogenous mRNA and protein expression levels of the two polycystins in MCF7, PC3, A549, HT29 and GOS3 cancer cell lines. We detected both mRNA ( Figure 1A) and protein ( Figure 1B) levels of PC1 and PC2 in all cell lines apart from PC2 protein in MCF7 cells. There were discrepancies between Pkd1 mRNA levels and PC1 protein levels in some cancer cell lines, as well as discrepancies between Pkd2 mRNA levels and PC2 protein levels.
For example, in MCF7 cells the Pkd2 gene expression is increased but the PC2 protein expression is negligible. These differences may be due to post-transcriptional and post-translational regulatory mechanisms.
In addition, we compared the mRNA and protein levels in the cancer cell lines to the levels in normal cell lines from the same embryonic origin. Our results show that PC1 protein levels were higher in prostate cancer cells (PC3) compared to normal cells (HPrEc) and lower in glioblastoma cells (GOS3) compared to normal brain cells (CHLA-259).
PC2 protein levels were found to be higher in CRC (HT29) and prostate cancer (PC3) cells compared to normal cells (CACO2 and HPrEc respectively), while they were lower in breast cancer cells (MCF7) compared to normal breast cells (MCF10A) ( Figure 1B). There were no differences observed in the mRNA levels of PC1 and PC2 between cancer and normal cell lines.

| Effect of antibody-mediated PC1 inhibition on cell migration and proliferation in cancer cell lines
Next, we wanted to explore if PC1 affects cancer cell behaviour.
Thus, we decided to investigate whether PC1 affects cell migration and proliferation in cancer cell lines by incubating them with a blocking antibody, IgPKD1, raised against the Ig-like domains of extracellular PC1. 16 Even though the function of PC1 remains obscure, and hence, there is still no specific assay to show that PC1 is inhibited, the IgPKD1 antibody is a valid method of inhibiting PC1. IgPKD1 has been used to block PC1 in murine, canine and human kidney epithelial cells, 16,17,44 bone cells, 45,46 CRC cells and xenografts 43 and endothelial cells. 47 We found that in A549 cells, IgPKD1 treatment led to increased cell migration with the greatest effect observed at a 1:50 dilution of the IgPKD1 antibody ( Figure 2C). Conversely, in These data indicate that PC1 enhances cell proliferation in GOS3 cells but hinders it in MCF7, A549 and HT29 cells.  IgPKD1 is the inhibitory antibody against PC1. Mock represents cells that have been incubated with non-immune rabbit serum (without the IgPKD1 antibody). The images were analysed using TScratch software. Bars represent mean areas ± SD. *P < 0.05, **P < 0.01 versus mock cells ( Figure 5E,J). According to these results, we were not able to clearly identify whether mTOR signalling is up-or down-regulated in each cancer cell line; of the mTOR pathway-related proteins that we analysed in individual cell lines, some demonstrated increased phosphorylation while others showed decreased phosphorylation.

| Effect of antibody-mediated PC1 inhibition on mTOR pathway in cancer cell lines
We could probably state that mTOR signalling is up-regulated in A549 cells, as there is an increase in both mTOR and p70S6K phosphorylation; however, these two proteins were the only ones to show a significant change in their phosphorylation after PC1 inhibition. Our difficulty in determining mTOR pathway activity after PC1 inhibition may be due to the complexity of its regulation which includes activating or inhibitory inputs from several other pathways. our results further support that PC1 interacts with the mTOR pathway in cancer cells.

| Effect of Pkd1 silencing on Jak pathway in cancer cell lines
To determine the effect of PC1 on the Jak pathway in our cancer cells, we silenced PC1 protein expression through siRNA. The knockdown F I G U R E 3 Effect of PC1 inhibition on cancer cell proliferation. (A-E) Cell proliferation assay in MCF7, PC3, A549, HT29 and GOS3 cells. IgPKD1 is the inhibitory antibody against PC1. Mock represents cells that have been incubated with non-immune rabbit serum (without the IgPKD1 antibody). Each bar represents mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001 versus mock at 24 hours. # P < 0.05, ## P < 0.01, ### P < 0.001 versus mock at 48 h Bars represent means ± SD. *P < 0.05, **P < 0.01 versus non-target. siPKD1 represents cancer cells transfected with siRNA targeting the Pkd1 mRNA; non-target represents cancer cells transfected with a non-targeting siRNA; mock represents cancer cells transfected with only transfection reagents (without siRNA) efficiency of the Pkd1 siRNA was confirmed by qRT-PCR ( Figure S1).
Our data demonstrate that the phosphorylation of Jak2 is increased in MCF7 (Figure 6A,F) and GOS3 ( Figure 6E,J) cells treated with siRNA targeting the Pkd1 mRNA (siPKD1) compared to MCF7 and GOS3 cells treated with non-targeting siRNA (non-target), while it is decreased in PC3 ( Figure 6B,G) and A549 ( Figure 6C,H) cells treated with siRNA targeting the Pkd1 mRNA (siPKD1) compared to PC3 and A549 cells treated with non-targeting siRNA (non-target). According to this evidence, PC1 down-regulates Jak signalling in MCF7 and GOS3 cells, whereas it up-regulates Jak signalling in PC3 and A549 cells. These results suggest that PC1 interacts with the Jak pathway in cancer cells.

| D ISCUSS I ON
Considering the common cellular features and signalling pathways between PKD and cancer, we speculated whether polycystins PC1 and PC2 play a role in cancer biology. First, we sought to evaluate the mRNA and protein levels of polycystins in cell lines from different types of cancer, including glioblastoma, prostate, lung, breast and CRC. We detected both mRNA and protein of PC1 and PC2 in the cancer cell lines. Our results also showed that there were differences in PC1 and PC2 protein levels between cancer cells (PC3, to our study, PC1 and PC2 protein expression has been detected in normal developing brain, breast ductal epithelium, colonic epithelium, prostate epithelium and lung epithelium. [48][49][50] Nevertheless, immunohistochemistry data on polycystin protein expression in human tissues have not been consistent. Although in the present study we evaluated the protein expression of PC1 and PC2 in cancer and normal cell lines, it is also important to determine the subcellular localization of both polycystins because it is essential to their function. The subcellular localization of polycystins is complex and still debated by researchers. Results based on renal epithelial cells demonstrate that PC1 and PC2 are found on the primary cilia and in other subcellular compartments and membrane domains. Their localization, particularly PC2, has been found to be regulated by chemical chaperones, proteasome inhibitors, protein-protein interactions and phosphorylation. In addition, given that PC1 and PC2 physically interact via their cytoplasmic C-terminal tails, it is possible that they modulate each other's subcellular localization, but this remains controversial. 51 Similarly to renal epithelial cells, all cell lines used in our study, apart from glioblastoma cells (GOS3), are epithelial in origin.
However, whether PC1 and PC2 localization in our cancer epithelial cell lines follows the same pattern as in normal renal epithelial cells needs to be investigated in future studies.
For our next experiments, we focused on PC1 because of its large size, flexible nature, participation in cell-cell and cell-matrix contacts and known communication with many downstream signalling pathways via its intracellular C-terminal tail. 7-15 First, we evaluated the effect of PC1 on two important cellular features, cell proliferation and migration, which are commonly deregulated in cancer. 28 Increased cell proliferation is a major feature of a polycystic kidney; cysts have even been characterized as 'neoplasia in disguise'. 52 Several studies have reported that PC1 also regulates cell migration. [53][54][55][56][57] We found that blocking PC1 in vitro with IgPKD1 affected both cell proliferation and migration in cancer cells in a cell type-dependent manner. According to our results, PC1 functions as a tumour-suppressor protein in A549 cells inhibiting cell migration. In contrast, PC1 probably acts as an oncogene protein in GOS3 cells enhancing cell migration. Since PC1 has been reported to promote cell migration, could it be that GOS3 glioblastoma cells hijack this function of PC1 and turn it into a malignant signal that enhances their migratory and invasive abilities?
Concerning cell proliferation, our results show that PC1 might be a tumour-suppressor protein in MCF7, A549 and HT29 cells that impedes cell proliferation. Conversely, in GOS3 cells PC1 appears to be an oncogene that promotes cell proliferation. Because PC1 has been shown to inhibit cell proliferation in non-cancerous cells, 12,39,51,58 we wondered whether MCF7, A549 and HT29 cancer cells are deregulated in such a way that interferes with the normal PC1-mediated inhibition of proliferation. Do cancer cells achieve this effect by abolishing the function of the PC1 protein to transmit inhibitory signals to the cell's interior or by making the targets of these signals insensitive to inhibition? More study is required to confirm the influence of PC1 on cancer cell proliferation and migration and to uncover the mechanisms of this effect.
Next, we wanted to determine if PC1 regulates signalling pathways that are constitutively activated in cancer. Cancer and PKD are frequently accompanied by aberrant activation of the mTOR pathway. [29][30][31][32][33][34] Previous data have demonstrated that PC1 overexpression in SW480 colon cancer cells leads to down-regulation of mTOR signalling. 43 Jak signalling also becomes up-regulated in cancer [35][36][37] and studies have shown that PC1 activates Jak signalling in PKD. [38][39][40][41][42] Therefore, we investigated if the mTOR and Jak pathways are   Bars represent means ± SD. *P < 0.05, **P < 0.01 versus non-target. siPKD1 represents cancer cells transfected with siRNA targeting the Pkd1 mRNA; non-target represents cancer cells transfected with a non-targeting siRNA; mock represents cancer cells transfected with only transfection reagents (without siRNA) signalling in GOS3 cells, while it suppresses mTOR signalling in PC3 and HT29 cells. The mTOR suppression observed in HT29 colon cancer cells is consistent with the previous finding that PC1 down-regulates mTOR signalling in SW480 colon cancer cells. 43 These results also suggest that the effect of PC1 on mTOR signalling in cancer is cell type-dependent. Moreover, these data contribute to our knowledge on the regulation of mTOR and Jak signalling in cancer. It should be noted that the conclusions on whether the mTOR cascade is activated or inhibited in cancer cells after changes in PC1 function are based only on the phosphorylation of a single mTOR-related protein.
Likewise, we focused only on Jak2 phosphorylation and did not evaluate any downstream effectors or target genes as surrogate markers of Jak pathway activity. Therefore, our conclusions in terms of Jak signalling activation are solely based on the phosphorylation status of Jak2. The inhibitory or activating effect of PC1 on the two cascades has to be validated through further studies.
A challenge that we encountered in the present study was the following: the two methods used to inhibit PC1 activity, PC1 Bars represent means ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 versus mock. Mock represents cells that have been incubated for 3 h with non-immune rabbit serum; 1 h, 3 h, and 6 h represent time points of cell harvesting after 3 h incubation of cancer cells with the IgPKD1 antibody knockdown with siRNA and PC1 inhibition with IgPKD1, did not generate the same results for most cancer cell lines as far as Jak pathway activity is concerned. The siRNA experiment data propose that PC3 and A549 cells use PC1 in order to activate Jak signalling, while MCF7 and GOS3 cells use PC1 to suppress Jak signalling. In contrast, data from the antibody-mediated PC1 inhibition experiment imply that PC3, A549, HT29 and GOS3 cells use PC1 to down-regulate Jak signalling, whereas MCF7 cells use PC1 to up-regulate Jak signalling.
These discrepancies could be due to limitations inherent in the two methods used to perturb the function of PC1; in contrast to RNAi where the PC1 protein is absent, the IgPKD1-inhibited PC1 protein may lack certain activities but may still execute other activities and/ or interact with other proteins. Moreover, both methods can have substantial off target effects.
In summary, our study demonstrates that PC1 regulates cell proliferation and migration and interacts with mTOR and Jak signalling in various cancer cell lines. Given that there is a lack of prior research on the subject of polycystins and cancer biology, this study represents the first steps towards understanding the function of polycystins in the pathophysiology of cancer. As we expected, our research prompted more question than answers. Future research should focus on the mechanism through which PC1 promotes or inhibits cell proliferation and migration, and the molecular details of the interaction between PC1 and mTOR and Jak signalling.
Moreover, future studies on polycystins and cancer should explore whether polycystins are associated with any other signalling pathways in cancer cells. It would also be interesting to evaluate the clinical relevance of polycystins in cancer by studying human cancer tissues. All the above will reveal the significance of polycystins in cancer biology and may lead to the identification of new therapeutic targets or prognostic markers in cancer.

CO N FLI C T O F I NTE R E S T
The authors declare that there is no conflict of interest.