Cholesterol and different components in cholesterol transport and metabolism pathways have become attractive targets for novel diagnostic and therapeutic strategies to treat various cancers, including PDAC (16). This is not only because intracellular cholesterol flux is so critical for the invasive behavior of cancer cells and cancer progression in general, but also because many cholesterol-regulating drugs, such as HMGCR inhibitors (statins), cholesteryl ester transfer protein (CETP) inhibitors, and niacin derivatives have already been developed and are used clinically for other disorders such as hypercholesterolemia. In this study, we found that PDAC cells have increased levels of cellular FC but reduced ability to store CE. Disturbing cholesterol flux at the levels of cholesterol esterification, CE lipolysis and cholesterol synthesis generally suppressed cell growth and migration in the examined PDAC cells. However, some of the inhibitors of cholesterol flux also promoted cell invasion, depending on cell type and the pathway inhibited, reflecting heterogeneity in cholesterol flux among the different PDAC cell lines. Interestingly, blocking cholesterol flux by SOAT1 inhibition altered the expression of a broad range of proteins in PDAC cells, far beyond components involved exclusively in lipid metabolism. Many of these changes in the PDAC proteome and in cell behavior induced by blocking cholesterol flux were restored by OA, underlining the importance of fatty acids for cellular cholesterol balance in PDAC (Fig. 7).
The expression pattern of proteins involved in lipid flux regulation not only differs between HPDE and PDAC cells, but varies also among different PDAC cell lines, indicating a complex heterogeneity in their lipid metabolism. PDAC cells expressed higher levels of LDLR as well as HMGCR and SQLE compared to HPDE cells, implying their increased cholesterol uptake and synthesis, respectively. This pattern is in accordance with the notion that cancer cells have an increased demand for cholesterol (16). The level of FC was higher in PDAC cells than in HPDE cells, which is consistent with findings from previous studies (28, 29). LDs are hubs for intracellular lipid turnover, including fatty acids and cholesterol. FC can be esterified with fatty acids and stored in LDs as CE, while CE can be degraded via lipolysis or lipophagy to release FC (44). When all cells were grown in the same medium supplemented with 10% FBS, the CE content was lower in PDAC cells compared to HPDE cells. In addition, expression of SOAT1, the enzyme responsible for CE synthesis, was lower in the three PDAC cell lines than in HPDE cells, consistent with reduced CE storage in PDAC. In contrast, a study by Li et al. reported that CE content is higher in PDAC tissue than in normal pancreatic tissue, and higher in PDAC cell lines compared to HPDE cells (34). However, in their study, PDAC cells were grown in medium supplemented with FBS, while HPDE cells were grown in serum-free medium. As serum is rich in both FA and cholesterol, and we also found that LD content in PDAC and HPDE cells is largely affected by exogenous lipids, this discrepancy might be due to different culture conditions. In addition, it might be that CE storage is higher in normal pancreatic duct epithelial cells than in pancreatic acinar cells, which may explain the discrepancy in relative CE content when comparing CE levels in cultured PDAC cells with HPDE cells, or PDAC tissue with normal pancreatic tissue. As PDAC cells have reduced LD storage and CE content, but elevated intracellular levels of FC, these findings suggest that PDAC cells have a reduced capacity to buffer intracellular lipid flux by incorporating them into LDs. Moreover, it seems that the presence of exogenous cholesterol could prominently increase LD storage (mainly CE), more than the presence of fatty acids alone (mainly TAG), indicating that synthesis and/or turnover of LDs are closely associated with cellular cholesterol turnover in both PDAC and HPDE cells.
LDs harbor a continuous turnover of cellular cholesterol through simultaneous esterification and lipolysis. This seemingly futile circle fine tunes the regulation of intracellular cholesterol flux (45, 46). Except for hepatocytes and small intestine epithelial cells, the main enzyme for intracellular esterification of FC is SOAT1 (47). It was recently reported that cholesterol esterification by SOAT1 prevents the negative feedback on cholesterol synthesis elicited by FC in PDAC cells, thereby promoting the mevalonate pathway (38). This may explain why abrogation of cholesterol esterification either by SOAT1 knockdown or by the SOAT1 inhibitor avasimibe suppresses PDAC growth and metastasis (34). Moreover, increased CE storage is suspected to contribute to chemotherapy resistance, while SOAT1 inhibitor synergistically suppresses PDAC growth when given together with gemcitabine (48). In line with these findings, we found that inhibition of SOAT1 by avasimibe prominently increased cell death and inhibited cell proliferation in all the three PDAC cell lines examined and suppressed cell migration in MIA PaCa-2 and PANC-1 cells. However, it seems that SOAT1 inhibition promoted cell migration and invasion in BxPC-3 cells. These findings suggest that the effect of suppressing cholesterol esterification is cell-type dependent. Whether this is related to effects of oncogenic KRAS mutations (e.g., in MIA PaCa-2 and PANC-1, but not BxPC-3), or connected to other unknown mechanisms, needs further study. SOAT1 inhibition may well increase the intracellular cholesterol burden. This is supported by the observed increase in cellular FC content and aggregation, increased expression of ABCA1, and decreased expression of HMGCR.
Lipolysis of CE are less studied in PDAC. Several candidate enzymes may function as neutral CE esterase, such as HSL (49) and NCEH1 (50), yet it is unclear whether they play a role in cholesterol homeostasis in PDAC. HSL catalyzes lipolysis of a broad range of substrates, including TAG, DAG, MAG, and CE (51). CAY10499, an inhibitor of HSL/MGLL, has been reported to suppress PDAC cell invasion by reducing the availability of fatty acids that would otherwise fuel cancer metastasis (52). However, whether HSL inhibition affects cholesterol flux in PDAC has not been investigated previously. We found that CAY10499, but not BAY599435 (a HSL-specific inhibitor), suppressed PDAC. Thus, it is possible that the observed suppressive effect of CAY10499 in PDAC depends on inhibition of MGLL in addition to inhibition of HSL. NCEH1 (also known as KIAA1363) is known to lipolyze CE in macrophages (53) and is overexpressed in various cancers (54). NCEH1 inhibition has been reported to suppress prostate cancer, but it is unclear if this suppression disturbs cellular cholesterol balance (55–57). NCEH1 expression levels are inversely correlated with disease-free survival of PDAC patients (58–60); however, it remains unclear whether NCEH1 regulates cholesterol balance in PDAC. We found that NCEH1 inhibitor JW480 significantly increased CE content in the presence of FA, indicating that NCEH1 might be an important enzyme responsible for CE lipolysis in PDAC. Furthermore, we observed that inhibition of NCEH1 suppressed PDAC cell growth and migration, but apparently increased cell invasion.
HMGCR is a rate-limiting enzyme in the cholesterol synthesis pathway. HMGCR inhibitors such as statins have been widely used clinically for their cholesterol-lowering effect (61). Statins have been shown to exhibit tumor-suppressive properties in multiple cancer types (62, 63). However, the effects of statins on PDAC seem to be complex and controversial (36, 39, 64–66). It has recently been reported that the therapeutic effect of statins on PDAC is dependent on tumor differentiation grade (67). In our study, inhibition of HMGCR with simvastatin suppressed cell survival and proliferation in all three PDAC cell lines, and inhibited cell migration (except in BxPC-3), and invasion. Although inhibition of HMGCR slightly reduced cellular cholesterol content, it also induced aggregation of FC and increased the expression of ABCA1, indicating disturbed intracellular cholesterol flux.
We found that the impact of inhibitors targeting intracellular cholesterol flux was influenced by the availability of external lipids. Upon inhibition of LD lipolysis (HSL/MGLL and NCEH1), cholesterol esterification (SOAT1) or cholesterol synthesis (HMGCR), we observed decreased HMGCR, increased ABCA1, and reduced PDAC cell survival and proliferation. Most of these alternations could be further aggravated by addition of FC in the culture medium, but surprisingly, these inhibitor-induced changes were completely reversed by addition of OA. Previously, OA has been reported to inhibit cholesterol synthesis in glioma cells through downregulation of HMGCR expression and activity (68). In our study, OA slightly reduced cellular FC levels (independent of lipid flux inhibitors) and reduced cholesterol aggregation in PDAC cells, without obvious change in HMGCR expression, indicating that fatty acids may promote intracellular cholesterol balance at multiple levels. Since OA did not significantly increase CE content in the presence of these inhibitors (except for NCEH1 inhibitor), cholesterol esterification may not be the main mechanism by which OA restores the cholesterol balance. As fatty acids can affect cellular membrane components, one possibility would be that they may interfere with the endosomal sorting system and ER membranes (69), and thus promote intracellular cholesterol transport.
As discussed above, SOAT1 is a key regulator of cholesterol balance and PDAC cell growth. To our knowledge, avasimibe-induced changes in the PDAC cell proteome have not been reported previously. Noteworthy, the present study demonstrates a substantially altered proteome of PANC-1 cells following the blocking of cholesterol flux by SOAT1 inhibition. The induced changes affected approximately one third of the cell proteome, extending far beyond proteins involved in the cholesterol metabolism per se. The changes mainly involve proteins regulating lipid metabolism, cell survival and growth, as well as interactions with the extracellular matrix. Notably, most of these changes were restored by OA. Overall, these findings indicate that fatty acids regulate cholesterol balance in PDAC cells and thereby may have an impact on PDAC cell behavior.