Abstract
Pancreatic cancer remains one of the most difficult to treat human cancers despite recent advances in targeted therapy. Inhibition of isoprenylcysteine carboxylmethyltransferase (ICMT), an enzyme that posttranslationally modifies a group of proteins including several small GTPases, suppresses proliferation of some human cancer cells. However, the efficacy of ICMT inhibition on human pancreatic cancer has not been evaluated. In this study, we have evaluated a panel of human pancreatic cancer cell lines and identified those that are sensitive to ICMT inhibition. In these cells, ICMT suppression inhibited proliferation and induced apoptosis. This responsiveness to ICMT inhibition was confirmed in in vivo xenograft tumor mouse models using both a small-molecule inhibitor and shRNA-targeting ICMT. Mechanistically, we found that, in sensitive pancreatic cancer cells, ICMT inhibition induced mitochondrial respiratory deficiency and cellular energy depletion, leading to significant upregulation of p21. Furthermore, we characterized the role of p21 as a regulator and coordinator of cell signaling that responds to cell energy depletion. Apoptosis, but not autophagy, that is induced via p21-activated BNIP3 expression accounts for the efficacy of ICMT inhibition in sensitive pancreatic cancer cells in both in vitro and in vivo models. In contrast, cells resistant to ICMT inhibition demonstrated no mitochondria dysfunction or p21 signaling changes under ICMT suppression. These findings not only identify pancreatic cancers as potential therapeutic targets for ICMT suppression but also provide an avenue for identifying those subtypes that would be most responsive to agents targeting this critical enzyme. Mol Cancer Ther; 16(5); 914–23. ©2017 AACR.
Introduction
Pancreatic carcinomas are among the most difficult cancers to treat, and the 5-year survival rate remains as low as 5% (1). Isoprenylcysteine carboxylmethyltransferase (ICMT) is the enzyme that catalyzes the last step of posttranslational prenylation-dependent modification of proteins. Most ICMT substrate proteins contain a C-terminal CAAX consensus motif. The carboxylmethylation by ICMT is essential for the proper function of CAAX proteins by regulating their subcellular localization, protein–protein interactions, and/or protein stability (2–5). ICMT has shown promise as a therapeutic target, as suggested by genetic and inhibitor studies (6–9). Recent investigations have indicated that inhibition of ICMT leads to metabolic disarray and suppression of cell anabolism and proliferation in some cancers (10–12). However, the role of ICMT in human pancreatic cancer tumorigenesis has not been directly addressed, despite confusing genetic studies in mice (6, 13, 14).
CDKN1A (often referred to as p21Cip1/Waf1 or p21) belongs to the Cip and Kip family of CDK inhibitors that bind to and inhibit the function of G1 cyclin/CDK complexes (15, 16). p21 has been mostly studied as a factor that mediates the downstream signaling of wild-type p53 as tumor suppressor, particularly in response to DNA damage, to cause cell-cycle arrest (17). However, a number of studies have suggested that p21 performs tumor-suppressive functions independent of p53 (18–23). Apart from inhibiting cell-cycle progression as a CDK inhibitor, p21 also regulates gene transcription and apoptosis (18). Many cancers have altered or loss of p53 function, and p21 tumor suppressor functions are of particular interest in these cancers, among which many are of pancreatic origin (24).
Intrinsic or acquired resistance to apoptosis is a major reason for treatment failure of pancreatic cancers, and this has been frequently associated with the dysregulation of BCL2 family proteins (25–27). BNIP3 is a mitochondrial member of proapoptotic BCL2 family protein containing a motif similar to the BH3 domain (26, 28–30). BNIP3 can interact with the prosurvival BCL2 family members BCL2 and BCLXL and thereby facilitates the induction of apoptosis (31–33). In addition to its reported role in apoptosis, BNIP3 has been identified as a regulator of autophagy (34, 35). Supporting its role in cancer survival, BNIP3 expression was found to be downregulated in various types of cancers, including in pancreatic adenocarcinoma (PDAC), in comparison to normal tissues (36, 37). Loss of BNIP3 correlates with poorer survival, and downregulation of BNIP3 results in increased resistance of pancreatic cancer to cytotoxic drug treatment (33, 38). Furthermore, BNIP3 expression was found to sensitize pancreatic cancer cells to apoptosis (39).
In this study, we investigated the impact of ICMT inhibition on human pancreatic cancer cells. We found that a subset of these cells are sensitive to ICMT inhibition through the inhibition of mitochondria function and induction of an energy-depleted state, which results in the elevation of p21 and p21-dependent BNIP3 expression, leading to cell-cycle arrest and apoptosis.
Materials and Methods
Cells and ICMT inhibitor
ATCC pancreatic cancer cell lines MiaPaCa2, AsPC-1, PANC-1, BxPC-3, PANC-10.05, CAPAN-2, and HPAF-II were obtained from Duke University Tissue Culture Facility (Durham, NC) in 2011. These cell lines have been cultured per ATCC guidelines. The ICMT inhibitor cysmethynil was synthesized by the Duke Small Molecule Synthesis Facility via established methods (7, 40). Cysmethynil treatment of cells was performed as described in prior publications from the laboratory (8, 10).
Cell culture, viability study, and soft-agar colony formation assay
Cells were seeded in standard DMEM containing 10% FBS and allowed to attach overnight. Cells were treated with various agents and collected for protein and mRNA analysis at the time points indicated in the respective figure legends. For glucose starvation studies, culture media were replaced with glucose-free DMEM (GIBCO) supplemented with 10% FBS. Cell viability was determined using CellTiter 96 AQueous One Solution cell proliferation assay (Promega) according to the manufacturer's instructions. Soft-agar colony formation was determined as described (7).
Flow cytometric analysis
Saturating propidium iodide/DAPI staining was used for analyzing cell cycle and apoptotic cell death as described (8). Flow cytometry was done using Miltenyi Biotec MACSQuant VYB Flow instrument, and data were analyzed by FlowJo software (FlowJo).
Oxygen consumption rate analysis
Cell and mitochondrial respiration was determined using Seahorse XF24 analyzer (Seahorse Bioscience), as previously described (12). Briefly, cells were seeded in XF24 cell culture plates at densities appropriate for each cell line and incubated under standard cultural condition. One hour before performing the oxygen consumption rate (OCR), the media were replaced with XF assay medium (Seahorse Bioscience) and incubated at 37°C in a CO2-free environment. Oligomycin, FCCP, and rotenone plus antimycin A were added sequentially to measure OCR, as described.
Quantitative RT-PCR, Western blot, siRNA knockdown, and cloning
These procedures are described in our previous report (41). Antibodies for p21 (2947), p-p53, cleaved caspase-7 (9492), cyclin D1 (2978), p-Rb (S795) (9301), and PARP (9542) were from Cell Signalling. The anti-LC3 and anti-p53 were from Abgent (Ab1802a) and Santa Cruz (sc-s28), respectively. The siRNA sequences used are: p53-1, 5′-GUAAUCUACUGGGACGGAATT-3′ and p53-2, 5′-GGUGAACCUUAGUACCUAATT-3′; ATG5: 5′-AUUCCAUGAGUUUCCGAUUGAUGGC-3′.
Cloning and virus production
BNIP3 shRNA was designed and cloned into pSuper retroviral vector according to the User Manual (OligoEngine). The targeting sequences for BNIP3 were: (i) 5′-CACGAGCGTCATGAAGAAA-3′ and (ii) 5′-TACTGCTGGACGCACAGCA-3′. MCherry-expressing p21 shRNA retroviral vector with the target sequence 5′-CTAGGCGGTTGAATGAGAG-3′ was a gift from Dr. Mathijs Voorhoeve. ICMT shRNA–expressing plasmids were constructed in lentiviral vector PLL3.7. The ICMT target sequences were: (i) 5′-CCCTGTCATTGTTCCACTATT-3′ and (ii) 5′-CTTGGTTTCGGCATCCTTCTT-3′. For expression, BCL-XL cDNA was cloned into retroviral vector pMSCV. HEK293T transfection and viral production were done according to the standard calcium phosphate protocol (42).
Mouse xenograft studies
MiaPaCa2 xenograft tumors were developed by subcutaneous injection of 10 million MiaPaCa2 cells into the flanks of SCID mice that were 6 to 10 weeks old and weighed 18 to 20 g. Drug treatment, tumor measurement and euthanization were performed as described previously (8), in accordance with IACUC guidelines.
Statistical analysis
GraphPad Prism (GraphPad) and Instat (GraphPad) software were used for data analysis and presentation. To calculate the statistical significance, experimental groups were compared to the control group using Dunnet test one-way ANOVA to generate P values. All experimental data are presented as mean ± SD. Differences were considered statistically significant at P < 0.05.
Results
Suppression of ICMT inhibits proliferation and induces apoptosis in human pancreatic cancer cells
Treatment of a panel of pancreatic cancer cell lines including MiaPaCa2, AsPC-1, PANC-1, BxPC-3, CAPAN-2, PANC-10.05, and HPAF-II with the ICMT inhibitor cysmethynil resulted in dose-dependent inhibition of proliferation and reduction of viability (Fig. 1A, top; Supplementary Fig. S1). The study identified ICMT inhibition–sensitive cell lines and those relatively resistant, with MiaPaCa2 being the most responsive line (Fig. 1A, top; Supplementary Fig. S1). The steep apparent dose–response curves are the result of serum-binding property of cysmethynil (43). The inhibition of ICMT enzymatic activity can be followed by analyzing the accumulation of prelamin A (Fig. 1A, bottom), which directly correlates with the unmethylated prelamin A level (44). Indeed, the range of cysmethynil concentrations causing progressively increasing prelamin A levels is consistent with that of decreasing cell viability in MiaPaCa2 cells (Fig. 1A, top), supporting the ICMT-specific action of cysmethynil. Immunoblot analysis of sensitive cells MiaPaCa2 and AsPC-1 treated with increasing concentrations of cysmethynil showed increased levels of cyclin kinase inhibitor p21 and apoptosis marker cleaved caspase-7 and decreased level of cyclin D1 (Fig. 1B), suggesting the induction of cell-cycle arrest and apoptosis. In contrast, no similar molecular changes were observed in the resistant cell lines HPAF-II and CAPAN-2 under the same treatment by cysmethynil (Fig. 1B). Consistent with the viability study (Fig. 1A) and the immunoblot apoptosis markers (Fig. 1B), flow cytometric analysis on MiaPaCa2 cells showed significant increase in apoptotic cell population (i.e., the sub-G1 population) under inhibitor treatment (Fig. 1C). These responses to ICMT inhibition were confirmed by use of shRNA to target ICMT in the sensitive MiaPaCa2 cells and resistant HPAF-II cells. ICMT knockdown resulted in increased levels of p21, cleaved PARP, and caspase-7 and LC3 autophagy marker only in MiaPaCa2, but not HPAF-II cells (Fig. 1D).
Suppression of ICMT inhibits proliferation and induces apoptosis and autophagy in multiple pancreatic cancer cells. A, Top, Cell viability curve of MiaPaCa2, AsPC-1, PANC-1, BxPC-3, PANC-10.05, CAPAN-2, and HPAF-II pancreatic cancer cells treated with cysmethynil at concentrations range from 10 to 40 μmol/L with increments of 2.5 μmol/L for 48 hours. Bottom, Immunoblot study on MiaPaCa2 cell lysates for prelamin A and loading control GADPH. The calculated ratios of prelamin A and GAPDH are shown. B, Immunoblot analysis on lysates of MiaPaCa2, AsPC-1, CAPAN-2, and HPAF-II cells prepared after 48 hours of treatment with 0, 20, 22.5, and 25 μmol/L cysmethynil. C, Flow cytometric quantification of apoptotic (sub-G0) population of MiaPaCa2 cells after 24-hour cysmethynil treatment at 0 or 22.5 μmol/L. D, Immunoblot analysis of the lysates of MiaPaCa2 and HPAF-II cells expressing either control shRNA or that targeting ICMT, 96 hours after infection by shRNA expressing lentiviruses (top). The ICMT knockdown efficiency is assessed by qPCR analysis (bottom). E, Colony formation assay of MiaPaCa2 cells treated with the indicated concentration of cysmethynil for 14 days. F, Colony formation assay of MiaPaCa2 cells expressing either control shRNA or that targeting ICMT; qPCR analysis of ICMT expression levels are presented on the right. G, Growth of MiaPaCa2 xenograft tumors under every other day treatment by vehicle, 100 or 150 mg/kg cysmethynil. n = 6 for each group. ***, P < 0.001 between different groups. H, Image to demonstrate the contralateral growth of xenograft MiaPaCa2 tumors derived from cells expressing either control shRNA (right side of mouse, pointed by black arrow) or that targeting ICMT (left side of mouse, pointed by red arrow). I, Analysis of xenograft data on tumors expressing control shRNA or that targeting ICMT. n = 3 for each dosing group. ***, P < 0.001 between treatment groups and control. The qPCR analysis of ICMT expression for the 2 groups of cells used for the study was done before implantation and presented on the right side of I. A–F, Data shown are from a single experiment that has been repeated 3 times with similar results.
To further evaluate the role of ICMT in tumorigenesis, we assessed the impact of ICMT inhibition on MiaPaCa2 cell growth in soft agar and on xenograft tumor formation in mice. ICMT inhibition by either cysmethynil or by the expression of shRNA targeting ICMT reduced soft agar colony formation of MiaPaCa2 (Fig. 1E and F). In the xenograft study, low- and high-dose cysmethynil treatment led to tumor growth inhibition and tumor regression, respectively (Fig. 1G). Similarly, when MiaPaCa2 cells expressing either control shRNA or shRNA targeting ICMT were injected into contralateral sides of the same mice, significant suppression of xenograft tumor formation was observed in the ICMT-knockdown group (Fig. 1H and I). In summary, both genetic and pharmacologic suppression of ICMT showed efficacy against MiaPaCa2 pancreatic cancer cells in vitro and in vivo.
Upregulation of p21, induced by ICMT inhibition, plays an active role in the induction of apoptosis in a p53-independent manner
As noted above, treatment of several pancreatic cancer cell lines with cysmethynil markedly increased p21 protein levels, similarly observed with ICMT knockdown (Fig. 1), which suggests a specific link between ICMT inhibition and p21 induction. While the inhibitory function of p21 in cell-cycle progression has been extensively studied (15), its role in apoptosis is not well understood (18). To investigate the functional role of p21 elevation that occurs with ICMT inhibition, 2 MiaPaCa2 clones with stable knockdown of p21 were generated by introducing a retroviral vector expressing shRNA targeting p21. As expected, these cells exhibited significantly reduced p21 levels both at baseline and upon cysmethynil treatment, compared with parental cells (Fig. 2A). Importantly, flow cytometric analysis demonstrated that apoptosis was significantly attenuated in the p21-knockdown cells when subjected to cysmethynil treatment, in comparison to the parental cells similarly treated (Fig. 2B). Consistent with these findings, viability assays showed that p21-knockdown clones survived better under cysmethynil treatment (Fig. 2C). Further investigation showed that p21-knockdown cells were also much more resistant to ICMT inhibition in the soft-agar colony formation assay compared with control cells (Fig. 2D). In xenograft mouse model studies, while the tumors from cells expressing control shRNA were responsive to cysmethynil treatment (Fig. 2E), similar to the parental MiaPaCa2 cells (Fig. 1G), tumors from cells expressing shRNA targeting p21 were resistant to cysmethynil treatment (Fig. 2F), despite their faster growth rate due to the loss of p21. Hence, both in vitro and in vivo studies indicate that p21 elevation is functionally important for the efficacy of ICMT inhibition in sensitive pancreatic cancer cells.
p21, induced by suppression of ICMT, promotes apoptosis of MiaPaCa2 cells in vitro and inhibits tumor formation in vivo. A, p21 levels were assessed in parental MiaPaCa2 cells (EV) and 2 derived clones expressing shRNA targeting p21 (sh-p21 B and Q), after 48 hours of treatment with concentrations of cysmethynil of 0, 17.5, 20, and 22.5 μmol/L. B, Flow cytometric analysis of parental and p21 shRNA expressing MiaPaCa2 cells, after 48-hour treatment with vehicle or 22.5 μmol/L of cysmethynil. C, Viability of MiaPaCa2 parental and p21-knockdown clones after 48-hour treatment of cysmethynil at concentrations of 0, 17.5, 20, and 22.5 μmol/L. D, Soft-agar colony formation was evaluated for parental MiaPaCa2 and a stable p21-knockdown clone (sh-p21), under the treatment by vehicle or 20 μmol/L cysmethynil for 14 days. Samples from technical repeats were analyzed and the data are presented on the right of D. E and F, In vivo efficacy study of cysmethynil treatment in xenograft tumor model of parental MiaPaCa2 cells (E) and stable p21-knockdown cells (sh-p21; F). Animals were dosed with vehicle or cysmethynil at 150 mg/kg every other day. ***, P < 0.001. A–D, Data shown are from a single experiment that has been repeated 3 times with similar results.
Although p21 production is subject to multiple modes of regulation, p53 is considered the most prominent regulator (17). However, cysmethynil treatment, while inducing p21 and autophagosome protein LC3 in a dose-dependent fashion, did not elicit significant changes in either the total amount of p53 or its Ser15-phosphorylated nuclear form in MiaPaCa2 cells (Supplementary Fig. S2A). Furthermore, knockdown of p53 had no effect on either the increases in the levels of p21 or LC3 resulting from cysmethynil treatment (Supplementary Fig. S2B). Hence, the induction of p21 and its link to cell responsiveness to ICMT suppression are independent of p53 function.
p21 elevation induced by ICMT suppression promotes the expression of autophagy and apoptosis genes
We next evaluated the impact of ICMT inhibition on apoptosis and autophagy in MiaPaCa2 cells. We found that cysmethynil treatment increased p21, LC3, ULK1, and BNIP3 mRNA levels (Fig. 3A). Similarly, MiaPaCa2 cells expressing shRNA targeting ICMT demonstrated elevated p21, LC3, ULK1, and BNIP3 mRNA levels, albeit at more modest levels compared with that from inhibitor treatment (Fig. 3B). The differences are likely due to the requirement for high efficiency and prolonged knockdown to completely suppress the cellular function of ICMT. Worth noting, analysis of mRNA levels in the MiaPaCa2 xenografts consistently showed higher expression of p21, LC3, ULK1, and BNIP3 in the tumors exposed to cysmethynil, in a dose-dependent fashion, compared with those treated with control vehicle (Fig. 3C). Furthermore, the xenograft tumors derived from cells expressing shRNA targeting ICMT also demonstrated higher levels of expression of these genes compared with tumors containing control shRNA (Fig. 3D). These in vitro and in vivo results strongly support the notion that ICMT inhibition induction of these genes is functionally important for the antitumor efficacy. LC3, ULK1, and BNIP3 proteins promote cellular autophagy (45), and BNIP3 also functions as a proapoptosis BCL family member. As the evidence demonstrated the importance of p21 in ICMT inhibition in targeting tumor cell survival and proliferation (Fig. 2), we investigated the role of p21 in these autophagy and apoptosis gene expression. We found that p21 knockdown significantly attenuated the induction of ULK1, LC3, and BNIP3 transcription arising from cysmethynil treatment (Fig. 3E), placing p21 as an upstream regulator for the transcription of these genes, which is an underexplored role for p21, in contrast to its well-known function in regulating cell cycle (15).
ICMT inhibition–induced p21 promotes transcription of autophagy and apoptosis genes. All panels show qPCR expression analysis of the indicated genes. RNA samples were prepared from (A) MiaPaCa2 cells after 48-hour treatment with 0, 20, or 22.5 μmol/L cysmethynil, (B) MiaPaCa2 cells 96 hours after infection with lentivirus expressing either control shRNA or that targeting ICMT, (C) xenograft tumor samples obtained from mice treated every other day with either vehicle or cysmethynil at 100 and 150 mg/kg, respectively, (D) xenograft tumor samples derived from MiaPaCa2 cells expressing control shRNA or that targeting ICMT, and (E) MiaPaCa2 cells selected to express either control shRNA (−) or shRNA targeting p21 (+) after 48-hour treatment with cysmethynil at 0, 17.5, 20, or 22.5 μmol/L. All studies have been repeated 3 times with similar results.
Induction of BNIP3 plays an essential role in pancreatic cancer cell apoptosis induced by ICMT inhibition
Among the genes whose expression is induced by ICMT inhibition, LC3 and ULK1 have well-established roles in promoting cellular autophagy. BNIP3, on the other hand, has been reported to have dual roles of promoting autophagy and apoptosis (46, 47). Furthermore, recent studies suggest that downregulation or silencing of BNIP3 may be an important step in tumorigenesis and tumor maintenance in several human cancer types, most notably pancreatic cancers (36, 37). To assess the importance of upregulation of BNIP3 and it role in apoptosis and autophagy in the sensitivity of pancreatic cancer cells to ICMT inhibition, we studied the responsiveness of MiaPaCa2 cells to cysmethynil when BNIP3 expression was silenced. Knockdown of BNIP3 by shRNA (Fig. 4A) significantly reduced cysmethynil-induced apoptosis, as assessed by the sub-G1 population from flow cytometry (Fig. 4B) and consistent reduction of the apoptosis markers cleaved PARP and caspase-7 (Fig. 4C).
BNIP3-stimulated apoptosis, but not autophagy, promotes cell death resulted from ICMT inhibition. A, qPCR analysis of BNIP3 expression in MiaPaCa2 cells expressing either control shRNA (−) or shRNA targeting BNIP3 (+) after 48-hour treatment with 0, 20, or 22.5 μmol/L cysmethynil. B, Flow cytometric analysis of MiaPaCa2 cells with or without BNIP3 knockdown after 48-hour treatment with either vehicle or 22.5 μmol/L cysmethynil. C, Immunoblot analysis of the autophagy and apoptosis markers on cell lysates from (A). D, Immunoblot analysis of autophagy and apoptosis markers on MiaPaCa2 cells treated with cysmethynil at 0, 20, and 22.5 μmol/L, with or without Atg5 knockdown as indicated. E, Viability of MiaPaCa2 cells treated with 0, 17.5, 20, and 22.5 μmol/L cysmethynil for 48 hours, with or without Atg5 knockdown as indicated. F, Flow cytometric analysis of MiaPaCa2 cells treated with vehicle or 22.5 μmol/L cysmethynil for 48 hours, with or without of Atg5 knockdown as indicated. For all studies, data shown are from a single experiment that has been repeated 3 times with similar results.
Consistent with the notion that BNIP3 is a positive regulator of cellular autophagy (34, 35, 48), suppression of BNIP3 expression also reduced the level of autophagy marker LC3 (Fig. 4C). Autophagy is known to be double edged in cancer cell survival. Studies have demonstrated that excessive stimulation of autophagy can also lead to cell death (8). We then investigated the role of autophagy in the sensitivity of MiaPaCa2 cells to ICMT inhibition by concurrently suppressing autophagy using Atg5 siRNA and treating with cysmethynil. As expected, cellular autophagy was impaired in Atg5 knockdown cells (Fig. 4D). However, the reduction of cell viability and induction of apoptosis resulted from ICMT inhibition were not significantly different between cells transfected with control or Atg5 siRNA (Fig. 4D–F), suggesting that autophagy induction is not a pro-death mechanism in MiaPaCa2 cells under ICMT inhibition. In contrast, overexpression of BCL-XL confers significant resistance of MiaPaCa2 cells to cysmethynil treatment (Supplementary Fig. S3), supporting the role of apoptosis in the cell death induced by ICMT inhibition. In summary, these data indicate that induction of p21 and p21-stimulated expression of proapoptotic BNIP3 mediates the pancreatic cancer cell apoptosis induced by ICMT inhibition.
ICMT inhibition–sensitive, but not the resistant, pancreatic cancer cells demonstrated characteristic alterations of AMPK-mTOR, cell proliferation, and apoptosis signaling profiles in response to ICMT inhibition
Cell viability assays have identified MiaPaCa2 and AsPC-1 cells to be more sensitive (Fig. 1A; Supplementary Fig. S1), while HPAF-II, CAPAN-2, and BxPC-3 are more resistant, to cysmethynil treatment (Fig. 1A; Supplementary Fig. S1). To assess determinants of sensitivity to ICMT inhibition in pancreatic cancer cells, we compared the molecular responses of MiaPaCa2 and HPAF-II cells to ICMT inhibition. MiaPaCa2 cells responded to ICMT inhibition with increased p21 and BNIP3 expression (Fig. 5A, top), whereas HPAF-II showed little response (Fig. 5A, bottom). Analysis of major regulatory pathways showed different cell signaling profiles between MiaPaCa2 and HPAF-II cells in several categories, including cell-cycle regulators, mTOR and AMPK signaling, and apoptosis indicators (Fig. 5B). MiaPaCa2 cells presented with increased level of p21 and decreased levels of Cyclin D1 and pRb, indications for G1 cell-cycle arrest. In contrast, these markers were unchanged in the resistant HPAF-II cells under the same cysmethynil treatment.
Suppression of ICMT inhibits cell-cycle progression and induces metabolic stress and cell death in sensitive pancreatic cancer cells. A, qPCR analysis of BNIP3 and p21 mRNA levels in MiaPaCa2 (top) and HPAF-II (bottom) cells after 48-hour treatment with DMSO control or 22.5 μmol/L cysmethynil. B and C, Immunoblot analysis of the indicated proteins in MiaPaCa2 and HPAF-II cells following 48-hour treatment with DMSO control (−) or 22.5 μmol/L (+) of cysmethynil (B), or MiaPaCa2 cells 96 hours after infection with lentivirus-expressing either control shRNA (−) or that targeting ICMT (+) (C).
MiaPaCa2 cells also showed increased phosphorylation of AMPK and ACC under cysmethynil treatment, suggesting that the cells may be under metabolic stress. Consistent with this notion, mTOR signaling was suppressed by cysmethynil treatment in MiaPaCa2 cells, as demonstrated by hypophosphorylation of 4EBP1 and S6 (Fig. 5B). Together, these data suggest that ICMT inhibition activated catabolic and suppressed anabolic signaling, consistent with the observation of elevated autophagy. In contrast, HPAF-II cells showed no significant changes in these markers. Finally, while apoptosis markers such as cleaved PARP and caspase-7 were elevated in MiaPaCa2 cells under cysmethynil treatment, they were not observed in HPAF-II cells (Fig. 5B), consistent with the outcome of viability study (Fig. 1A; Supplementary Fig. S1). It is important to note that these signaling changes were also observed in MiaPaCa2 cells under the condition of shRNA knockdown of ICMT, including cell-cycle and apoptosis marker changes (Fig. 1D) and AMPK and mTOR signaling changes (Fig. 5C). Further testing of the signaling profile demonstrates that ICMT inhibition induced similar changes in the other sensitive cell line such as AsPC-1 but not in the more resistant BxPC-3 cells (Supplementary Fig. S4), consistent with their sensitivity profile.
Activated p21/BNIP3 signaling, in response to mitochondria respiratory deficiency and metabolic stress, leads to apoptotic cell death
We recently discovered that ICMT inhibition leads to mitochondria dysfunction in PC3 prostate and MDA-MB-231 breast cancer cells, which in turn results in cell metabolic distress (12). We were particularly interested in determining whether sensitive and resistant pancreatic cancer cells respond differently to ICMT inhibition in terms of mitochondrial respiration and cell metabolism, which seemed a likely scenario based on the differences observed in AMPK and mTOR signaling in these lines as described above. To this end, OCRs were measured for MiaPaCa2 and HPAF-II cells treated with either vehicle or cysmethynil. MiaPaCa2 cells responded to cysmethynil with dose-dependent reduction of cell respiration, mitochondria basal respiration, and maximum respiratory capacity (Fig. 6A). HPAF-II cells, in contrast, demonstrated no changes in mitochondria respiration under the same treatment conditions (Fig. 6B). Notably, non-mitochondrial oxygen consumption was not changed under ICMT inhibition in either cell line (Fig. 6A and B), confirming that the reduction of cellular respiration is of mitochondria origin.
Metabolic stress–induced p21 elevation inhibits cell proliferation and promotes autophagy and apoptosis. A and B, Cell respiration analysis of MiaPaCa2 (A) and HPAF-II (B) cells for intact cell, basal mitochondrial, and maximum mitochondrial OCR (O2 consumption) and non-mitochondrial oxygen consumption (ROX) following 24 hours of treatment with 0 μmol/L (DMSO), 20 μmol/L (low-dose cysm), or 22.5 μmol/L (high-dose cysm) cysmethynil. C, Immunoblot analysis of the indicated proteins in MiaPaCa2 cells after growth in normal (+) or glucose-deprived (−) media. D, qPCR analysis of mRNA levels of BNIP3 and p21 in the cells from (C). A–D, Data shown are from a single experiment that has been repeated 3 times with similar results. E, Schematic model summarizing the findings that energy depletion, induced by ICMT inhibition or nutrient deprivation, leads to transcriptional activation of p21 and p21-dependent inhibition of cell proliferation and induction of ULK1, LC3, and BNIP3, which results in further induction of autophagy and apoptosis in susceptible pancreatic cancer cells.
The studies described above suggested that the induction of p21 is an important mediator of the responses of pancreatic cancer cells to energy depletion and metabolic distress arising from ICMT inhibition–induced mitochondria dysfunction. We hypothesized that p21 and apoptosis induction would also occur in other energy-depleted metabolic stress. To investigate whether the regulation by p21 and BNIP3 represented a cellular response mechanism not limited to ICMT inhibition but rather to energy depletion in general, we compared cell signaling in MiaPaCa2 cells grown either under glucose deprivation or normal growth condition. Glucose is a fuel molecule for electron transport chain (ETC) that can be supplied in growth media. Indeed, pAMPK and LC3II levels were increased, and pS6 level was reduced in cells growing in glucose-deprived medium, consistent with energy stress (Fig. 6C), and these cells showed significant upregulation of p21 and BNIP3 expression (Fig. 6D). In the same cells, increased p21 and reduced cyclin D1 and pRb protein levels were observed, suggesting inhibition of cell proliferation (Fig. 6C). Finally, there was consistent increase in the levels of cleaved PARP and caspase-7 under glucose deprivation, demonstrating the induction of apoptosis (Fig. 6C).
In summary, our findings identified a previously undescribed regulatory function of p21, as modeled in Fig. 6E, to be a coordinator for the control of cell proliferation, catabolism, and cell death in response to metabolic stress, which can result from nutrient deprivation or mitochondria dysfunction as in the case of ICMT inhibition. These perturbations can activate p21 and its downstream signaling, leading to cell catabolism, proliferation inhibition, and apoptosis.
Discussion
In this study, we provide a strong case for suppressing ICMT as a method to inhibit cell proliferation and induce cell death in pancreatic cancer cells. In sensitive cells, suppression of ICMT resulted in mitochondria dysfunction and metabolic stress, induction of p21 independent of p53, followed by a constellation of changes that included cell-cycle inhibition, autophagy, and apoptosis. The consistent finding that p21 accumulates in these sensitive cells points to its involvement not only in the cell-cycle arrest but also in autophagy and apoptosis induction. It makes biologic sense that coordinated cell adaptations to energy depletion would include (i) cessation of growth and proliferation that depend on energy-rich condition and anabolic activity such as protein synthesis and lipid synthesis, (ii) compensatory increase in catabolic activity, such as autophagy, which provides emergent production of fuel and building block molecules, and (iii) initiation of cell death process when the energy-depleted state prolongs or gets more severe. This study provides strong evidence that p21 is likely a coordinator for these adaptations. We also identified BNIP3, a BH3 domain–containing BCL2 family protein that promotes both autophagy and apoptosis, as a downstream effector of p21 to regulate autophagy and apoptosis.
It remains unclear why some cancer cells are more sensitive to ICMT inhibition–induced respiratory suppression and resultant energy depletion. One possible explanation for the differences can be that the expression, and therefore the function, of particular ICMT substrate(s) that regulate mitochondria function vary between different cells types. Another possibility is that mitochondria function in different cancer cells have different vulnerability to ICMT suppression mediated by the enzyme's substrates. While the exact mechanism that accounts for the sensitivity of particular cancer cells to ICMT inhibition will depend on further studies, the signatures we have identified of sensitive cells—reduction of ETC function, energy and metabolic distress, p21 elevation and subsequently signaling—should be useful in the identification of cancer types that can be targeted by ICMT inhibition.
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
Authors' Contributions
Conception and design: K.A. Manu, T.F. Chai, M. Wang
Development of methodology: K.A. Manu, T.F. Chai, P.J. Casey, M. Wang
Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): K.A. Manu, J.T. Teh, W.L. Zhu
Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): K.A. Manu, T.F. Chai, J.T. Teh, M. Wang
Writing, review, and/or revision of the manuscript: K.A. Manu, T.F. Chai, P.J. Casey, M. Wang
Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): K.A. Manu, T.F. Chai, W.L. Zhu, M. Wang
Study supervision: M. Wang
Other (secured grant support for the project): M. Wang
Grant Support
Financial support for this work is from Singapore Ministry of Health and Ministry of Education awarded to M. Wang.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Acknowledgments
The authors would like to thank Drs. Koji and Yoko Itahana of Duke-NUS Medical School for the selection of p53 antibodies and PCR primers.
Footnotes
Note: Supplementary data for this article are available at Molecular Cancer Therapeutics Online (http://mct.aacrjournals.org/).
- Received October 23, 2016.
- Revision received January 24, 2017.
- Accepted January 24, 2017.
- ©2017 American Association for Cancer Research.