Prostate Apoptosis Response 4 (Par-4), a Novel Substrate of Caspase-3 during Apoptosis Activation

Abstract

Prostate apoptosis response 4 (Par-4) is a ubiquitously expressed proapoptotic tumor suppressor protein. Here, we show for the first time, that Par-4 is a novel substrate of caspase-3 during apoptosis. We found that Par-4 is cleaved during cisplatin-induced apoptosis in human normal and cancer cell lines. Par-4 cleavage generates a C-terminal fragment of ∼25 kDa, and the cleavage of Par-4 is completely inhibited by a caspase-3 inhibitor, suggesting that caspase-3 is directly involved in the cleavage of Par-4. Caspase-3-deficient MCF-7 cells do not show Par-4 cleavage in response to cisplatin treatment, and restoration of caspase-3 in MCF-7 cells produces a decrease in Par-4 levels, with the appearance of a cleaved fragment. Additionally, knockdown of Par-4 reduces caspase-3 activation and apoptosis induction. Site-directed mutagenesis reveals that Par-4 cleavage by caspase-3 occurs at an unconventional site, EEPD¹³¹↓G. Interestingly, overexpression of wild-type Par-4 but not the Par-4 D131A mutant sensitizes cells to cisplatin-induced apoptosis. Upon caspase-3 cleavage, the cleaved fragment of Par-4 accumulates in the nucleus and displays increased apoptotic activity. Overexpression of the cleaved fragment of Par-4 inhibits IκBα phosphorylation and blocks NF-κB nuclear translocation. We have identified a novel specific caspase-3 cleavage site in Par-4, and the cleaved fragment of Par-4 retains proapoptotic activity.

INTRODUCTION

Caspases belong to a family of cysteine proteases which participate in the cleavage of aspartic acid-containing motifs (43). Caspases are classified into two groups. Upstream “initiator” caspases, which include caspase-6, -8, -9, and -10, function as initiators of a proteolytic cascade by activating the procaspases to amplify the death signal. The substrate specificity of initiator caspases is V/LEXD, a site similar to that found in procaspase (43). The second group, consisting of caspase-2, -3, and -7, are known as downstream “effector” caspases; they are activated by initiator caspases and have substrate specificity for the DEXD motif, a cleavage site similar to that found in many target proteins (21). In most cases caspase activation is indispensable for complete cell apoptosis (37). Over 400 caspase substrates have been identified, including proteins involved in apoptosis, DNA metabolism and repair, and regulation of the cell cycle and proliferation (15), and the number is still increasing with the addition of tumor suppressor proteins (29, 41).

Par-4, the product of the proapoptotic gene Par-4, was originally identified in prostate cancer cells undergoing apoptosis (39). Since its discovery, Par-4 has been shown to possess a remarkable apoptotic activity in response to numerous stimuli in various cellular systems (5, 12, 33). Previous studies (reviewed in reference 42) suggest that the role of Par-4 in apoptosis is cancer cell selective in that (i) overexpression of Par-4 triggers apoptosis in various cancer cell lines but not in normal and primary cells (13), (ii) depletion of Par-4 by RNA interference (RNAi) confers resistance in cancer cells, but not in primary fibroblasts, to various apoptotic agents (1), and (iii) Par-4 displays proapoptotic functions in cells transformed with oncogene Ras but not in normal cells (33). Recently, Par-4 was shown to be secreted by mammalian cells and, through interaction with the cell surface receptor GRP78, to induce cancer cell apoptosis in a specific manner (4).

Human Par-4 protein consists of 342 amino acids, whereas rat Par-4 has 332 amino acids and mouse Par-4 has 333 amino acids. Par-4 contains conserved functional domains which include (i) two putative nuclear localization sequences (NLS), designated NLS1 (with reference to rat Par-4, amino acid residues 20 to 25) and NLS2 (residues 137 to 153) in the N-terminal region, (ii) a leucine zipper domain spanning amino acids 290 to 332 in the C-terminal region, and (iii) a nuclear export sequence in the C terminus (12). Studies have provided evidence that NLS2 is essential for nuclear localization and induction of apoptosis by Par-4 (13). Analysis of several deletion mutants of the Par-4 protein led to the identification of a unique core domain (residues 137 to 195), which induces apoptosis specifically in cancer cells and therefore is called the SAC domain (selective for apoptosis in cancer cells) (13). All these domains are 100% conserved in the human, rat, and mouse Par-4s. In fact, Par-4 also contains potential nontypical caspase cleavage sites, raising the possibility that it could be cleaved during apoptosis. Here, we hypothesized that tumor suppressor Par-4 could also be a substrate of caspase-3 during cisplatin-induced apoptosis.

In the present study, we found that cisplatin causes a decrease of full-length Par-4 in multiple cell lines, with a concomitant increase of a cleaved fragment of Par-4. Additionally, we demonstrated that Par-4 is a direct substrate of caspase-3. Likewise, Par-4 is not cleaved in cisplatin-treated caspase-3-deficient MCF-7 cells; this effect could be reversed upon restoration of caspase-3 in MCF-7 cells. Using site-directed mutagenesis, we have identified that EEPD¹³¹↓G is the caspase-3 cleavage site in Par-4. Further, we demonstrated that the caspase-3-generated cleaved fragment of Par-4 exhibits increased apoptotic activity. Taken together, our data point to a novel caspase-3-mediated cleavage of Par-4 during apoptosis induction.

MATERIALS AND METHODS

Cell culture and reagents.

Human ovarian cancer cell lines, including A2780-S, A2780-CP, and OVCAR-3, human endometrial carcinoma cell lines, including KLE, Ishikawa, and Hec-1-A, human cervical carcinoma cell line HeLa, human prostate carcinoma cell line PC3, human breast cancer cell lines, including MCF-7 and MDA-468, human uterine epithelial cells (HIEEC), and human uterine stromal cells (HIESC) were used for the present study (HIEEC and HIESC were kindly provided by Michel A. Fortier, Université Laval, Québec, Canada). Cisplatin, doxorubicin, Hoechst 33248, and MG132 were obtained from Sigma-Aldrich (St. Louis, MO). Human and mouse recombinant active caspase-3 and caspase inhibitors were purchased from Calbiochem (La Jolla, CA). All the antibodies were obtained from Cell Signaling Technology (Danvers, MA) except for anti-Par-4 against the C terminus and GAPDH (glyceraldehyde-3-phosphate dehydrogenase) (Abcam, Cambridge, MA) and anti-rabbit secondary antibody (Bio-Rad, Hercules, CA).

Reverse transcriptase PCR (RT-PCR).

To measure the transcript level, total RNA was isolated from cells using TRIzol reagent (Invitrogen, Carlsbad, CA) according to the manufacturer's instructions. The RNA (400 ng) from each group was reverse transcribed using Moloney murine leukemia virus (M-MLV) reverse transcriptase (Invitrogen, Carlsbad, CA) and oligo(dT) primers. The reverse-transcribed RNA was then amplified by PCR using specific primers. GAPDH was used as an internal control to minimize intersample variations. Human Par-4 was amplified using sense primer 5′-GCCGCAGAGTGCTTAGATGAG-3′ and antisense primer 5′-GCAGATAGGAACTGCCTGGATC-3′ (fragment length, 136 bp). For GAPDH, expression was determined using sense primer 5′-GTCAGTGGTGGACCTGACCT-3′ and antisense primer 5′-TGAGCTTGACAAAGTGGTCG-3′ (fragment length, 212 bp). Each reaction mixture (final volume, 50 μl) contained 1× buffer, cDNA or water for negative control (5 μl), MgCl₂ (1.5 mM), deoxynucleoside triphosphates (dNTPs; 0.2 mM), primers (10 μM each), and Taq polymerase (1 unit). PCRs were performed in an MJ Research thermal cycler (model PTC-100) using the following parameters: 30 s at 94°C, 30 s at 58°C, and 1 min at 72°C for 35 cycles (Par-4) or 25 cycles (GAPDH). The PCR products obtained were electrophoresed on a 1% agarose gel and visualized using SYBR-Safe (Invitrogen, Carlsbad, CA) staining.

Western blot analysis.

Floating and attached cells were harvested and washed with phosphate-buffered saline (PBS). Cells were lysed in cold radioimmunoprecipitation assay (RIPA) lysis buffer containing protease inhibitors (Complete; Roche Applied Science), followed by three freeze-thaw cycles. Protein content was determined using the Bio-Rad DC protein assay. Equal amounts of protein were separated onto 10 to 15% polyacrylamide gels and transferred to nitrocellulose membranes (Bio-Rad, Hercules, CA). The membranes were blocked with 5% milk in PBS with 0.05% Tween 20 for 1 h at room temperature and probed with primary antibody (overnight at 4°C). Following incubation with primary antibody, membranes were washed in PBS with 0.05% Tween 20 and incubated with horseradish peroxidase-conjugated secondary antibody (Bio-Rad, Hercules, CA). Detection was performed using SuperSignal West Femto substrate (Thermo Scientific, Rockford, IL), as described by the manufacturer.

Immunofluorescence microscopy analysis.

Fluorescence microscopy was performed on cells grown in 6-well plates containing sterile coverslips. On the day of analysis, culture media were removed and cells were washed twice with PBS, fixed with 4% paraformaldehyde for 10 min, and permeabilized for 10 min using 0.1% Triton X-100 in 0.1% sodium citrate. After being blocked with Dako blocking serum for 1 h, cells were incubated with anti-Myc antibody (1:100 dilution) or isotypic control antibody for 1 h. After being washed with PBS, cells were incubated with fluorescent-tag-conjugated secondary antibodies (as mentioned in the figure legends; 1:200 dilution) for 30 min in the dark. Rhodamine-phalloidin, which stains the actin cytoskeleton, is used to visualize the shape and integrity of the cells. Cells were counterstained with Hoechst 33248 (0.25 μg/ml) for 5 min, and slides were mounted using Slowfade gold antifading reagent (Invitrogen, Carlsbad, CA) and viewed under a Carl Zeiss Axio observerZ1 microscope.

Hoechst nuclear staining.

The treated cells were collected, washed twice in PBS, resuspended at an approximate density of 2 × 10⁵ cells/ml in PBS containing Hoechst 33258 (Sigma-Aldrich, St. Louis, MO), and incubated for 24 h at 4°C before fluorescence microscopy analysis of apoptotic cells. At least 200 cells were counted for each sample, and a percentage of apoptotic cells was calculated as the ratio of apoptotic cells (with characteristic apoptotic morphology such as nuclear shrinkage and condensation) to total cell count.

Flow cytometry analysis.

For measuring apoptosis, transfected cells were dual stained with propidium iodide (PI) and Alexa Fluor 488-annexin V using an Alexa Fluor 488-annexin V Dead Cell apoptosis kit (Invitrogen, Carlsbad, CA) according to the manufacturer's protocol. Stained cells were analyzed by flow cytometry (FC 500 MPL system; Beckman Coulter).

Inhibition of caspase activity.

To assess the role of caspase-3 in Par-4 cleavage, inhibition of caspase activity was achieved in vivo using a caspase inhibitor. Cells were preincubated for 4 h with the caspase-3 inhibitor (25 μM), followed by cisplatin treatment (10 μM). After 24 h, cells were harvested and analyzed by Hoechst nuclear staining and Western blotting.

Transfection with constructs and small interfering RNA (siRNA).

Wild-type (WT) caspase-3 (pcDNA3-Casp3-Myc) and caspase-3 catalytic mutant (pcDNA3-Casp3 C163A-Myc) plasmids were purchased from Addgene (Cambridge, MA). Par-4 (pCMV Entry-Par-4-Myc-DDK) and empty vector (pCMV Entry-Myc-DDK) plasmids encoding a Myc tag at the C terminus were purchased from Origene (Rockville, MD). One day before transfection, cells were plated at 3 × 10⁵/well to achieve a confluence of ∼70% after 24 h. Cells were transfected with 1 to 2 μg of expression vector using Fugene 6 (Roche, Indianapolis, IN) according to the manufacturer's instructions. Cells were incubated for 48 h at 37°C, and the medium was replenished with fresh medium containing 10 μM cisplatin. The plates were incubated for an additional 24 h before the cells were harvested.

For silencing Par-4 expression, cells were transiently transfected with Par-4-specific siRNA or control nonsilencing (NS) siRNA (Santa Cruz Biotechnology, Santa Cruz, CA). Exponentially growing cells were transfected with 50 nM Par-4 or control siRNA using TransIT-TKO transfection reagent (Mirus, Madison, WI) in accordance with the manufacturer's instructions. Cells were cultured for 48 h, and the medium was replenished with fresh medium containing 10 μM cisplatin. Following 24 h, cells were harvested for further analysis.

In vitro cleavage assay.

Cell lysates (50 μg of total protein) were incubated for 6 h at 37°C in PIPES [piperazine-N,N′-bis(2-ethanesulfonic acid] assay buffer {PIPES (20 mM), NaCl (100 mM), dithiothreitol (DTT; 10 mM), EDTA (1 mM), 3-[(3-cholamindopropyl)-dimethylammonio]-1-propanesulfonic acid (CHAPS; 0.1% [wt/vol]), and sucrose (10% [wt/vol]; pH 7.2)} containing 10 units of recombinant active caspase-8, caspase-9, and caspase-3 in the presence or absence of a caspase inhibitor (25 μM). The incubation was terminated with the addition of an equal volume of 2× sample buffer (Tris-HCl [100 mM; pH 6.8], DTT [200 mM], SDS [4% {wt/vol}], glycerol [20% {vol/vol}], and bromphenol blue [0.2%]). Par-4 cleavage was assessed by Western blotting.

Site-directed mutagenesis.

Site-directed mutagenesis of human Par-4 cDNA in pCMV Entry-Myc-DDK vector was carried out by a commercial source (BioPioneer Inc.). The Asp residues at positions 110 (RSED¹¹⁰), 126 (PQRD¹²⁶), and 131 (EEPD¹³¹) were replaced by Ala. The resulting mutants were denoted D110A, D126A, and D131A, respectively. The identities of all the mutants were confirmed by sequencing. Transient transfection was done as described before.

Cloning of the C-terminal Par-4 fragment.

The C-terminal fragment of Par-4 was cloned from the Par-4 sequence between bp 566 and the end of the open reading frame; a Myc tag was also included at the C terminus. Amplification of the C-terminal Par-4 fragment was done by PCR using primers specific for the region of interest with 100 pg of pCMV Entry-Par-4-Myc-DDK plasmid (Origene, Rockville, MD) as the DNA template. PCR was conducted using 25 cycles with an annealing temperature of 60°C with the Pfx50 DNA polymerase (Invitrogen, Carlsbad, CA). The amplified product size was confirmed by 1% agarose gel electrophoresis, and the DNA fragment was purified on a MinElute gel extraction kit (Qiagen, Valencia, CA). To proceed with TA cloning, 3′-A overhangs were added by incubating the eluted DNA fragment encoding the Par-4–Myc C terminus with 0.1 U Taq DNA polymerase (New England BioLabs, Ipswich, MA) for 15 min at 68°C with appropriate buffer supplemented with dATPs (100 mM). The DNA fragment encoding the Par-4–Myc C terminus was cloned in pcDNA3.1-V5-His-TOPO vector (Invitrogen, Carlsbad, CA) using a standard manufacturer's protocol. T7 promoter and BGH reverse primers were used to confirm the accuracy of the DNA sequence.

Animals.

Sprague-Dawley female rats, 200 to 225 g, were obtained from Charles River Laboratories Canada. Animals were maintained on standard chow and water, which were available ad libitum, in animal facilities illuminated between 6:00 a.m. and 20:00 p.m. All procedures were performed in accordance with guidelines of the Canadian Council on Animal Care for the handling and training of laboratory animals and the Good Health and Animal Care Committee of the Université du Québec à Trois-Rivières. Stages of the estrous cycle were confirmed by vaginal smears. Rats with three regular cycles of 4 days were used in these experiments and killed at various stages of the estrous cycle (diestrus, proestrus, estrus, and metestrus). Four different rats were used for each time of the estrous cycle. Uteri were collected, and endometrial protein extracts were collected for Western blot analysis.

Statistical analysis.

Experiments were performed in triplicate to verify the reproducibility of the findings. Statistical analyses were carried out with GraphPad (San Diego, CA) PRISM software, version 3.03. Differences between experimental groups were determined using the one-way analysis of variance (ANOVA) followed by the post hoc Tukey's test. Statistical significance was accepted when P was <0.05.

RESULTS

Par-4 protein content is decreased during cisplatin-induced apoptosis.

To determine whether Par-4 participates in the cisplatin-induced apoptotic pathway, we monitored Par-4 expression in the presence of cisplatin (10 μM) in a variety of cell lines. The results revealed that cisplatin reduced Par-4 protein content in these cell lines except for A2780-CP (cisplatin-resistant ovarian cancer cell line) and PC-3 (prostate cancer cell line) (Fig. 1A). To evaluate whether the decrease in Par-4 was related to downregulation of mRNA expression in cisplatin-treated cells, we measured the mRNA expression level of Par-4 by RT-PCR. As shown in Fig. 1B, Par-4 mRNA levels remain constant with cisplatin treatment in all cell lines tested, suggesting that the degradation of Par-4 protein is regulated by posttranscriptional mechanisms in cisplatin-treated cells. However, cisplatin significantly induced apoptosis in all the cell lines except for A2780-CP (Fig. 1C), as determined by Hoechst nuclear staining. Apoptosis levels correlated with the activation of caspase-3 and concomitant cleavage of poly(ADP-ribose) polymerase (PARP), which is a known substrate of caspase-3 (Fig. 1A). We further tested whether the decrease in Par-4 levels in response to cisplatin treatment occurs selectively in cancer cells. Additional cancer cell lines, including MCF-7, MDA-468, Hec-1-A, and OVCAR-3, and normal cell lines, including HIESC and HIEEC, were treated with cisplatin. In the presence of cisplatin, Par-4 protein levels decreased in all cell lines except for MCF-7 (Fig. 1D). Moreover, a recent study provides evidence that intracellular Par-4 is spontaneously secreted by mammalian cells and therefore is also present in the extracellular compartment, e.g., in cell culture conditioned medium or circulating in the serum (4). We sought to determine whether the decrease in Par-4 protein levels in the presence of cisplatin could be due to increased secretion in the conditional medium. The results showed that there was no increase in the secreted Par-4 levels in the conditional medium of cisplatin-treated cells (data not shown).

Fig 1.

Par-4 expression during cisplatin-induced apoptosis. Cells were incubated in the presence or absence of 10 μM cisplatin for 24 h. Total protein and RNA were extracted for Western blot (A) and RT-PCR (B) analysis. Cl., cleaved. (C) Cells were collected for Hoechst staining to count apoptotic cells. Data represent means ± standard deviations (SD) for three independent experiments. ∗, P < 0.05 compared to control cells. (D) Cells were incubated in the presence or absence of 10 μM cisplatin for 24 h. Total proteins were extracted for Western blot analysis using antibodies against Par-4 (molecular mass: 41 kDa), cleaved PARP (89 kDa), and cleaved caspase-3 (12, 17, and 19 kDa). GAPDH (36 kDa) was used as a loading control.

Reduction in Par-4 protein level parallels its degradation and activation of caspase-3 during cisplatin-induced apoptosis in vivo.

To establish the mechanism underlying Par-4 reduction, further experiments were performed with regard to inhibition of protein synthesis and acceleration of protein degradation. A2780-S cells were pretreated for 1 h with cycloheximide (CHX) to inhibit de novo protein synthesis, followed by cisplatin treatment for 24 h. The results revealed that Par-4 levels were decreased by 51% with cisplatin treatment compared to those in untreated cells. However, upon pretreatment with CHX, a reduction of 86% was observed (Fig. 2A), indicating that the degradation occurs at the protein level.

Fig 2.

Cleavage of Par-4 during cisplatin-induced apoptosis. (A) A2780-S cells were preincubated with 10 μg/ml of cycloheximide (CHX) for an hour and exposed to 10 μM cisplatin. After 24 h, Par-4 expression was analyzed by Western blotting. (B) A2780-S and A2780-CP cells were incubated with 10 μM cisplatin for 0, 8, 16, or 24 h. Equal amounts of proteins were separated by SDS-PAGE and immunoblotted with different antibodies (Par-4, total Akt [60 kDa], pAkt [60 kDa], XIAP [54 kDa], cleaved caspase-8 [10 kDa], cleaved caspase-9 [35 kDa], cleaved PARP, and cleaved caspase-3). (C) MCF-7, PC-3, KLE, and HeLa cells were treated with cisplatin (10 μM) for 24 h. After 24 h, Par-4 expression was analyzed by Western blotting using anti-Par-4 antibody. It is noteworthy that the cleaved fragment of Par-4 could be seen only in KLE and HeLa cells. (D) HeLa cells were treated with cisplatin (10 μM) or doxorubicin (1 μM) for 24 h. Treatment with both the drugs induced Par-4 cleavage. The arrow indicates the cleaved fragment of Par-4. GAPDH was used as a loading control.

To evaluate whether the decrease in Par-4 protein levels was time dependent, A2780-S (cisplatin-sensitive) and A2780-CP (cisplatin-resistant) cell lines were treated with 10 μM cisplatin for 0 h, 8 h, 16 h, and 24 h. Western blot analysis revealed a time-dependent decrease in Par-4 protein levels in A2780-S cells, which is clearly evident from the fold change (Fig. 2B). However, the levels of Par-4 in resistant A2780-CP cells remain unchanged. Interestingly, in A2780-S cells, Par-4 degradation was followed by the appearance of an ∼25-kDa polypeptide fragment (corresponding to Par-4, as revealed by anti-Par-4 antibody) after 16 h of cisplatin treatment; thereafter, the fragment remained stable throughout the time course experiment. Considering time course analysis, the decrease in Par-4 protein levels and appearance of the cleaved fragment of Par-4 occurred at the same time as the cleavage of PARP. This indicates that Par-4 cleavage is an apoptotic event that occurs as early as PARP cleavage. In A2780-S cells, cisplatin treatment induced the activation of caspase-8, caspase-9, and caspase-3. There was a decrease of both XIAP content and Akt phosphorylation (pAkt) in A2780-S cells upon cisplatin treatment. The reduction in pAkt levels could be due to the caspase-3-dependent cleavage of Akt as we have shown previously (2). In order to examine whether cisplatin-induced cleavage of Par-4 protein is cell specific, several other cell lines, including HeLa, KLE, MCF-7, and PC-3, were treated with cisplatin. Notably, the cleaved fragment of Par-4 could be seen only in KLE and HeLa cells (Fig. 2C). Next, we tested whether the cleavage of Par4 is specific to cisplatin treatment or a general response to apoptosis induction. As a further test, A2780-S cells were treated with doxorubicin (1 μM) for apoptosis induction. Western blot analysis revealed that Par-4 cleavage occurred simultaneously with caspase-3 activation and PARP cleavage upon doxorubicin treatment. This experiment provided an unequivocal demonstration that cleavage of Par-4 is not specific for the type of apoptotic stimulus (Fig. 2D). To understand whether the cleavage of Par-4 is a result of ubiquitin-mediated proteasomal degradation, A2780-S cells were pretreated with MG132, a proteasome inhibitor, followed by cisplatin treatment. The cleavage of Par-4 was not inhibited in the presence of MG132 (data not shown). However, MG132 pretreatment itself resulted in a small amount of Par-4 cleavage, suggesting that the cleavage is a result of apoptosis induction by MG132. These data demonstrate that the cleavage of Par-4 is relevant to apoptosis.

Par-4 is a substrate of caspase-3 in vitro.

As is well known, the caspase family is the major protease group involved in apoptosis, and the findings that cisplatin treatment induced the activation of caspase-8, caspase-9, and caspase-3 prompted us to investigate which particular caspase is responsible for the cleavage of Par-4 (Fig. 3). Toward this end, total cell lysates of HeLa cells were incubated with recombinant caspase-8, recombinant caspase-9, and recombinant caspase-3 (10 units) for 6 h in the presence or absence of their respective inhibitors (25 μM). After 6 h, the protein fragments were resolved and compared with those obtained in vivo after cisplatin treatment. As shown in Fig. 3A, only incubation with recombinant caspase-3 in vitro resulted in an ∼25-kDa fragment that mirrored the fragment observed in vivo after apoptosis induction with cisplatin, doxorubicin, and MG132. The generation of the cleaved fragment was completely blocked in the presence of a caspase-3 inhibitor.

Fig 3.

In vitro cleavage of Par-4 by recombinant caspases. (A) HeLa cell lysates were incubated with human recombinant caspase-8, caspase-9, and caspase-3 for 6 h in the presence or absence of their respective inhibitors (25 μM). Samples were then analyzed by SDS-PAGE and immunoblotting using Par-4 and cleaved PARP antibodies. Cleavage of PARP was analyzed to monitor caspase activity. (B) A2780-S, A2780-CP, PC-3, MCF-7, and Ishikawa cell lysates were incubated with human recombinant caspase-3 for 6 h in the presence or absence of caspase-3 inhibitor. Samples were then analyzed by SDS-PAGE and immunoblotting using Par-4 and cleaved PARP antibodies.

The finding that Par-4 is a caspase-3 substrate in vitro prompted us to investigate whether caspase-3 might also be responsible for cleavage of Par-4 in vitro in MCF-7 cells, which are caspase-3 deficient, and PC-3 and A2780-CP cells, which showed less caspase-3 activation upon cisplatin treatment (as shown in Fig. 1). Therefore, we incubated total cell lysates of A2780-S, A2780-CP, PC-3, MCF-7, and Ishikawa cells with recombinant caspase-3 in the presence or absence of a caspase-3 inhibitor. Following in vitro caspase-3 activation, a cleaved Par-4 band (∼25 kDa) appeared in all the lysates tested (Fig. 3B). Interestingly, shorter fragments appeared in A2780-S and PC-3 cells; this could be the result of cleavage of Par-4 at multiple sites. Cleaved PARP was used as a positive control in all the experiments to monitor caspase activity. Together, these results suggest that Par-4 is a substrate of caspase-3 in vitro.

Cleavage of Par-4 by caspase-3 occurs during apoptosis in vivo.

Based on the evidence that Par-4 is a substrate of caspase-3 in vitro, we tested whether the cleavage of Par-4 by caspase-3 is also executed in vivo. Thus, A2780-S cells were pretreated with a caspase-3 inhibitor (25 μM) for 4 h, followed by cisplatin treatment. Par-4 expression decreased together with the generation of the cleaved fragment in the presence of cisplatin. The generation of the ∼25-kDa fragment correlated with the activation of caspase-3 and cleavage of PARP (Fig. 4A). The induction of apoptosis was determined by counting the apoptotic cells by fluorescence microscopy after staining with Hoechst 33258 dye, and the results were plotted as the percentage of apoptotic cells (Fig. 4B). The appearance of the apoptotic cells (irregular Hoechst nuclear staining with multiple bright specks of chromatin fragmentation and condensation) is shown in Fig. 4C. However, the activation of caspase-3, cleavage of PARP, decrease in full-length Par-4, and generation of the ∼25-kDa cleaved fragment of Par-4 were completely inhibited by pretreatment of A2780-S cells with a caspase-3 inhibitor prior to cisplatin treatment. In parallel, the caspase-3 inhibitor significantly blocked cisplatin-induced apoptosis. Therefore, these results suggest that Par-4 is a physiological substrate of caspase-3 in vivo.

Fig 4.

Caspase-3 inhibitor attenuated cisplatin-induced cell apoptosis and blocked Par-4 cleavage. Four hours before cisplatin treatment (10 μM), A2780-S cells were pretreated with a caspase-3 inhibitor (25 μM). After 24 h, cells were harvested for Western blot analysis (A) and Hoechst staining (B). A minimum of 200 cells/treatment group were counted in each experiment. Letters above bars (a and b) indicate significant differences from other groups (P < 0.001). (C) Representative images of Hoechst staining in control (a), caspase-3 inhibitor-treated (b), cisplatin-treated (c), and cisplatin- and caspase-3 inhibitor-treated (d) A2780-S cells. Arrows represent cells with typical apoptotic nuclear morphology.

Restoration of caspase-3 expression in caspase-3-deficient MCF-7 and PC-3 cells promotes cisplatin-induced Par-4 cleavage.

Considering that the caspase-3-mediated cleavage of Par-4 occurs in vitro, we further evaluated whether overexpression of caspase-3 in MCF-7 and PC-3 cells could facilitate the cleavage of Par-4 in these cells in vivo. To address this issue, MCF-7 and PC-3 cells were transiently transfected with empty pcDNA3 and wild-type (WT) caspase-3–Myc plasmid. In the present study, a mutant caspase-3 C163A–Myc plasmid was also used as an additional control. After 48 h of transfection, cells were incubated in the presence or absence of cisplatin for 24 h. As shown in Fig. 5A, empty vector-transfected MCF-7 cells did not express any detectable procaspase-3 whereas mutant caspase-3 C163A–Myc- and WT caspase-3–Myc-transfected MCF-7 cells contained significant amounts of this proenzyme. However, Par-4 was found to be cleaved only in WT caspase-3-transfected cells treated with cisplatin. On the other hand Par-4 was completely resistant to caspase-3-dependent cleavage in empty vector- or mutant caspase-3 C163A–Myc-transfected cells in the presence of cisplatin. PARP was still degraded in the presence of cisplatin in control cells. This is consistent with the evidence for caspase-3-independent mechanisms for PARP cleavage during apoptosis (14). However, the cleavage of PARP was pronounced in the WT caspase-3–Myc-transfected MCF-7 cells. Similar results were also observed in PC-3 cells following transfection with empty vector, WT caspase-3–Myc, or mutant caspase-3 C163A–Myc plasmid and cisplatin treatment (Fig. 5B).

Fig 5.

Restoration of caspase-3 expression in caspase-3-deficient MCF-7 cells promotes cisplatin-induced Par-4 cleavage. (A) MCF-7 cells were transiently transfected with 1 μg of control plasmid (pcDNA3; empty), mutant caspase-3 C163A–Myc plasmid (caspase-3 Mut), or WT caspase-3–Myc plasmid (caspase-3 WT) using Fugene 6. After 48 h, cells were incubated in the presence or absence of 10 μM cisplatin. Following cisplatin treatment for 24 h, cells were harvested for Western blotting. (B) PC-3 cells were transiently transfected with 1 μg of control plasmid (pcDNA3; empty), mutant caspase-3 C163A-Myc plasmid (caspase-3 Mut), or WT caspase-3-Myc plasmid (caspase-3 WT) using Fugene 6. After 48 h, cells were incubated in the presence or absence of 10 μM cisplatin. Following cisplatin treatment for 24 h, cells were harvested for Western blotting. (C and D) MCF-7 cells were transiently transfected with control plasmid (pcDNA3; empty) or WT caspase-3–Myc plasmid (caspase-3 WT) using Fugene 6. After 48 h, cells were pretreated with caspase-3 inhibitor for 4 h followed by incubation in the presence or absence of 10 μM cisplatin for 24 h. At the end of the treatment period, cells were harvested for Western blotting and Hoechst staining. ∗, P < 0.05 compared to control cells.

In the next set of experiments, empty vector- and WT caspase-3-Myc-transfected MCF-7 cells were treated with cisplatin in the presence or absence of a caspase-3 inhibitor. Western blot analysis confirmed that the caspase-3 inhibitor completely blocked the cleavage of Par-4 in WT caspase-3–Myc-transfected MCF-7 cells treated with cisplatin (Fig. 5C and D). Concurrently, pretreatment with the caspase-3 inhibitor significantly reduced the number of cells entering apoptosis and inhibited the activation of caspase-3 and PARP cleavage. Collectively, these results therefore unambiguously confirm that Par-4 cleavage was an event downstream of caspase-3 activation during apoptosis.

Par-4 cleavage by caspase-3 occurs at an unconventional site, EEPD¹³¹↓G.

Next, we sought to determine the caspase-3 cleavage site in Par-4. We used a Par-4 antibody corresponding to amino acids 324 to 342 of human Par-4. The ∼25-kDa cleaved fragment of Par-4 could be easily detected with this antibody, suggesting that the cleaved fragment of Par-4 contains the C-terminus sequence (Fig. 6A). Therefore, we predicted that the caspase-3 cleavage site might be located at ∼25 kDa from the C terminus. Analysis of the amino acid sequence of Par-4 revealed that aspartic acid residues at RSED¹¹⁰↓E, PQRD¹²⁶↓E, and EEPD¹³¹↓G could be the likely candidates for caspase-3-mediated cleavage of Par-4. All these sites are located ahead of NLS2 and the SAC domain of Par-4, as shown in Fig. 6B. To confirm the caspase-3 cleavage site, Par-4 mutants in which D¹¹⁰, D¹²⁶, or D¹³¹ was mutated to alanine (A) were constructed and the effect of these mutations on the cleavage of Par-4 by caspase-3 was analyzed in vivo and in vitro. HeLa cells were transiently transfected with WT Par-4 or mutant Par-4 (D110A, D126A, or D131A) in the pCMV Entry-Myc-DDK vector followed by cisplatin treatment. The results, as revealed by using Myc tag antibody, showed that the cleaved fragment of Myc-tagged Par-4 was observed in HeLa cells transfected with WT or D110A or D126A mutant Par-4 plasmids. However, Myc-tagged Par-4 expressed in D131A mutant Par-4-transfected cells was completely resistant to cleavage during cisplatin-induced apoptosis (Fig. 6C). Cleaved PARP and cleaved caspase-3 were also analyzed to monitor caspase-3 activation. These results were further confirmed by using recombinant caspase-3 in an in vitro assay. Activated caspase-3 was able to cleave Myc–Par-4 from WT or D110A or D126A mutant Par-4-transfected cell lysates but not from lysates obtained from D131A mutant Par-4 (Fig. 6D). These results demonstrate that the aspartic acid residue at 131 is the essential caspase-3 cleavage site in Par-4.

Fig 6.

Identification of the caspase-3 cleavage site in Par-4. (A) A2780-S and HeLa cells were treated with cisplatin for 24 h. The cleaved fragment of Par-4 was recognized by Western blot analysis using anti-Par-4 antibody specific for the C terminus of Par-4. (B) Schematic diagram representing Par-4. The nuclear localization sequences (NLS1 and NLS2), SAC (selective for apoptosis in cancer cells) domain, and leucine zipper (LZ) domain are depicted. The putative caspase-3 cleavage sites are indicated in boldface. Aspartic acid residues at 110, 126, and 131 were mutated to alanine. (C) HeLa cells were transiently transfected with 2 μg of WT or mutant Par-4 (D110A, D126A, or D131A) in the pCMV Entry-Myc-DDK vector. After 48 h, cells were treated with 10 μM cisplatin for 24 h. Following treatment, cells were collected for Western blot analysis. Anti-Myc tag antibody was used to reveal the cleaved fragment. Cleaved PARP and cleaved caspase-3 antibodies were used to monitor caspase-3 activation. (D) Lysates from HeLa cells expressing WT or mutant Par-4 (D110A, D126A, or D131A) in pCMV Entry-Myc-DDK vector were incubated with human recombinant caspase-3 for 6 h. Samples were then analyzed by SDS-PAGE and immunoblotting using Myc tag antibody. Cleavage of PARP was analyzed to monitor caspase-3 activity. (E) Expression of Par-4 and cleaved caspase-3 in the rat endometrium during the estrous cycle. Four cycling rats were sacrificed at each stage of the estrous cycle (D, diestrus; P, proestrus; E, estrus; M, metestrus). Endometrial proteins were collected and analyzed for Par-4 and cleaved caspase-3. (F) Densitometric analysis of results in panel E. Data represent means ± SD of four independent experiments. ∗, P < 0.05 compared to the other groups. (G) Protein lysates from mouse tissues were incubated with mouse recombinant caspase-3 for 6 h in the presence or absence of a caspase-3 inhibitor. Samples were then analyzed by SDS-PAGE and immunoblotting using Par-4 and cleaved PARP antibodies.

Protein sequence alignment of human, rat, and mouse Par-4 proteins shows that caspase-3 cleavage site EEPD¹³¹ is conserved in all three species (12); therefore, to corroborate our results that Par-4 is a substrate of caspase-3, we evaluated Par-4 expression under physiological conditions during the rat estrous cycle. Under reduced-sex-steroid conditions, endometrial epithelial and stromal cells undergo apoptosis in a cycling manner during different phases (diestrus, proestrus, estrus, and metestrus) of the estrous cycle. Previously, we have shown that the apoptotic levels were higher during estrus phase and lower during proestrus phase (8, 28). However, in the present study, higher levels of Par-4 were observed during the proliferative phase (proestrus) and lower levels were observed during the apoptotic/secretory phase (estrus) (Fig. 6E and F). Taking into account the proapoptotic role of Par-4, the lower expression during estrus phase was rather surprising. It is worth mentioning that higher activation/cleavage of caspase-3 is observed during the estrus phase (8, 28). Hence, these results suggest that the decrease in Par-4 levels during the estrus phase could be the result of caspase-3-mediated Par-4 cleavage. These findings further confirm that Par-4 is a substrate of caspase-3 even under normal physiological conditions. Besides, the caspase-3 cleavage site in rat and mouse Par-4s contains a serine residue at the P-1 site; however, human Par-4 contains a glycine at the same site. We therefore tested whether caspase-3-mediated cleavage of Par-4 also occurs in mouse tissues and whether the presence of serine at P-1 could prevent caspase-3-mediated cleavage of Par-4. Lysates from mouse liver and uterine tissues were treated in vitro with mouse recombinant caspase-3 in the presence or absence of caspase-3 inhibitor, and the proteins were analyzed by Western blotting. The cleaved fragment of Par-4 was clearly visible in the lysates treated with recombinant caspase-3. In contrast, the presence of the caspase-3 inhibitor completely prevented caspase-3-mediated cleavage of mouse Par-4 protein (Fig. 6G). Likewise, the size of cleaved fragments of Par-4 corresponds to the size of the C-terminal fragment generated in case of human Par-4. These results further suggest that the presence of serine at the P-1 site does not affect the specificity of caspase-3 cleavage site in Par-4.

Knockdown of endogenous Par-4 inhibits cisplatin-induced apoptosis.

The above finding that Par-4 is a substrate of caspase-3 prompted us to evaluate the potential role of Par-4 in the regulation of apoptosis. HeLa cells were transiently transfected with either control NS siRNA or Par-4-specific siRNA and then treated with cisplatin to examine the effect on apoptosis. Compared with NS siRNA transfection, transfection with Par-4 siRNA significantly reduced Par-4 expression. Western blot analysis further showed that suppression of Par-4 expression inhibited cisplatin-induced apoptosis, as evidenced by the activation of caspase-8, caspase-9, caspase-3, and PARP cleavage as well as by annexin V/PI staining (Fig. 7A to C). Knockdown of Par-4 significantly reduced cisplatin-induced cell death compared to the control. Notably, the cleaved fragment of Par-4 could be seen in NS siRNA-transfected and cisplatin-treated cells, which was not the case in Par-4 siRNA-transfected cells. The possible explanation could be the overall reduced expression of Par-4 due to knockdown by Par-4 siRNA. Together, these results imply that Par-4 is required for apoptosis induction and that its cleavage does not inhibit apoptosis.

Fig 7.

Knockdown of Par-4 using specific siRNA inhibits cisplatin-induced apoptosis. (A and B) HeLa cells were transfected with Par-4 siRNA or control nonsilencing (NS) siRNA. After 48 h, culture medium was replaced with fresh medium containing 10 μM cisplatin. Following 24 h of cisplatin treatment, cells were collected for Western blot analysis and flow cytometry using Alexa Fluor 488-annexin V/PI staining. B1, B2, B3, and B4 represent quadrants for dead, late apoptotic, viable, and early apoptotic cells, respectively. Panel B shows representative results of three independent experiments. (C) Histogram showing apoptosis induction (% total) in different groups. Data represent means ± SD of three independent experiments. Letters above bars (a and b) indicate significant differences from other groups (P < 0.05).

Caspase-3-mediated cleavage increases nuclear localization of Par-4 and promotes apoptosis.

To investigate whether caspase-3-mediated cleavage of Par-4 alters its apoptotic activity, we tested the sensitivity of HeLa cells expressing the empty vector, WT Par-4, and the uncleavable Par-4 D131A mutant to cisplatin-induced apoptosis by flow cytometry. Confirmation of the introduction of Myc-tagged protein and transfection efficiency was determined by immunofluorescence (Fig. 8A) and Western blot analysis (Fig. 8D). Annexin V/PI staining showed that ectopic expression of wild-type Par-4 slightly increased the apoptotic rates. Remarkably, overexpression of wild-type Par-4 enhanced cisplatin-induced apoptosis compared to that in empty vector-transfected HeLa cells (Fig. 8B and C). The introduction of the Par-4 D131A mutant resulted in a significant reduction in the number of apoptotic cells, suggesting that the Par-4 D131A mutant has a dominant negative effect on the apoptosis levels. These results clearly support the view that Par-4 cleavage promotes apoptosis.

Fig 8.

Overexpression of wild-type Par-4 promotes cisplatin-induced apoptosis. (A) HeLa cells were transiently transfected with 2 μg of the empty vector, WT Par-4, or the Par-4 D131A mutant in the pCMV Entry-Myc-DDK vector. After 48 h, cells were subjected to immunofluorescence for Myc–Par-4 (Alexa Fluor 594, green). The nuclei were stained with Hoechst 33258 (blue), and phalloidin (red), which stains the actin cytoskeleton, is used to visualize the shape and integrity of the cells. The localization of Myc–Par-4 was determined by confocal microscopy. Magnification, ×63. (B) HeLa cells were transiently transfected with 2 μg of the empty vector, WT Par-4, or the Par-4 D131A mutant in the pCMV Entry-Myc-DDK vector. After 48 h, cells were treated with 10 μM cisplatin for 24 h. Following treatment, the induction of apoptosis was analyzed by flow cytometry after Alexa Fluor 488-annexin V/PI staining. B1, B2, B3, and B4 represent quadrants for dead, late apoptotic, viable, and early apoptotic cells, respectively. Representative results from three independent experiments are shown. (C) Histogram showing apoptosis induction (% total) in the different groups. Data represent means ± SD of three independent experiments. Letters above bars (a and b) indicate significant differences from other groups (P < 0.01). E, empty vector; Cp, cisplatin. (D) Cells were transiently transfected with 2 μg of the empty vector, WT Par-4, or the Par-4 D131A mutant. After 48 h, total cell lysates were analyzed for Bcl-2 (28 kDa) and FLIP (55 kDa) protein levels by Western blotting. Anti-Myc tag antibody was used to reveal the extent of Par-4 overexpression. GAPDH was used as a loading control. (E) Histogram showing the fold changes in protein levels of blots in panel D for Bcl-2 and FLIP. Data represents means ± SD of three independent experiments. ∗, P < 0.05 compared to the other groups.

Previously, Par-4 was shown to induce apoptosis by suppressing nuclear factor κB (NF-κB) transcriptional activity (10) or by acting as a transcriptional repressor for WT1 (25). In both these cases, an inverse correlation between Par-4 and antiapoptotic molecules such as Bcl-2 was observed. The latter studies prompted us to investigate whether the overexpression of Par-4 mutant D131A results in reduced ability to inhibit NF-κB transcriptional activity. Protein lysates from empty vector-, WT Par-4-, and uncleavable Par-4 D131A mutant-transfected cells were analyzed for levels of Bcl-2 and FLIP, which are also known to be transcriptional targets of NF-κB. As expected, wild-type Par-4-expressing cells showed an approximately 2-fold decrease in Bcl-2 and FLIP protein levels relative to empty vector-transfected cells; the reduction from the empty vector-transfected cells was statistically significant (Fig. 8D and E). However, the protein levels of Bcl-2 and FLIP were not found to be decreased in cells expressing the Par-4 D131A mutant. Together, these experiments indicate that wild-type Par-4 induces apoptosis by reducing the cellular levels of antiapoptotic proteins.

Given that the proapoptotic activity of Par-4 is directly related to its nuclear localization (13), we investigated whether caspase-3-mediated Par-4 cleavage has an impact on its subcellular localization. We performed subcellular fractionation and immunofluorescence in HeLa cells transiently transfected with WT Par-4 followed by cisplatin treatment. Western blot analysis illustrates that cisplatin treatment reduced Par-4 levels in the cytoplasmic compartment, with no significant difference in the nuclear compartment (Fig. 9). Interestingly, the cleaved fragment of Par-4 was found to accumulate in the nuclear extracts as determined by using Myc tag antibody (Fig. 9A), suggesting that, upon caspase-3-mediated cleavage, Par-4 readily translocates to the nucleus. To further characterize the subcellular localization of the cleaved fragment of Par-4, we performed immunofluorescence microscopy analysis in cells using anti-Myc-tagged antibody. As shown in Fig. 9B, Par-4 is present in cytosolic and nuclear compartments in untreated cells. However, intense nuclear staining for Myc-tagged Par-4 was revealed after cisplatin treatment, which could be due to the increased nuclear translocation of the cleaved fragment of Par-4. Overall, these results suggest that caspase-3-mediated cleavage of Par-4 promotes its nuclear entry and thereby apoptosis.

Fig 9.

C-terminal fragment of Par-4 localizes to the nucleus. (A) HeLa cells were transiently transfected with 2 μg of WT Par-4 in the pCMV Entry-Myc-DDK vector. After 48 h, cells were treated with 10 μM cisplatin for 24 h. Following treatment, cells were collected for cytoplasmic and nuclear extraction, followed by Western blot analysis. Anti-Myc tag antibody was used to reveal Par-4 and its cleaved fragment. PARP (116 kDa) and GAPDH were used as loading controls for nuclear and cytoplasmic extracts, respectively. (B) HeLa cells were transiently transfected with 2 μg of WT Par-4 in the pCMV Entry-Myc-DDK vector. After 48 h, cells were treated with 10 μM cisplatin for 24 h. Cells were subjected to immunofluorescence for Myc–Par-4 (Alexa Fluor 594, green). The nuclei were stained with Hoechst 33258 (blue) and visualized for localization of Par-4 by confocal microscopy. Note the intense nuclear staining of Par-4 upon cisplatin treatment.

The C-terminal cleaved fragment of Par-4 retains proapoptotic activity.

To gain further insight into how the proapoptotic activity of Par-4 could be maintained even after caspase-3-mediated cleavage, we generated a Myc-tagged C-terminal fragment of Par-4. Interestingly, D131 is located ahead of NLS2 and the SAC and leucine zipper domains; thus, the cleaved fragment contains all the domains necessary for the induction of apoptosis. We therefore hypothesized that the C-terminal fragment could induce apoptosis when overexpressed in the cells. To test this hypothesis, HeLa cells were transfected with the empty vector or C-terminal fragment of Par-4 expression constructs. Immunofluorescence studies were performed to assess the cellular localization of the cleaved fragment of Par-4. The result showed that the cleaved fragment of Par-4 is predominantly localized in the nuclear compartments of the cells (Fig. 10A). The expression of the Myc-tagged C-terminal fragment was also confirmed by Western blot analysis (Fig. 10D and E). To investigate whether the cells expressing the C-terminal Par-4 fragment undergo apoptosis, cells were collected after 48 h of transfection and analyzed by flow cytometry for annexin V/PI staining. The results revealed that the cells expressing the C-terminal Par-4 fragment underwent significant apoptosis compared to empty vector-transfected cells (Fig. 10B and C). As mentioned above, the C-terminal Par-4 fragment possesses all the necessary domains required for apoptotic activity; we tested whether the C-terminal fragment follows a mechanism of apoptosis induction similar to that of the full-length WT Par-4. Among the mechanisms by which Par-4 triggers apoptosis, the best-characterized one is through the inhibition of the protein kinase C (PKC)/NF-κB pathway (9, 18). As shown previously, the overexpression of WT full-length Par-4 reduced the protein levels of Bcl-2 and FLIP, which are known to be the transcriptional targets of NF-κB; therefore, we evaluated the levels of Bcl-2 and FLIP in C-terminal Par-4 fragment-expressing cells. Western blot analysis revealed that the overexpression of the C-terminal Par-4 fragment indeed reduced the protein levels of Bcl-2 and FLIP compared to those in empty vector-transfected cells (Fig. 10D). Overexpression of the C-terminal Par-4 fragment also reduced the levels of phosphorylated IκBα; however, there was no change in the levels of total IκBα. Previous evidence that phosphorylation of IκBα regulates the nuclear localization of NF-κB and thereby its activity prompted us to examine the nuclear localization of NF-κB p50 and p65 in cells expressing the C-terminal Par-4 fragment. Western blot analysis was performed on the cytoplasmic and nuclear extracts of cells transfected with empty vector and C-terminal fragment of Par-4 expression constructs. The C-terminal fragment of Par-4 prevented the nuclear mobilization of NF-κB (p65) compared to results for empty vector-transfected cells. On the other hand, NF-κB (p50) protein levels were found to be decreased in cells expressing the C-terminal Par-4 fragment, suggesting that the C-terminal fragment of Par-4 possesses proapoptotic activity and regulates apoptotic induction by inhibiting the nuclear localization and activity of NF-κB in a manner similar to that of wild-type full-length Par-4. Next, we evaluated whether the C-terminal fragment of Par-4 is more efficient at inducing apoptosis than the wild-type Par-4 and Par-4 D131A mutant. HeLa cells were transfected with the empty vector, WT Par-4, the Par-4 D131A mutant, or the C-terminal fragment of Par-4. The results from annexin V/PI staining showed that cells expressing the wild type underwent apoptosis but cells expressing the C-terminal fragment of Par-4 showed up to 20% enhancement of the apoptotic rates (Fig. 10F). To conclude, our data suggest that caspase-3-mediated cleavage of Par-4 results in nuclear localization of the C-terminal Par-4 fragment with enhanced proapoptotic activity.

Fig 10.

The C-terminal fragment of Par-4 retains apoptotic activity. (A) HeLa cells were transiently transfected with 2 μg of the empty vector and the vector expressing the C-terminal cleaved fragment of Par-4 (Cl. Par-4). After 48 h, cells were subjected to immunofluorescence for Myc–Cl. Par-4 (Alexa Fluor 594, green). The nuclei were stained with Hoechst 33258 (blue), and phalloidin (red), which stains the actin cytoskeleton, is used to visualize the shape and integrity of the cells. The localization of Myc-tagged Cl. Par-4 was determined by confocal microscopy. Note the intense nuclear localization of the C-terminal cleaved fragment of Par-4. Magnifications, ×63 (top two rows) and ×100 (bottom two rows). (B) HeLa cells were transiently transfected with the empty vector and the vector expressing Cl. Par-4. After 48 h, cells were harvested and the induction of apoptosis was analyzed by flow cytometry using Alexa Fluor 488-annexin V/PI staining. B1, B2, B3, and B4 represent quadrants for dead, late apoptotic, viable, and early apoptotic cells, respectively. Representative results from three independent experiments are shown. (C) Histogram showing apoptosis induction (% total) in different groups. Data represent means ± SD of three independent experiments. ∗, P < 0.001 compared to the other group. (D and E) HeLa cells were transiently transfected with the empty vector and the vector expressing Cl. Par-4. After 48 h, cells were harvested for total cell lysates and cytoplasmic (C) and nuclear (N) extracts, followed by Western blot analysis using antibodies against Myc tag, Bcl-2, FLIP, p-IκBα (40 kDa), IκBα (39 kDa), NF-κB p65 (65 kDa), and NF-κB p50 (50 kDa). PARP and GAPDH were used as loading controls. (F) HeLa cells were transiently transfected with the empty vector, WT Par-4, the Par-4 D131A mutant, and Cl.Par-4. After 48 h, cells were collected and the induction of apoptosis was analyzed by flow cytometry using Alexa Fluor 488-annexin V/PI staining. The histogram shows the fold changes in apoptosis induction (% total) in the different groups compared to that for empty vector-transfected cells. Data represent means ± SD of three independent experiments. Letters above bars (a, b, and c) indicate significant differences from other groups (P < 0.05).

DISCUSSION

Par-4 has been proposed as an important target of therapeutic intervention since it does not cause apoptosis of normal or nontransformed cells on its own (42). Par-4 plays a crucial role in the activation of apoptosis and inhibition of cell survival. However, investigations are still going on to decipher the mechanisms by which Par-4 functions in the cells. The present study demonstrated for the first time that proapoptotic tumor suppressor Par-4 is a novel substrate of caspase-3 during apoptosis. A noncanonical caspase-3 cleavage site (EEPD¹³¹↓G) in Par-4 has been identified.

Earlier studies indicate that Par-4 is strongly upregulated in cells undergoing apoptosis (5, 12). In contrast, our data showed that the Par-4 is degraded in cisplatin-treated cells. However, the earlier notion of Par-4 upregulation during apoptosis still holds true for androgen-independent PC-3 cells in our study. Cleavage of Par-4 during cisplatin-, doxorubicin-, and MG132-induced apoptosis, thus resulting in the generation of an ∼25-kDa C-terminal fragment, suggests that this cleavage is a common and universal feature during apoptosis induction. In addition to caspase-3, caspase-8 and caspase-9 were also found to be activated during cisplatin-induced apoptosis. However, the involvement of caspase-3 in Par-4 cleavage was confirmed by an in vitro cleavage assay using recombinant caspase-3, -8, and -9 in the presence or absence of a caspase inhibitor, where Par-4 was found to be cleaved only with caspase-3. The in vitro cleavage assay revealed that Par-4 was cleaved even in MCF-7, PC-3, and A2780-CP cell lysates using recombinant caspase-3 (Fig. 3). Nevertheless, a substantial amount of the full-length protein remained uncleaved even after 6 h of incubation with recombinant caspase-3. Thus, Par-4 seems to belong to a group of proteins that undergo incomplete caspase-mediated cleavage during apoptosis, as observed in the case of actin (31). Further, this could be possibly due to the lower activity of recombinant caspase-3 in vitro. However, the patterns of proteolytic cleavage of Par-4 were slightly different in vivo, in vitro, and within the cell lines (with the appearance of additional bands together with ∼25-kDa fragment), suggesting that additional caspases or their respective regulators may be involved. Based on these findings, Par-4 can be added to the growing list of tumor suppressor proteins that are cleaved by caspases, including the transcription factor PU.1/spi-1 (45). Further, the involvement of caspase-7 in Par-4 cleavage can be ruled out based on the findings that Par-4 was not cleaved in caspase-3-deficient MCF-7 during cisplatin-induced apoptosis, during which caspase-7 is known to be activated (19). Our results demonstrated that Par-4 was resistant to caspase-3-mediated cleavage in PC-3 cells. Given the established role of Par-4 in the activation of apoptosis in prostate cancer cells, the resistance of Par-4 to caspase-3-mediated cleavage is not surprising. This could be explained by the findings of the previous studies, which showed that the expression of caspases, particularly caspase-3, is significantly reduced during prostate cancer progression (3, 24, 30). Par-4 does not induce apoptosis in the LNCap cell line, which is a hormone-dependent prostate cancer cell line (12), suggesting that the requirement of Par-4 for apoptosis is cell type specific. More importantly, Par-4 could also exacerbate caspase-3 activation (23). On the other hand, knockdown of Par-4 expression attenuates caspase-3 activation and apoptosis induction (Fig. 7). We therefore cannot rule out the possibility that there is an intermittent upregulation of Par-4 in response to cisplatin treatment, as observed with other apoptotic stimuli (5, 12), but the increased levels of Par-4 are responsible for pronounced activation of caspase-3, which is further responsible for the cleavage of Par-4 through a feedback loop.

Next, we were interested to know the cleavage site in Par-4. We therefore used a C-terminus-specific Par-4 antibody which was able to detect the ∼25-kDa fragment. It should be emphasized that the ∼25-kDa fragment was the primary cleavage product of all cells tested in this work. There is, however, no DEVD site (typical caspase-3 cleavage site) in the Par-4 amino acid sequence that would generate an ∼25-kDa fragment following cleavage. Our results using Par-4 mutants confirmed that caspase-3 cleaves Par-4 at amino acid 131 (EEPD¹³¹↓G) (Fig. 6). The involvement of caspase-3 in Par-4 cleavage is also substantiated from our observations in the rat estrous cycle (Fig. 6E and F), where Par-4 levels were found to be lowest during estrus phase, where maximum caspase-3 activation was observed (28). Further, protein sequence alignment of human, rat, and mouse Par-4s shows that caspase-3 cleavage site EEPD¹³¹ is conserved in all three species (12), suggesting the importance of this site.

Caspase-3 usually cleaves substrates with consensus motif (DXXD) (P4-P3-P2-P1), and the P1 position has absolute specificity for Asp. Although Par-4 contains two consensus sites, DLDD at 313 to 316 and DIED at 316 to 319, the ∼25-kDa cleaved fragment cannot be the result of cleavage at these sites. Nevertheless, there is evidence that several caspase-3 substrates lack the typical consensus DXXD motif. Examples of caspase-3 activity at unexpected sites are the cleavages of p21 at SMTD/F (35), topoisomerase I at EEED/G (38), calpastatin at DFTC/G (26), human recombinase Rad51 (HsRAd51) at AQVD/G (16), and HLJ1 (member of heat shock protein 40 family) at MEID/G (29). Additionally, a recent review reported a preference for glycine, alanine, threonine, serine, and asparagine in P1, and the invariant preference of the P3 position is glutamic acid for all mammalian caspases (44). These results suggest that other parameters such as the tertiary structure and posttranslational modification of proteins could also be involved in the specificity of cleavage by caspases.

One of the most prominent structural features of the Par-4 C terminus is a leucine zipper domain, which is necessary for homodimerization and heterodimerization (17, 25). Interestingly, the N-terminal peptide (amino acids 1 to 36) of Par-4 is found to prevent the interaction of Par-4 with itself and possibly with other binding partners (17). Although Par-4 contains a basic leucine zipper domain characteristic of transcription factors, its role as a transcription factor is yet to be determined. However, Par-4 can regulate transcription by acting as a corepressor and binding partner of various transcription factors (6, 9, 20, 25, 27, 34, 36) and proteins such as PKC (32). Mutation and deletions in the leucine zipper domain abolish the proapoptotic activity of Par-4 (22, 40). The proapoptotic action of Par-4 involves the inhibition of the PKC/NF-κB pathway (18). In addition, overexpression of Par-4 reduces the levels of antiapoptotic Bcl-2 protein (6). Our findings of Bcl-2 downregulation in WT Par-4-expressing cells are in accordance with these studies. Similarly, the decrease in FLIP protein levels in WT Par-4-expressing cells could also be due to the inhibition of NF-κB transcriptional activity (Fig. 8). The important question that remains unanswered is why Par-4 undergoes cleavage during apoptosis induction. Many proteins are known to be cleaved by caspases in order to liberate their regulatory domain from their catalytic domain during apoptosis. This could, however, lead to either inactivation or the constitutive activation of these proteins (7). In the context of Par-4, this could be explained by the results from a previous study (17), which indicates that the N-terminal peptide of Par-4 functions in dominant negative fashion for transactivation of androgen receptor, which is one of the binding partners of Par-4. The other key finding of the same study showed that the N-terminal peptide of Par-4 prevents its intramolecular interaction by blocking the leucine zipper domain and does not let the C terminus of Par-4 interact with one of its binding partners, like WT1 or PKC-ς, to regulate apoptosis induction. Thus, the cleavage of Par-4 at D131 is crucial for the dissociation of the N-terminal domain from the C-terminal domain and enhanced apoptotic induction. The caspase-3 cleavage site in Par-4 is strategically located and is ahead of NLS2 and the SAC, and leucine zipper domains. Therefore, the cleaved fragment of Par-4 retains all the key domains required for the induction of apoptosis (Fig. 6B). Functional and localization studies suggest that the nuclear entry of Par-4 is crucial for apoptosis induction (13). It can therefore be speculated that cleavage of Par-4 markedly induces nuclear targeting of the C-terminal fragment, suggesting that the Par-4 N terminus might contain cytoplasmic targeting signals. This is clearly evident from the immunofluorescence and Western blot analyses showing that the cleaved fragment is largely localized in the nuclei of cells transfected with the C-terminal Par-4 fragment (Fig. 10).

How does the C-terminal fragment of Par-4 induce apoptosis? Release of the C-terminal fragment from the N terminus could potentiate a proapoptotic response in a couple of ways: (i) by increasing the translocation of SAC and leucine zipper domains, which could easily interact with and bind to the transcription factors and binding partners with higher affinity, and (ii) by disrupting the Akt1 phosphorylation site present in the N terminus, which prevents Par-4 nuclear translocation, thus retaining it in the cytoplasm and rendering it incapable of causing apoptosis (20). The overexpression of the C-terminal Par-4 fragment significantly induced apoptosis by reducing the protein levels of antiapoptotic Bcl-2 and FLIP through the IκBα/NF-κB pathway (Fig. 10). While this paper was in preparation, another group also observed the presence of the cleaved fragment of Par-4 (11). However, the identity of the cleaved fragment was not confirmed. In that study, the authors showed that RASSF2 promotes Par-4 nuclear translocation. Immunoprecipitation using RASSF2 antibody followed by blotting with Par-4 antibody revealed full-length Par-4 and a shorter fragment, which the authors speculated to be the cleaved fragment of Par-4, suggesting that both Par-4 and its cleaved fragment interact with RASSF2, which could promote their nuclear translocation (11). However, future studies are required to elucidate additional roles of the cleaved fragment of Par-4 in the regulation of apoptosis and the underlying molecular mechanism involved.

In summary, the present study provided the first evidence that Par-4 is a physiological substrate of caspase-3 and is cleaved during apoptosis. Moreover, Par-4 degradation itself may contribute to the process of apoptosis induction by enhancing the nuclear localization of the C-terminal fragment. Present studies have explored a new, yet-unexplained function of Par-4 in regulating cell death. These novel findings raise a number of questions regarding the cellular functions of Par-4 and its cleaved fragments produced in vivo. This observation may serve as a powerful stimulus for the investigation of the various unseen roles of Par-4.