DIABLO transcriptional upregulation

DIABLO transcriptional upregulation

Cellular Signalling 19 (2007) 1212 – 1220 www.elsevier.com/locate/cellsig Apoptosis induced by cAMP requires Smac/DIABLO transcriptional upregulation...

913KB Sizes 5 Downloads 49 Views

Cellular Signalling 19 (2007) 1212 – 1220 www.elsevier.com/locate/cellsig

Apoptosis induced by cAMP requires Smac/DIABLO transcriptional upregulation Moises Martinez-Velazquez, Jorge Melendez-Zajgla ⁎, Vilma Maldonado ⁎ Molecular Biology Laboratory, Subdireccion de Investigacion Basica, Instituto Nacional de Cancerologia, Mexico City, Mexico Received 20 May 2006; received in revised form 20 December 2006; accepted 8 January 2007 Available online 21 January 2007

Abstract Smac/DIABLO is a mitochondrial protein that participates in apoptotic cell death by means of sequestering several members of the inhibitor of apoptosis protein family. This action allows caspase activation, cleavage of key cellular substrates and death. Release from mitochondria is considered the main regulatory step of Smac/DIABLO activity. Nevertheless, the fact that at least one isoform, Smac-beta, does not reside in this organelle implies that transcriptional regulation could also be important. cAMP is a well known second messenger with important apoptotic effects. To analyze if cAMP could be involved in Smac/DIABLO gene regulation, we analyzed 2903 base pairs upstream of the coding sequence and characterized the minimal promoter, which contains a consensus CRE site. We found that cAMP/PKA/CREB pathway is indeed an important regulator of Smac/DIABLO transcription, since exposure to the cAMP analog 8-CPT-cAMP, the adenylyl cyclase activator forskolin, the inhibitor of phosphodiesterase isobutylmethylxanthine or by hindering PKA activation with H89, regulated the promoter activity, as shown by gene reporter and RT-PCR assays. Additionally, the results of site-directed mutagenesis revealed that the consensus CRE site was biologically functional and required for cAMP-induced promoter activity in human HeLa cells. Supporting these results, a negative dominant version of the protein kinase A responsive factor, KCREB, reduced basal Smac/DIABLO expression and rendered the promoter unresponsive to cAMP. Reducing Smac expression using an antisense approach blocked the apoptosis effects of cAMP in cervical cancer cells. These results show that cAMP is an important modulator of the apoptotic threshold in cancer cell by means of regulating Smac/DIABLO expression. © 2007 Elsevier Inc. All rights reserved. Keywords: Apoptosis; PKA; cAMP; CREB; Promoter

1. Introduction Smac (Second mitochondria derived activator of caspase) and its murine ortholog DIABLO (Direct IAP binding with Low pI) are mitochondrial proteins encoded by nuclear DNA which are released into the cytosol in response to apoptotic stimuli that disrupt the integrity of the mitochondria. Smac/ DIABLO participates in the two main apoptotic pathways, the intrinsic or mitochondrial pathway [1,2] and the extrinsic or death receptor pathway [3,4]. After apoptotic stimuli, released Smac acts as a dimer in the cytosol, activating caspases by means of sequestering and neutralizing members of the inhibitor ⁎ Corresponding authors. Molecular Biology Laboratory, División de Investigación Básica, Instituto Nacional de Cancerología, Av. San Fernando # 22, Tlalpan 14000, México, D.F. Mexico. Tel.: +52 55 56280439; fax: +52 55 56280432. E-mail addresses: [email protected] (J. Melendez-Zajgla), [email protected] (V. Maldonado). 0898-6568/$ - see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.cellsig.2007.01.001

of apoptosis proteins family (IAPs) [5,6]. Although the exact mechanism that regulates Smac mitochondrial release is unknown, there is evidence that some cellular transduction signal pathways are able to regulate it [7–11]. In addition, as with cytochrome c, Smac release is modulated by members of the Bcl-2 family (Bcl-2, BID and Bcl-w) [12]. Interestingly, mitochondrial release of Smac is blocked by a broad-spectrum caspase inhibitor [13–15], showing the presence of a positive cellular feedback loop. Aside from the mitochondrial release of Smac/DIABLO, additional regulation mechanisms have been far less studied. Modulation of Smac mRNA and protein has been found in some systems. For example, folic acid and tumor necrosis factor increases Smac/DIABLO mRNA in kidney tubular cells [16]. This increase correlates with apoptosis in vivo. It has also been found that Smac is deregulated in cervical cancer [17], sarcomas [18], lung cancer carcinomas [19] and renal tumors [20]. The reason for the differences in the expression of this molecule in

M. Martinez-Velazquez et al. / Cellular Signalling 19 (2007) 1212–1220

normal cells versus cancer cells is unknown. Although the possible implications of these differences have not been thoroughly studied, alterations in Smac/DIABLO expression are able to influence the apoptotic threshold. For example, ectopic overexpression of Smac/DIABLO sensitizes hepatic cancer cells to apoptosis induced by anti-neoplasic drugs [21,22]. For these reasons, Smac/DIABLO is an interesting target for therapy. Indeed, it has been shown that peptides corresponding to the IAP recognition motif of Smac sensitize cancer cells to tumor necrosis factor-related apoptosis inducing ligand (TRAIL), and chemotherapeutic agents [23–25]. The second messenger cAMP is able to induce growth arrest and apoptosis in diverse cancer cell lines [26]. cAMP can also potentate chemotherapeutic agents such as paclitaxel, an important antineoplasic agent used in a variety of tumors [27]. The higher susceptibility to cAMP-mediated apoptosis of

1213

certain tumor cells compared to their normal counterpart has spurred interest in developing analogs as anti-cancer drugs [28]. The mechanisms of these effects are poorly understood. In the present paper we have investigated the role of cAMP in the transcriptional regulation of Smac/DIABLO as a potential mechanism for the second messenger apoptosis effects. 2. Materials and methods 2.1. Cloning of the Smac/DIABLO genomic 5′ untranslated region Genomic DNAwas prepared with DNAzol (Invitrogen, MD USA) from a pool of human leukocytes isolated by Ficoll density gradient centrifugation. In silico analysis was used to identify the 5′untranslated region from the chromosome 12 clone RP11-512M8 (GenBank accession no. AF62240 and GenBank accession no. AC048338). Two fragments of 2903 and 1398 bp of the 5′untranslated flanking region of the DIABLO gene were amplified by nested PCR using the proofreading

Fig. 1. DIABLO non-coding 5′ region. (A) DIABLO 5′ untranslated region was amplified by long range PCR using information derived from the published GenBank genome data. 1519 bp are shown. The absent region is represented by the crossed nucleotides. The minimal promoter region is marked in bold. The arrow indicates the putative transcription initiation site. (B) Similarity plot of the 5′ non-coding region of Smac/DIABLO from four mammalian species. Plot was generated using VectorNTI software with data from Homo sapiens, Mus musculus, Rattus norvegicus and Canis familiaris.

1214

M. Martinez-Velazquez et al. / Cellular Signalling 19 (2007) 1212–1220

Pfu polymerase (Stratagene, CA USA). The primers were: Forward 1.4 kb: 5′ CCGCCTCTCGAGGGAACGCCTGTGCGCAGCTCCCTG3′, 2.9 Kb: 5′ CAGGCTCGAGGTTGCTCAAACTCCTGGCCTCAAGAG3′; Reverse 1: 5′AGCCAACTCTTCAGAGCCGCCAT3′, Reverse 2: 5′GAGCCGAATTCGTGCAGCGCGCGGACGCCAGACGC3′. PCR was performed in presence of 5% DMSO to enhance the amplification of the GC-rich DIABLO promoter region. The PCR fragment was sequenced in both directions using an ABI PRISM DNA sequencing facility and compared with the chromosome 12 clone RP11-512M8. Promoter regions were predicted using Promoter Inspector software (Genomatix, Germany). Potential transcription factor binding sites were predicted using TESS (Transcription Element Search Software) programs (http://www.cbil.upenn.edu/ tess) [29]. The transcription start site was predicted by Promoter 2.0 predictor program (http://www.cbs.dtu.dk/services/Promoter/) [30] and Eponine (http:// servlet.sanger.ac.uk%3A8080/eponine/) [31].

2.2. Cell culture HeLa cells were obtained from the American Type Culture Collection and maintained in DMEM medium supplemented with 8% fetal bovine serum at 37 °C in a humidified 5% CO2 incubator. All cell culture reagents were obtained from Invitrogen (Maryland, USA). 8-(4-chlorophenylthio)-adenosine 3′,5′cyclic monophosphate (8-CPT-cAMP), H89, forskolin and 3-isobutyl-1methylxanthine (IBMX) were obtained from Sigma-Aldrich (Missouri, USA).

2.3. Cellular viability and cytotoxic effects Cells were seeded in 24-well dishes and exposed to drug for 24 h. The cells were then fixed in 70% ethanol at − 20 °C, washed in PBS 1X (1.7 mM KH2PO4, 5.2 mM Na2HPO4, 150 mM NaCl, 2.7 mM KCl) and stained with crystal violet (1% in water). After washing, the stain was solubilized in 33% acetic acid and the absorbance determined in an ELISA reader at 570 nm. Analyses were performed at least by triplicate in four independent experiments. Statistical analysis was performed using Bonferroni's T test.

2.4. Transfections and SEAP assays Cells were seeded in 12-well culture dishes. Quadruplicate wells were transfected using FuGENE 6 Reagent (Roche, IN USA) with 25 ng of reporter plasmid together with 1 ng of a plasmid containing a β-galactosidase gene under the control of a constitutive cytomegalovirus promoter (pβGAL), in order to normalize transfection efficiency. The DIABLO promoter-SEAP reporter plasmid was constructed by cloning the products of PCR into the pSEAP2Basic Vector (Clontech, CA USA), which uses a secreted form of human placental alkaline phosphatase as a reporter molecule to monitor activity of the promoter. Chemiluminescent detection of SEAP activity was assayed 48 h post transfection using a commercial kit (Clontech, CA, USA). β-galactosidase activity was determined by a commercial kit (Promega, WI USA). Experiments were performed by triplicate in three independent experiments and results normalized against β-galactosidase activity. The HeLaSEK-AL cell line was generated by transfecting HeLa cells with a pcDNA 3 plasmid containing a mutated version of MEKK4 (SEK) fused with a hemagglutinin epitope. The SEK-AL plasmid was a kind gift from Dr. Brent Zanke. To establish a stable cell line, after transfection, the cells were selected using 800 μM of G418 for 4 weeks. A pool of five clones was then used for subsequent experiments. Expression of the transgene was assessed by a western blot assay using antihemagglutinin antibody (3F10, Roche, Indiana, USA). For comparison, a control cell line was created by transfecting HeLa cells with pcDNA 3 (Invitrogen, MD, USA) and selecting with G418 for 4 weeks. The same approach was used to produce a HeLa cell line expressing the dominant negative version of CREB, HeLaKCREB. KCREB is unable to bind to DNA due to mutation in the binding motif of the transcription factor [32]. This plasmid was a kind gift from Dr. R.H. Goodman. To inhibit Smac-alpha expression, a complete open reading frame was amplified by RT-PCR from HeLa cells and cloned in an antisense orientation in pcDNA3.1 (Invitrogen, Maryland, USA). This construct was transfected, as described previously. As a control, empty pcDNA3.1 was used.

2.5. Promoter deletions Deletions were constructed using stepwise unidirectional cleavage with exonuclease III/Nuclease S1. Briefly, the 1.3 Kb Smac DNA fragment was linearized with Xho I and Kpn I to generate 3′recessed ends and a controlled unidirectional degradation of single strand overhang DNA was performed with exonuclease III, followed by removal of the remaining single strand overhang with S1 nuclease. The deletion reaction was performed at 30 °C with 1 min timed aliquot removal. Religation was performed with T4 DNA ligase. All reagents were obtained from Invitrogen, MD USA. All deletions were verified by sequencing.

2.6. Semiquantitative RT PCR assays RNA was isolated with TRIzol (Invitrogen, MD USA) as recommended by the manufacturer. cDNA was synthesized using the thermoscript RT-PCR system (Invitrogen, MD USA). Five micrograms of DNAse I pre-treated RNA was incubated at 50 °C for 1 h with random hexamers, dNTPs and reverse transcriptase. The RNA template was digested with RNAse H. To verify equal cDNA loading, primers 5′CCCCTTCATTGACCTCAACT-3′ and 5′ TTGTCATGGATGACCTTGGC-3′ were used to amplify a small fragment of the house-keeping gene GAPDH using Advantage Taq polymerase (Clontech, CA USA). Smac cDNA was amplified using the primers: 5′ AGCTGGAAACCACTTGGATG-3′ and 5´CAGCTTGGTTTCTGCTTTCC3′. The PCR products were subjected to electrophoresis in 1% agarose gels. In order to obtain amplification in the log phase, a curve was constructed using densitometric data derived from amplicons produced with increasing PCR cycles and the data normalized using GAPDH data. Densitometric analysis was performed using Image J software (Wayne Rasband, NIH. http://rsb.info. nih.gov/ij/).

2.7. Site-directed mutagenesis Site-directed mutagenesis of the CRE (cAMP-Responsive Element) site was performed using a PCR-based method as described by Fisher and Pei [33]. Briefly, the reporter plasmid was amplified with the primers: GCTCACGAAGCTGCAGTCCGGCGTGTG and CACACGCCGGAAGTGCTGCAGCTTCGTGAGC using Advantage Polymerase (Clontech, CA, USA) with 16 two-step cycles. Parental DNA was degraded using Dpn I and competent DH5 alpha E. coli, transformed with 1 μl of the reaction. Colonies were screened by restriction analysis and sequencing.

3. Results 3.1. Cloning and analysis of the human DIABLO promoter region In order to investigate if Smac/DIABLO transcription could be regulated by cAMP and to characterize the mechanism of this regulation, we first isolated and analyzed the human promoter for this gene. Based on similarity between the Smac/DIABLO sequence (GenBank accession no. AF262240) and the chromosome 12 clone RP11-512M8 (GenBank accession no. AC048338), the human DIABLO gene has been mapped to chromosome 12, region 12q24.31 [34]. Using high fidelity long range PCR and information derived from GenBank contig NT_009755 in chromosome 12, we amplified two different sized fragments from the 5′-region of the gene. Both nucleotide fragments were fully sequenced and found to agree with the reported GenBank contig (NC_000012, position 121214228 to 121217228) with the exception of a 169 base pairs absent segment from − 387 to − 218 (Fig. 1A). To further verify this discrepancy, two other clones obtained from the long range

M. Martinez-Velazquez et al. / Cellular Signalling 19 (2007) 1212–1220

1215

Fig. 2. Transcriptional activity of DIABLO non-coding 5′ region. Transcriptional activity of deletion constructs of DIABLO promoter in HeLa cells. A schematic representation of DIABLO promoter deletion constructs is shown on the left. Transcriptional activity of the corresponding constructs is indicated on the right. Transfection efficiency has been controlled with p βGAL plasmid cotransfections, and the results are presented as relative values of promoter activity. The activity of the full-length construct, 3K (− 2903 bp), is set to 1. Error bars indicate standard deviation.

PCR reaction were sequenced, confirming the difference. In silico analysis [30] defined the promoter to − 254 bp from the initial ATG and a putative consensus transcriptional initiation element at − 37 bp (Fig. 1A). In agreement with this, an alignment of the 5′ non-coding region of Smac/DIABLO from four mammalian species showed a high degree of similarity in this region (Fig. 1B). To perform an initial characterization of the promoter, both fragments were cloned in pSEAP2-basic, a vector that allows insertion of promoter sequences upstream the gene for secreted alkaline phosphatase (SEAP). Fig. 2 shows that similar transcriptional activity was induced by the 3 kb and the 1.5 kb fragments in HeLa cells. Interestingly, most of the activity could be derived from the 1.4 kb fragment. Analysis of the putative promoter showed consensus sites for Sp1, ELK, c-Rel, Myb and a proximal AP1/CREB. To map the minimal promoter, we generated thirteen 5′ deletion constructs using the 1398 base pair fragment using exonuclease III and nuclease S1 (D1–D13) (Fig. 2). The first nine deletions (D1–D9) showed slight increases in activity over the complete 1.4 or 2.9 kb promoter (Fig. 2B). Deletions D10 to D12, which correspond to − 376 to − 193 bp (D10 = 376 bp; D11 = 288 bp; D12 = 193 bp) presented higher activity than the rest of the constructs. The D13 deletion, which only contains the first 38 nucleotides upstream of the Smac/DIABLO first exon, did not support promoter activity. To further map the minimal promoter, an additional internal deletion construct was made by deleting a segment corresponding to 253 bp upstream of the first exon by PCR. This construct (D253) was not able to activate transcription of the reporter gene. These results mapped the minimal promoter to the first 193 bp upstream of DIABLO first exon, consistent with in silico analysis (Fig. 1).

3.2. Regulation of Smac/DIABLO by cAMP As previously mentioned, in silico analysis of the minimal promoter using the TESS program [29] revealed several putative consensus sequences (Fig. 3A) for transcription factors. An alignment of four different putative promoters from mammalian species showed a high degree of conservation in the sequences provided by TESS (Fig. 3B) Among these, the CRE (cAMP-Responsive Element) site (GCTGCGTCACT) positioned at − 46 to− 35 was particularly attractive as a possible functional element, since an alignment of the 5′ non-coding region of several mammalian species showed that this site was conserved (Fig. 3C). It is noteworthy that the core CRE sequence (CGTCA), essential for the transcription factor binding, was identical in all the sequences. To elucidate if this CRE site could be functional, HeLa cells transfected with the D12 construct were exposed to a cell permeable cAMP analog (8-CPT-cAMP), known to activate PKA and modulate several genes via CRE sites. As shown in Fig. 4A, 8-CPT-cAMP exposure increased DIABLO promoter activity by 25%. This effect was substantially higher (four times the basal activity) when the cells were cultivated in serum-free conditions (Fig. 4B) previous to the addition of the analog. To support these results, we exposed HeLa cells to the selective PKA inhibitor H-89. This compound is a member of the isoquinolinesulfoamide group of protein kinase competitive inhibitors that selectively inhibit PKA in vivo [35,36]. Exposure of HeLa cells to the inhibitor reduced the reporter activity by 40% (Fig. 4C). Similarly, exposure of these cells to the activator of adenylyl cyclase forskolin (Fig. 4D) or to the inhibitor of phosphodiesterase 3isobutyl-1-methylxanthine (IBMX) also increased the transcriptional activity of the Smac/DIABLO promoter. To further

1216

M. Martinez-Velazquez et al. / Cellular Signalling 19 (2007) 1212–1220

Fig. 3. Sequence analysis of DIABLO promoter. (A) Putative transcription factor consensus sites are marked. Arrows indicate the 5′ of the 12th and 13th deletions (193 bp and 38 bp, respectively). (B) Absolute complexity plot derived from the alignment of the 5′ non-coding region of Homo sapiens, Mus musculus, Rattus norvegicus and Canis familiaris Smac/DIABLO genes. This analysis, performed using VectorNTI software, indicates how conserved is each nucleotide in the region. (C) Alignment of putative CRE sites (shadowed bases) from the non-coding 5′ region of four different mammalian DIABLO genes. The core CRE sequence is marked in bold.

corroborate these results, we produced a promoter-reporter vector in which the core CRE site of the Smac promoter was mutated. Fig. 4E shows that the activity of this mutated promoter after addition of the cAMP analog was dramatically reduced, when compared to the wild type version. These results clearly define the binding site for CREB located at position (− 46/− 35) as essential to confer responsiveness to cAMP, Interestingly, the basal activity of the mutated promoter in HeLa cancer cells was strongly reduced to less of one-tenth of the wild type promoter, although some residual activity was still found. In order to ascertain if the transcriptional regulation by cAMP was accompanied with an increase in Smac mRNA, we used RTPCR analysis in HeLa cells exposed to 8-CPT-cAMP, forskolin and IBMX. As expected, all the compounds induced an increase in total Smac mRNA (Fig. 5A, B). Finally, to verify that CREB was required for basal transcription of Smac and for the regulation induced by cAMP, we compromised CREB signaling by using a dominant negative version of the factor (KCREB)

[32]. Blocking CREB activity reduced steady-state DIABLO mRNA levels by 85% (Fig. 5D) and rendered cells insensitive to the cAMP analog effect (Fig. 5E). These results strongly suggest that expression of DIABLO in cancer cells is regulated by the cAMP pathway via CRE-binding protein. 3.3. JNK does not regulate Smac/DIABLO expression The previous results showed that up regulation of Smac/ DIABLO by cAMP was strongly enhanced after serum deprivation. It has been shown that serum withdrawal induces apoptosis in cancer cells by means of activating JNK [37]. Death induced by this stimulus was suppressed by the expression of a dominant negative version of the JNK upstream kinase MEKK4 (also known as SEK). Additionally, it has been shown that JNK is able to induce apoptosis by activating AP1 transcription factor [38]. Since the CRE site found in the minimal promoter was a composite motif also containing an AP1 consensus, a logical

M. Martinez-Velazquez et al. / Cellular Signalling 19 (2007) 1212–1220

1217

Fig. 4. Effects of cAMP on DIABLO promoter transcriptional activity. HeLa cells were transfected with a plasmid containing the gene for Secreted Alkaline Phosphatase (SEAP) under the control of the DIABLO minimal promoter. Cells were exposed to cAMP analog 8-CPT-cAMP (A) in the presence or (B) absence of serum or (C) in the presence of a 10 μM concentration of the PKA inhibitor H89. (D) Cells were exposed to a100 μM concentration of forskolin or a 200 μM concentration of 3-isobutyl-1-methylxanthine (IBMX) (E) Transcriptional activity of the wild type or mutated Smac/DIABLO promoter exposed to vehicle (control) or the cAMP analog 8-CPT-cAMP. Experiments were performed in triplicate in three independent experiments. ⁎p b 0.05. Statistical analyses were performed using Bonferroni's T test.

Fig. 5. Effects of cAMP on Smac/DIABLO expression. (A) HeLa cells were exposed to a 50, 100 or 200 μM concentration of the cell permeable cAMP analog 8-CPTcAMP and Smac/DIABLO mRNA expression was analyzed in semiquantitative RT-PCR assays. Equal loading was verified by amplification of the housekeeping gene GAPDH. Smac/DIABLO expression was increased in cells exposed to 100 and 200 μM concentration of 8-CPT-cAMP by 4.8 (±0.9) and 4.2 (±0.7) times the control, respectively. (B) HeLa cells were exposed to a100 μM concentration of forskolin, a 200 μM concentration of 3-isobutyl-1-methylxanthine (IBMX) or a 100 μM concentration of 8-CPT-cAMP and Smac/DIABLO mRNA expression was analyzed in semiquantitative RT-PCR assays. Equal loading was verified by amplification of the housekeeping gene GAPDH. Smac/DIABLO expression was increased in cells exposed to forskolin, IBMX or 8-CPT-cAMP by 5.2 (±0.6), 4.9 (± 0.8) or 4.7 (±0.8) times the control, respectively. (C) Analysis of CREB expression in HeLaKCREB or parental cell line transfected with an empty vector to demonstrate overexpression of the transgene. (D) HeLaKCREB cell line or a control parental cell line transfected with an empty vector were analyzed for Smac/DIABLO expression using semiquantitative RT-PCR. (E) HeLaKCREB cell line or a control parental cell line transfected with an empty vector were exposed to a 50 μM concentration of the cAMP analog 8-CPT-cAMP. After 24 h, the cells were analyzed for Smac/DIABLO expression using semiquantitative RT-PCR. The expression of Smac/DIABLO was increased only in cells transfected with the empty vector by 3.6 (±0.4) times. Expression of Smac/DIABLO was reduced to 0.3x in HeLaKCREB cells. Procedures were performed in triplicate as stated in Materials and Methods.

1218

M. Martinez-Velazquez et al. / Cellular Signalling 19 (2007) 1212–1220

Fig. 6. Effects of Smac downregulation on the apoptosis induced by cAMP. (A) HeLa cells exposed to vehicle (left panel) or 8-CPT-cAMP (right panel) and visualized using phase contrast microscopy at a 10× magnification. Bar: 30 microns. Inset: 40× magnification. Arrow shows a cell undergoing condensation and membrane convolution. (B) HeLa cells exposed to vehicle (left panel) or 8-CPT-cAMP.(right panel) and nuclei visualized using ethidium bromide stain as described in Material and Methods. Nuclei are condensed and fragmented. (C) HeLa cells were transfected with a plasmid containing a Smac/DIABLO full-length cDNA in antisense orientation or an empty vector. Expression levels were measured by RT-PCR assay. (D) HeLa cells transfected with a vector containing an antisense version of Smac or an empty vector were exposed to 8-CPT-cAMP and viability assays were performed as described in Materials and Methods. Procedures performed by quintuplicate. ⁎p b 0.05 by Bonferroni's T-test.

candidate for the serum effect was JNK. For this reason, we created HeLa cells stably expressing a dominant negative version of the upstream JNK kinase, SEK-AL [39]. This mutant derivative renders the pathway insensitive to extracellular signals. As shown in the Supplementary Fig. 1 neither Smac promoter activity nor its expression was affected in these cells. Supporting these results, we found that Smac/DIABLO expression was not modified after exposure of HeLa cells to the cell-permeable JNK inhibitor, SP600125. These results exclude JNK as the pathway responsible for the transcriptional regulation of Smac/DIABLO mediated by serum. In addition, inhibition of the ERK pathway using the MEK inhibitor U0126 did not produce noticeable changes in the expression of Smac/ DIABLO in these cells (see Supplementary Fig. 2), excluding also the participation of this signaling cascade. 3.4. Apoptosis effects of cAMP are mediated by Smac/DIABLO It has been previously shown that cAMP is able to induce apoptosis in a variety of cancer cell lines.[26,40–42]. To test if this was also true for cervical cancer, we exposed HeLa cells to 8-CPT-cAMP. Fig. 6A shows that the cAMP analog induced morphologic changes compatible with apoptosis, including shrinkage and apoptotic body' formation. In addition, clear nuclear condensation and fragmentation were evident (Fig. 6B). In order to investigate if the previously reported apoptotic effect

of cAMP could be mediated by the upregulation of Smac, we transfected HeLa cells with a plasmid containing an antisense version of this molecule (Fig. 6C). Decreased Smac/DIABLO expression protected cells to the apoptotic effects of 8-CPTcAMP (Fig. 6D). These results show that Smac/DIABLO upregulation is required for the apoptotic effects of cAMP. 4. Discussion The mechanism of cAMP signal transduction in mammalian cells is one of the best understood biochemical pathways. Phosphorylation of CREB mediated by the cAMP signaling pathway can be initiated by a plethora of physiological stimuli and is critically involved in the regulation of metabolism, cell growth and differentiation, apoptosis, and gene expression [43]. Transcriptional regulation by cAMP is mediated by a group of nuclear factors that bind to and regulate the expression of genes containing the consensus cAMP-responsive element (CRE) in their promoters. Phosphorylation of these CRE binding proteins (CREBs) by PKA modulates their activity [44]. Even with considerable knowledge of the pathways that mediate the proliferative effects of cAMP, the precise determinants of its death effect are poorly understood. It has recently been shown that cell death induced by cAMP analogs occurs via PKA and CREB-dependent apoptosis [45,46]. In hematopoietic cells, upregulation of the BH3-only

M. Martinez-Velazquez et al. / Cellular Signalling 19 (2007) 1212–1220

protein Bim by this pathway is a key event of the apoptotic response to cAMP analogs and glucocorticoids [46]. However, this mechanism has not been explored for solid tumors, which present divergent mechanisms for death induction by chemotherapeutic drugs [47]. In addition to direct induction of apoptosis, cAMP can also sensitize cells to anti-neoplasic drugs, such as paclitaxel [27]. This drug induces death via a mitochondrial-dependent pathway by means of releasing proapoptotic factors, such as Smac/DIABLO into the cytosol, after disruption of the integrity of this organelle [48]. Given the reported importance of Smac/DIABLO in apoptosis induced by paclitaxel [49], and the fact that the death effect of this drug is enhanced by cAMP, [27] we sought to investigate if apoptosis induced by this second messenger could be mediated by a transcriptional mechanism acting upon the DIABLO promoter. For this, we cloned and analyzed a 2.9 kilobase fragment from the 5′ end of the gene, describing the minimal promoter required for basal expression in a cancer cell line. Interestingly, in silico analysis demonstrated a composite consensus site for AP1/CREB within this region. As expected by this result, cAMP was able to induce transcription of a reporter gene controlled by this minimal promoter. Unexpectedly, this effect was noticeably increased after serum deprivation. This observation could be due to additional transcriptional factors being negatively regulated by a serum component that could act on the promoter to enhance transcription. Alternatively, a repressor modulated by serum could also be involved. Since the consensus site included an AP1 motif, a logical candidate would be c-jun activated kinase pathway (JNK). Nevertheless, our results using a dominant negative version of the upstream kinase of this pathway (SEKAL) clearly excluded this possibility. We are currently investigating additional factors. It has been shown that several signal transduction pathways lead to phosphorylation and activation of CREB [50]. Using pharmacological inhibitors, we analyzed the participation of MAPK (U0126, PD98059), p38 MAPK (SB203580, not shown), JNK/SAPK (SP600125) and Akt/PKB (LY-294002, not shown) in the transcriptional regulation of Smac/DIABLO. Neither of these inhibitors modified the basal or induced expression of this gene, supporting the crucial role of PKA in the activation of CREB and subsequent regulation of Smac/ DIABLO expression. Nevertheless, we cannot exclude the participation of other unidentified kinases in the regulation of Smac/DIABLO. A more directed and thorough analysis using a panel of inhibitor should help to solve this question. In addition to being able to induce DIABLO transcription, the cAMP-induced pathway is also involved in the basal transcription of this gene, since a commonly used cell-permeable inhibitor also blocked its basal expression. This conclusion is further supported by the striking decrease found in the DIABLO basal promoter activity when the CRE site was mutated and the fact that a dominant negative version of the downstream transcriptional effector of cAMP, KCREB, blocked both the basal and induced expression of Smac/DIABLO. Further analysis using primary cell culture should be performed to establish if this is also the case for non-transformed cells.

1219

Even though the main regulatory step in the activity of Smac/ DIABLO relies on its release from mitochondria, expression levels could be an important additional regulatory mechanism. This step could mediate the apoptotic threshold in a particular cellular or tissue context. An example for an apoptotic molecule in which expression regulation is also important is APAF-1, the principal downstream effector of the mitochondrial death pathway. This protein is a transcriptional target for the suppressor gene p53 and is modulated in several tumors and in particular in melanoma [51,52]. Although not extensively explored, differences in expression of Smac/DIABLO have also been found in cancer. It has been reported that gastric [20], renal [53], pulmonary [19] and cervical tumors [54] have a deregulated expression of this molecule. More important, expression of Smac in renal and lung tumors correlates with prognosis, pointing toward a possible clinical relevance [19,53]. These results show that Smac/DIABLO is dynamically regulated in cancer progression. The existence of at least one Smac cytosolic isoform that is still able to induce apoptosis even without a mitochondrial localization signal, Smac-beta, adds further support to the importance of expression modulation as a regulatory mechanism [55]. The data presented here provides the first analysis of the transcriptional regulation of this interesting molecule, showing that cAMP could be one of the most important determinants of Smac/DIABLO regulation. The cellular consequences of Smac/DIABLO expression modulation have not been studied in a physiological context. Nevertheless, in the case of cancer, overexpression studies have implicated Smac in chemo-and radioresistance. Recent reports have shown that apoptotic threshold can be readily modulated by ectopically increasing Smac/DIABLO expression levels. In leukemia cells, sensitivity to stimuli that activate both intrinsic and extrinsic pathways was significant increased by Smac/DIABLO overexpression [56]. Similarly, transfection of a plasmid encoding Smac sensitizes hepatocarcinoma, myogenic, gastric adenocarcinoma and ovarian cells to a varied array of chemotherapeutic and physical agents, including gamma radiation [21,22,57–59]. In vivo Smac gene transfer to intracranial malignant gliomas sensitize them to apoptosis induced by chemotherapeutic agents and TRAIL [25]. Our results suggest that modulating CREB activity via cAMP analogs should be an attractive strategy to enhance not only chemotherapy, but also radiosensitivity, by means of modulating Smac/DIABLO expression. For these reasons, determining the importance of Smac/DIABLO regulation by the cAMP pathway in cancer and other human diseases should be an important concern for future research. In conclusion, we provide evidence that the cAMP analog 8-CPT-cAMP upregulates the proapoptotic gene Smac/DIABLO via a CREB-dependent mechanism acting upon its promoter. This effect is mediated by the increase of Smac mRNA transcription, as shown by RT-PCR and gene reporter analyses. The importance of this mechanism is underscored by the requirement of Smac/DIABLO upregulation for apoptosis induced by cAMP. The characterization of a novel mechanism for the regulation of Smac/DIABLO expression should aid the identification of potential targets for therapy.

1220

M. Martinez-Velazquez et al. / Cellular Signalling 19 (2007) 1212–1220

Acknowledgements This work was supported by grants from the Consejo Nacional de Ciencia y Tecnologia (CONACyT) A2039/A1 to VM and SEP-2004-C01-45728 to JMZ. Moises MartinezVelazquez was also supported by a graduate fellowship from CONACyT. We thank Dr. Brent Zanke for his kind gift of SEKAL plasmid and Dr. Richard Goodman for the kind gift of KCREB plasmid. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.cellsig.2007.01.001. References [1] A.M. Verhagen, P.G. Ekert, M. Pakusch, J. Silke, L.M. Connolly, G.E. Reid, R.L. Moritz, R.J. Simpson, D.L. Vaux, Cell 102 (2000) 43. [2] P.G. Ekert, J. Silke, C.J. Hawkins, A.M. Verhagen, D.L. Vaux, J. Cell Biol. 152 (2001) 483. [3] S.M. Srinivasula, P. Datta, X.J. Fan, T. Fernandes-Alnemri, Z. Huang, E.S. Alnemri, J. Biol. Chem. 275 (2000) 36152. [4] Y. Deng, Y. Lin, X. Wu, Genes Dev. 16 (2002) 33. [5] C.S. Duckett, Biochem. J. 385 (2005) e1. [6] J.C. Wilkinson, A.S. Wilkinson, F.L. Scott, R.A. Csomos, G.S. Salvesen, C.S. Duckett, J. Biol. Chem. 279 (2004) 51082. [7] L.L. Zhou, L.Y. Zhou, K.Q. Luo, D.C. Chang, Apoptosis 10 (2005) 289. [8] R.T. Uren, G. Dewson, C. Bonzon, T. Lithgow, D.D. Newmeyer, R.M. Kluck, J. Biol. Chem. 280 (2005) 2266. [9] T.M. Hansen, D.J. Smith, P. Nagley, Cell Death Differ. (2005) 1. [10] F.J. Mendoza, E.S. Henson, S.B. Gibson, Biochem. Biophys. Res. Commun. 331 (2005) 1089. [11] D. Chauhan, G. Li, T. Hideshima, K. Podar, C. Mitsiades, N. Mitsiades, N. Munshi, S. Kharbanda, K.C. Anderson, J. Biol. Chem. 278 (2003) 17593. [12] X.L. Hu, T. Olsson, I.M. Johansson, T. Brannstrom, P. Wester, Eur. J. Neurosci. 20 (2004) 1177. [13] C. Adrain, E.M. Creagh, S.J. Martin, EMBO J. 20 (2001) 6627. [14] S. Li, Y. Zhao, X. He, T.H. Kim, D.K. Kuharsky, H. Rabinowich, J. Chen, C. Du, X.M. Yin, J. Biol. Chem. 277 (2002) 26912. [15] M. Madesh, B. Antonsson, S.M. Srinivasula, E.S. Alnemri, G. Hajnoczky, J. Biol. Chem. 277 (2002) 5651. [16] P. Justo, A. Sanz, C. Lorz, D. Gomez-Garre, S. Mezzano, J. Egido, A. Ortiz, Kidney Inter., Suppl. (2003) S52. [17] M. Espinosa, D. Cantu, C.M. Lopez, J.G. De la Garza, V. Maldonado, J. Melendez-Zajgla, BMC Cancer 4 (2004) 84. [18] N.J. Yoo, H.S. Kim, S.Y. Kim, W.S. Park, C.H. Park, H.M. Jeon, E.S. Jung, J.Y. Lee, S.H. Lee, Apmis 111 (2003) 382. [19] A. Sekimura, A. Konishi, K. Mizuno, Y. Kobayashi, H. Sasaki, M. Yano, I. Fukai, Y. Fujii, Oncol. Rep. 11 (2004) 797. [20] Y. Yan, C. Mahotka, S. Heikaus, T. Shibata, N. Wethkamp, J. Liebmann, C.V. Suschek, Y. Guo, H.E. Gabbert, C.D. Gerharz, U. Ramp, Br. J. Cancer 91 (2004) 1349. [21] J. Zhao, J. Jin, X. Zhang, M. Shi, J. Dai, M. Wu, R. Wang, Y. Guo, Cancer Gene Ther. (2006) 420. [22] L.D. Zheng, Q.S. Tong, L. Wang, J. Liu, W. Qian, World J. Gastroenterol. 11 (2005) 79. [23] H. Sun, Z. Nikolovska-Coleska, C.Y. Yang, L. Xu, Y. Tomita, K. Krajewski, P.P. Roller, S. Wang, J. Med. Chem. 47 (2004) 4147. [24] K.M. Bockbrader, M. Tan, Y. Sun, Oncogene (2005) 7381.

[25] S. Fulda, W. Wick, M. Weller, K.M. Debatin, Nat. Med. 8 (2002) 808. [26] R. Boe, B.T. Gjertsen, S.O. Doskeland, O.K. Vintermyr, Br. J. Cancer 72 (1995) 1151. [27] V. Maldonado, J. De Anda, J. Melendez-Zajgla, J. Biochem. Toxicol. 11 (1996) 183. [28] A. Lerner, D.H. Kim, R. Lee, Leuk. Lymphoma 37 (2000) 39. [29] J. Schug, G.C. Overton, TESS: Transcription Element Search Software on the WWW. Technical Report CBIL-TR-1997-1001-v0.0, University of Pennsylvania, 1997. [30] S. Knudsen, Bioinformatics 15 (1999) 356. [31] T.A. Down, T.J. Hubbard, Genome Res. 12 (2002) 458. [32] K.M. Walton, R.P. Rehfuss, J.C. Chrivia, J.E. Lochner, R.H. Goodman, Mol. Endocrinol. 6 (1992) 647. [33] Fisher CL, Pei G.K. Biotechniques 23(1997)570-571, 574. [34] S.E. Antonarakis, 605219. Second Mitochondria-Derived Activator of Caspase. OMIM Database, John Hopkins, 2000. [35] H. Hidaka, M. Inagaki, S. Kawamoto, Y. Sasaki, Biochemistry 23 (1984) 5036. [36] Y.S. Lee, C.M. Chuong, J. Cell. Physiol. 170 (1997) 153. [37] Y. Huang, D. Hutter, Y. Liu, X. Wang, M.S. Sheikh, A.M. Chan, N.J. Holbrook, J. Biol. Chem. 275 (2000) 18234. [38] G.V. Putcha, S. Le, S. Frank, C.G. Besirli, K. Clark, B. Chu, S. Alix, R.J. Youle, A. LaMarche, A.C. Maroney, E.M. Johnson Jr., Neuron 38 (2003) 899. [39] B.W. Zanke, E.A. Rubie, E. Winnett, J. Chan, S. Randall, M. Parsons, K. Boudreau, M. McInnis, M. Yan, D.J. Templeton, J.R. Woodgett, J. Biol. Chem. 271 (1996) 29876. [40] G. Fassina, M.G. Aluigi, S. Gentleman, P. Wong, T. Cai, A. Albini, D.M. Noonan, Int. J. Cancer 72 (1997) 1088. [41] S.N. Kim, Y.H. Ahn, S.G. Kim, S.D. Park, Y.S. Cho-Chung, S.H. Hong, Int. J. Cancer 93 (2001) 33. [42] H.M. Wang, N.G. Zheng, J.L. Wu, C.C. Gong, Y.L. Wang, World J. Gastroenterol. 11 (2005) 6538. [43] Y.S. Cho-Chung, Cancer Treat. Res. 115 (2003) 123. [44] M.R. Montminy, G.A. Gonzalez, K.K. Yamamoto, Trends Neurosci. 13 (1990) 184. [45] L. Yan, V. Herrmann, J.K. Hofer, P.A. Insel, Am. J. Physiol., Cell Physiol. 279 (2000) C1665. [46] L. Zhang, P.A. Insel, J. Biol. Chem. 279 (2004) 20858. [47] I. Herr, E. Ucur, K. Herzer, S. Okouoyo, R. Ridder, P.H. Krammer, M. von Knebel Doeberitz, K.M. Debatin, Cancer Res. 63 (2003) 3112. [48] N. Andre, D. Braguer, G. Brasseur, A. Goncalves, D. Lemesle-Meunier, S. Guise, M.A. Jordan, C. Briand, Cancer Res. 60 (2000) 5349. [49] Z. Song, X. Yao, M. Wu, J. Biol. Chem. 278 (2003) 23130. [50] B. Mayr, M. Montminy, Nat. Rev., Mol. Cell Biol. 2 (2001) 599. [51] J.D. Allen, X.D. Zhang, C.L. Scott, G.M. Boyle, P. Hersey, A. Strasser, Cell Death Differ. 12 (2005) 680. [52] M.S. Soengas, R.M. Alarcon, H. Yoshida, A.J. Giaccia, R. Hakem, T.W. Mak, S.W. Lowe, Science 284 (1999) 156. [53] Y. Mizutani, H. Nakanishi, K. Yamamoto, Y.N. Li, H. Matsubara, K. Mikami, K. Okihara, A. Kawauchi, B. Bonavida, T. Miki, J. Clin. Oncol. 23 (2005) 448. [54] M. Espinosa, D. Cantu, C.M. Lopez, J.G. De la Garza, V.A. Maldonado, J. Melendez-Zajgla, BMC Cancer 4 (2004) 84. [55] D.L. Roberts, W. Merrison, M. MacFarlane, G.M. Cohen, J. Cell Biol. 153 (2001) 221. [56] F. Guo, R. Nimmanapalli, S. Paranawithana, S. Wittman, D. Griffin, P. Bali, E. O'Bryan, C. Fumero, H.G. Wang, K. Bhalla, Blood 99 (2002) 3419. [57] I.A. McNeish, S. Bell, T. McKay, T. Tenev, M. Marani, N.R. Lemoine, Exp. Cell Res. 286 (2003) 186. [58] B. Jiang, W. Xiao, Y. Shi, M. Liu, X. Xiao, Free Radic. Biol. Med. 39 (2005) 658. [59] S. Giagkousiklidis, M. Vogler, M.A. Westhoff, H. Kasperczyk, K.M. Debatin, S. Fulda, Cancer Res. 65 (2005) 10502.