Background
Tight regulation of apoptosis is essential for maintaining tissue homeostasis. Altering this balance in favor of apoptosis resistance is a common feature in cancer cells [
1], and overcoming cell death barriers is a primary goal of many chemotherapeutic treatments used in the clinic today. One proven strategy to induce apoptosis is through blocking components of the ubiquitin proteasome system (UPS).
The ubiquitin proteasome system is the major cellular pathway for regulated protein turnover. Ubiquitination and protein degradation allows the cell to respond quickly to extra- or intracellular signals, thereby maintaining cellular homeostasis. The addition of multiple ubiquitin moieties to a targeted protein "tags" the protein for degradation by the 26S proteasome. The mechanism by which multiple ubiquitin molecules are added to targeted substrates can be depicted as a three-step process. First, the ubiquitin activating enzyme (E1) catalyzes the formation of a C-terminal thiol ester in an ATP dependent reaction. The ubiquitin is then transferred to an E2 ubiquitin-conjugating enzyme via the formation of a thiol ester bond. Finally, the ubiquitin ligase enzymes (E3) catalyze the transfer of ubiquitin from the E2 to a lysine residue on the targeted protein via an isopeptide linkage. A polyubiquitinated protein is then recognized by the 19S regulatory subunit of the 26S proteasome and degraded by the 20S core particle.
Several strategies to block UPS components in the treatment of cancer have been devised. The initial approach was the development of the 26S proteasome inhibitor Velcade (bortezomib), which has been shown to induce apoptosis and is approved to treat multiple myeloma [
2]. Subsequently, investigators have sought to promote the stability of numerous apoptosis-associated proteins, such as the p53 interaction with the E3 ligase Mdm2 [
3]. Preventing proteasomal degradation of active caspases is another area that has shown promise. Polyubiquitination and subsequent degradation of active caspases by the inhibitor of apoptosis protein (IAP) family can downregulate the apoptotic response. In some systems, IAP self-ubiquitination and degradation is required for full apoptosis activation [
4], while in others free Smac/DIABLO released from the mitochondria can bind IAPs and inhibit their anti-apoptotic activity [
5]. The ability of Smac mimetic compounds that target IAP activity to sensitize cancer cells to chemotherapeutic agents is currently being tested in clinical trials [
6].
There are two major apoptotic-inducing pathways, intrinsic (mitochondrial) and extrinsic (death receptor-mediated). The involvement of ubiquitination in the regulation of apoptosis has largely been assessed in the intrinsic apoptotic pathway (i.e., Mdm2/p53, Smac/IAP). In contrast, much less is known regarding the ubiquitination machinery or ubiquitin substrates that regulate cell-extrinsic apoptosis. Death receptor-mediated apoptosis is induced by the binding of death ligands to their corresponding death receptors at the plasma membrane [
7]. Death ligands are members of the tumor necrosis factor (TNF) family of ligands such as Fas Ligand (FasL) and TNF related apoptosis inducing ligand (TRAIL). Clustering of death receptors upon ligation allows for the formation of the death inducing signaling complex (DISC) [
8]. Procaspase-8 present at the DISC, and possibly in other high molecular weight aggregates, is autocatalytically processed to form an active p10/p18 homodimer [
9]. Active caspase-8 can cleave effector caspases-3 and -7, as well as other death substrates, fully inducing the downstream apoptotic pathway.
Here, using a high throughput siRNA screen we identified two E3 ubiquitin ligases, Siah2 (Seven in absentia homologue) and SH3RF1 (SH3 domain containing RING finger 1, also known as POSH for plenty of SH3 domains), as negative regulators of death receptor mediated apoptosis through the modulation of caspase-8 activity. We also defined a physical interaction between these two E3s suggesting Siah2 and POSH may function through shared or similar signaling pathways to regulate extrinsic apoptosis.
Conclusion
The primary finding presented here is the ability of two E3 ubiquitin ligases to weaken signaling through the death receptor apoptotic pathway. By utilizing RNA interference technologies we have successfully established Siah2 and POSH as two UPS components that can reduce caspase activity in response to death ligands in prostate cancer cells.
Siah2 is a homolog of the
Drosophila Sina gene, a critical component in photoreceptor development. The protein contains an N-terminal RING-finger domain which provides E3 ubiquitin ligase activity. The RING-domain is followed by two zinc finger-domains and a C-terminal substrate binding-domain [
14]. Current literature provides evidence for Siah2 function in the hypoxic response. One proposed mechanism is by regulating HIF-1α protein levels by the ubiquitination and subsequent degradation of PHD3 (prolyl hydroxylating domain-containing 3), a prolyl hydroxylase that controls HIF-1α accumulation in hypoxic environments [
15]. A second possible hypoxia regulatory point is through the degradation of HIP2K (homeodomain-interacting protein kinase 2), a member of the HIPK family of proline-directed kinases [
16]. Under normoxia, the ability of Siah2 to modulate cell proliferation in lung cancer cells has been detailed [
17]; a finding we have confirmed in prostate cancer cells (Additional File
3).
POSH (plenty of SH3s) is a large scaffold molecule containing an N-terminal RING-finger domain followed by four SH3-domains and a region known to bind Rac1-GTP. The RING-domain has been shown to display ubiquitin ligase activity and is thought to regulate trans-Golgi network transport [
18]. The scaffolding function of POSH is most closely associated with c-Jun N-terminal kinase signaling [
19]. Although multiple investigators have reported transient POSH overexpression can induce apoptosis [
19,
20], we did not observe death in our prostate cancer models upon lower, stable expression (data not shown). Conversely, we have shown that down regulation of POSH can sensitize cells to TRAIL-dependent apoptosis induction.
Providing mechanistic data to explain Siah2-dependent apoptosis regulation is ongoing. Initial Siah2 characterizations in the death receptor pathway have revealed clues to possible mechanisms. As shown in Figure
4a, the E3 activity of Siah2's RING-domain is required for its anti-apoptotic function, suggesting the E3 ligase enzymatic activity is responsible for the degradation of a pro-apoptotic protein. Proposed Siah2 substrates include TRAF2, PHD3 and HIPK2 [
15,
16,
21]. We have failed to confirm the interaction with PHD3 or TRAF2 or Siah2-mediated regulation of NF-κB activity in our prostate models (data not shown). However, the recently identified pro-apoptotic, Siah2-specific substrate HIPK2 is a promising candidate. HIPK2 is a serine/threonine kinase known to phosphorylate p53 at serine 46, thereby promoting p53 acetylation and p53-dependent gene expression [
22]. Inhibition of HIPK2 has been shown to protect lung cancer cells against UV-induced apoptosis while overexpression of HIPK2 sensitizes cells to UV-induced apoptosis and decreases cell proliferation [
22,
23]. Together, these data suggest a possible scenario in which Siah2 targeting of HIPK2 protects cancer cells from apoptosis induction while inhibition of Siah2 could increase HIPK2 abundance thus sensitizing cancer cells to cell death agents.
Regulation of cell death by Siah2 was found to be selective, but not specific for the death receptor pathway. This finding, along with data demonstrating Siah2 silencing induces caspase-8 activity without altering procaspase-8 processing, suggested to us that Siah2 was acting further downstream in the death receptor pathway. Cellular localization analysis of Siah2 revealed a predominately cytoplasmic staining pattern, consistent with previous reports [
16,
24]. High levels of Siah2 co-localization with the 20S core proteasome was also observed and may provide insight into Siah2's regulation of cell death. Interestingly, several E3-ligases such as Hul5, Parkin, Ubr1, and VHL are known to localize to complexes in association with the 26S proteasome [
25], adding validity to this novel finding.
In contrast to Siah2, the ability of POSH to regulate apoptosis induction in response to TRAIL was independent of E3 activity, suggesting protein-protein interactions (possibly mediated through one of four SH3-domains) are important for this activity. Surprisingly, it has been demonstrated that POSH in an E3-independent fashion interacts with Siah2 and can stabilize Siah proteins [
13]. We have successfully confirmed a physical interaction between E3 ligases Siah2 and POSH. By utilizing a yeast-two-hybrid mapping approach we have defined the spacer region of POSH, more specifically the RPxAxVxP Siah2 substrate binding motif encompassing amino acids 601-607, to be the site of Siah2 binding. POSH has been shown to bind both Siah1 and Siah2 suggestive of possible overlapping, POSH-dependent functions of these two E3s. The RPxAxVxP motif within POSH is highly conserved from
Xenopus to humans suggesting a Siah-POSH binding event is critical for the proper function of a highly conserved biochemical pathway. Taken together, the ability of Siah2 to interact with such a large signaling molecule such as POSH, as well as its capacity to dimerize, could provide opportunities for Siah2 to be recruited to key protein signaling complexes where it can carry out its E3 ligase function. For example, monomeric Siah2 could be recruited to signaling complexes via a degradation-independent interaction with POSH. Free monomeric Siah2 could then associate with the complex through dimerization. POSH, through one of four SH3 domains, could recruit possible target substrates allowing Siah2 to carry out its polyubiquitination/degradation function.
As Siah2 and POSH suppress cell death, their upregulation may be advantageous in cancer cells. We, therefore, speculated that these E3s would be overexpressed in a subset of cancers. Oncomine analysis revealed Siah2 to be significantly overexpressed in both B-cell acute lymphoblastic leukemia [
26] and squamous non-small-cell lung carcinoma [
27], while POSH is one of the most frequently overexpressed genes in prostate cancer [
28‐
30]. At this point, however, there is no evidence that either gene is capable of cellular transformation.
Together, these data suggest that development of second generation proteasome inhibitors that block E3 ubiquitin ligases, such as Siah2 and POSH, could be beneficial for cancer treatment. Recently, major advances in targeted E3 therapies for the intrinsic apoptotic pathway have been made. Most notably, generation and characterization of Smac mimetic compounds that inhibit IAP function. Smac mimetic compounds have been shown to sensitize cancer cells to chemotherapeutic agents as wells as strongly enhance the anti-tumor activity of death ligands such as TRAIL in vivo [
6]. It remains to be tested if targeting of the extrinsic apoptotic pathway via Siah2 or POSH inhibition in combination with current intrinsic targeting compounds is a promising cancer therapeutic strategy.
Methods
Cell Culture and Reagents
All prostate cancer cell lines were obtained from ATCC (Manassas, VA) and grown in DMEM containing 10% FBS. Velcade (Millenium Pharmaceuticals, Cambridge, MA) was donated by the University of Kentucky, Markey Cancer Center Pharmacy. The GST-TRAIL fusion protein was purified by affinity chromatography as described [
31]. Fas ligand was expressed in cells as described [
10]. Staurosporine, doxorubicin and thapsigargin were purchased from EMD Biosciences (San Diego, CA).
Cell Viability, Apoptosis and Proliferation Assays
Cell viability was measured using a colorimetric MTT assay. Media was removed from the cells and a 1 mg/ml solution of thiazolyl blue tetrazolium bromide (MTT) (Alfa Aesar, Ward Hill, MA) in 1 × PBS was added. The samples were read on a spectrophotometer at 570 nm minus 690 nm. All assays were conducted in replicates of four to six. Annexin V staining was carried out as described by the manufacturer (Invitrogen, Carlsbad, Ca). Cells were immediately analyzed by flow cytometry (FACSCalibur, BD Immunocytometry Systems, San Jose, CA). CellQuest Pro software (BD Immunocytometry Systems) was used to quantify percentage of annexin V positive and viable cells in 20,000 gated events.
BrdU labeling and flow cytometric analysis was carried out as previously described [
32]. The percentage of BrdU positive cells (10,000 gated events) was determined using CellQuest software (BD). All assays were carried out in replicates of three. All statistical assessments were made using a Student's t-test.
Western Blot Analysis
Cell lysates were prepared in ubiquitin protein extraction buffer (UPEB; 150 mM NaCl, 50 mM Tris-HCl (pH 7.5), 5 mM EDTA, 1% (v/v) NP-40, 0.1% (w/v) SDS, 0.5% (w/v) sodium deoxycholate). Western blot analysis was carried out as previously described [
33]. Antibodies were obtained from the following sources: β-actin (Sigma, St. Louis, MO), caspase-8, 1C12 (Cell Signaling Technology, Danvers, MA), myc, 9B11 (Cell Signaling Technology), cytochrome-C (Santa Cruz Biotechnology, Santa Cruz, CA), PDI (Stressgene, Ann Arbor, MI).
Caspase Activity
In vivo caspase-8 activity was measured with the luminescent-based Caspase-Glo 8 Assay kit while caspase-3/7 activity was measured by the addition of the luminescent-based Caspase-Glo 3/7 Assay kit (Promega, Madison, WI). At 16 hr post-treatment, three wells in each treatment group were pre-incubated for 30 min with the cell permeable caspase-8 inhibitor Z-IETD-fmk or the pan caspase inhibitor Z-VAD fmk (25 mM, R&D Systems, Minneapolis, MN). Each treatment group consisted of seven wells, four without fmk inhibitor and three with the inhibitor. The Caspase-Glo reagent was added to the cells and activity measured with a Lmax luminometer (Molecular Devices Corporation, Sunnyvale, CA). Data was analyzed by SOFTmax Pro (Version 1.1L). Specific caspase activity was calculated by subtracting the activity derived from cells treated with Z-IETD/VAD-fmk.
E3-ligase siRNA Screen
The siARRAY E3 ubiquitin-ligase library containing 239 total E3 specific siRNAs was purchased from Dharmacon (Lafayette, CO). The predispensed dehydrated siRNA was first rehydrated with a solution of DMEM containing DharmaFECT transfection reagent. DU145 cells (10,000) were seeded directly into the 96-well plates containing the rehydrated siRNA. 48 hours after siRNA transfection, cells were treated with GST-TRAIL (50 ng/mL) for 16 hours. The addition of 50 mM Velcade in combination with TRAIL served as a positive control. Caspase-8 activity was measured as previously discussed by the addition of Caspase-GLO 8 pro-luminescent reagent. All E3 ligase siRNA experiments were performed in triplicate.
Quantitative Reverse-transcriptase PCR
DU145, PC-3 and PPC-1 cells were seeded into 6-well tissue culture plates. Cells were transfected with non-targeting siRNA or siRNA specific for Siah2 or POSH. 48 hours after siRNA transfection, RNA was harvested using the RNeasy kit (Qiagen, Valencia, CA). cDNA was generated using SuperScript II Reverse Transcriptase (Invitrogen). qRT-PCR was carried out using primer sets for 18S, Siah2 or POSH in the presence of FastStart Universal SYBR Green Master Mix (Roche, Nutley, NJ). PCR was performed and analyzed using a 4300 Real Time PCR System (Applied Biosystems, Foster City, CA). Percent mRNA remaining was determined by the comparative CT method (ΔCT) for relative quantification.
E3 Ubiquitin-Ligase Cloning and Expression
Wild-type human Siah2 (NM_0056067) and RING mutant Siah2 (H98A, C101A) were generated by PCR, cloned into a retroviral pBabe puro expression vector (Addgene, Cambridge, MA) and verified by sequence analysis. Retrovirus was generated as previously described [
34] and added to PC-3 and PPC-1 cells for 24 hours. Cells were selected by treatment with 2 μg/mL of puromycin (EMD Biosciences, San Diego, CA). Selection was carried out for one week and expression was confirmed by Western blotting. Similarly, wild-type POSH (NM_020870) and RING mutant POSH (H28A, C30A) were generated.
Sub-cellular Fractionation
Cytoplasmic, microsomal and mitochondrial fractions were obtained from PPC-1 cells stably overexpressing rmSiah2-myc using the Qproteome Mitochondria Isolation Kit (Qiagen). Cells were first treated +/- GST-TRAIL (100 ng/mL) for 2 hours. Cells were collected and protein harvested in provided lysis buffer. The cytoplasmic fraction was obtained by centrifuging total cell lysate at 1000 × g for 10 min at 4°C and the collecting supernatant. Nuclei and unbroken cells were collected by pelleting lysate in the disruption buffer at 1000 × g for 10 min at 4°C. A microsomal fraction was obtained by centrifugation of protein lysate (in disruption buffer) at 6000 × g for 10 min at 4°C. Finally, the mitochondrial fraction was collected in provided mitochondrial purification buffer by sequential centrifugation at 14,000 × g for 15 min then 8000 × g for 10 min both at 4°C. The purified mitochondrial pellet was resuspended in the provided mitochondrial storage buffer. 50 μg of each fraction was added to SDS loading buffer and Western blot analysis was performed.
Immunocytochemistry
PPC-1 cells stably expressing rmSiah2-myc were seeded into 4-well chamber slides. Cells were fixed with 4% paraformaldehyde for 20 minutes and permeabilized with saponin (1 mg/mL) in Hank's Buffered Saline Solution for 15 minutes. Following blocking in 2% BSA for 30 minutes, primary antibodies were added for 1 hour followed by PBST washing. Fluorescently conjugated secondary antibodies were added for 45 minutes. Slides were mounted in Vectashield + DAPI (Vector Laboratories, Burlingame, CA) and visualized by a Leica TSP SP5 confocal microscope. Antibodies were obtained from the following sources: myc, 9B11 (Cell Signaling Technology), 20S proteasome core subunits (Calbiochem, San Diego, CA).
Myc-POSH Immunoprecipitation
Cells expressing myc-tagged constructs were incubated at 37°C for 48 hrs. Protein was harvested in IP buffer (10 mM Tris pH 7.4, 1% Triton X-100, 0.5% NP-40, 150 mM NaCl, 20 mM NaF, 1 mM EDTA) and incubated on ice for 30 min. Cell debris was pelleted by centrifugation at 10,000 × g for 15 min at 4°C. 50 μl μMACS anti-myc MicroBeads (Miltenyi Biotech, Auburn, CA) was added to the lysate and incubated on ice for 30 min. myc-POSH was purified by passing cell lysate + microbeads through a μMACS column attached to the μMACS magnetic Separator. Column was washed with 800 μl IP buffer and protein was eluted by adding 20 μl pre-heated (95°C) elution buffer (provided), incubating for 5 min and adding another 50 μl elution buffer. Eluted protein was subjected to SDS-PAGE followed by Western blot analysis.
Yeast Two-Hybrid
Full length POSH was cloned into pGBKT7 (Clontech, Mountain View, CA) as a Gal4-DNA-binding domain fusion. Full length Siah2 as cloned into pACT2 (Clontech) as a Gal4-Activiation domain fusion. The fusion constructs or empty vector controls were co-transformed into the AH-109 yeast strain. Briefly, a single colony of AH-109 yeast (Clontech) was inoculated into 5 mL liquid YPD and incubated overnight at 30°C. The overnight culture was inoculated into 50 mL liquid YPD and allowed to grow to a cell density of 2 × 107 cell/mL. The 50 mL culture was harvested by centrifugation at 3000 × g for 5 min. Yeast were resuspended in 1 ml of 100 mM LiAc. Cells were pelleted at 10,000 rpm in a table top centrifuge for 5 sec and the LiAc was removed. Yeast cells were resuspended to a final cell density of 2 × 109 cells/mL in 100 mM LiAc. 50 μl of resuspended yeast were used per co-transformation. A transformation mix containing 240 μl PEG (Bainbridge Island, WA) (50% w/v), 36 μl 1.0 M LiAc, 25 μl single stranded DNA (2.0 mg/mL), plasmid DNA (2 μg per construct) and ddH2O up to 350 μl was made. The transformation mix was added directly to the resuspended yeast, vortexed and incubated at 30°C for 30 min. Yeast were heat shocked at 42°C for 25 min and microcentrifuged at 8,000 rpm for 15 sec. Transformation mix was removed and yeast were resuspended in 400 μl ddH2O. Transformed yeast were plated onto selective agar plates (200 μl per plate).
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
PC carried out the E3 screen and subsequent Siah2 and POSH characterization and drafted the manuscript. MF carried out the JNK phosphorylation studies. SS conceived of the study and participated in its design and coordination. All authors read and approved the final manuscript.