Targeting Hsp90 with small molecule inhibitors induces the over-expression of the anti-apoptotic molecule, survivin, in human A549, HONE-1 and HT-29 cancer cells

To determine whether the inhibition of Hsp90 with small molecule inhibitors was able to affect survivin expression, we treated human cancer cells with Hsp90 inhibitors geldanamycin and 17-AAG. 17-AAG is a selective Hsp90 inhibitor that exhibited therapeutic activities in various cancers and is currently undergoing phase III clinical trials [17, 21‐24]. To ensure that both of our selected Hsp90 inhibitors were functioning normally at the molecular level, HeLa cells (positive control) were incubated with 17-AAG and geldanamycin for 24 h and Western blot analysis was used to determine the amount the survivin presented in cells. Consistent with other studies, targeting Hsp90 with 17-AAG and geldanamycin (100 nM, 250 nM and 500 nM) reduced the amount of survivin expressed in HeLa cells (Figure 1A) [5]. The effectiveness of Hsp90 inhibitors was subsequently tested with the use of the three-dimensional cell culturing system. Western blot analysis revealed that the expression of survivin was also down-regulated in three-dimensionally cultured HeLa cells treated with 1 μM of 17-AAG for 24 h (Additional file 1).

To determine whether Hsp90 inhibitors were able to down-regulation survivin in other cancer cell lines, human A549, HT-29 and HONE-1 cancer cells were used. In A549 cells, targeting Hsp90 with 17-AAG and geldanamycin slightly induced the baseline expression of Hsp90 as previously reported (Figure 1B) [5]. The same treatment also induced down-regulation of both Akt and phosporylated-Erk in human A549 (p53-mutant) lung carcinoma cells as expected (Figure 1B) [7, 25]. Together, these results indicated that both Hsp90 inhibitors were functioning normally at the molecular level. Surprisingly, targeting Hsp90 with 17-AAG and geldanamycin did not induce survivin down-regulation in A549 cancer cells. Instead, Western blot analysis revealed that survivin expression was induced by Hsp90 inhibitors in A549 cells in a concentration-dependent manner (Figure 1B). In addition, over-expression of survivin was shown in cells treated with 17-AAG in a time-dependent manner (Figure 1C). Since clinical study of 17-AAG revealed that the maximum peak serum level of this compound could reach 2-3 μM, A549 cells were further treated with high concentrations of 17-AAG and the expression of survivin was determined. Over-expression of survivin was also found in A549 cells treated with high concentrations (1 μM and 2 μM) of 17-AAG (Figure 1D). Moreover, Western blot analysis revealed that the expression of survivin was up-regulated in three-dimensionally cultured A549 cells treated with 1 μM of 17-AAG for 24 h (Additional file 1).

The expression of survivin in Hsp90 inhibitors-treated HONE-1 and HT-29 cells was determined by Western blotting. Western blot analysis revealed that both 17-AAG and geldanamycin treatments were able to induce the over-expression survivin in HONE-1 (p53 wildtype) nasopharyngeal carcinoma cells in a concentration-dependent manner (Figure 2A). 17-AAG or geldanamycin treated HT-29 (p53 mutant) colon adenocarcinoma cells also over-expressed survivin (Figure 2B). In contrast, 17-AAG treatment was able to reduce the expression of Akt in a concentration-dependent manner as previously reported (Figures 2A and 2B) [7, 25]. Taken together, these data indicate that the surprising effect of Hsp90 inhibitors were specific to survivin in this study, as 17-AAG and geldanamycin did not modulate the expression of Hsp90, Akt and phosphorylated-Erk in A549 cells and the expression of Akt in HONE-1 and HT-29 cells differently compared to the results of other studies. In three-dimensional culture situations, Western blot analysis revealed that 17-AAG treatment increased the amount of survivin presented in HONE-1 and HT-29 cells (Additional file 1). These results suggest that down-regulation of survivin is not a definite therapeutic effect induced by Hsp90 inhibitors, as over-expression of survivin was observed in cells treated with 17-AAG and geldanamycin in this study.

It has been widely demonstrated that the expression of survivin is tightly regulated during cell cycle and maximized during the G₂/M phase. To determine whether the over-expression of survivin in Hsp90-targeted cells was a downstream result caused by cell cycle arrest at the G₂/M phase, flow cytometric analysis was performed. Interestingly, 17-AAG treatment did not induce a uniform cell cycle response among A549, HONE-1 and HT-29 cells. Experimental results demonstrated that 500 nM of 17-AAG induced cell cycle arrest of A549 cells at the G₂/M phase after 24 h (Figure 3). In contrast, the same treatment induced S-phase arrest and G₀/G₁ phase cell cycle arrest in HONE-1 and HT-29 cells respectively (Figure 3). Taken together, results from both Western blot analysis and flow cytometric analysis indicated that changes in survivin expression in Hsp90-targeted cells were not downstream results caused by cell cycle arrest at the G₂/M phase.

Since 17-AAG induced over-expression of survivin was not an indirect result caused by cell cycle arrest, detailed molecule mechanisms that govern survivin over-expression were investigated. It is widely believed that the amount of protein presented in cells is tightly regulated through the process of gene transcription, protein translation and protein degradation. To determine whether the over-expression of survivin in Hsp90 inhibitor-treated A549, HONE-1 and HT-29 cells was caused by changes at the level of gene transcription, quantitative real-time PCR was performed after 24 h post-treatment. In contrast to the result of Western blot analysis, a dose-dependent decrease in the amount of survivin mRNA transcript in 17-AAG-treated A549 cells was shown by real-time PCR (Figure 4A). A general decrease in the amount of survivin mRNA transcript was also shown in geldanamycin-treated cells (Figure 4A). Cells treated with geldanamycin and 17-AAG showed reduced amount of survivin mRNA by 45% (500 nM) and 75% (500 nM) respectively in A549 cells, as compared to the control (Figure 4A). However, phase-contrast microscopy did not reveal morphological signs of death in Hsp90 inhibitors-treated cells (Figure 4B). Therefore, both geldanamycin and 17-AAG induced decreases in survivin mRNA instead of cell death-induced global decreases at the total intracellular mRNA level. These results also indicated that 17-AAG/geldanamycin treatment increased the level of survivin protein in cells possibly through transcription-independent mechanism in A549 cancer cells. In contrast, 17-AAG and geldanamycin treatment did not decrease the amount of survivin RNA transcript present in HT-29 and HONE-1 cells (Figure 4C and 4D). Instead, the same treatment increased the level of survivin RNA in cells in a dose-dependent manner. Thus, the effect induced by Hsp90 inhibitors on the level of survivin gene transcription seems to depend on the cellular context. Furthermore, the results suggest that increases in survivin protein in response to Hsp90-targeted therapy in HT-29 and HONE-1 cells are possibly due to increases in survivin gene transcription. Taken together, these results indicated that targeting Hsp90 with small molecule inhibitors might interfere with survivin gene transcription differently in different cancer cells.

Since the above results revealed that changes at the level of gene transcription did not contribute to the increase of survivin protein in 17-AAG treated A549 cells, possible post-transcriptional mechanisms such as protein translation and 26S proteasome-dependent protein degradation were investigated in this cell line. To determine whether 17-AAG induced over-expression of survivin through protein translation, translation inhibition study was performed. Human A549 cancer cells were pre-incubated with 500 nM of cycloheximide for three hours and subsequently treated with various concentrations of 17-AAG for 24 hours. Cycloheximide is a translational inhibitor that exerts its effect by interfering with the translocation step in protein synthesis, thus blocking translational elongation. Western blot analysis revealed dose-dependent increases of survivin in the 17-AAG-treated A549 cancer cells (Figure 5A). Interestingly, cycloheximide treatment completely abolished the dose-dependent pattern of survivin expression in 17-AAG-treated cells (Figure 5A). However, the amount of survivin in cycloheximide/17-AAG co-treated cells did not decrease to the baseline level of the negative control (Figure 5A). These results suggest that 17-AAG induces over-expression of survivin in A549 cancer cells partially through the regulation of protein translation. Moreover, other post-translational mechanisms may also contribute to the over-expression of survivin in 17-AAG-treated cells.

To determine whether 17-AAG may interfere with the stability of survivin protein in A549 cells, the rate of survivin protein degradation was determined in cells treated with/without 17-AAG. Cells were pre-incubated with 500 nM of cyclohexmide for one hour and subsequently co-incubated with 17-AAG for various time. Western blot analysis revealed that the degradation rate of survivin protein was slightly slower in 17-AAG treated A549 cells than the untreated cells (Figure 5B). The amount of survivin protein presented in 17-AAG untreated-cells was reduced by 70% after 4 h of protein synthesis inhibition (Figure 5C). In contrast, the amount of survivin presented in 17-AAG-treated cells was reduced by only 20% at the same time point (Figure 5C). Thus, the reduced rate of protein degradation might also contribute to the increased amount of survivin protein presented in 17-AAG treated A549 cancer cells.

Survivin is normally degraded through the proteasomal degradation pathway and it has been shown that the 26S proteasome is responsible for this process [26]. On the other hand, Hsp90 plays a role in the assembly and maintenance of the 26S proteasome [20]. Furthermore, reduced proteasomal activity has been shown in 17-AAG and geldanamycin-treated cells [27, 28]. Here, proteasome-dependent protein degradation pathway was investigated to determine whether proteasome plays a role in the reduced survivin degradation rate in 17-AAG treated A549, HONE-1 and HT-29 cells. In our study, Western blot analysis revealed that 17-AAG reduced the amount of 26S proteasome presented in A549 cells (Figure 6A). In addition, proteasomal activity assay also revealed that the activity of 26S proteasome was reduced by 25-30% in cells treated with 17-AAG (Figure 6B). To demonstrate that inhibition of 26S proteasome would subsequently affect the amount of survivin expressed in cells, the 26S proteasome inhibitor MG-132 was used. Western blot analysis clearly revealed that inhibition of the proteasomal activity by MG-132 induced survivin over-expression in a concentration-dependent manner (Figure 6C). Together with the results from the protein degradation experiment, these data suggest that increased survivin levels in A549 cells cannot simply be attributed to the induction of protein translation. Post-translational mechanisms such as the regulation of the 26S proteasome-dependent protein degradation may also play a role in survivin over-expression in 17-AAG treated A549 cells.

To determine whether proteasomal activity was also affected in 17-AAG treated HT-29 and HONE-1 cells, proteasomal activity of drug-treated cells was measured. Proteasomal activity assay revealed that the activity of 26S proteasome was reduced by ~20% in HT-29 and HONE-1 cells treated with 17-AAG (Figure 7A). In addition, Western blot analysis revealed co-treatment of 17-AAG and the proteasome inhibitor, MG-132, induced synergistic increases in the level of survivin present in both cell lines (Figure 7B). These results indicated that the inhibition of 26S proteasome might also play a role in the up-regulation of survivin in 17-AAG treated HT-29 and HONE-1 cells.

It has been shown that silencing of survivin gene by small interfering RNAs produces supra-additive growth suppression in combination with 17-AAG in human prostate cancer cells [18]. To determine the functional importance of survivin in interfering with drug sensitivity to Hsp90 inhibitors in A549, HONE-1 and HT-29 cells, survivin was down-regulated by siRNA and cell viability was measured by MTT assay. Cells were treated with survivin-specific siRNA oligomer (siR-S) or scramble oligomer (siR-C) for 48 h and sub-subsequently incubated with/without 250 nM of 17-AAG for 24 h. Cells viability assay revealed that A549, HT-29 and HONE-1 cells treated with 250 nM of 17-AAG did not show reduced viability as compared to cells treated with DMSO (Figure 8). In addition, down-regulation of survivin by siR-S significantly reduced cell viability by ~30% in both cell lines as compared to cells transfected with control oligomer, siR-C (Figure 8). Interestingly, siR-S/17-AAG combination treatment further reduced the cell viability of A549, HT-29 and HONE-1 as compared to 17-AAG mono-treatment (Figure 8). Taken together, our results indicate that survivin plays an important role in the sensitivity to the Hsp90 inhibitor, 17-AAG, in our tested cancer cell lines.

The human lung carcinoma (A549), nasopharyngeal carcinoma (HONE-1) and colorectal adenocarcinoma (HT-29) cells were purchased from the American Type Culture Collection (ATCC, Manassas, VA). A549 cells were cultured in RPMI 1640 medium (Gibco, Grand Island, NY), supplemented with 10% fetal bovine serum, penicillin (100 U/mL), streptomycin (100 μg/mL) and L-glutamine (0.29 mg/mL), at 37°C. HONE-1 and HT-29 cells were cultured in RPMI 1640 medium (Gibco, Grand Island, NY), supplemented with 5% fetal bovine serum, penicillin (100 U/mL), streptomycin (100 μg/mL) and L-glutamine (0.29 mg/mL), at 37°C. The antibodies used in this study included a mouse anti-Actin antibody (Santa Cruz Biotechnology, Santa Cruz, CA), a rabbit anti-Survivin antibody (R&D Systems, Minneapolis, MN), a rabbit anti-Akt antibody (Cell Signaling Technology, Danvers, MA) and a mouse anti-26S proteasome antibody (abcam, Cambridge, UK). Hsp90 inhibitors used in this study included: 17-AAG (Calbiochem, Darmstadt, Germany), geldanamycin (Calbiochem, Darmstadt, Germany) and cycloheximide (Calbiochem, Darmstadt, Germany).

Expression level of survivin transcript was determined by real-time reverse transcriptase (RT)-polymerase chain reaction (PCR) using a LightCycler instrument (Roche, Indianapolis, IN). Primers and Taqman probes were designed by Probe Finder™ http://​www.​universalprobeli​brary.​com. Taqman probes were from the Universal Probe Library: survivin and hGAPDH. Specific primers with following sequences were used: survinin forward, 5' GCCCAGTGTTTCTTCTGCTT; Survivin reverse, 5'CCGGACGAATGCTTTTTATG; hGAPDH forward, 5' AGCCACATCGCTCAGACAC and hGAPDH reverse, 5' GCCCAATACGACCAAATCC. The real-time PCR condition was as follows: 1 cycle of initial denaturation at 95°C for 10 min, 45 cycles of amplification at 95°C for 10 s, 60°C for 30 s, and 72°C for 1 s, with a single fluorescence acquisition. hGAPDH gene was used as an internal control. All experiments have been repeated twice.

Cells were lysed with ice-cold lysis buffer (10 mM Tris, 1 mM EDTA, 1 mM DTT, 60 mM KCl, 0.5% NP-40 and protease inhibitors). Total cell lysates, fractions of supernatant or pellet were resolved on 10% and 12% polyacrylamide SDS gels under reducing conditions. The resolved-proteins were electrophoretically transferred to PVDF membranes (Amersham Life Science, Amersham, U.K.) for Western blot analysis. The membranes were blocked with 5% non-fat milk powder at room temperature for two hours, washed twice with PBST (1% Tween) and then incubated with primary antibody for 90 minutes at room temperature. The membranes were washed twice with PBST then subsequently incubated with a horseradish peroxidase-conjugated secondary antibody (dilution at 1:10000, Santa Cruz Biotechnology, Santa Cruz, CA). Immunoreactivity was detected by Enhanced Chemiluminescence (Amersham International, Buckingham, U.K.) and autoradiography. All experiments have been repeated twice.

Cells exposed to various concentrations of 17-AAG and MG-132 for 24 h were washed twice with PBS and lysed with TNESV buffer [50 mM Tris-HCl (pH 7.5), 1% NP40 detergent, 2 mM EDTA, 100 mM NaCl, 10 mM sodium orthovanadate] without protease inhibitors. Cell lysate were assayed for proteasome chymotrypsin activity using the synthetic fluorogenic peptide chymotrypsin substrate, N-Succinyl-Leu-Leu-Val-Tyr-AMC. Fluorescent signals were measured with a 96-well plate reader with an excitation wavelength of 380 nm and emission wavelength of 460 nm. All experiments have been performed as triplicate and repeated twice.

Target-validated siRNA oligos (Santa Cruz Biotechnology, Santa Cruz, CA) were transfected into cells using the Lipofectamine-2000 reagent (Invitrogen, Carlsbad, CA). Briefly, cells were seeded onto 96-well plates or chamber-slides, and cultured overnight in 100 μl of antibiotic-free RPMI media. siRNA oligomers (8 pmol in 0.4 μl) were diluted in 25 μl of Opti-MEM^® I medium (Invitrogen, Carlsbad, CA) without serum, and then mixed with 0.2 μl of Lipofectamine-2000 transfection reagent for 25 min at room temperature. Cells were overlaid with the transfection mixture, and incubated for various times.

Cells seeded onto 96-well plates were transfected with/without survivin-specific siRNA oligomer for 48 h and subsequently treated with 17-AAG for 24 h. 25 μl of MTT (5 mg/mL) was added to each sample and incubated for 4 hours, under 5% CO₂ and 37°C. 100 μl of lysis buffer (20% SDS, 50% DMF) was subsequently added into each sample and further reacted for 16 hours.