Targeted inhibition of heat shock protein 90 disrupts multiple oncogenic signaling pathways, thus inducing cell cycle arrest and programmed cell death in human urinary bladder cancer cell lines

Abstract

Background: Geldanamycin (GA) can be considered a relatively new component with a promising mode of action against human malignancies. It specifically targets heat shock protein 90 (Hsp90) and interferes with its function as a molecular chaperone.

Methods: In this study, we have investigated the effects of geldanamycin on the regulation of Hsp90-dependent oncogenic signaling pathways directly implicated in cell cycle progression, survival and motility of human urinary bladder cancer cells. In order to assess the biological outcome of Hsp90 inhibition on RT4 (grade I) and T24 (grade III) human urinary bladder cancer cell lines, we applied MTT assay, FACS analysis, Western blotting, semi-quantitative (sq) RT-PCR, electrophoretic mobility shift assay (EMSA), immunofluorescence and scratch-wound assay.

Results: We have herein demonstrated that, upon geldanamycin treatment, bladder cancer cells are prominently arrested in the G1 phase of cell cycle and eventually undergo programmed cell death via combined activation of apoptosis and autophagy. Furthermore, geldanamycin administration proved to induce prominent downregulation of several Hsp90 protein clients and downstream effectors, such as membrane receptors (IGF-IR and c-Met), protein kinases (Akt, IKKα, IKKβ and Erk1/2) and transcription factors (FOXOs and NF-κΒ), therefore resulting in the impairment of proliferative -oncogenic- signaling and reduction of cell motility.

Conclusions: In toto, we have evinced the dose-dependent and cell line-specific actions of geldanamycin on cell cycle progression, survival and motility of human bladder cancer cells, due to downregulation of critical Hsp90 clients and subsequent disruption of signaling -oncogenic- integrity.

Figure 1

Reduction of bladder cancer cell cycle activity in response to geldanamycin. (A) Representative FACS analysis of RT4 and T24 human urinary bladder cancer cells. Total cell sub-population percentages at each cell cycle phase are shown. FACS experiments were repeated three times. (B) Detection of protein (Cdk4) and mRNA expression (Cyclin D1) levels of the Cyclin/Cdk complex -critical- components responsible for the G1→S transition of cell cycle in RT4 and T24 cells. Left panel: images of protein (top) and RNA transcript (bottom) expression profiles after geldanamycin administration. Right panel: protein (top) and RNA transcript (bottom) densitometric quantification bars, denoting the drug-induced alterations of Cdk4 and Cyclin D1 expression levels compared to control conditions, using Actin (upper left panel) and GAPDH (lower left panel) as protein and gene of reference, respectively. (C) pRb and E2F1 protein expression profiles in response to 24 hours of geldanamycin treatment in RT4 and T24 bladder cancer cells. Left panel: images of protein expression profiles after geldanamycin administration. Right panel: protein densitometric quantification bars, denoting the drug-induced alterations of pRb (top) and E2F1 (bottom) expression levels compared to control conditions, using Actin (left panel) as protein of reference. Western blotting and sqRT-PCR experiments were carried out three times, one of which is respectively shown here. Standard deviation values are depicted as error bars on top of each value.

Figure 2

Geldanamycin-induced cytotoxicity. MTT toxicity assays were performed in RT4 and T24 cells after 24- (left) and 48-hour (right) treatment with increasing concentrations of geldanamycin. All assays were carried out three times. Standard deviation values are depicted as error bars on top of each value, while asterisks denote statistical significance of the observed averaged differences. P < 0.05 .

Figure 3

Caspase-dependent activation of programmed cell death. Expression and proteolytic processing patterns of apoptosis-related critical proteins in response to 24 hours of geldanamycin treatment in RT4 and T24 bladder cancer cells. Western blottings of members of the caspase signaling cascade (caspase-8, -9 and −3) and the caspase repertoire substrates PARP and Lamin A/C are shown. Actin has been used as protein of reference. Data are obtained from three different experimental trials, one of which is presented here.

Figure 4

Autophagy-mediated programmed cell death. Characteristic images of RT4 and T24 bladder cancer cell activation of autophagy in response to 1 μM of geldanamycin treatment (scale bars: 10 μm). Cells were stained with Lysotracker Red that specifically recognizes active lysosomes (white arrows), therefore allowing the detection of an autophagic cell death program. Images were taken under a Nikon EZ-C1 confocal microscope. All cell staining experiments were conducted three times, while a representative image collection is shown here.

Figure 5

Hsp90 harm after exposure of bladder cancer cells to geldanamycin. Western blottings (upper left panel) of critical members of the eukaryotic chaperosome (Hsp90 and Hsp70) after 24-hour geldanamycin administration in RT4 and T24 bladder cancer cells. A cell line-specific Hsp90 proteolytic processing is observed, as documented by the production of a ~65 kDa Hsp90-like protein fragment. Detection of Hsp90α and Hsp90β mRNA levels via sqRT-PCR (lower left panel), proving the absence of Hsp90 transcriptional regulation in response to the drug. Right panel: protein (top and middle) and RNA transcript (bottom) densitometric quantification bars, denoting the drug-induced alterations of Hsp90/Hsp70/α-Tubulin and Hsp90α/Hsp90β expression levels compared to control conditions, using Actin (upper left panel) and GAPDH (lower left panel) as protein and gene of reference, respectively. Western blotting and sqRT-PCR experiments were executed three times, with a respective characteristic image collection presented here. Standard deviation values are depicted as error bars on top of each value.

Figure 6

Oncogenic signal transduction downregulation (I). Geldanamycin-induced Hsp90 inhibition results in functional impairment of Akt-dependent signaling in human urinary bladder cancer cells. (A) Western blotting experiments reveal drastic downregulation of Akt signaling activity, evolving through the IGF-IR/Akt/IKK signal transduction axis, as evidenced by the eradication of both total and constitutively phosphorylated protein forms of all examined Hsp90 protein clients (IGF-IR, Akt, IKKα and IKKβ), in RT4 and T24 bladder cancer cell lines. Actin was used as protein of reference. All Western blottings were carried out three times, with a representative image collection shown here. (B) Protein densitometric quantification bars, denoting the drug-induced expression level alterations of the IGF-IR/Akt/IKK signaling axis components, and their respective phosphorylated forms, compared to control conditions, using Actin (A) as protein of reference. Standard deviation values are depicted as error bars on top of each value. (C) Immunofluorescence images illustrating the cellular localization of Hsp90 (red color) and Akt kinase (green color) in RT4 and T24 bladder cancer cells, in the presence or absence of geldanamycin. Upon drug administration (10 μΜ), a pronounced depletion of cytoplasmic Akt protein is detected, in contrast to what is observed under control conditions. The bright orange color (white arrows) indicates the cellular areas where Akt and Hsp90 are likely co-localized (associated), in the absence of drug conditions. Images were taken under a Nikon EZ-C1 confocal microscope. All immunofluorescence experiments were conducted three times, while a characteristic image collection is presented here (scale bars: 10 μm).

Figure 7

NF-κΒ inactivation after exposure of human bladder cancer cells to geldanamycin. NF-κB transcriptional capacity is diminished upon 24-hour geldanamycin administration in RT4 and T24 bladder cancer cells. (A) Upper panel: Western blottings displaying NF-κB protein expression levels and compartmentalization responses to drug treatment in both cytoplasmic and nuclear protein extracts of RT4 and T24 cells. Protein sample quantification of cytoplasmic and nuclear extracts was evaluated using Actin and Lamin A/C as respective (compartment-specific) reference markers, while the absence of detectable NF-κB in nuclear extracts derived from drug-treated cells ensured samples’ purity. Lower panel: cytoplasmic and nuclear NF-κB densitometric quantification bars, denoting its drug-induced compartmentalization change compared to control conditions, using Actin and Lamin A/C as respective proteins of reference. Standard deviation values are depicted as error bars on top of each value. Experiments were carried out three times, while characteristic Western blotting results are presented here. (B) Electrophoretic mobility shift assay (EMSA) conducted on RT4 and T24 nuclear protein extracts. Binding assays were performed using NF-κB-specific and 5´ -biotin labeled, double-stranded (annealed), oligonucleotides. EMSA experiments were repeated three times, one of which is shown here. (C) Left panel: transcriptional expression profiles of four critical anti-apoptotic NF-κB target genes (Survivin, cIAP-1, cIAP-2 and XIAP), in response to 24-hour geldanamycin administration, in RT4 and T24 bladder cancer cells. Right panel: RNA transcript densitometric quantification bars, denoting the drug-induced expression level alterations of the NF-κB target genes examined, compared to control conditions, using GAPDH (left panel) as gene of reference. All sqRT-PCR reactions were carried out three times, while a characteristic assembled profiling is presented here. Standard deviation values are depicted as error bars on top of each value.

Figure 8

Oncogenic signal transduction downregulation (II). Left panel: Western blotting experiments presenting protein expression levels of total and constitutively phosphorylated FOXO transcription factors and p44/42 MAP kinases, in RT4 and T24 bladder cancer cells, upon geldanamycin administration for 24 hours. Right panel: protein densitometric quantification bars, denoting the drug-induced expression level alterations of the FOXO and MAPK family members examined, and their respective phosphorylated forms, compared to control conditions, using Actin (left panel) as protein of reference. All experiments were repeated three times, with a typically obtained image collection shown here. Standard deviation values are depicted as error bars on top of each value.

Figure 9

Impairment of bladder cancer cell motility and invasion mechanisms. (A) Top panel: total and phosphorylated protein expression levels of the hepatocyte growth factor receptor (c-Met), upon 24-hour incubation of RT4 and T24 bladder cancer cells with increasing doses of geldanamycin, as shown by Western blotting. Bottom panel: protein densitometric quantification bars, denoting the drug-induced downregulation of total and phosphorylated c-Met expression levels, compared to control conditions, using Actin (top panel) as protein of reference. All Western blottings were performed three times, with a characteristic image collection presented here. Standard deviation values are depicted as error bars on top of each value. (B) Scratch-wound assays were carried out in RT4 and T24 bladder cancer cells under control conditions or treatment with 10 μM of geldanamycin for 24 hours. Observations were made under an inverted microscope (Carl Zeiss Axiovert 25) and pictures were taken at 20x magnification. All assays were repeated three times, while a representative image collection is shown here.

Figure 10

Hsp90 inhibition eliminates the “hallmark traits” of cancer. The data presented here clearly demonstrate that the geldanamycin-induced inhibition of Hsp90 results in the combinatorial and multi-step impairment of critical oncogenic pathways likely implicated in the “hallmark traits” of bladder cancer initiation and progression.