Ex vivo assays to predict enhanced chemosensitization by hyperthermia in urothelial cancer of the bladder

Abstract

Introduction: Bladder cancer (urothelial carcinoma) is a common malignancy characterized by high recurrence rates and intense clinical follow-up, indicating the necessity for more effective therapies. Current treatment regimens include intra-vesical administration of mitomycin C (MMC) for non-muscle invasive disease and systemic cisplatin for muscle-invasive or metastatic disease. Hyperthermia, heating a tumor to 40-44°C, enhances the efficacy of these chemotherapeutics by various modes of action, one of which is inhibition of DNA repair via homologous recombination. Here, we explore whether ex vivo assays on freshly obtained bladder tumors can be applied to predict the response towards hyperthermia.

Material and methods: The cytochrome C release assay (apoptosis) and the RAD51 focus formation assay (DNA repair) were first established in the bladder cancer cell lines RT112 and T24 as measurements for hyperthermia efficiency, and subsequently tested in freshly obtained bladder tumors (n = 59).

Results: Hyperthermia significantly increased the fraction of apoptotic cells after cisplatin or MMC treatment in both RT112 and T24 cells and in most of the bladder tumors (8/10). The RAD51 focus formation assay detected both morphological and numerical changes of RAD51 foci upon hyperthermia in the RT112 and T24 cell lines. In 64% of 37 analyzed primary bladder tumor samples, hyperthermia induced similar morphological changes in RAD51 foci.

Conclusion: The cytochrome C assay and the RAD51 focus formation assay are both feasible on freshly obtained bladder tumors, and could serve to predict the efficacy of hyperthermia together with cytotoxic agents, such as MMC or cisplatin.

Fig 1. Survival responses of tumor cell lines to MMC and cisplatin.

A)-C) Colony survival assays of bladder cancer cell lines RT112 and T24. Cells were treated without (blue) or with hyperthermia (red), and exposed to either A) irradiation up to 6 Gy, B) MMC up to 0.4 μg/ml and C) cisplatin up to 6 μM. D) and E) Graphs presenting percentages of cells with cytochrome C-release 24 hours after treatment with hyperthermia and indicated doses of D) MMC or E) Cisplatin. Cells were treated with caspase-inhibitor Q-VD-OPh to prevent full completion of the apoptosis-program. The arrow in the Fig embedded in panel D) T24 indicates a cell scored as Cytochrome C-release-positive. At least 500 cells were analyzed for each condition, bars represent proportion ± standard error of the proportion. Asterisk indicate significance as determine by Students’ t-tests: * p<0.05; ** 0.001<p<0.01; *** p<0.001. Exact p-values can be found in S2 Table.

Fig 2. Cytochrome C release in dissociated bladder tumors.

Graphs representing the percentage of Cytochrome-C-positive cells in dissociated bladder tumors. Tumor cells were cultured for 1–2 days before being subjected to hyperthermia in combination with 40 μg/ml MMC or 40 μg/ml cisplatin. After treatment, medium was refreshed and the caspase-inhibitor Q-VD-OPh was added. After 24 hours, cells were fixed, stained and scored for Cytochrome C-release positivity. A)-H) all represent different tumors. N.d. = not determined. The number of cells analyzed are indicated in each bar, which represent proportion ± standard error of the proportion. Asterisk indicate significance as determine by Students’ t-tests: * p<0.05; ** 0.001<p<0.01; *** p<0.001. Exact p-values can be found in S3 Table.

Fig 3. Hyperthermia induces distinct changes in RAD51 foci in RT112 and T24 cell lines.

A) Immunoblots of samples treated with or without hyperthermia, probed for BRCA2 and PARP-1 as a loading control. B) RT112 cells were treated with or without hyperthermia and subsequently irradiated with the indicated dose. Cells were then incubated at 37°C, and fixed at indicated time-points afterward irradiation. EdU was present for one hour prior to fixation to identify S phase cells. Cells were stained for EdU and RAD51. Pictures are a maximum projection of a Z-stack, and the dotted lines denote the perimeter of EdU-positive nuclei. C) The 2-dimensional graph expresses a quantified classification of RAD51 foci. Each dot in the graph represents one cell, plotted based on the number of RAD51 foci in its nucleus (y axis) and their integrated density (x axis); a derivate of intensity and area per nucleus. Cells without foci are presented in the embedded figure noting ‘0 foci’. D) and E) As panel B and C, but for T24 cells.

Fig 4. Classification of RAD51-focus morphology upon hyperthermia in human bladder cancer.

A) Two hours after DNA-damage induction irradiation or zeocin, tumors were fixed and prepared for immunohistochemistry and subsequently stained for GMNN (green) and RAD51 (red). Pictures presented are examples of both undissociated and dissociated tumors exhibiting the two distinct morphological responses of RAD51-foci. Panels below the blue thermometer received no treatment before damage induction; panels below red thermometers were first treated with hyperthermia. The images are maximum projections of Z-stacks from representative areas of the tumors. B) Graph representing the frequency distribution of the morphological response of RAD51-focus upon hyperthermia treatment. The white part of the bar represents foci which have become smaller and more dispersed upon hyperthermia, the black part represents foci that did not change in morphology. Each bar contains the number (n) of analyzed tumors below. C) Frequency distribution based on the tumor preparation step (undissociated or dissociated) or on the method of damage induction. Two tumors are classified in both methods of tumor preparation, because both were performed on the same tumor, and had the same result. No differences were found in the foci-response distribution between the technical methods (Fisher’s Exact test, p-values stated in the Fig). D) Distribution of RAD51-foci response upon hyperthermia based on clinical data (available for 36 tumors). No differences were found in the frequency distribution based on tumor grade (G1 and G2 vs G3, Fisher’s exact test), tumor stage (Ta, T1 and T2, Chi-square), primary versus recurrent tumors (Fisher’s exact test), or in the EAU risk classification (intermediate or high [43], Fisher’s exact test). p-values are indicated in the figures.