Introduction

Several antineoplastic protocols have been extensively evaluated with the aim of improving patient survival and quality of life. A phase III trial of bladder cancer tumors indicated the methotrexate, vinblastine, adriamycin and cisplatin (MVAC) chemotherapy has an efficacy that was similar to that of a 2-drug regimen that included cisplatin and gemcitabine [1]. More recently, Yuh et al. [2] confirmed that neoadjuvant treatment with cisplatin and gemcitabine may warrant further prospective assessment. The mechanisms of action of cisplatin and gemcitabine are different. Cisplatin induces DNA cross-linking and causes severe lesions that can lead to apoptosis [3]. However, gemcitabine is a deoxycytidine analog that forms an active dFdCTP metabolite when phosphorylated. It is then incorporated into DNA, where it blocks replication [4].

Transcriptome studies performed over the last decade have helped us to identify molecular mechanisms that are related to tumorigenesis in different organs and to promote a more comprehensive understanding of chemotherapeutic protocols [5]. The characterization of genes known to facilitate tumor sensitivity to different antineoplastic drugs has played a key role in selecting the most optimal therapeutic strategies [6]. Using a PCR array, da Silva et al. [7] showed that in vitro treatment with only gemcitabine modulated several molecular pathways, including DNA repair, G1/S transition, apoptosis and transcription factors, in bladder tumor cell lines. Furthermore, Sávio et al. [8] used RT-qPCR to demonstrate that allyl isothiocyanate (AITC), an abundant compound that is found in cruciferous vegetables and mustard seeds, possesses promising antineoplastic potential. These authors showed that AITC had an important effect on the BAX/BCL2 axis in wild-type TP53 bladder tumor cells and on ANLN and S100P gene expression in a TP53-mutated cell line. These data suggested that AITC might induce cell cycle arrest. In addition to gene expression profiling, morphological data are also important for establishing diagnoses and indicating prognoses while treating a patient for cancer. Changes in cell morphology and growth that were characteristic of neoplastic development were detected in Vero cells that were treated with cisplatin [9]. Moreover, morphological changes, including rounding and shrinkage in a glioma cell line (GBM-8401), were visualized after 24 and 48 h of incubation with AITC, suggesting that this agent is an effective treatment option [10].

The present study was designed to answer the need to identify biomarkers that can be used to anticipate the effectiveness of antineoplastic treatments. Because combined treatment with gemcitabine/cisplatin has been supported as a promising chemotherapeutic protocol for bladder cancer and because TP53 mutations are among the most frequent gene modifications observed in bladder cancer cells [11], the aim of this study was to investigate the effect of administering these two drugs in combination on cell cycle-related gene expression and cell morphology in bladder tumor cell lines with two different TP53 mutation statuses. PCR-arrays and transmission and scanning electron microscopy were used.

Material and Methods

Cell Lines and Test Compounds

The established cell lines RT4, 5637 and T24, which were obtained from human bladder urothelial cell carcinomas (UCCs), were purchased from the Cell Bank of the Federal University of Rio de Janeiro, Brazil. The RT4 cell line (wild type TP53) is derived from a low grade papillary bladder tumor. Lineage 5637 harbors two TP53 mutations, one in codon 72 (Arg-Pro) and other in codon 280 (Arg-Thr). The T24 cell line, which was obtained from an invasive tumor, encodes an in-frame deletion of tyrosine 126 in the TP53 allele. The cells were maintained as previously described by Da Silva et al. [12] and recently tested for contamination.

The antineoplastic drugs that were used include gemcitabine (dFdC, Gemzar), which was obtained from Eli Lilly Laboratory (Eli Lilly and Company, USA), and cisplatin (CDDP), which was obtained from Sigma-Aldrich USA). Ultra-pure sterilized water was used for all dilutions.

Experimental Design

The cells were seeded into 12-well culture plates at 1.5 × 104 cells per plate. After 24 h, the cultures were treated with cisplatin (0.5, 1.0, 2.5, or 5.0 μmol/L), gemcitabine (0.78, 1.56, 3.12, or 6.25 μmol/L) or both drugs simultaneously (0.5 μmol/L cisplatin +0.78 μmol/L gemcitabine, 0.5 μmol/L cisplatin +1.56 μmol/L gemcitabine, 1.0 μmol/L cisplatin +0.78 μmol/L gemcitabine, or 1.0 μmol/L cisplatin +1.56 μmol/L gemcitabine) and incubated for an additional 24 h. These drug concentrations were chosen based on a protocol described in a previous study [12]. Untreated cells were cultured in parallel and used as the negative control. After 24 h of incubation with the drugs, the cells were washed with Hank’s solution (0.4 g KCl, 0.06 g KH2PO4, 0.04 g Na2HPO4, 0.35 g NaHCO3, 1 g glucose and 8 g NaCl in 1000 mL H2O), and complete fresh medium was then added. The cells were collected for apoptosis assays immediately after 24 h of treatment was completed (designated 0 h) and at 24 and 48 h after treatment was ended. To analyze morphological alterations and gene expression profiles, cells were treated for 24 h with 1.56 μmol/L gemcitabine and 1.0 μmol/L cisplatin. As shown in comet assays and trypan blue and XTT tests, these two concentrations were genotoxic but not cytotoxic in the two cell lines that were used [13].

Scanning Electron Microscopy

Cells were seeded under normal conditions in 12-well culture plates and placed under coverslips. After the cells were treat with the drugs, they were washed four times with 0.5 M sodium cacodylate buffer and fixed with 0.1% glutaraldehyde and 4% formaldehyde for 30 min. Then, the cells were washed again with 0.5 M cacodylate buffer, postfixed in 1% osmium tetroxide for 20 min, dehydrated in a graded series of ethanol solutions, critical-point dried (Balzers CDT 030) and coated with gold for 60 s using a sputter coater (Balzers CDT 050). The samples were analyzed using a scanning electron microscope (JEOL JSM model 6460). These experiments were performed in triplicate.

Transmission Electron Microscopy

Cells were seeded under normal conditions in 12-well culture plates and placed under coverslips. The cells were then fixed with glutaraldehyde and washed with phosphate buffer three times. Osmium tetroxide (1%) was added, and 2 h later, the cells were washed with distilled water. Aftare 2 h of incubation in 0.5% uranyl acetate, the cells were dehydrated in 50%, 70%, 90% and then 100% acetone. Then, an Araldite® resin with acetone mixture was added, and the plates were incubated overnight. Pure resin was then added, and the plates were incubated at 37 °C for 2 days. Ultra-thin sections were cut from the blocks containing the cells, and the blocks were stained with uranyl acetate and lead citrate. Then, the cells were analyzed using a Tecnai Spirit by Fei Company transmission electron microscope [14]. These experiments were performed in triplicate.

Apoptosis Assay

Three fluorescent dyes were used to evaluate the rates of apoptosis and necrosis in 5637 and T24 cells to check whether mutational site can influence the type of cell death: Hoechst 33,342 (stains DNA in viable cells), propidium iodide (identifies ruptured cell membranes) and fluorescein diacetate (indicates viable cells). After drug treatments, the cells were incubated with 10 μl Hoechst 33,342 (1 mg/ml), 25 μl propidium iodide (1 mg/ml) and 50 μl fluorescein diacetate (1.5 mg/ml) for 5 min at 37 °C. Then, the cells were examined using a fluorescent microscope with a triple filter for DAPI, FITC and rhodamine. A total of 500 cells from three independent experiments were scored. The following criteria were adopted: 1) normal cells: blue intact nucleus and green cytoplasm; 2) apoptotic cells: blue nucleus with green cytoplasm and apoptotic bodies; 3) necrotic cells: red intact nucleus and cytoplasm.

RNA Extration and PCR ARRAYS

To evaluate changes in the gene expression profile in 5637 and T24 cell lines, gemcitabine and cisplatin were used at concentrations of 1.56 μmol/L and 1.0 μmol/L, respectively.

Total RNA was extracted using RNeasy Mini kits (Qiagen) according to the manufacturer’s instructions and stored at −80 °C. RNA integrity and quality were evaluated using 2% denaturing agarose gels and Nanovue equipment (Thermo Fischer), respectively. A cell cycle pathway PCR array (PAHS-020a — Human Cell Cycle PCR Array - 84 genes - SA Biosciences) was used for the RT-qPCR analysis. RNA was reverse transcribed using an RT2 first strand kit (SA Biosciences) according to the manufacturer’s instructions. An aliquot of the diluted first-strand synthesis reaction was added to Sybr Green/Rox Master Mix (SA Biosciences) and nuclease-free water, in accordance with the PCR array system’s user manual. Then, 25 μl of the cDNA/master mix cocktail was added to each well of the pathway-specific RTq-PCR microplate. The plate was sealed and transferred to the Real Time-PCR equipment, where the cycling profile was performed as described by the manufacturer instructions. Data were analyzed using the method of ΔΔCT. In this case, the Ct values ​​were calculated and melting curves were constructed. The quality controls used to evaluate genomic DNA contamination, reverse transcription efficiency and PCR amplification efficiency were then analyzed. The RT-qPCR array data were normalized using the arithmetic mean values of two reference genes (GAPDH and ACTB) for 5637 and T24 cell lines. All the arrays were performed in triplicate. Analyses of significant gene clusters were performed using bioinformatics tools available on page http://pcrdataanalysis.sabiosciences.com/pcr/arrayanalysis.php. Information about the biological functions of the genes was obtained from FATIGO (http://babelomics.bioinfo.cipf.es/).

Statistical Analysis

One-way ANOVA was used to analyze the apoptosis data. The gene expression data were evaluated as fold changes. In this case, p values were calculated using Student’s t-test with triplicate values for each gene in the control and treatment groups. p ≤ 0.05 was considered to indicate statistical significance.

Results

Scanning Electron Microscopy

After cells were treated according to each antineoplastic drug protocol, the reduction in the number of cells was determined in three cell lines (RT4, 5637 and T24). Figures 1, 2 and 3 show the irregular morphology and reduced number of microfilaments that were visualized in RT4, 5637 and T24 cells after treatment with gemcitabine or a combination of gemcitabine/cisplatin. Characteristic apoptotic cells were observed in the 5637 cultures that were treated with 1.0 μmol/L cisplatin +1.56 μmol/L gemcitabine (Fig. 2 D).

Fig. 1
figure 1

Photomicrographs of RT4 cells were obtained using scanning electron microscopy: (a) untreated (controls), (b) treated with cisplatin (1.0 μmol/L), (c) treated with gemcitabine (1.56 μmol/L), or D) treated with cisplatin + gemcitabine (1.0 μmol/L cisplatin +1.56 μmol/L gemcitabine). Black arrows indicate microfilaments, and white arrows indicate irregular morphology. Scale bar: 10 μm. These experiments were performed in triplicate

Full size image
Fig. 2
figure 2

Photomicrographs of 5637 cells were obtained using scanning electron microscopy: (a) untreated (controls), (b) treated with cisplatin (1.0 μmol/L), (c) treated with gemcitabine (1.56 μmol/L), or (d) treated with cisplatin + gemcitabine (1.0 μmol/L cisplatin +1.56 μmol/L gemcitabine). Black arrows indicate microfilaments, and white arrows indicate apoptotic cells. Scale bar: 10 μm. These experiments were performed in triplicate

Full size image
Fig. 3
figure 3

Photomicrographs of T24 cells were obtained using scanning electron microscopy: (a) untreated cells (control), (b) cells treated with cisplatin (1.0 μmol/L), cells treated with gemcitabine (1.56 μmol/L), and (d) cells treated with cisplatin + gemcitabine (1.0 μmol/L cisplatin +1.56 μmol/L gemcitabine). Black arrows indicate microfilaments, and white arrows indicate irregular morphology. Scale bar: 50 μm. These experiments were performed in triplicate

Full size image

Transmission Electron Microscopy

Initially, an analysis using transmission electron microscopy showed that there were differences between the untreated (control) RT4, 5637 and T24 cells. T24 showed more cytoplasmic prolongations with irregular shapes, clearer nucleoli, different organelles (e.g., endoplasmic reticulum, mitochondria and Golgi apparatus) in the cytoplasm, electron-dense granules (probably protein-derived material) and larger nuclei than the RT4 and 5637 cells (Fig. 4). However, the three cell lines had similar morphological characteristics after treatment and showed a decrease in the size of the nucleus. Increases in the number and size of nuclear chamfers and a decreased number of organelles were observed mainly in TP53 mutated cells (Figs. 5 and 6).

Fig. 4
figure 4

Photomicrographs of RT3 (a), 5637 (b) and T24 (c) cells were obtained using transmission electron microscopy. * Decondensed chromatin; cytoplasmic prolongations and nucleolus are indicated by black and white arrows, respectively. White arrowheads indicate nuclear chamfers, while black arrowheads indicate electron-dense granules. Scale bar: 10 μm. These experiments were performed in triplicate

Full size image
Fig. 5
figure 5

Photomicrographs of the following groups of T24 cells were obtained using transmission electron microscopy: (a) untreated cells (control), (b) cisplatin-treated cells (1.0 μmol/L), (c) gemcitabine-treated cells (1.56 μmol/L), and (d) cells treated with cisplatin + gemcitabine (1.0 μmol/L cisplatin +1.56 μmol/L gemcitabine). Black arrows indicate smaller nuclei and a larger number and extended size of nuclear chamfers in comparison to the untreated cells. Scale bar: 5 μm. These experiments were performed in triplicate

Full size image
Fig. 6
figure 6

Photomicrographs of the following groups of RT4 and 5637 cells were obtained using transmission electron microscopy: (a) untreated RT4 cells (control), (b) RT4 cells treated with cisplatin + gemcitabine (1.0 μmol/L cisplatin +1.56 μmol/L gemcitabine), (c) untreated 5637 cells (control) and (d) 5637 cells treated with cisplatin + gemcitabine (1.0 μmol/L cisplatin +1.56 μmol/L gemcitabine). Black arrows indicate smaller nuclei and a larger number and extended size of nuclear chamfers in comparison to the untreated cells. Scale bar: 5 μm. These experiments were performed in triplicate

Full size image

Apoptosis Assay

A statistically significant increase in the frequency of cell death (via apoptosis and necrosis) was observed in the 5637 cultures that were treated with cisplatin and/or gemcitabine. The highest percentage of apoptotic cells was detected at 48 h after treatment (15.86% for 5.0 μmol/L cisplatin and 16.6% for 6.25 μmol/L gemcitabine). However, the highest percentage of necrotic cells was observed immediately after (0 h) treatment with cisplatin (5.66% for 5.0 μmol/L) or gemcitabine (2.32% for 0.78 μmol/L) and at 48 h after treatment with cisplatin + gemcitabine (19.8%). However, treatment with a combination of these drugs increased the number of necrotic cells at the three timepoints that were tested (Table 1). In T24 cells, a significant increase was observed in the proportion of apoptotic cells after 24 h (3.20% for 3.125 μmol/L and 4.0% for 6.25 μmol/L) and 48 h (3.60% for 6.25 μmol/L) after treatment with gemcitabine. However, the highest rates of apoptosis were observed at 24 and 48 h after cells were simultaneously treated with gemcitabine + cisplatin at various combinations that were tested. With regard for necrotic cells, an increase in the percentage of necrotic cells was observed immediately (3.0% for 1.0 μmol/L and 3.80% for 2.5 μmol/L) and at 48 h (5.0% for 2.5 μmol/L and 3.60% for 5.0 μmol/L) after treatment with cisplatin. For gemcitabine, an increase in the percentage of necrotic cells was observed at 48 h after treatment (3.80% for 1.56 μmol/L, 5.20% for 3.125 μmol/L and 5.0% for 6.25 μmol/L). When cells were simultaneously treated with both drugs, an increase in the percentage of necrotic cells was observed at 24 h after treatment with all tested concentrations (Table 2 and Fig. 7).

Table 1 The frequency (%) of apoptotic and necrotic 5637 cells after 0, 4 and 48 h of treatment with cisplatin and/or gemcitabine
Full size table
Table 2 Frequency (%) of apoptotic and necrotic T24 cells after 0, 24 and 48 h of treatment with cisplatin and/or gemcitabine
Full size table
Fig. 7
figure 7

The morphology of 5637 (a) and T24 (b) cells at 72 h after treatment with cisplatin + gemcitabine was evaluated using Hoechst 33,342, propidium iodide and fluorescein diacetate. Red arrows: intact nuclei, blue arrows: apoptotic cells, green arrows: necrotic cells (1000 X). These experiments were performed in triplicate

Full size image

PCR Arrays

When compared to the controls, 7/84 (8.3%) and 19/84 (22.6%) of the evaluated genes were differentially expressed (fold change ≥2) in 5637 (Fig. 8a and Table 3) and T24 cells (Fig. 8b and Table 4) after the cells were simultaneously treated with gemcitabine and cisplatin. CDKN2B, GADD45A and SERTAD1 were overexpressed in both cell lines after combined treatment.

Fig. 8
figure 8

Heat map showing the gene expression profile of the 5637 (a) and T24 (b) human bladder transitional carcinoma cell lines after the cells were simultaneously treated with gemcitabine + cisplatin (1.0 μmol/L + 1.56 μmol/L, respectively). The data from three independent cultures are shown. Red, green and black spots indicate upregulated, downregulated and non-modulated genes, respectively (fold change ≥2)

Full size image
Table 3 Significantly (p ≤ 0.05; fold change ≥2.0) modulated genes in 5637 cells after simultaneous treatment with cisplatin (1.0 μmol/L) and gemcitabine (1.56 μmol/L)
Full size table
Table 4 Significantly (p ≤ 0.05; fold change ≥2.0) modulated genes in T24 cells after simultaneous treatment with cisplatin (1.0 μmol/L) and gemcitabine (1.56 μmol/L)
Full size table

Discussion

Determining the relationships between gene expression profiles and tumor sensitivity to antineoplastic drugs is fundamentally important for developing effective therapeutic approaches [15]. Therefore, this study was performed to evaluate the effect of cisplatin and gemcitabine activity on cell cycle-related genes in two bladder urothelial carcinoma cell lines that possess a mutated TP53 gene. Before we analyzed the transcriptome, the effects of gemcitabine and cisplatin on cell morphology were assessed. The morphological differences that were observed between the three cell lines indicate that T24 cells display increased metabolic activity, potentially because they originated from a higher degree tumor (grade 3). Previously, Vaezy and Zderic [16] suggested that the degree of metabolic activity in brain tumors may be correlated with the tumor’s degree of malignance. Moreover, Rajcevic et al. [17] suggested that the general cellular metabolism of glioblastomas is more active in angiogenic than non-angiogenic phenotype glioblastomas.

Scanning and transmission microscopy also revealed that morphological alterations (e.g., in the number of microfilaments and organelles, irregular contours, and the size of the nucleus) occurred after treatment. We suspect that the changes that were detected in cell shape and microvilli numbers resulted from impaired actin cytoskeletal polymerization and because there was a lower amount of contact between the cells themselves and between the cells and the extracellular matrix. This may be because these chemotherapeutic agents were purposefully designed to target cell membranes and the cytoskeleton, to induce cytotoxicity and to alter cytoadherence [18]. The actin network is a complex and important structural and functional system that provides the basic infrastructure that maintains cell morphology, adhesion, motility, exocytosis, endocytosis and cell division. Actin polymerization and remodeling play pivotal roles in regulating the morphology and phenotype of a malignant cell. Moreover, these actin filaments play important roles in mitosis and cell signaling [9]. Thus, alteration or decrease in cellular microfilaments could have compromised cellular viability. Additionally, the observed decreases in the size of the nucleus and the number of organelles might indicate a reduction in cellular metabolism. These events could contribute to or be a repercussion of cell cycle arrest or the early stages of cell death.

Scattered cells were also detected after cells were treated with cisplatin and gemcitabine, and these findings confirm that this drug combination exerts an antiproliferative effect, probably by stimulating apoptosis and necrosis, which were observed in the fluorescent apoptosis assay. Recently, Wang et al. [19] demonstrated that TP53 stimulates a wide network of signals that act through two major apoptotic pathways: 1) the death receptor pathway, which triggers the activation of a caspase cascade, and 2) a mitochondrial pathway that shifts the balance in the expression of Bcl-2 family members towards pro-apoptotic members, thereby promoting the formation of the apoptosome and, consequently, caspase-mediated apoptosis. Nevertheless, Wang et al. [20] previously reported that cancer cells lacking a functional TP53 were also killed by cisplatin via a TP53-independent mechanism that was similar to replicative senescence. Furthermore, there is evidence showing that genes other than TP53 also participate in regulating DNA repair and apoptotic mechanisms [21]. Similarly to our previous findings [12], Fencher et al. [22] also observed that apoptosis occurred independently of TP53 status in bladder urothelial carcinoma cell lines.

To gain a better understanding of the mechanisms underlying the activities of cisplatin and gemcitabine, we evaluated their effects on gene expression profiles. The data showed that CDKN2B, GADD45A and SERTAD1 were upregulated. According to the gene ontology (GO) analysis, these genes are primarily involved in negatively regulating the cell cycle and cell cycle arrest (CDKN2B and GADD45A), G1/S transition (CDKN2B), DNA repair and apoptosis (GADD45A) and the regulation of transcription (SERTAD1). Additionally, some of these genes are targets of cancer treatments. While GADD45A expression has been associated with the efficiency of 5-aza-CdR to treat different pancreatic cancer cell lines [23], CDKN2B, a cyclin-dependent kinase inhibitor, is a tumor suppressor gene that inhibits cell cycle progression [24]. Interestingly, SERTAD1, which is known to inhibit p16INK4a activity [25], is overexpressed in several cancers and associated with oncogenic potential [26]. However, L’Espérance et al. [27] have also shown that SERTAD1 is upregulated in ovarian cancer cells after treatment with Topotecan.

It is important to emphasize that differences in gene expression (fold change ≥2) were detected between the two cell lines. Four genes were significantly modulated in only the 5637 cells, while 16 genes were significantly modulated in only the T24 cells. In a previous study we have also suggested that MRE11A (which is involved in the response to a DNA damage stimulus and the cellular response to stress) and SKP2 (involved in the negative regulation of the cell cycle) downregulation seemed be responsible for the synergistic therapeutic effects of cisplatin and gemcitabine in wild type TP53 cell line [28]. This discrepancy in the expression profile intensities of these genes supports the notion that complex mechanisms are involved in carcinogenesis. In conclusion, gemcitabine and cisplatin caused morphological alterations that are supportive of their antiproliferative potential and upregulated GADD45A, CDKN2B and SERTAD1. These effects were independent of the type of TP53 mutation in the cells and the tumor grade. The identification of gemcitabine- and cisplatin-responsive genes provides insight into their anti-cancer mechanisms and shows that these genetic profiles can be used in a clinical setting to predict responses to chemotherapies.