Cigarette Smoke Containing Acrolein Contributes to Cisplatin Resistance in Human Bladder Cancers through the Regulation of HER2 Pathway or FGFR3 Pathway

Bladder cancer is the most common cancer of the urinary track and predominantly occurs in men worldwide (1). The majority of bladder cancers are urothelial cell carcinomas (UC), and more than 70% of patients are diagnosed with non–muscle invasive bladder cancer (NMIBC; ref. 2). Frequent recurrence in patients with NMIBC is a serious problem (3) that is observed in more than 50% of patients within 2 years, with 10% to 30% of recurrent patients progressing to muscle invasive bladder cancer (MIBC; ref. 4). Cisplatin-based chemotherapy is extensively used as a first-line treatment for many cancers, including bladder cancer (5); however, its efficacy is limited due to the development of drug resistance (6). Therefore, understanding the molecular mechanisms underlying cisplatin resistance is critical for improving bladder cancer treatment and prognosis.

The currently known mechanisms of bladder cancer cisplatin resistance are overactivation of growth factor signaling pathways, including FGFR and ErbB/HER, which results in diverse biological processes, including cell proliferation, survival, and tumorigenesis (7, 8). Members of the FGFR family are frequently activated by a mutation in bladder cancer and overexpressed in many bladder cancer cell lines, where they increase the activation of many survival pathways, such as MAPK/ERK (9). PD173074, a small molecular inhibitor, was reported to show both high affinity and selectivity for the FGFR family and was also shown to inhibit FGF-driven neoangiogenesis (10, 11). Previous studies have shown that PD173074 is extremely effective at potentiating the effect of cisplatin in lung cancer (12); however, similar work has not been carried out on bladder cancer thus far. Moreover, overexpression and/or amplification of the transmembrane tyrosine kinase (TK) receptor HER2, which is a member of the EGFR family, has been observed in bladder cancer (13), and HER2 overexpression is frequently detected in chemotherapy- and radiation-resistant bladder cancer (14). Anti-HER2 agents are able to induce cellular apoptosis in vitro and in vivo (15); however, multiple clinical trials of HER2-targeting agents generally in unselected patients with bladder cancer have not shown definitive clinical efficacy (16). Therefore, understanding the molecular alterations in bladder cancer has significant clinical implications in that they not only define radically different subgroups of patients with bladder cancer but also significantly affect tumor susceptibility to either conventional or target therapy and thus impact patient prognosis.

Smoking is the most common etiological risk factor for bladder cancer, and it accounts for up to 60% of all cases (17). Smoking intensity and duration have been shown to be closely associated with an increased risk of bladder cancer development (18). Furthermore, smoking cessation has also been suggested to be a cancer-specific benefit for bladder cancer prognosis (19, 20). Emerging evidence indicates that cigarette smoking is associated with the response to standard chemotherapeutic agents, including cisplatin (21, 22). Continued smoking during therapy is associated with low response rates to chemotherapy and radiation in many types of cancers, including bladder cancer (21, 23). However, little is known about the mechanism by which cigarette smoking enhances chemoresistance in bladder cancer. Acrolein, an α,β-unsaturated aldehyde, is abundant in cigarette smoking, cooking fumes, and automobile exhaust fumes (24). Acrolein can also be produced endogenously through lipid peroxidation, resulting from free radical damage to polyunsaturated fatty acids (25). Our previous studies showed that acrolein contributes to bladder carcinogenesis through the induction of DNA damage and inhibition of DNA repair (26). In addition, our current studies have shown that acrolein acts as an endogenous uremic toxin and contributes to urothelial carcinoma formation in patients with chronic kidney diseases (27). In this study, we investigated the molecular mechanisms by which acrolein induced cisplatin resistance in two well-established NMIBC and MIBC cell lines, RT4 and T24. Furthermore, the effectiveness of targeted therapies on cisplatin-resistant bladder cancer was analyzed in vitro and in vivo.

The following chemicals were purchased: acrolein from Sigma–Aldrich; cisplatin from Merck; trastuzumab from MedChemExpress; and PD173074 from MedChemExpress.

Human urinary bladder cancer cell lines RT4 (HTB-2) and T24 (HTB-4) were purchased from ATCC and maintained in McCoy 5A medium supplemented with 10% FBS and DMEM supplemented with 10% FBS, respectively. The absence of Mycoplasma in these cell lines was tested every month using an EZ-PCR Mycoplasma Detection Kit (Biological Industries). Acrolein stock solution was freshly prepared before use. Cells at 70% confluency were treated with acrolein (10 μmol/L) in complete culture medium for 1 month at 37°C in the dark, acrolein-containing medium was changed every 2 days, and selected clones were named Acr-clone#. For the dose or time effects, cells were treated with different concentrations of acrolein (0–10 μmol/L) for 24 hours or acrolein (7.5 μmol/L) for 3 to 24 hours, respectively.

Cell viability was determined using a modified 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT; Sigma–Aldrich) assay and a lactate dehydrogenase (LDH) leakage assay, as previously described (28). Briefly, cells (5,000/well) were seeded in 96-well plates overnight and treated with cisplatin (0–30 μg/mL), trastuzumab (1 μg/mL), or PD173074 (5 μmol/L) for 24 hours. For the MTT assay, the resulting formazan dissolved in DMSO was measured at 570 nm, and the results are presented as the percentage of the control values. For the LDH leakage assay, LDH activities were measured using an LDH detection kit (Sigma–Aldrich) as described in the manufacturer's protocol. The extent of cellular damage was determined to be dependent on the percent LDH activity in the supernatants with respect to that in the cell lysates. These experiments were performed at least three times in triplicate.

Cell viability was determined using a modified MTT assay as described in the cell viability assay. Briefly, cells (1,000/well) were seeded in 96-well plates overnight and measured every day for 7 days. These experiments were performed at least three times in triplicate.

To evaluate anchorage-independent cell growth, a soft agar colony formation assay was performed as described previously (29). Briefly, bottom layers were formed with a 3-mL aliquot of 1.2% agar in culture medium in six-well plates. Then, the top layer containing 1,000 cells/well of parental cells or Acr-clones was mixed with 3 mL of 0.35% agar in medium and plated on solidified bottom agar. When the top agar solidified, the dishes were transferred to an incubator and cultured for 14 days. Two or three drops of the medium were added to each dish three times a week. Colonies were stained with 0.005% crystal violet, photographed, and counted. All of these experiments were performed in triplicate and repeated independently at least three times.

The cell migration assay was carried out in a Transwell apparatus with 8-μm pore size membranes (Corning) as previously described (30). Briefly, parental or Acr-clones (5 × 10⁴/well, six-well plates) were added to the upper chamber, and the lower chamber contained growth medium supplemented with 10% FBS. After 24 hours of incubation at 37°C in a 5% CO₂ incubator, cells on the upper filter surface were removed by wiping with a cotton swab. Filters were then fixed in methanol and stained with crystal violet. The number of cells was counted in six random fields under a microscope at 200× magnification. These experiments were performed at least three times in triplicate.

Total RNA were prepared and subsequent real-time RT-PCR analysis of cDNA was analyzed as described previously (31). The primers (5′-3′) were AAGCTTACCATGGAGCTGGCGGCCTTG and CTCGAGCACTGGCACGTCCAGACC for HER2; CCACTGTCTGGGTCAGGAT and AGGATGGAGCGTCTGTCAC for FGFR3; CCGTCTAGAAAAACCTGCC and GCCAAATTCGTTGTCATACC for GAPDH. To calculate the relative RNA expression, GAPDH was used as an internal control for all qRT-PCR reactions and compared with control groups. These experiments were performed at least three times in duplicate.

Cell lysates were prepared and analyzed as described previously (31). Briefly, blots were blocked with 5% nonfat milk and hybridized with primary antibodies overnight at 4°C. The antibodies against phospho-HER2/ErbB2 (Tyr1221/1222; 1:1000, Cell Signaling Technology #2249), HER2 (1:1000, Cell Signaling Technology #9212), phospho-FGFR3 (Tyr724; 1:1000, Abcam #ab155960), FGFR3 (1:1000, Novus Biologicals #JM81–10), phospho-p38 (Thr180/Tyr182; 1:1000, Cell Signaling Technology #9211), p38 (1:1000, Cell Signaling Technology #9212), phospho-p44/42 MAPK (Erk1/2; Thr202/Tyr204; 1:1000, Cell Signaling Technology #9101), p44/42 MAPK (Erk1/2; 1:1000, Cell Signaling Technology #9102), PARP-1 (1:1000, Cell Signaling Technology #9542), caspase-9 (1:1000, Cell Signaling Technology #9502), caspase-3 (1:1000, Cell Signaling Technology #9662), E-cadherin (1:1000, Cell Signaling Technology #3195), EpCAM (gifted from Prof. Ying Chih Chang from Genomic Research Center at Academia Sinica, Taipei, Taiwan), vimentin (1:1000, Taiclone # TCEA19744), Snail (1:1000, Cell Signaling Technology #C15D3), and GAPDH (1:5000, Cell Signaling Technology #5174). The immunodetection was performed using Enhanced Chemiluminescence (ECL; Millipore Corporation).

Forty 6-week-old male Balb/c nude mice weighing 20 to 25 g were used. All animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC) of National Yang Ming Chiao Tung University (Taipei, Taiwan) and carried out according to the Guidelines for Animal Research of National Yang Ming Chiao Tung University (IACUC #1070208rr). T24 parental cells or T24 Acr-clone#1 cells (5 × 10⁶ in 50% Matrigel, BD Bioscience) were subcutaneously inoculated into the right flank of mice. After 1 week, mice with palpable xenografts were randomly divided into four groups: vehicle (PBS, 50 μL, i.p., once weekly; and corn oil, 50 μL, i.p. once weekly), cisplatin (5 mg/kg in PBS, 50 μL, i.p. once weekly), cisplatin (5 mg/kg in PBS, 50 μL, i.p. once weekly) combined with PD173074 (10 mg/kg in corn oil, 50 μL, i.p. once weekly) and cisplatin (5 mg/kg in PBS, 50 μL, i.p. once weekly) combined with PD173074 (10 mg/kg in corn oil, 50 μL, i.p. once weekly). To obtain the tumor growth curve, the tumor dimensions were monitored every 3 days. The perpendicular diameter was measured with a digital caliper, and the volume was calculated based on the (length × width²)/2. Body weight was also assessed twice weekly. The animals were sacrificed after 3 weeks of the drug treatment. Tumor samples were collected after sacrifice. Each sample was cut in half, with one half preserved in 4% paraformaldehyde and the other flash frozen in liquid nitrogen and stored at −80°C until further use.

For histologic staining, sections (4 µm) were deparaffinized in xylene, rehydrated in a descending ethanol series, and stained with hematoxylin and eosin (H&E). Cell death by apoptosis was analyzed using an in situ cell death detection kit (POD, Roche), as described in the manufacturer's protocol. Briefly, paraffin-embedded tumor sections were dewaxed by heating at 60°C, followed by washing in xylene and rehydration. Endogenous peroxidase activity was eliminated by incubation in hydrogen peroxide. Then, tumor sections were incubated with proteinase K solution (20 μg/mL in 10 mmol/L Tris-HCl, pH 7.4) for 20 minutes at room temperature. Following washing with PBS, tumor sections were permeabilized in 0.1% Triton X-100 and 0.1% sodium citrate for 8 minutes at room temperature and incubated in 50 µL of terminal deoxynucleotidyl transferase–mediated dUTP nick end labeling (TUNEL) reaction mixture for 1 hour at 37°C in the dark. Tumor sections were washed with PBS and incubated with 50 µL Converter-POD for 30 minutes at 37°C. After rinsing in PBS, the sections were incubated for 10 minutes with 100 µL diaminobenzidine (DAB; Sigma–Aldrich) substrate in the dark. Finally, the samples were mounted and analyzed under a light microscope at 400× magnification, where apoptotic cells could be seen as condensed shrinking dark brown cells. For IHC analysis, sections were incubated in Tris–EDTA, pH 6.0, and boiled for 12 minutes to enhance immunoreactivity. Endogenous peroxidase activity was eliminated by incubation in hydrogen peroxide. Incubation with primary antibodies for FGFR3 antibody (1:300, Novus Biologicals #JM81–10) was performed overnight at 4°C in 1% BSA in PBS. Bound antibodies were visualized with DAB used as a chromogen, and omission of the primary antibody served as a negative control. Assessment of FGFR3 immunoexpression was performed by light microscopy at ×400 magnification.

Descriptive statistics are presented as the mean ± SD or as the number (percentage). All calculated P values were two-tailed. Statistical significance was defined as *P < 0.05. A minimum of three independent replicate experiments was performed to justify the use of statistical tests.

To investigate whether cigarette smoke acrolein contributes to cisplatin resistance in human bladder cancer, we treated the NMIBC and MIBC cell lines, RT4 and T24, respectively, with a physiologically relevant dose of acrolein (10 μmol/L) for 1 month and selected three different clones, namely, RT4 Acr-clone# and T24 Acr-clone#. The results showed that RT4 Acr-clone#3 (IC₅₀ = 25.13 ± 3.01 μg/mL) had a higher IC₅₀ than the parental RT4 (IC₅₀ = 12.04 ± 1.86 μg/mL; Fig. 1A and B), and T24 Acr-clone#1 (IC₅₀ = 18.57 ± 3.95 μg/mL) had a higher IC₅₀ than the parental T24 (IC₅₀ = 9.46 ± 0.64 μg/mL; Fig. 1C and D) using MTT and LDH analysis. The results suggest that acrolein increased the IC₅₀ of cisplatin by two-fold in both RT4 and T24 cells.

We further analyzed cell proliferation, soft agar colony formation activity, and cell migration ability in acrolein-induced cisplatin-resistant RT4 and T24 clones. The results showed that increased cell proliferation (Fig. 2A and B), soft agar colony formation (Fig. 2C and D), and cell migration ability (Fig. 2E and F) were increased in RT4 Acr-clone#3 and T24 Acr-clone#1 cells compared with parental cells. We also analyzed epithelial–mesenchymal transition (EMT) markers in both cell clones. The results showed decreased E-cadherin but increased EpCAM in RT4 Acr-clone#3 cells compared with parental cells and increased vimentin and Snail in T24 Acr-clone#1 cells compared with the parental cells (Supplementary Fig. S1). These results suggest that acrolein induced cancer progression, including cisplatin resistance, in both high-grade and low-grade bladder cancer cell lines.

The currently known mechanisms of bladder cancer cisplatin resistance are overactivation of growth factor signaling pathways, including FGFR and ErbB/HER, which results in downstream MAPK/ERK activation and diverse biological processes, including cell proliferation differentiation, survival, and tumorigenesis (7, 8). We further analyzed the FGFR3 pathway and HER2 pathway in both acrolein-induced cisplatin-resistant RT4 and T24 clones. The results showed increased mRNA and protein expression of HER2 in RT4 Acr-clone#3 cells compared with the parental RT4 cells (Fig. 3A and C). Moreover, increased mRNA and protein expression of FGFR3 in T24 Acr-clone#1 cells were observed compared with that in the parental T24 cells (Fig. 3B and C). However, the phosphorylation of HER2 or FGFR3 was not increased in these cells (Supplementary Fig. S2). HER2- or FGFR3-downstream MAPK pathways, including the p38 and ERK pathways, were activated in RT4 Acr-clone#3 and T24 Acr-clone#1 cells compared with their parental cells (Fig. 3A and B). Furthermore, short-term treatment with acrolein induced protein expression of HER2 in both RT4 and T24 cells, while it increased protein expression of FGFR3 in T24 cells (Fig. 3D and E). These results suggest that acrolein induced cisplatin resistance in RT4 and T24 cells through HER2- or FGFR3-downstream MAPK pathways.

To verify the activation of HER2 and FGFR3 signaling in RT4 Acr-clone#3 and T24 Acr-clone#1 cells, trastuzumab, an anti-HER2 mAb, and PD173074, an FGFR3 inhibitor, were utilized. The results showed that trastuzumab reversed cisplatin resistance in RT4 Acr-clone#3 cells (Fig. 4A). The combination of trastuzumab with cisplatin enhanced cellular apoptosis (Fig. 4B) and decreased the expression of HER2 in RT4 Acr-clone#3 cells compared with cisplatin alone (Fig. 4C). This phenomenon was not observed in the T24 Acr-clone#1 cells treated with the combination of trastuzumab and cisplatin (Fig. 4D–F). Moreover, PD173074 reversed cisplatin resistance in T24 Acr-clone#1 cells (Fig. 4G). The combination of PD173074 with cisplatin enhanced cellular apoptosis in T24 Acr-clone#1 cells (Fig. 4H) and decreased FGFR3 compared with vehicle control (Fig. 4I). However, this phenomenon was also observed in the combination treatment of PD173074 with cisplatin in RT4 Acr-clone#3 cells (Figs. 4J–L). Since RT4 cells contain FGFR3 fusions, they have been shown to be more sensitive to FGFR inhibition (32, 33). Consistently, we found that PD173074 alone induced higher cytotoxicity and cellular apoptosis in both parental RT4 cells and Acr-clone#3 cells (Fig. 4J and K). These results further confirm that acrolein induced activation of HER2 signaling and FGFR3 signaling, resulting in cisplatin resistance in RT4 and T24 cells, respectively.

We further investigated the potential of PD173074 combined with cisplatin in cisplatin-resistant T24 cells using a xenograft mouse model. T24 parental cells or T24 Acr-clone#1 cells in Matrigel were subcutaneously injected into the flanks of nude mice, and once the tumors were measurable, animals were randomized to receive carrier buffer in the absence or presence of PD172074 (10 mg/kg i.p.) once a week with or without cisplatin (5 mg/kg i.p.) once a week. After 3 weeks, the mice were sacrificed and tumors were collected for further analysis. The results showed that tumor volumes in the cisplatin group did not show a significant difference from those of the vehicle group, while PD173074 as a single agent was as efficient at decreasing the tumor size compared to vehicle treatment (Fig. 5A and B). Furthermore, PD173074 combined with cisplatin led to significantly better impairment of tumor growth than either agent alone (P < 0.01; Fig. 5A and B). However, this phenomenon was not observed in mice inoculated with parental T24 cells (Supplementary Fig. S3). Slight weight loss occurred in mice treated with PD173074 combined with cisplatin (Fig. 5C). In addition, the morphologic changes in the kidney and liver tissues were assessed using H&E staining. The results showed that kidneys from mice in either the vehicle control or PD173074 alone group displayed normal renal architecture. As expected, cisplatin treatment alone resulted in obvious renal tubular damage characterized by tubular degeneration and dilatation, extensive epithelial vacuolization and necrosis, and the formation of hyaline casts (Supplementary Fig. S4). However, cisplatin combined with PD173074 significantly attenuated the pathologic changes in cisplatin-treated mice. These results suggest that PD173074 may have a renoprotective effect in cisplatin-treated mice. In addition, there were no obvious morphologic changes in the liver among the four groups (Supplementary Fig. S4). We further analyzed apoptosis in the tumor samples using the TUNEL assay, and the results showed that PD173074 combined with cisplatin significantly increased the number of TUNEL-positive apoptotic cells in xenografts compared with either agent alone (Fig. 5D and E). In addition, we examined the expression of FGFR3 using IHC staining assay and found that PD173074 combined with cisplatin significantly decreased the FGFR3 expression in xenografts compared with vehicle control or cisplatin alone (Fig. 5D and F). This finding is consistent with the in vitro results (Fig. 4G–I), which suggests that the FGFR3 pathway contributes to cisplatin resistance in T24 Acr-clone#1 cells and combining PD173074 with cisplatin induced more cellular apoptosis in T24 Acr-clone#1 cells.

Cigarette smoking is among the most important risk factors for bladder cancer (17), and continued smoking during therapy is associated with lower response rates to chemotherapy or radiation in patients with bladder cancer (21). Cisplatin-based chemotherapy is the most effective therapeutic regimen for bladder cancer (5); however, the development of chemoresistance in patients with bladder cancer is a critical problem in the management of advanced bladder cancer (6). Our previous studies showed that cigarette smoke–containing acrolein contributes to bladder carcinogenesis through the induction of DNA damage and inhibition of DNA repair (26). In this study, our results showed that long-term exposure to acrolein induced cisplatin resistance and tumor progression in both the NMIBC and MIBC cell lines, RT4 and T24, respectively (Figs. 1–2). The physiologically relevant dosage of acrolein is approximately 10 μmol/L based on previous studies showing that the acrolein exposure level from an average smoker is approximately 1.7 to 2 mg/day (34). Assuming a 60-kg adult with 4 L of blood volume, the exposure concentration of acrolein in smokers is approximately 7.6 to 9 μmol/L. Differential molecular alterations, including HER2 and FGFR3, were observed in acrolein-induced cisplatin-resistant RT4 and T24 cells (Fig. 3). Furthermore, HER2- and FGFR3-targeted therapy reversed cisplatin resistance in RT4 and T24 cells, respectively (Figs. 4 and 5). Cisplatin-resistant bladder cancer models are typically established by treatment with increasing concentrations of cisplatin until they become tolerant to relatively high concentrations of cisplatin (generally approximately 12 µmol/L; ref. 7). Multiple mechanisms of cisplatin resistance have been reported, including the reduction of cellular cisplatin accumulation, changes in DNA repair mechanisms, and alterations in apoptotic cell death pathways, and growth factor signaling (35). This model was used to study the molecular mechanism of cisplatin resistance induced by long-term cisplatin treatment in bladder cancer; however, our study aimed to understand the molecular mechanism underlying cigarette smoking–induced chemoresistance in bladder cancer.

In previous studies, T24 cells were derived from a patient with undifferentiated grade 3 carcinoma, they were found to be representative cells of a cancer with a poor prognosis (36). In comparison, RT4 cells were derived from a patient with a well-differentiated grade 1 papillary tumor, and they were found to be representative cells of a cancer with good prognosis (36). In addition, RT4 cells showed strong expression of epithelial markers (E-cadherin and EpCAM), while T24 cells exhibited strong expression of mesenchymal markers (vimentin and Snail; Supplementary Fig. S1), which is consistent with previous studies (37). Therefore, these two cell lines could represent two discrete “epithelial” and “mesenchymal” subsets, with the latter consisting entirely of muscle-invasive tumors (38). Consequently, we would be able to differentiate distinct molecular mechanisms of cisplatin resistance induced by cigarette smoke–containing acrolein using these two cell models. Previous studies have shown that EMT coordinately regulates drug resistance and muscle invasion/metastasis in cancers (39). EMT is characterized by a loss of cell-to-cell adhesion and cell polarity and increases in metastasis and cell progression of bladder cancer (40). The most common change that occurs during EMT is downregulation of surface E-cadherin expression, increased expression of several transcriptional repressors of E-cadherin expression (Zeb-1, Zeb-2, Twist, Snail, and Slug) and increased expression of vimentin, fibronectin, and S100 proteins (41). Our results showed that E-cadherin is lost in RT4 Acr-clone#3 cells and vimentin and Snail are increased in T24 Acr-clone#1 cells compared with that in the parental cells (Supplementary Fig. S1). Consistently, we also found that cell migration activity was increased in both cisplatin-resistant clones (Fig. 2E and F). Interestingly, we found increased expression of EpCAM in RT4 Acr-clone#3 cells compared with that in its parental cells (Supplementary Fig. S1). EpCAM is an epithelial cell adhesion molecule (42), although it may also function to enhance the mobility of cells and tissues (43). Previous results have shown that EpCAM is overexpressed in many epithelial malignancies, including bladder carcinoma in situ and high-grade and advanced stage bladder cancers (44). Therefore, it is possible that EpCAM also contributes to increased cell proliferation and soft agar colony formation in RT4 Acr-clone#3 cells (Fig. 2A and C).

Overactivation of growth factor signaling pathways, including FGFR and ErbB/HER, result in diverse biological processes, including cell proliferation, survival, and tumorigenesis, and these pathways are known mechanisms of bladder cancer cisplatin resistance (7, 8). In this study, we found that acrolein induced distinct molecular alterations resulting in cisplatin resistance in two bladder cancer cell lines, RT4 and T24. Higher mRNA and protein expression levels of HER2 were observed in RT4 Acr-clone#3, while higher RNA and protein expression levels of FGFR3 were observed in T24 Acr-clone#1 (Fig. 3A–C). Previous studies have shown that RT4 belongs to low-grade papillary bladder tumors containing FGFR3 fusion and is more sensitive to FGFR inhibition (32, 33). Therefore, it is postulated that RT4 becomes addicted to FGFR signaling during the process of tumor formation. Consistently, we found that PD173074 alone induced higher cytotoxicity and cellular apoptosis in both parental RT4 cells and Acr-clone#3 cells (Fig. 4J and K). In this study, we found increased HER2 expression but not FGFR3 expression in RT4 Acr-clone#3 cells compared with that in the parental cells (Fig. 3A and C). In addition, short-term treatment with acrolein induced HER2 expression and downstream ERK and p38 activation in RT4 cells (Fig. 3D), and trastuzumab reversed cisplatin resistance in RT4 cells (Fig. 4A). Moreover, acrolein upregulated FGFR3 in T24 cells, which are high-grade muscle invasive bladder tumors containing wild-type FGFR3 (Fig. 3B, C, and E). In addition, PD173074 reversed cisplatin resistance in T24 cells (Fig. 4G). These results indicate that acrolein induced distinct molecular alterations that underlie cisplatin resistance in RT4 and T24 cells with different genetic backgrounds. Our previous studies have shown that acrolein induces DNA damage and inhibits DNA repair activity, thus leading to bladder carcinogenesis (26). However, the molecular mechanisms by which acrolein induces upregulation of HER2 or FGFR3 expression still need further investigation. In addition, only a single cell line was utilized to represent each subset of NMIBC and MIBC in this study; hence, further studies are required to confirm the findings.

Previous studies have shown that HER2 overexpression is frequently detected in chemotherapy- and radiation-resistant bladder cancer (14, 45). Therefore, anti-HER2 agents are able to induce cellular apoptosis in vitro and in vivo (15). In this study, we found that the anti-HER2 antibody trastuzumab combined with cisplatin could induce more cellular apoptosis and reverse cisplatin resistance in RT4 Acr-clone#3 cells (Fig. 4A and B) and inhibit HER2 downstream signaling, including the activation of ERK and p38 (Fig. 4C). These results indicate that upregulation of HER2 contributes to cisplatin resistance in RT4 cells induced by acrolein. Since HER2 alterations are common in urothelial bladder cancer (13), HER2 has been targeted by numerous agents in multiple prospective clinical trials (16). Trastuzumab was suggested to have clinical efficacy in advanced urothelial cancer (46, 47). Furthermore, several clinical trials also showed that trastuzumab could be feasibly combined with chemotherapy for advanced urothelial cancer overexpressing HER2 (48, 49). A two-phase clinical trial showed that patients in the trastuzumab arm who received cisplatin-based chemotherapy had superior outcomes, although the higher than expected rate of cardiac toxicity was concerning (50). Therefore, our results may provide insights into the molecular mechanisms of cisplatin resistance in advanced bladder cancer.

Activating mutations of FGFR3 have been described in approximately 75% of low-grade papillary bladder tumors and 20% of MIBC (51). Furthermore, overexpression of FGFR3 is observed in approximately half of MIBC (52). Previous studies have shown that the FGFR3^S249C mutation promotes chemoresistance in bladder cancer cells (53). Therefore, FGFR3 is a particularly promising target for therapy in bladder cancer. Previous studies have shown that FGFR inhibition (with the small molecule inhibitor PD173074) is extremely effective at potentiating the effect of cisplatin in lung cancer (12); however, similar work has not been carried out in bladder cancer thus far. In this study, we found that PD173074 combined with cisplatin could induce more cellular apoptosis, reverse cisplatin resistance in T24 Acr-clone#1 cells (Fig. 4G and H), and downregulate FGFR3 expression (Fig. 4I). Furthermore, we found that PD173074 induced intratumoral apoptosis and potentiated the effect of cisplatin in bladder cancer using a xenograft mouse model with acrolein-induced T24 Acr-clone#1 (Fig. 5). These results together suggest that upregulation of FGFR3 contributes to cisplatin resistance in T24 cells induced by acrolein and that PD173074 could reverse cisplatin resistance in vitro and in vivo.

In conclusion, our results show that long-term exposure to cigarette smoke–containing acrolein induced cisplatin resistance in NMIBC and MIBC cell lines, RT4 and T24, respectively. Distinct molecular mechanisms, including upregulation of HER2 or FGFR3 expression and downstream signaling pathways, were induced in these two bladder cancer cell lines. Our study indicates that different molecular alterations behind cisplatin resistance in two subsets of bladder cancer cell lines significantly alter the effectiveness of targeted therapy combined with chemotherapy.

No disclosures were reported.

J.-H. Hong: Funding acquisition, writing–original draft. Z.-J. Tong: Data curation, investigation, methodology. T.-E. Wei: Data curation, investigation, methodology. Y.-C. Lu: Data curation, methodology. C.-Y. Huang: Data curation, methodology. C.-Y. Huang: Resources, supervision, funding acquisition. C.-H. Chiang: Data curation, methodology. F.-S. Jaw: Data curation, methodology. H.-W. Cheng: Data curation, investigation, methodology. H.-T. Wang: Data curation, validation, methodology, writing–original draft, project administration, writing–review and editing.

We thank Prof. Y.C. Chang from Genomic Research Center at Academia Sinica for providing EpCAM antibody. This work was supported by National Health Research Institutes, Taiwan (under NHRI-EX110-11027PI and NHRI-EX111-11027PI to H.-T. Wang), Ministry of Science and Technology, Taiwan (under grant no. 109-2320-B-010-024, to H.-T. Wang), Yen Tjing Ling Medical Foundation (CI-110-11, to H.-T. Wang), Veterans General Hospitals and University System of Taiwan Joint Research Program (VGHUST111-G1-4-2, to H.-T. Wang), National Yang Ming Chiao Tung University Far Eastern Memorial Hospital Joint Research Program (NYCU-FEMH 110DN01 and 111DN01 to H.-T. Wang), and National Taiwan University Hospital (grant no. 110-N4915, to J.H. Hong), and by grants from Taiwan Health Foundation.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.