Prognostic significance of 5-fluorouracil metabolism-relating enzymes and enhanced chemosensitivity to 5-fluorouracil by 5-chloro 2,4-dihydroxy-pyridine in urothelial carcinoma

Background Recently, S-1, a novel 5-fluorouracil (5-FU)-based agent containing the strong dihydropyrimidine dehydrogenase (DPD) inhibitor, 5-chloro-2,4-dihydropyrimidine (CDHP) has been clinically used to treat various non-urothelial carcinomas (UC). High levels of thymidylate synthase (TS), the target enzyme of 5-FU and DPD which degrades the majority of 5-FU, are associated with poor prognosis in some cancers. However, only a few reports have dealt with this in UC. The aim of this study was to investigate the clinical significance of TS and DPD in upper tract urothelial carcinoma (UTUC) and evaluate the role of TS and DPD on the sensitivity of 5-FU in UC cell lines and the anti-tumor effect of S-1 in UC xenograft model. Methods Firstly, we evaluated the immunohistochemical expression of TS and DPD in 176 patients with UTUC to determine their prognostic significance. Secondly, the levels of TS and DPD in UC cell lines were measured by ELISA and real-time PCR. Furthermore, the association between their levels and the sensitivity to 5-FU was examined using the small interfering RNA (siRNA) specific for TS and DPD. Thirdly, the anti-tumor effect of S-1 was evaluated in UC xenograft model. Results Immunohistochemical evaluation of TS and DPD in UTUC human samples revealed that TS expression was significantly associated with stage, grade, and lymphovascular invasion and DPD expression was significantly associated with grade. Multivariate analysis revealed that high TS expression was an independent predictor of disease-specific survival in them. In in vitro study using UC cell lines, high levels of TS and DPD were associated with low response to 5-FU and these associations were confirmed with siRNA specific for TS and DPD. In in vivo study using UC xenograft model, S-1 treatment dramatically inhibited tumor growth compared to controls, tegafur, or UFT in UC tumor with a high level of DPD. Conclusions TS plays an important role in the prognosis of UTUC and S-1 may be a key agent for UC tumor, especially with a high level of DPD.


Background
Urothelial carcinoma (UC) occurs throughout the urinary tract, including the upper urinary tract, bladder, and urethra, however, most cases of UC involve the bladder. Upper tract urothelial carcinoma (UTUC) is very rare, accounting for approximately 5% of all UC [1]. The prognosis of UC patients with metastasis is poor and 5year survival rates for bladder cancer with lymph node metastasis and UTUC with it were reported to be 18-29% and 35%, respectively [2,3]. Systemic chemotherapy with a cisplatin-containing regimen is often proposed for patients with metastatic UC. Cisplatin-based chemotherapy has a short-term therapeutic effect against metastatic UC with a response rate of about 50%, however, longer survival after receiving systemic chemotherapy is low, with a 5-year survival rate of only 13-15% [4][5][6]. Thus, a novel chemotherapeutic regimen for treating highly aggressive UC is urgently needed. 5-Fluorouracil (5-FU), an antitumor pyrimidine, has frequently been used clinically in patients with various cancers, including UC. 5-FU is converted to 5fluorodeoxyuridine monophosphate and this inhibits thymidylate synthase (TS), which is the enzyme that catalyzes the methylation of deoxyuridine monophosphate to deoxythymodine monophosphate, by forming a stable ternary complex with methylene tetrahydrofolate, and the ternary complex leads to the inhibition of DNA synthesis [7,8]. Several studies in some cancers including bladder cancer have shown that TS activity was greater in cancerous tissue specimens than in normal tissue samples and that the TS activity level was correlated with stage progression [9,10]. Furthermore, it was reported that high TS expression was associated with a poor prognosis in gastric, colon, and bladder cancers [11][12][13]. However, to our knowledge, there have not been any reports which showed the clinical significance of TS in UTUC.
It was reported that the response-limiting factor for 5-FU was the plasma level of 5-FU. Therefore, continuous infusion was necessary to obtain higher responses because most administered 5-FU was degraded through a catabolic pathway by dihydropyrimidine dehydrogenase (DPD) [14]. Previous studies in some cancers including bladder cancer showed that DPD activity was greater in cancerous tissue specimens than in normal tissue samples and that DPD activity level was correlated with stage progression [15,16]. Furthermore, it was reported that high DPD expression was associated with a poor prognosis in gastric and colon cancers [17,18]. However, no report has ever demonstrated the clinical significance of DPD in UTUC.
Previous reports in gastric and colon cancers showed that high TS expression was significantly related to a low response to 5-FU [19][20][21]. Other reports in gastric and colon cancer patients found that a high DPD level resulted in a low sensitivity to 5-FU [22][23][24]. However, to the best of our knowledge, no report has showed an association between TS or DPD level and the sensitivity to 5-FU in UC cells. Recently, DPD-inhibitory fluoropyrimidine (DIF) compounds such as UFT and S-1 have been developed in an attempt to resolve the problem of rapid reduction of 5-FU by DPD. S-1 is a new oral formulation of DIF developed in Japan and consists of a strong DPD inhibitor, 5-chloro-2,4-dihydropyrimidine (CDHP; gimeracil), which is approximately 180 times more potent than the DPD inhibitor uracil, which is a component of UFT. Thus, S-1 results in higher concentrations of 5-FU in the blood and tumor tissue than UFT [25]. According to a previous report, UFT has been used successfully in the study of bladder cancer [26]. Because S-1 is thought to be more potent than UFT with respect to the biochemical modulation effect, one might expect a stronger antitumor effect by using S-1 in UC. However, only a few in vitro reports have showed that CDHP enhanced the antitumor activity of 5-FU in bladder cancer cell lines [16,27]. Furthermore, no in vivo study has demonstrated the enhancement of antitumor activity of 5-FU by CDHP, i.e., the efficacy of S-1 in UC.
Therefore, in the present study, we evaluated 1) the association between the tumor characteristics of 176 cases of UTUC and the expression of TS and DPD by immunohistochemistry with slides re-reviewed by genitourinary pathologists to determine the clinical role of TS and DPD expression in tumor progression and survival in UTUC. We also examined 2) the level of TS and DPD in UC cell lines and the association between TS or DPD level and the sensitivity to 5-FU in vitro. Finally, we evaluated 3) the enhancement of the antitumor activity to 5-FU by CDHP in vitro and in vivo.

Immunohistochemical evaluation of TS and DPD in UTUC human samples
Surgical specimens from 176 patients who had been surgically treated for UTUC at Keio University Hospital from 1986 to 2007 were examined. The median followup was 45 months and the median patient age was 67 years (range, 36-89 years). The patients did not undergo any chemotherapy or radiation therapy prior to the surgery. Patients with distant metastasis at the time of diagnosis and incomplete clinical data were excluded from the study. A nephroureterectomy with the removal of the bladder cuff was the most common procedure (n = 172), while a partial ureterectomy was performed in the remaining 4 patients. Regional lymphadenectomy was generally performed in patients with suspicious lymph nodes on preoperative axial imaging or with adenopathy detected during an intraoperative examination.
Extended lymphadenectomy was not routinely performed. The patients were followed postoperatively with urinary cytology every 3 months for 2 years and every 6 months thereafter. Computed tomography as well as cystoscopy and magnetic resonance imaging were performed every 6 months for 5 years and annually thereafter.
Tissue samples were obtained from consented patients in this study which was approved by Keio University Ethics Committee. All specimens were fixed in 10% formalin and embedded in paraffin, and all slides were rereviewed by genitourinary pathologists. Tumors were staged according to the American Joint Committee on Cancer-Union Internationale Contre le Cancer TNM classification. Tumor grading was assessed according to the 1998 WHO/International Society of Urology Pathology consensus classification [28]. Lymphovascular invasion (LVI) was defined as the presence of tumor cells within an endothelium-lined space without underlying muscular walls.
Sections (4 μm) of formalin-fixed and paraffinembedded material were analyzed. The sections were deparaffinized in xylene and rehydrated in graded alcohols and distilled water. After antigen retrieval with citric acid (pH 6.0), endogenous peroxidase activity was blocked with 1% hydrogen peroxide for 30 minutes followed by washing with distilled water. To bind nonspecific antigens, the sections were incubated with 5% skim milk for 15 minutes. The sections were incubated with either an anti-TS rabbit monoclonal antibody (1:100 dilution, Taiho Pharmaceutical Co, Tokyo, Japan) at room temperature for 1 hour or an anti-DPD mouse polyclonal antibody (1:100 dilution, Taiho Pharmaceutical Co.) at room temperature for 1 hour. After washing with phosphate-buffered saline (PBS), they were incubated with secondary antibodies against rabbit IgG conjugated to a peroxidase-labeled dextran polymer (no dilution, anti-rabbit Envision, Dako Japan, Tokyo) or against mouse IgG conjugated to a peroxidase-labeled dextran polymer (no dilution, antimouse Envision, Dako Japan) for 1 hour. Color was developed with 3,3'-diaminobenzamine tetrahydrochloride in 50 nmol/L Tris-HCl (pH 7.5) containing 0.005% hydrogen peroxidase. The sections were counterstained with hematoxylin.
To evaluate TS and DPD staining, cancer cells with positive staining in the cytoplasm were counted in at least 10 representative fields selected randomly by a uropathologist and staining intensity that was stratified from 0 to 3 (0, no staining; 1, weak staining; 2, moderate staining; 3, strong staining) was estimated. Cases with more than 25% positive tumor cells (moderate and strong staining) in a section were regarded as positive expression as previously described [29]. The evaluation of immunostaining was made by a uro-pathologist who was unaware of the clinico-pathological data and clinical outcomes of the patients.
5-FU was synthesized by Wako Pure Chemical Industries (Osaka, Japan) and CDHP, tegafur, UFT and S-1 were kindly supplied by Taiho Pharmaceutical Co., Ltd (Tokyo, Japan). 5-FU and CDHP were dissolved in culture medium to prepare a 100 μg/ml solution and subsequently diluted in culture medium. Tegafur, UFT and S-1 were dissolved in distilled water with 0.5% hydroxypropylmethylcellulose (HPMC).

Cell growth assay
Briefly, 2 × 10 4 cells were seeded into each well of 96well plates and allowed to grow overnight. The cells were then treated with various concentrations of 5-FU with or without CDHP. After 72 hours of incubation, cytotoxicity was determined using WST-1; 4-[3-(4lodophenyl)-2-(4-nitrophenyl)-2 H-5-tetrazolio]-1,3-benzene disulfonate (Takara Bio Inc, Shiga, Japan). The absorbance value of each well was determined at 450 nm with a 650 nm reference beam by a microplate reader (Bio-Rad Laboratories, Inc, Tokyo, Japan). As previously described [25], the 5-FU concentration causing 50% growth inhibition compared with the control (IC 50 ) was calculated from the regression line. These experiments were repeatedly performed at different days, where new cells were grown.

Enzyme-linked immunosorbent assasy (ELISA)
Each sample (1.0 × 10 8 cells or xenograft tumors) was homogenized in a 10-fold volume of sample weight of the diluting solution (20 mM PBS which contained 0.05% Tween 20) and centrifuged at 105,000× g, 4°C for an hour. The supernatant (100 μl) was then dispensed onto an anti-human TS or DPD polyclonal antibody (Solid phase antibody: Mitsubishi Chemical Medience, Tokyo, Japan) immobilized plate and incubated for an hour at room temperature. After the wells were washed four times with PBS, 100 μl aliquots of horseradish peroxidase were conjugated to anti-human TS or DPD polyclonal antibody (Label antibody: Mitsubishi Chemical Medience, Tokyo, Japan). After the wells were washed four times with PBS, 100 μl aliquots of 0.1 M acetate buffer (pH 5.5; color-developing solution) containing 3 mg/ml orthphenylenediamine and 0.75 mM hydrogen peroxide were added, followed by incubation for 30 minutes in the dark. Finally, 100 μl aliquots of 1 M sulfuric acid were added to terminate the reaction and the measurements were conducted with the measuring wavelength of an ELISA plate reader set at 490 nm.

Real-time quantitative PCR (RT-PCR)
The cells were lysed with RNAiso reagent (Takara Bio Inc, Shiga, Japan) according to the manufacturer's directions for total RNA extraction. RNA was quantitated by the ratio of absorbance at 260/280 nm. Reverse transcription of RNA to cDNA was carried out using a High Capacity cDNA Archive Kit (Applied Biosystems, Tokyo, Japan). Next, real time PCR was carried out in a final volume of 20 μl containing cDNA template, TS, DPD or GAPDH primers (Applied Biosystems, Tokyo, Japan) and TaqMan W Universal PCR Master Mix (Applied Biosystems, Tokyo, Japan), and DNase RNAase free water, using a StepOne real time PCR system (Applied Biosystems, Tokyo, Japan) according to the manufacturer's protocol. Cycling conditions were 50°C for 10 minutes, 95°C for 10 minutes, and then 40 cycles at 95°C for 15 seconds and 60°C for 1 minute. The data were then quantified using the comparative C t method for relative gene expression compared with GAPDH as endogenous control. The primers and TaqMan probe sets for TS (TYMS) (Hs00426591_m1), DPD (DPYD) (Hs00559279_m1), and human GAPDH endogenous control (Hs99999905_m1) were purchased from Applied Biosystems.

Small interfering RNA (siRNA)
Three predesigned siRNAs for DPYD, TYMS and nontargeting control (NTC) siRNA (AllStars Negative Control siRNA) were obtained from Invitrogen Co (Tokyo, Japan). UMUC-3 cells (1.5 × 10 5 per well) were cultured in antibiotic-free medium overnight at 37°C in 5% CO 2 and then transfected with each siRNA for DPYD or TYMS or nontargeting control, using Lipofectamine Max (Invitrogen Co, Tokyo, Japan). Forty-eight hours later, the transfected cells were washed and used for subsequent experiments.

Treatment in vivo
All of the procedures involving animals and their care in this study were approved by the Animal Care Committee of Keio University in accordance with institutional and Japanese government guidelines for animal experiments. Female BALB/c-nu/nu mice were obtained from Sankyo Lab Service Co. (Tokyo, Japan). UMUC-3 cells (2 × 10 6 ) were implanted subcutaneously into the flank of each nude mouse. When a mouse developed a palpable tumor, it was randomly assigned to one of 4 groups. Treatment groups were dosed at the maximal tolerable dose; tegafur 180 mg/kg/day, UFT 20 mg/kg/ day, and S-1 8.3 mg/kg/day [25]. Each drug was administered daily by gavage. Control animals only received vehicle by gavage. Each experimental group consisted of 10 mice. The mice were carefully monitored, and tumor size and body weight were measured every 4 days. Tumor volume was calculated according to the formula a 2 × b × 0.52, where a and b are the smallest and largest diameters, respectively. Eight weeks after tumor cell implantation, the mice were sacrificed and the tumors were collected. The levels of TS and DPD in the tumors were then measured by ELISA as described previously.

Statistical analysis
The association between TS and DPD, and clinicopathological features was assessed using χ 2 test. Diseasespecific survival (DSS) and progression-free survival (PFS) were calculated by the Kaplan-Meier method and analyzed by the log-rank test. Cox proportional hazards regression analysis was used to assess the prognostic indicators that included age, gender, tumor location, pT, grade, nodal involvement, LVI, TS expression, and DPD expression for survival. The difference between two groups in in vitro study and in the animal model was assessed with the Mann-Whitney U-test. The level of statistical significance was set at P <0.05. These analyses were performed with an SPSS Version 16.0 statistical software package (SPSS Corporation).

Results
The clinical role of TS and DPD expression in UTUC human samples TS and DPD expression in UTUC All the specimens were fixed in 10% formalin and embedded in paraffin, and all slides were re-reviewed by genitourinary pathologists. To elucidate the clinical significance of TS and DPD in UTUC, we examined the immunohistochemical expression of TS and DPD ( Figure 1A-D). TS staining was weak in pTa and low grade UTUC ( Figure 1A). On the other hand, in the pT3 and high grade UTUC, strong TS staining was observed in many of the cancer cells ( Figure 1C). DPD staining was weak in the low grade UTUC ( Figure 1B). On the other hand, in the high grade UTUC, strong DPD staining was observed in many of the cancer cells ( Figure 1D).

Association between TS and DPD expression and clinico-pathological features
We analyzed the association between TS and DPD ex- UTUC samples ( Table 1). The TS expression levels were significantly associated with pT-stage (p < 0.001), grade (p < 0.001), LVI (p < 0.001), and lymph node metastasis (p < 0.001). DPD expression was significantly associated with tumor grade (p < 0.001).

Prognostic significance of TS expression in overall UTUC patients
To examine the prognostic value of TS and DPD expression in UTUC patients, univariate and multivariate analyses were performed (Additional file 1). Univariate analysis revealed that the pT-stage, grade, LVI, lymph node metastasis, and TS were all significant predictors of the PFS and DSS. Multivariate Cox regression analysis showed that pT-stage, LVI, and TS expression were independent prognostic indicators of PFS and DSS.

The in vitro study using UC cell lines The levels of TS and DPD in UC cell lines
The levels of TS and DPD in three UC cell lines are shown in Additional file 2. The median TS protein level in T24 cells was 2375 (range:2283-2475) ng/mg, which was the highest among the three cell lines. The relative TS mRNA levels exhibited the same pattern as the protein levels. The median DPD protein and relative DPD mRNA levels in UMUC-3 cells (232.6 ng/mg protein, 13.0; DPD/GAPDH ratio) were the highest among the three cell lines.

Cytotoxic effects of 5-FU treatment with or without CDHP
In T24 and UMUC-3 cells, 5-FU at concentrations of 10 μg/ml or higher had significant cytotoxic effects (Figure 2Aa, b). Meanwhile, in 5637 cells, significant growth inhibition by 5-FU was seen at concentrations of 0.6 μg/ml or higher (Figure 2Ac). CDHP alone did not demonstrate a cytotoxic effect in bladder cancer cells tested at concentrations of 30 μg/ml or lower (data not shown).
Additional file 2 shows that the values of IC 50 for 5-FU with or without CDHP (1 μg/ml) in the three cell lines. It was also observed that the value of IC 50 of 5-FU in UMUC-3 cells treated with 5-FU and CDHP (25.0 ± 0.8 μg/ml) was significantly lower than that in UMUC-3 cells treated with 5-FU alone (53.4 ± 1.4 μg/ml) (p < 0.05). Meanwhile, there were no significant differences in the IC 50 between the combination treatment of 5-FU with CDHP (54.5 ± 2.1 and 2.8 ± 0.3 μg/ml) and treatment with 5-FU alone (58.3 ± 1.8 and 3.0 ± 0.3 μg/ml) in T24 and 5637 cells, respectively. The addition of 1 μg/ml of CDHP enhanced the growth inhibitory effect of 5-FU . Cell viability was measured by WST assay. Each value represents the mean derived from at least three individual experiments; bars ± S.E.*, more statistically significant than controls (p < 0.05). by 2.1-fold in UMUC-3 cells. Also, no significant cytotoxic enhancement by CDHP was observed in 5-FU treatment in T24 and 5637 cells.
Effects of siRNA for DPD and TS on the sensitivity to 5-FU in UMUC-3 cells Transfection of siRNA for DPD reduced the mRNA level of DPD to 10% in UMUC-3 cells (p < 0.05) whereas no significant difference was observed by transfection with NTC compared to vehicle control ( Figure 3A). 5-FU (3 μg/ml or higher) treatment in UMUC-3 cells transfected with siRNA for DPD resulted in significantly higher levels of cell growth inhibition compared to that in UMUC-3 cells transfected with NTC or treated with vehicle control ( Figure 3B). Also, the IC 50 value for 5-FU in UMUC-3 cells transfected with siRNA for DPD (9.6 ± 0.4 μg/ml) was significantly lower than that treated with vehicle control (53.4 ± 1.4 μg/ml) or transfected with NTC (52.1 ± 1.2 μg/ml).
Transfection of siRNA for TS reduced the mRNA level of DPD to 10% in UMUC-3 cells (p < 0.05) whereas no significant difference was observed by transfection with NTC compared to control ( Figure 3C). 5-FU (3 μg/ml or higher) treatment in UMUC-3 cells transfected with siRNA for TS resulted in significantly higher levels of cell growth inhibition compared to that in UMUC-3 cells transfected with NTC or treated with vehicle control ( Figure 3D). Furthermore, the IC 50 value for 5-FU in UMUC-3 cells transfected with siRNA for TS (4.5 ± 0.5 μg/ml) was significantly lower than that treated with vehicle control (53.6 ± 1.2 μg/ml) or transfected with NTC (52.3 ± 1.0 μg/ml).

Discussion
The clinical role of TS and DPD expression in UTUC human samples Recently, some reports about the association between the level of TS and DPD and prognosis in some cancers including UC have ever been found [11][12][13][16][17][18]29]. In our present study, we retrospectively analyzed the impact of TS and DPD protein expression by immunohistochemistry especially in patients with UTUC. The results indicated that TS expression was closely related to tumor grade, stage, and LVI. On the other hand, DPD expression was closely related to tumor grade. To our knowledge, this is the first study examining the expression pattern of TS and DPD in case-matched large UTUC specimens. Recently, histologic evidence of TS expression in human tissue samples has also been reported in various cancers and it has been reported that TS expression was a significant prognostic indicator in gastric and colon cancers [11,12]. Additionally, it has also been reported that the level of TS is correlated with both progression of the stage and increase in the grade, and that TS expression was a significant prognostic indicator of recurrence free survival in patients with bladder cancer [13,29]. In this study, we demonstrated that higher TS expression was significantly associated with a lower PFS and DSS and was an independent prognostic factor in UTUC. These results indicate that the role of TS in UTUC might be of special importance for its progression and the establishment of metastasis.
On the other hand, histologic evidence for DPD expression in human tissue samples has also been reported in various cancers and it has been reported that DPD expression was a significant prognostic indicator in breast and colon cancers [17,18]. Additionally, it has also been reported that the level of DPD was correlated with both progression of the stage and increase in the grade of bladder cancer. However, according to previous reports, DPD expression was not a significant prognostic indicator of recurrence free survival in patients with bladder cancer [16]. In the present study, we found a significant association between DPD expression and grade in UTUC. However, we showed that DPD expression was not significantly associated with PFS and DSS in UTUC. This result was almost the same as that for bladder cancer in a previous report [16].

The in vitro study using UC cell lines
There is some evidence in the literature indicating that an association between the levels of TS and DPD and the sensitivity to 5-FU. Previous reports found that the TS and DPD levels were significantly correlated with 5-FU sensitivity, with high TS and DPD levels resulting in low sensitivity to 5-FU in various cancers [19][20][21][22][23][24]. Our study showed that the TS level was the highest in T24 cells and the DPD level was the highest in UMUC-3 cells. Furthermore, the sensitivity to 5-FU in T24 and UMUC-3 cells was lower than that in 5637 cells. Therefore, the high TS and DPD levels in the present study may have resulted in low sensitivity to 5-FU in UC cells.
In the present study, targeting of TS and DPD mRNA expression in UMUC-3 cells with siRNA down-regulated the mRNA expression and enhanced sensitivity to 5-FU using WST-assay. Thus, from these results, we demonstrated that there was a close correlation between the levels of TS and DPD and the sensitivity to 5-FU in UC cells.

The in vivo study using UC xenograft model
Recently, it was reported that S-1 containing CDHP had a significantly higher anti-tumor effect than UFT in various cancers [25]. In fact, S-1 has been clinically used in various cancers, such as of the stomach, colon, lung and pancreas. However, in UC, S-1 is not clinically used at present. Our in vitro study demonstrated the enhancement of sensitivity to 5-FU by CDHP in UMUC-3 cells. Furthermore, in in vivo study, S-1 had more anti-tumor effect than UFT. Thus, these results indicate that S-1 is more effective than UFT, particularly in UC tumor with a higher DPD level. With respect to the levels of TS and DPD in tumors treated with 5-FU, it was reported that the TS level was significantly higher and the DPD level was significantly lower than those in control in patients with colorectal cancer [30]. Our study also showed that the level of TS protein in tumors treated with S-1 was significantly higher than that in tumors treated with vehicle alone, tegafur or UFT. On the other hand, the level of DPD in tumors treated with S-1 was significantly lower than that in tumors treated with vehicle alone, tegafur or UFT. Thus, the result of our present study exhibited almost the same pattern as that of the previous study in patients with colorectal cancer. S-1 511*,**,*** (487-531) 351*,**,*** (331-382) *p < 0.05 compared with control; **p < 0.05 compared with tegafur. ***p < 0.05 compared with UFT. TS: thymidylate synthase; DPD: dihydropyrimidine dehydrogenase.