Identification of thymidylate synthase as a potential therapeutic target for lung cancer

Background: Thymidylate synthase (TS), a key enzyme in the de novo synthesis of thymidine, is an important chemotherapeutic target for malignant tumours including lung cancer. Although inhibition of TS has an antiproliferative effect in cancer cells, the precise mechanism of this effect has remained unclear. Methods: We examined the effects of TS inhibition with an RNA interference-based approach. The effect of TS depletion on the growth of lung cancer cells was examined using colorimetric assay and flow cytometry. Results: Measurement of the enzymatic activity of TS in 30 human lung cancer cell lines revealed that such activity differs among tumour histotypes. Almost complete elimination of TS activity by RNA interference resulted in inhibition of cell proliferation in all tested cell lines, suggestive of a pivotal role for TS in cell proliferation independent of the original level of enzyme activity. The antiproliferative effect of TS depletion was accompanied by arrest of cells in S phase of the cell cycle and the induction of caspase-dependent apoptosis as well as by changes in the expression levels of cyclin E and c-Myc. Moreover, TS depletion induced downregulation of the antiapoptotic protein X-linked inhibitor of apoptosis (XIAP), and it seemed to activate the mitochondrial pathway of apoptosis. Conclusion: Our data provide insight into the biological relevance of TS as well as a basis for clinical development of TS-targeted therapy for lung cancer.

Thymidylate synthase (TS) is an essential enzyme that catalyses the transfer of a methyl group from methylenetetrahydrofolate to dUMP to generate dTMP (Carreras and Santi, 1995). The subsequent phosphorylation of dTMP to dTTP provides a direct precursor for DNA synthesis. The level of TS expression is increased in highly proliferative cells, and an increased abundance of TS in a broad range of tumours is associated with a poor treatment response and prognosis (Costi et al, 2005). Transfection of nontransformed cells with an expression vector for TS has also been found to confer transformed-like behaviour, suggesting that TS itself is a potential oncoprotein (Rahman et al, 2004). These findings have led to TS being considered as a molecular target for cancer therapy. To date, the antiproliferative effect of TS inhibition has been examined mostly with the use of drugs such as 5-fluorouracil and its active metabolite 5-fluoro-dUMP, the former of which is used in cancer chemotherapy. Although studies with antisense oligodeoxynucleotides have also shown that depletion of TS results in growth inhibition in human tumour cells (Ferguson et al, 1999;Lin et al, 2001;Flynn et al, 2006), the underlying mechanism of the antiproliferative effect of specific TS inhibition has remained largely unknown.
Lung cancer is one of the most common forms of cancer worldwide. Advanced lung cancer is treated by combination chemotherapy, although further improvement in such therapy is warranted. High levels of TS in tumours have been associated with a poor response to TS-targeting agents in individuals with advanced lung cancer (Oguri et al, 2005;Kubota et al, 2009;Ozasa et al, 2009), although the biological relevance of TS in lung cancer has remained to be well established. We have now abrogated both the expression and activity of TS in lung cancer cells by the application of RNA interference (RNAi). With this approach, we investigated the precise mechanism of the antiproliferative effect of TS depletion in lung cancer cells and further examined the potential role of TS as a target for chemotherapeutic agents in these cells. Our results provide a basis for the further development of TS-targeted therapy in lung cancer patients.

Assay of TS activity
Thymidylate synthase activity was quantified using tritiated 5-fluoro-dUMP binding assay (Spears et al, 1984). Cells were harvested, diluted in 0.2 M Tris-HCl (pH 7.4) containing 20 mM 2-mercaptoethanol, 15 mM CMP, and 100 mM NaF, and disrupted by ultrasonic treatment. The cell lysate was centrifuged at 1600 g for 15 min at 41C, and the resulting supernatant was centrifuged at 105 000 g for 1 h at 41C. A portion (50 ml) of the final supernatant was mixed consecutively with 50 ml of Buffer A (600 mM NH 4 HCO 3 buffer (pH 8.0), 100 mM 2-mercaptoethanol, 100 mM NaF, and 15 mM CMP) and with 50 ml of [6-3 H]5-fluoro-dUMP (7.8 pmol) plus 25 ml of cofactor solution (50 mM potassium phosphate buffer (pH 7.4), 20 mM 2-mercaptoethanol, 100 mM NaF, 15 mM CMP, 2% bovine serum albumin, 2 mM tetrahydrofolic acid, 16 mM sodium ascorbate, and 9 mM formaldehyde). The resulting mixture was incubated at 301C for 20 min, after which the reaction was terminated by the addition of 100 ml of 2% bovine serum albumin and 275 ml of 1 M HClO 4 and by centrifugation at 1630 g for 15 min at 41C. The resulting precipitate was suspended in 2 ml of 0.5 M HClO 4 , and the mixture was subjected to ultrasonic treatment followed by centrifugation at 1600 g for 15 min at 41C. The final precipitate was solubilised with 0.5 ml of 98% formic acid, mixed with 10 ml of ACS II scintillation fluid, and assayed for radioactivity.

Growth inhibition assay in vitro (MTT assay)
Cells were plated at 50 -60% confluence in 25 cm 2 flasks and then incubated for 24 h before transient transfection with an siRNA specific for TS mRNA or a control siRNA as described above. The cells were then isolated by exposure to trypsin, transferred to 96-well flat-bottom plates, and cultured for 72 h before the addition of TetraColor One (5 mM tetrazolium monosodium salt and 0.2 mM 1-methoxy-5-methyl phenazinium methylsulfate; Seikagaku, Tokyo, Japan) to each well and incubation for an additional 3 h at 371C. The absorbance at 490 nm of each well was measured using Multiskan Spectrum instrument (Thermo Labsystems, Boston, MA, USA), and absorbance values were expressed as a percentage of that for nontransfected control cells.

Subcellular fractionation
A cytosolic fraction was isolated from cells by centrifugation. In brief, cells were washed, resuspended in homogenisation buffer (0.25 M sucrose, 10 mM HEPES-NaOH (pH 7.4), and 1 mM EGTA), and homogenised by 50 strokes in a Dounce homogeniser. The homogenate was centrifuged at 1000 g for 15 min at 41C to remove nuclei and intact cells, and the resulting supernatant was centrifuged at 10 000 g for 15 min at 41C. The final supernatant (cytosolic fraction) was subjected to immunoblot analysis.

Statistical analysis
Data were analysed using Student's two-tailed t-test. A P-value of o0.05 was considered statistically significant.

TS activity varies among histotypes of lung cancer cells
We first examined the enzymatic activity of TS in 30 human lung cancer cell lines (Table 1). The median TS activity in small cell lung cancer (SCLC) lines was significantly higher than that in non-SCLC (NSCLC) lines. Among NSCLC cell lines, the values for squamous cell carcinoma were higher than those for non-squamous cell carcinoma ( Figure 1). There was no significant correlation between TS activity and cell proliferation rate among these lung cancer cell lines (data not shown). These data thus suggested that TS activity varies according to histotype among lung cancer cell lines.

TS depletion induces growth inhibition regardless of original TS activity level in lung cancer cells
We next examined the effect of TS depletion by RNAi on the growth of lung cancer cell lines. The abundance of TS was markedly decreased as a result of cell transfection with either of three different siRNAs targeted to TS mRNA ( Figure 2A). Given that the TS-1 siRNA induced the most pronounced downregulation of TS expression, we selected this siRNA for use in subsequent experiments. In all tested lung cancer cells, transfection with TS-1 resulted in marked depletion of TS, whereas no such effect was observed in cells transfected with a nonspecific siRNA ( Figure 2B). Moreover, transfection of cells with TS-1 resulted in a 490% decrease in TS activity compared with that in corresponding cells transfected with a nonspecific siRNA or in untreated cells ( Figure 2C), regardless of the original levels of TS expression and activity. The antiproliferative effect of TS depletion was evaluated using the MTT assay. Depletion of TS resulted in pronounced inhibition of proliferation in all tested cells compared with the corresponding cells transfected with a nonspecific siRNA or untreated cells ( Figure 2D), and this antiproliferative effect was found to be time dependent ( Figure 2E). These data thus suggested that the almost complete elimination of TS activity resulted in marked inhibition of the proliferation of lung cancer cells regardless of the original level of such activity.

TS depletion induces S-phase arrest in lung cancer cells
To investigate the mechanism by which TS depletion inhibits lung cancer cell growth, we examined the cell cycle profile by flow cytometry. Transfection with TS-1 siRNA increased the proportion of cells in S phase of the cell cycle and reduced that of cells in G 1 or G 2 -M phases in all tested cell lines regardless of histotype ( Figure 3A). Immunoblot analysis of proteins implicated in regulation of the G 1 -S transition revealed that TS depletion increased the abundance of cyclin E in all tested cell lines ( Figure 3B) without affecting that of cyclins D or A (data not shown). In addition, TS depletion induced downregulation of c-Myc ( Figure 3B), a transcription factor that activates the expression of several cell cycle-related genes. However, expression of c-Myc was not detected in H69 cells, consistent with previous observations (Plummer et al, 1993). These results thus suggested that the S-phase arrest induced by TS depletion in lung cancer cells was related to upregulation of cyclin E and downregulation of c-Myc.

TS depletion induces caspase-dependent apoptosis in lung cancer cells
We next examined the effect of TS depletion on apoptosis in lung cancer cells. Flow cytometric analysis revealed that TS depletion induced a time-dependent increase in the size of the sub-G 1 (apoptotic) cell population ( Figure 4A). Depletion of TS also induced the cleavage of PARP ( Figure 4B), a characteristic of apoptosis, in the cell lines examined. Furthermore, the activity of caspase-3 in cell lysates was increased as a consequence of TS depletion ( Figure 4C), and previous exposure of lung cancer cells to the pan-caspase inhibitor ZVAD-FMK significantly inhibited the increase in the size of the sub-G 1 cell population induced by depletion of TS ( Figure 4D). These data thus indicated that TS depletion induces caspase-dependent apoptosis in lung cancer cells.

TS depletion activates the mitochondrial pathway of apoptosis and induces downregulation of XIAP
To elucidate further the mechanism of apoptosis induced by TS depletion, we examined the expression of members of the Bcl-2   (Hengartner, 2000). Although depletion of TS did not substantially affect the expression levels of Bcl-2, Bcl-x L , Bax, Bak, and survivin, it resulted in a substantial decrease in the abundance of X-linked inhibitor of apoptosis (XIAP) ( Figure 5A). The activity of XIAP is siRNA:  modulated by mitochondrial proteins such as cytochrome c, Smac (also known as Diablo), and Omi (also known an HtrA2) (Hengartner, 2000;Srinivasula et al, 2003;Martinez-Ruiz et al, 2008). To investigate the mechanism of the downregulation of XIAP induced by TS depletion, we examined the release of these mitochondrial proteins into the cytosol. Immunoblot analysis revealed that the amounts of these mitochondrial proteins in the cytosol were increased by TS depletion in a time-dependent manner ( Figure 5B). These data thus suggested that TS depletioninduced apoptosis is mediated, at least in part, by the mitochondrial signalling pathway.

DISCUSSION
Studies of TS-targeted therapy as well as the role of TS in DNA synthesis have provided the rationale for consideration of this enzyme as a prime therapeutic target. However, the precise mechanism by which inhibition of TS results in inhibition of tumour cell growth has remained incompletely understood.
The aim of this study was therefore to investigate the underlying mechanism of the antiproliferative effect of specific TS inhibition in lung cancer cells with the use of an siRNA-based approach.
We first examined TS activity in lung cancer cell lines of different histotypes. Thymidylate synthase activity was determined with the use of the well-established 5-fluoro-dUMP binding assay. We found that TS activity was significantly higher in SCLC lines than in NSCLC lines, and that, among the latter, TS activity was higher in squamous cell carcinoma lines than in non-squamous cell carcinoma lines. A previous microarray analysis showed that mRNAs for proliferation-related proteins including TS were more abundant in SCLC lines than in NSCLC lines (Bhattacharjee et al, 2001). In addition, recent studies showed that the amount of TS mRNA was higher in resection specimens from patients with squamous cell carcinoma of the lung than in those from individuals with other histotypes of NSCLC (Ceppi et al, 2006;Ishihama et al, 2009;Monica et al, 2009). Given that TS activity in lung cancer cell lines was proportional to the abundance of TS protein in the present study (data not shown), our data showing a differential profile of TS activity among histotypes of lung cancer are consistent with these previous findings. The cell line SCLC differs from NSCLC in terms of its faster growth and earlier spread (Allen and Jahanzeb, 2008), and recent clinical trials in NSCLC patients have revealed a poorer prognosis for squamous cell carcinoma than for adenocarcinoma (Asamura et al, 2008). The differential activity of TS among histotypes of lung cancer is thus suggestive of a role for this enzyme in promoting cell proliferation, with TS activity being a potential marker of tumour aggressiveness in lung cancer, although TS activity was not correlated with cell proliferation rate among the lung cancer cell lines examined in this study. We induced downregulation of both TS abundance and enzymatic activity in lung cancer cell lines by RNAi. The almost complete elimination of TS activity was associated with a marked antiproliferative effect in all tested lung cancer cell lines, including those with an original relatively low level of TS activity. These data suggest that TS is important for tumour cell proliferation in a manner independent of the original activity level. We found that depletion of TS induced S-phase arrest and caspase-dependent apoptosis in lung cancer cells. Previous studies have shown that TS inhibition results in an imbalance between the amounts of dUTP and dTTP and a consequent decrease in the efficiency of DNA synthesis (Curtin et al, 1991;Houghton et al, 1993). Furthermore, this dUTP -dTTP imbalance results in misincorporation of dUTP into DNA and consequent DNA damage (Curtin et al, 1991;Houghton et al, 1993). In this study, we examined the effect of TS depletion on DNA damage as determined by immunofluorescence imaging of histone g-H2AX foci, a sensitive and specific marker of DNA double-strand breaks (Burma et al, 2001;Stiff et al, 2004). Such foci were detected in B90% of lung cancer cells transfected with TS siRNA (Supplementary Figure S1). Given that DNA damage or a reduced rate of DNA synthesis triggers the S-phase checkpoint mechanism (Sclafani and Holzen, 2007), the observed S-phase arrest induced by TS depletion likely results from activation of the S-phase checkpoint. Cellular responses to DNA damage are important for maintenance of genomic stability and cellular integrity (Bunz et al, 1998;Hirao et al, 2000). Cells either repair DNA damage or, if it is too severe for repair, initiate the cell death program (Zhao et al, 2001). Our data thus suggest that cells that arrest in S phase after TS depletion subsequently undergo apoptosis as a result of the accumulation of unrepairable DNA damage. We further showed that TS depletion resulted in upregulation of cyclin E and downregulation of c-Myc. Both cyclin E and c-Myc contribute to the transition of cells from G 1 to S phase (Wang et al, 2008;Malumbres and Barbacid, 2009)   S-phase arrest induced by DNA damage or inhibition of DNA synthesis in tumour cells (Mazumder et al, 2000;Leonce et al, 2001;Lu et al, 2009;Sankar et al, 2009). The effects of TS depletion on the abundance of cyclin E and c-Myc therefore likely contribute to the associated S-phase arrest and caspase-dependent apoptosis in lung cancer cells. Our present data thus suggest that the antiproliferative effect of TS depletion is attributable to S-phase arrest and the induction of caspase-dependent apoptosis in these cancer cells.
Our investigation of the mechanism by which TS depletion led to caspase-dependent apoptosis revealed that elimination of TS resulted in downregulation of XIAP, a member of the IAP family of proteins. Activation of the mitochondrial signalling pathway for apoptosis results in inhibition of IAP proteins and consequent promotion of caspase-dependent apoptosis (Hengartner, 2000;Takasawa et al, 2005;Yu et al, 2007). We also found that TS depletion resulted in the release of mitochondrial proteins, including cytochrome c, Smac/Diablo, and Omi/HtrA2, into the cytosol, suggestive of a link between activation of the mitochondrial pathway and downregulation of XIAP in lung cancer cells depleted of TS. Activation of the mitochondrial pathway is induced by a variety of stimuli including DNA damage (Hengartner, 2000). Given that TS depletion induced DNA double-strand breakage, our data suggest that loss of TS may contribute to activation of the mitochondrial pathway of apoptosis. We found that TS depletion did not affect the expression level of the IAP protein survivin. Further study will thus be needed to elucidate the precise mechanism by which XIAP is downregulated specifically in TS-depleted cells.
In conclusion, we have shown that the almost complete elimination of TS activity with an RNAi-based approach resulted in an apparently universal antiproliferative effect in lung cancer cells that was attributable to S-phase arrest and the induction of apoptosis. High levels of TS expression have been suggested to predict resistance to TS-targeted agents such as 5-fluorouracil (Johnston et al, 2003;Showalter et al, 2008). The new TS-targeted agent pemetrexed was found to have low activity in the treatment of SCLC (Ceppi et al, 2006;Socinski et al, 2009), possibly as a result of a high level of TS expression in such tumours. Our results now suggest that TS depletion inhibits the growth of lung cancer cells including SCLC cells with a high original activity of TS. This apparent discrepancy may be explained by the fact that 5-fluorouracil and pemetrexed inhibit TS activity by only B60% (van Triest et al, 1997(van Triest et al, , 1999Codacci-Pisanelli et al, 2002;Giovannetti et al, 2008), whereas our siRNA-based method inhibit TS activity almost completely. Our present results thus suggest that novel TS-targeted agents with an increased inhibitory efficacy might prove beneficial for the treatment of lung cancer regardless of histotype. They further provide a rationale for future clinical investigation of the therapeutic efficacy of TS-targeted agents for lung cancer patients. Supplementary