Characterization of a Thymidylate Synthase (TS)-inducible Cell Line: A Model System for Studying Sensitivity to TS- and non-TS-targeted Chemotherapies¹ Free

TS³ is a folate-dependent enzyme that catalyzes the conversion of dUMP and 5,10-methylene tetrahydrofolate to dTMP and dihydrofolate, providing the sole intracellular source of dTMP, which is essential for DNA replication and repair (1, 2, 3). Because of its central importance to DNA replication, TS has become an important target for fluoropyrimidine drugs, such as 5-FU, used in the treatment of gastrointestinal, breast, and head and neck cancers (2, 3, 4). More recently, inhibitors of TS have been designed that have high affinity for the folate binding site of TS, TDX, MTA, and ZD9331 (5, 6, 7, 8, 9). TDX and ZD9331 are TS specific, but MTA is also able to inhibit dihydrofolate reductase, as well as the purine biosynthetic enzyme glycinamide ribonucleotide formyltransferase (10). Intracellularly, MTA and TDX become polyglutamated by folylpolyglutamate synthase, which increases their potency as TS inhibitors by prolonging their cellular retention (5, 8). Because ZD9331 contains moieties that mimic glutamate residues, it does not require polyglutamation, which makes it more active than TDX and MTA in tumors that express low levels of folylpolyglutamate synthase (11).

In vitro and in vivo studies have shown a strong association between increased TS expression and the development of 5-FU and TDX resistance (12, 13, 14, 15, 16, 17). The molecular basis for the elevation in TS protein levels after 5-FU treatment was found to be the inhibition of a TS autoregulatory translational feedback loop (18, 19). TS is now known to form ribonucleoprotein complexes with a number of other mRNAs including p53 and c-myc (20, 21, 22). Increased TS expression also appears to play a significant role in the development of resistance to the folate-based TS inhibitors (15, 16). Taken together, these observations support a critical role for TS in defining the activity of fluoropyrimidines and antifolate-based TS inhibitors in cancer cells.

In this study, we have developed a tet-inducible TS expression system to examine the effect of specifically up-regulating TS expression on the response of malignant cells to the TS inhibitors 5-FU, TDX, ZD9331, and MTA, and the non-TS-specific chemotherapeutic agents cisplatin, oxaliplatin, CPT-11, and Taxol, which are commonly used in patient treatment in conjunction with 5-FU.

The TS coding region was PCR amplified from the pMHTS-1 construct (Ref. 23; a kind gift from Dr. K. Takeishi, Department of Immunology and Virology and Department of Biochemistry, Saitama Cancer Research Institute, Japan) with the introduction of SacII restriction sites at the 5′ and 3′ ends. The 3′-oligonucleotide primer was designed so that six extra codons that specify the Hemagglutinin (HA) epitope are translated at the COOH terminus. The modified TS coding region was ligated into the pUHD10-3 tet-inducible expression vector (Ref. 24; Clontech Laboratories, Palo Alto, CA) to generate a TS-inducible construct.

The MDA47 founder cell line was generated by the stable transfection of the human MDA435 breast cancer cell line with the pUHD15-1 tet transactivator construct (Clontech Laboratories) under selection with Geneticin (G418) as described previously (25). Ten μg of the TS-inducible construct and 1 μg of a plasmid that expresses the puromycin resistance gene were cotransfected into the MDA47 cell line using the lipofectin reagent according to the manufacturer’s instructions (Life Technologies, Inc., Paisley, Scotland). Transfected cells were selected in 1 μg/ml puromycin, and resistant colonies were isolated. Inducible expression of the TS trans-gene was assessed by Northern blotting, and the MTS-5 clone was selected.

MTS-5 cells were maintained in 5% CO₂/37°C in DMEM with 10% dialyzed bovine calf serum (both from Life Technologies, Inc.), supplemented with 1 mm sodium pyruvate, 2 mm l-glutamine, 50 μg/ml penicillin/streptomycin, 50 μg/ml G418 (all from Life Technologies, Inc.), 1 μg/ml puromycin, and 1 μg/ml tet (both from Sigma Chemical Co., Poole, Dorset, England). To induce expression of the TS trans-gene, MTS-5 cells were washed three times with 1× PBS and incubated in culture-medium lacking tet.

RNA was extracted from cultured cells using RNAStat60 (Biogenesis, Poole, Dorset, England) according to the manufacturer’s instructions. Twenty μg of each RNA sample was size fractionated on a 1% agarose formaldehyde gel, transferred to nitrocellulose filters (Hybond-N, Amersham, Little Chalfont, Buckinghamshire, England), and UV cross-linked. The resulting blot was hybridized overnight at 65°C in 7% SDS, 250 mm Na₂HPO₄ with an [α-³²P]-radiolabeled probe complementary to the entire TS coding region, which was obtained by SacII restriction digestion of the TS-inducible construct. The membrane was washed at 65°C twice in 5% SDS/20 mm Na₂HPO₄, and twice in 1% SDS/20 mm Na₂HPO₄ for 20 min each time, prior to detection.

Cells were washed twice in ice-cold 1× PBS, harvested, and resuspended in 20 mm Tris-HCl (pH 7.4), 150 mm NaCl, 1 mm EDTA (pH 8.0), 1% Triton X-100, and 0.1% SDS. Cells were then lysed by sonication using three 2-to-3-s bursts and were centrifuged at 10,000 × g for 15 min to remove cell debris. Protein concentrations were determined using the BCA Protein Assay Reagent (Pierce, Rockford, IL). Fifty μg of each protein sample were resolved by SDS-PAGE (10%) according to the method of Laemmli (26). The gels were electroblotted onto nitrocellulose membranes (Hybond-P, Amersham). Antibody staining was performed with a chemiluminescence detection system (ECL+, Amersham), using a 1:200 dilution of the mouse TS106 monoclonal primary antibody (NeoMarkers, Fremont, CA) and 1:1000 dilutions of cyclin E, cyclin A, and cdk2 mouse monoclonal antibodies (Santa Cruz Biotechnology, Santa Cruz, CA) in conjunction with a 1:2000 dilution of a horseradish peroxidase-conjugated sheep antimouse secondary antibody (Amersham). Membranes were stripped in 0.3% glycine-HCl (pH 2.8), 1 m NaCl for 1 h prior to reprobing with a 1:1000 dilution of a β-tubulin primary antibody (Sigma Chemical Co.) followed by a 1:2000 of the horseradish peroxidase-conjugated sheep antimouse secondary antibody.

Cells were harvested from T175 tissue culture flasks, washed three times in 1× PBS and resuspended in 230 μl of 80 mm KH₂PO₄/20 mm K₂HPO₄ prior to being lysed by sonication as described above. FdUMP binding activities were determined as described previously (4).

MTS-5 cells were seeded onto 24-well tissue culture plates at 1× 10⁵ cells per well. After 24 h, cells were washed three times in 1× PBS and incubated in either +tet or −tet culture medium, supplemented with the described concentrations of 5-FU, TDX, ZD9331, MTA, cisplatin, oxaliplatin, CPT11, and Taxol. After 72 h of continuous drug exposure, cells were harvested and counted using a Z2 particle and size analyzer (Coulter, Miami, Fl).

MTS-5 cells were seeded at 5× 10⁵ per well of 6-well tissue culture plates. After 24 h, cells were washed three times in 1× PBS and incubated in either +tet or −tet culture medium, supplemented with no drug or with 2.6 μm 5-FU, 2 nm TDX, 32 nm MTA, 2.2 μm oxaliplatin, or 140 nm CPT11. Triplicate samples were harvested after 24, 48, and 72 h, and the DNA content was evaluated after propidium iodide staining of cells as described previously (27). Fluorescence-activated cell sorting analysis was carried out using an EPICS XL flow cytometer (Coulter).

We have generated a tet-inducible (tet off) TS expression system in the p53 mutant MDA435 breast cancer cell line called MTS-5. Northern blot analysis demonstrated that removal of tet from the MTS-5 culture medium caused rapid transcriptional activation of the TS trans-gene, and TS mRNA levels continued to rise throughout the time course (Fig. 1,A). Western blot analysis of TS protein expression indicated that the trans-gene product migrated more slowly than endogenous TS because of the additional amino acid residues of the Hemagglutinin (HA) epitope (Fig. 1,B, −tet lanes). Exogenous TS protein levels were not significantly elevated until 12 h postinduction, and the maximum induction of exogenous TS protein was attained 48 h after the removal of tet, with levels 6-fold higher than those of control (Fig. 1,B). To assess the biological activity of the trans-gene product, we measured the capacities of lysates obtained from TS-induced and -uninduced cells to bind radiolabeled FdUMP (Fig. 1,C). The maximum binding capacity was observed at 24 h after TS induction, with 3.6 ± 0.2 pmol of FdUMP bound per mg of protein, compared with 1.1 ± 0.1 pmol/mg in the uninduced cells. Comparison of the growth rates of TS-induced and -uninduced cells indicated that TS overexpression in itself did not significantly alter the rate of cellular proliferation (data not shown). Examination of p53^mut levels in the MTS-5 cell line revealed that TS overexpression did not affect p53^mut expression (Fig. 1,D). Baseline p53^mut levels were found to be high in this cell line, and exposure to 20 J/m² UV radiation failed to enhance p53^mut expression (Fig. 1 D), both of which are indicative of mutant p53.

We assessed the impact of inducible TS expression on the sensitivity of MTS-5 cells to a range of chemotherapeutic agents including: the TS inhibitors 5-FU, TDX, ZD9331, and MTA; the DNA damaging agents cisplatin and oxaliplatin (28); the topoisomerase I inhibitor CPT-11, which inhibits DNA repair (29); and Taxol, which arrests cells during mitosis by binding to the dynamic microtubules (30). The growth of TS-induced (−tet) and -uninduced (+tet) cells was assessed after cotreatment with a range of concentrations of each of these drugs (Fig. 2).

The 5-FU IC₅₀ dose increased from 2.6 μm to 4.7 μm after inducible expression of TS (Fig. 2,A). For TDX and ZD9331, TS overexpression increased the IC₅₀ doses by 9- and 24-fold, respectively (Fig. 2, B and C). The IC₅₀ dose for MTA could not be determined after induction of TS because of insufficient growth inhibition at the highest concentration used (10 μm; Fig. 2,D). In the absence of TS trans-gene activation, the IC₅₀ doses of MTA and ZD9331 were very similar (32 nm and 39 nm, respectively). Our analysis of the DNA-damaging agents cisplatin, oxaliplatin, and the topoisomerase I inhibitor CPT-11 indicated that TS overexpression had no significant impact on cellular sensitivity to these drugs (Fig. 2, E and F). Similarly, the cytotoxic effect of Taxol was identical in TS-induced and -uninduced populations of MTS-5 cells (Fig. 2 G). Clonogenic assays confirmed that TS overexpression conferred increased resistance to TS-targeted therapies, but not to agents that did not inhibit TS (data not shown).

To investigate the effect of TS expression on cell cycle progression, we analyzed the cell cycle profiles of MTS-5 cells in which the TS trans-gene was inactive or active after treatment with no drug or with 5-FU, TDX, MTA, oxaliplatin, and CPT-11. Each drug was administered at a concentration corresponding to the IC₅₀ dose in TS-uninduced MTS-5 cells. Representative cell cycle profiles 48 h posttreatment are shown in Fig. 3. In the absence of drug, up-regulation of TS caused no significant change in the cell cycle profile (Fig. 3,A), with 16.6 ± 0.4% of TS-uninduced (+tet) cells and 16.4 ± 1.4% of TS-induced (−tet) cells in S phase. Treatment with 5-FU resulted in an accumulation of cells in S phase, such that 26.1 ± 1.5% of cells were in S phase (Fig. 3,B, +tet). This arrest was completely abrogated when the TS trans-gene was activated, with 15.5 ± 2.0% of cells in S phase (Fig. 3,B, −tet), similar to the S-phase percentages in the control populations (Fig. 3,A). Treatment of TS-uninduced cells with TDX and MTA also resulted in a dramatic S-phase arrest with 32.8 ± 3.4% and 31.9 ± 2.5% of cells, respectively, accumulating in S phase (Fig. 3, C and D, +tet). However, as with 5-FU, overexpression of TS caused complete abrogation of the TDX- and MTA-induced S-phase arrests (Fig. 3, C and D, −tet). In contrast, treatment with oxaliplatin resulted in equal S-phase accumulation in TS-induced (34.8 ± 1.0%) and -uninduced (35.1 ± 2.6%) cells (Fig. 3,E). Treatment with CPT-11 resulted in no change in the S-phase distribution; however, a significant similar number of TS-induced and -uninduced cells did arrest in G₂-M phase (Fig. 3 F). The magnitude of the CPT-11-induced G₂-M arrest was the same in the TS-uninduced (33.8 ± 1.6%) and -induced (34.3 ± 1.3%) populations.

To begin to characterize the molecular events underlying the S-phase arrest we assessed the expression of the key cell cycle genes cyclin E, cyclin A, and cdk2 in the MTS-5 system after treatment for 48 h with IC₅₀ doses of 5-FU, TDX, MTA, CPT-11, and oxaliplatin. Western blot analysis indicated that, in the absence of drug, cyclin E, cyclin A, and cdk2 were expressed at the same level in TS-uninduced (+tet) and -induced (−tet) cell populations (Fig. 4). Treatment with 5-FU in the absence of inducible TS expression resulted in increased levels of cyclin E, cyclin A, and cdk2 of ∼1.7-, 2.3-, and 3.9-fold, respectively, compared with TS-induced cells. Similarly, treatment of TS-uninduced cells with TDX resulted in increased expression of cyclin E, cyclin A, and cdk2 of ∼2.9-, 1.9-, and 3.4-fold respectively compared with cells expressing the TS trans-gene. Furthermore, treatment with MTA also resulted in increased expression of cyclin E by 3.4-fold, cyclin A by 3.3-fold, and cdk2 by 3.4-fold in TS-uninduced cells compared with induced cells. In contrast, after treatment with CPT-11 and oxaliplatin, the expression levels of cyclin E, cyclin A, and cdk2 did not alter between the TS -uninduced and -induced populations.

We have successfully generated a tet-inducible cell line, MTS-5, in which we can control expression of a TS trans-gene. Our analysis of the TS-inducible MTS-5 cell line has confirmed its usefulness as a physiologically relevant reagent with which to study the effects of elevated TS expression on the sensitivity of malignant cells to a wide range of chemotherapeutic compounds. Treatment of MTS-5 cells with 5-FU and antifolate-based chemotherapeutic inhibitors of TS demonstrated that up-regulation of TS significantly decreased cellular sensitivity to these agents. We found that the protective effect of TS overexpression was greater for TDX and ZD9331 than for 5-FU. MTA toxicity was significantly more sensitive to increased TS levels than that of either TDX or ZD9331, which is somewhat surprising because MTA can inhibit several other folate-dependent enzymes (10). However, the results of this study suggest that TS is the main target for MTA in this cell line.

We also analyzed the effect of TS overexpression on the toxicity of the chemotherapeutic agents cisplatin, oxaliplatin, CPT-11, and Taxol, which do not directly inhibit TS enzyme function. These agents are often used as second-line agents to treat tumors, such as colorectal cancer, that are refractory to 5-FU. Our data suggest that TS up-regulation does not affect the toxicity of cisplatin, oxaliplatin, and CPT-11, and that TS overexpression does not confer cross-resistance to these drugs.

Cell cycle analysis of MTS-5 cells treated with IC₅₀ doses of 5-FU, TDX, and MTA (and with ZD9331, data not shown) demonstrated that, in the absence of TS trans-gene expression, significant numbers of cells arrested in S phase 48 h posttreatment. When the TS trans-gene was activated, the S-phase arrest caused by each drug was completely abrogated, which indicated that TS overexpression was able to circumvent the S-phase block. The proportion of TS-uninduced cells in S phase after 5-FU treatment decreased to control levels 72 h posttreatment, whereas the arrests caused by TDX and MTA were sustained at this time point (data not shown), which indicated that the antifolates produce a more potent and sustained S-phase arrest than does 5-FU. This is likely to be the cause of the smaller increase in IC₅₀ dosage observed in TS-overexpressing cells treated with 5-FU compared with those treated with antifolates. Treatment with oxaliplatin resulted in equal levels of TS-induced and -uninduced cells accumulating in S phase, which indicated that TS overexpression is not able to overcome the S-phase arrest caused by this drug. Treatment with an IC₅₀ dose of CPT-11 did not cause an accumulation of cells in S phase; however, a significant G₂-M arrest was observed in TS-induced and -uninduced populations, which suggested that TS overexpression does not influence the cycling of CPT-11-treated MTS-5 cells.

Progression through the cell cycle is regulated by the coordinate action of cdks and their associated regulatory subunits, cyclins (31). Cyclin E-cdk2 activity is maximal near the G₁-S boundary (32) and is required for G₁ to S-phase transition (33). Once in S phase, cyclin E is degraded, and cyclin A becomes the principle cyclin associated with cdk2; cyclin A-cdk2 activity is required for S-phase progression (34). We analyzed the expression of these important regulatory proteins after treatment of MTS-5 cells with TS inhibitors. We found that in cells that did not overexpress TS and that significantly accumulated in S phase, expression of cyclin E, cyclin A, and cdk2 proteins was elevated by between 2- and 4-fold after treatment with IC₅₀ doses of 5-FU, TDX, and MTA compared with cells in which TS was overexpressed. In contrast, treatment with CPT-11 and oxaliplatin did not result in differential expression of these proteins in the TS-induced and -uninduced populations. Cyclin A-cdk2 can form stable complexes with the E2F-1 transcription factor, which eliminates E2F-1-dependent DNA binding and transactivation function (35). High cyclin A-cdk2 activity in S-phase arrested cells may act to inhibit the activity of E2F-1, which is required for S-phase traversal. In addition, because cyclin E-cdk2 and cyclin A-cdk2 have been shown to limit DNA replication to a single round per cell cycle (36), it is possible that elevated cdk2 activity inhibits reinitiation of interrupted DNA replication in S-phase arrested cells.

We found that TS overexpression did not affect expression of mutant p53 in the MTS-5 cell line. Ju et al. have reported that TS protein inhibits translation of p53 mRNA (37). It is possible that the high levels of mutant p53 in the MTS-5 cell line masks the negative effect of high TS levels on p53 expression, or that the extent to which TS is up-regulated in this cell line is not sufficient to significantly down-regulate p53 mRNA translation.

In conclusion, we have developed a TS-inducible system with which to study the patterns of sensitivity to various commonly used anticancer drugs. We have demonstrated a direct correlation between the levels of TS expression and the development of resistance to TS targeted drugs. We have also demonstrated that TS overexpression confers increased resistance to TS inhibitors by abrogating an S-phase arrest. Furthermore, the S-phase arrest induced by TS inhibitors is accompanied by increased expression of several key cell cycle proteins: cyclin E, cyclin A, and cdk2. In addition, we have identified other non-TS-targeted drugs to which TS induction does not confer resistance and that may be useful in tumors resistant to fluoropyrimidines or antifolates.