Yuzuru Takemura

Folate-based thymidylate synthase inhibitors in cancer chemotherapy.

Understanding the relationship between chemical structure and biological properties of folate analogs, particularly their interactions with the target enzymes, transport proteins and folate-metabolizing enzyme, folylpolyglutamate synthetase (FPGS), has enabled the rational design and development of the selective thymidylate synthase (TS) inhibitors with folate-based structures for clinical uses. These compounds specifically inhibit TS devoid of concomitant effects at other loci, unlike 5-fluorouracil (5-FU). ZD1694 (Tomudex) was designed as a non-nephrotoxic and highly active analog of N10-propargyl-5,8-dideazafolic acid (CB3717), which is a potent TS inhibitor but had unacceptable nephrotoxicity caused by its poor water solubility. The potent cytotoxic activity of ZD1694 is dependent upon active uptake into cells via the reduced folate carrier (RFC), and subsequent rapid and extensive metabolism to polyglutamate forms inside cells. Marked enhancement of the TS inhibitory activity has been noted as the glutamate chain is elongated. Polyglutamation is critical to the biological activity of ZD1694 against tumor and normal proliferating tissues. The retentive property of ZD1694 polyglutamates inside cells led to a single, infrequent administration schedule in clinical studies. ZD1694 has completed phase I and phase II evaluation with activity observed in several tumor types, particularly in colorectal cancer with a 26% objective response rate. A recent European phase III study of ZD1694, randomized against a 5-FU plus leucovorin regimen, demonstrated an equivalent response rate for advanced colorectal cancer (complete or partial responses; 20 versus 17%) and less toxicity than seen with the latter regimen. The newer selective TS inhibitors, which retain potency for TS inhibition but are not substrates for RFC and/or FPGS, are currently under clinical evaluation. These classes of compound may have benefits for circumvention of resistance by virtue of alterations in these protein functions and for the management of toxicity.

Relationship between tumor cell density and drug concentration and the cytotoxic effects of doxorubicin or vincristine: mechanism of inoculum effects

SummaryWhen tumor cell density increases, the cytotoxic activity of certain anticancer agents, such as vincristine (VCR) and doxorubicin (DXR), progressively decreases. This phenomenon is termed the inoculum effect. Since VCR and DXR are less active in an acidic environment, we questioned whether the inoculum effects could have resulted from acidification of the medium that may have developed due to the high cell density. However, measurements of the cytotoxic activity of these agents in a pH-controlled medium revealed only a minor correction of the inoculum effects. Second, we wondered whether the inoculum effects that occurred at the high cell density might have been attributable to insufficient amounts of drugs to bind all the binding sites of the cells. To test this hypothesis, we used drug-resistant sublines, which required higher VCR or DXR concentrations for cell killing than did the parent cell line. When higher drug concentrations were used, the dose-response curves generated for low- and high-density cell populations became closer and overlapped each other, resulting in virtual disappearance of the inoculum effects. Measurements of cellular drug levels revealed that at a high cell density, cells accumulated much smaller amounts of both VCR and DXR in parallel with the positive inoculum effect. In contrast, when high concentrations of the drugs were used in drug-resistant cells, differences in the cellular drug contents between low and high cell densities became narrow. Cisplatin (DDP) belongs to a group of drugs that do not produce inoculum effects, and DDPs cytotoxic effects were not influenced by the pH-controlled medium or by the use of drug-resistant cell lines. These observations indicate that the inoculum effects are the result of the unavailability of VCR or DXR molecules to all cellular binding sites when cells at high densities are exposed to drugs. The drug concentration relative to cell density was apparently the major determinant for the inoculum effects seen in VCR- or DXR-induced cell killing.

The influence of drug-exposure conditions on the development of resistance to methotrexate or ZD1694 in cultured human leukaemia cells.

The influence of drug‐exposure conditions on the development of resistance to methotrexate (MTX) or ZD1694 was studied by treating MOLT‐3 human lymphoblastic‐leukaemia cells in a continuous or a pulsatile (high‐dose, short‐term) drug‐exposure schedule. Continuous exposure of the cells to MTX with stepwise escalation of the drug concentrations resulted in a MTX‐resistant sub‐line (MOLT‐3/MTX10,000) with impaired reduced‐folate carrier (RFC) and increased dihydrofolate‐reductase (DHFR) activity. Conversely, a MTX‐resistant clone (MOLT‐3/MTX·P‐9) with unaltered RFC and DHFR activity, but with decreased cellular accumulation of antifolates, was selected by high‐dose short‐term treatment of the cells with MTX. MTX resistance in the latter cells was pronounced after short‐term rather than continuous‐exposure incubation with MTX, suggesting defective polyglutamation of the drug. On the other hand, 2 ZD1694‐resistant sub‐lines which were established by continuous (MOLT‐3/ZD1694·C) or by pulsatile drug‐exposure schedule (MOLT‐3/ZD1694·P‐9) demonstrated extremely low accumulation and poor retention of [3H]ZD1694, with no change of initial drug uptake and little or no increase of thymidylate‐synthase (TS) activity, irrespective of drug‐exposure conditions for their establishment. HPLC analysis displayed a virtual absence of ZD1694 polyglutamates in both ZD1694‐resistant sub‐lines and low accumulation in MOLT‐3/MTX·P‐9 as compared with the parent line. However, folylpolyglutamate‐synthetase (FPGS) mRNA was only moderately decreased in the 2 ZD1694‐resistant sub‐lines and to an even lesser extent in MOLT‐3/MTX·P‐9. In addition, γ‐glutamyl‐hydrolase (GGH) activity was not increased, but was slightly down‐regulated in the polyglutamation‐defective sublines. These results indicate that the mechanism(s) of the resistance developed may depend not only on drug‐exposure conditions while raising resistance but also on the biochemical properties of the drug.

Antitumor activities of new platinum compounds, DWA2114R, NK121 and 254-S, against human leukemia cells sensitive or resistant to cisplatin

Summary(R)-(-)-1,1-(2-amino-methylpyrrorodine)-platinum(II) (DWA2114R), cis-1,1-cyclobutanedicarboxylato(2R)-2-methyl-1,4-butanediammineplatinum(II) (NK121; CI-973) and glycolate-o,-o′-diammine platinum(II) (254-S; NSC375101D) are new platinum compounds developed in Japan. We studied the antitumor effects of these compounds on the cisplatin (cis-diamminedichloroplatinum, DDP)-resistant human leukemia cell line, K562/DDP. K562/DDP cells were 10-fold resistant to DDP, while the cells showed minimal cross-resistance to carboplatin (2.1-fold) and DWA2114R (3.3-fold), and were as sensitive to NK121 (1.6-fold) and 254-S (1.0-fold) as the parent cells. Increases in exposure time of K562 cells to DWA2114R resulted in progressive shifting of the dose-response curve to the left, or more effective cell growth inhibition of the cells. Time dependency indices (ID80 obtained from dose-response curve after 1 hr-exposure of K562 cells to drug followed by 72 hr-culture without drug/ID80 after 24 hr-exposure) of DDP, NK121 and 254-S were 10, 8 and 20, respectively. A multidrug resistant cell-line, MOLT-3/TMQ200, was as sensitive to platinum compounds as the parent MOLT-3 cells. Little or no influence of tumor cell density was observed in the growth inhibition of MOLT-3 or K562 cells induced by these new compounds even if cells were concentrated to a density of 108 cells/ml. These results indicate that NK121 and 254-S may overcome the drug resistance developed in the patients after treatment with DDP. The antitumor effect of DWA2114R is more dependent not only on drug-concentration but also on exposure time than that of DDP, suggesting that continuous infusion rather than bolus administration appears the favorable schedule in clinical trials.

The influence of tumor cell density on cellular accumulation of doxorubicin or cisplatin in vitro.

SummaryThe effect of tumor cell density on the cellular pharmacokinetics of doxorubicin (DXR) and cisplatin (CDDP) was studied using MOLT-3 human acute lymphoblastic leukemia cells. As determined by the MTT assay, the growth-inhibitory effect of DXR was approx. 40 times lower when cell density was increased from 106 to 108 cells/ml (positive inoculum effect), whereas little or no influence of cell density was observed in CDDP-induced cell-growth inhibition. As measured by high-performance liquid chromatography using a fluorescence detector, the cellular accumulation of DXR showed 6- and 18-fold decreases after 1 h incubation when the cells were concentrated from 106 to 107 and 108 cells/ml, respectively. Only at low cell density (106 cells/ml) did the amount of DXR in the cells increase with increasing exposure times of up to 6 h. The DXR concentration in the supernatant that was separated from a cell suspension showing a density of 108 cells/ml fell to 20% of that obtained at 106 cells/ml. The metabolites of DXR, including Adriamycinol and Adriamycinone, were not detectable in the cell extracts or supernatants at any cell density examined. In contrast, the cellular accumulation of CDDP calculated from the platinum concentration, which was measured with a flameless atomic absorption spectrophotometer, was essentially identical at all cell densities examined; moreover, extension of the exposure period resulted in a linear increase in the amount of CDDP in the cells. CDDP concentrations in the supernatants were equally retained, inrrespective of cell densities. These observations indicate that the positive inoculum effect shown in DXR-induced cell-growth inhibition results from the decreased cellular accumulation of the drug at high cell densities. We found no influence for cell density on the cellular accumulation of CDDP that might be relevant to the therapeutic potentiation of this drug at high tumor-cell density.

Retrovirus-mediated transfer of anti-MRD1 hammerhead ribozymes into multidrug-resistant human leukemia cells: Screening for effective target sites

One of the underlying mechanisms of multidrug resistance (MDR) is cellular over‐production of P‐glycoprotein (P‐gp), which acts as a drug efflux pump. P‐gp is encoded by a small group of related genes termed MDR; only MDR1 is known to confer drug resistance. To overcome P‐gp‐mediated drug resistance, we have developed two anti‐MDR1 hammerhead ribozymes driven by the β‐actin promoter. Upon transduction of the ribozymes into MDR cells, vincristine resistance was decreased. These two ribozymes were constructed, which showed different cleavage activities. In this study, to determine suitable target sites for the anti‐MDR1 ribozyme, the exon 1b–intron 1 boundary, the translation‐initiation site, the intron 1–exon 2 boundary and the exon 2–intron 2 boundary, codons 179 and 196 of the MDR1 gene were selected as candidates. To improve the ribozyme activity, a retroviral vector containing RNA polymerase III promoter was used. Stable retrovirus producer cells were generated by transfecting the retroviral vector plasmids carrying the ribozyme into the packaging cell line. Retroviral vector transduction of human leukemia cell lines expressing MDR1 was accomplished by co‐culturing these with virus producer cells. Stably transduced cells were selected by G418 and pooled to determine the efficacy of each ribozyme. These ribozyme‐transduced cells became vincristine‐sensitive concomitant with the decreases in MDR1 expression, P‐gp amount and drug efflux pump function. Among the ribozymes tested, the anti‐MDR1 ribozyme against the translation‐initiation site exhibited the strongest efficacy. This retrovirus‐mediated transfer of anti‐MDR1 ribozyme may be applicable to the treatment of MDR cells as a specific means to reverse resistance. Int. J. Cancer81:944–950, 1999.

Effect of hammerhead ribozyme against human thymidylate synthase on the cytotoxicity of thymidylate synthase inhibitors

One of the resistance mechanisms to folate‐based thymidylate synthase (TS) inhibitors is the increase in TS activity in tumor cells. Human B lymphoblastoid cell line (W1L2) was made resistant to a lipophilic non‐polyglutamatable TS inhibitor (ZM249148), and the subline (W1L2:R179) showed a 20‐fold increase in TS enzyme activity with concomitant overexpression of TS mRNA. To overcome the resistance, we designed a ribozyme that can cleave the CUC sequences in a triple tandemly repeated sequence of TS mRNA. Expression of this ribozyme in W1L2:R179 cells transfected with Epstein Barr virus‐based expression vector resulted in sensitization to TS inhibitors concomitantly with a decrease of TS expression. The ribozyme expressed in transfectants was shown to be functional in cleaving artificial TS RNA in vitro.

Cellular pharmacokinetics of ZD1694 in cultured human leukaemia cells sensitive, or made resistant, to this drug

We have analysed the cellular metabolism of a novel thymidylate synthase (TS) inhibitor, ZD1694, in MOLT-3 and K562 human leukaemia cell lines sensitive to or made resistant to ZD1694 by continuous exposure of the cells to ZD1694 with stepwise escalation of the drug concentration. The initial cellular uptake of [3H]ZD1694 was greater in K562 cells than in MOLT-3 cells and the drug accumulated approximately 3-fold more in the former cells following incubation with 0.1 μM ZD1694 at 37°C for 24 h. TS and dihydrofolate reductase activities were not significantly different between the two cell lines. After a 30-min incubation with the drug at 37°C, 85% of the total drug (2.3pmol/mg protein) in K562 cells was found as tri-to pentaglutamates, whereas MOLT-3 cells accumulated less drug in this time (0.83 pmol/mg protein) and polyglutamates of chain length greater than triglutamate were not found to a significant extent. When the incubation time was extended to 24 h, the polyglutamate profile in K562 cells was progressively shifted towards those of long glutamate chain length and 59% of the total cellular drug (204 pmol/mg protein) was identified as the penta form. In contrast, even distribution between tri-and pentaglutamate was observed in MOLT-3 cells. Total cellular polyglutamates were approximately 3-fold higher in K562 cells than in MOLT-3 cells, and this may explain the 2.5-fold difference in the sensitivity to ZD1694 between the two cell lines. Continuous exposure of MOLT-3 and K562 cells to ZD1694 up to 1 μM or 0.1μM resulted in 1600-and 4200-fold resistant sublines, respectively (MOLT-3/ZD1694·C and K562/ZD1694·C). The resistant MOLT-3 cells showed a markedly lower cellular accumulation and poor retention of [3H]ZD1694 with no significant change of initial drug uptake by 10 min and with a little increase of TS activity. HPLC analysis demonstrated that more than 90% of the3H co-eluted with the monoglutamate (parent drug) in the resistant MOLT-3 cells, indicating extremely diminished polyglutamation in the cells. On the other hand, cellular uptake of [3H]ZD1694 was extensively impaired in K562/ZD1694·C cells and cellular accumulation of the drug was only 2.5% of that in the parent cells following 24 h incubation with the drug. Neither an increase of TS or dihydrofolate reductase activity nor a change in the polyglutamate formation profile was observed in the resistant K562 cells. These results indicate that the cellular ability to produce the polyglutamate metabolites of ZD1694 must influence the sensitivity of the tumour cells to this drug, and development of mechanisms involved in the ZD1694 resistance may relate to the intrinsic biochemical properties of the cells.

Quantitative analysis of human multidrug resistance 1 (MDR1) gene expression by nonisotopic competitive reverse transcriptase polymerase chain reaction assay

We have established competitive reverse transcriptase polymerase chain reaction (RT‐PCR) assay for the quantification of MDR1 mRNA encoding P‐glycoprotein (P‐gp) by analyzing leukemia sublines of MOLT‐3 with various expression of MDR1. The expression was quantified by simultaneous RT‐PCR of cellular RNA with decreasing amounts of heterologous competitor RNA, which shares the MDR1 primer sequences with the cellular MDR1 mRNA, but yields a different‐sized PCR product. This allows resolution of the amplified cDNA fragments. The amounts of MDR1 mRNA measured by the assay were accurate and reproducible over wide range, and were determined as 31.6, 100, and 316 amol/μg total RNA in MOLT‐3/TMQ70, MOLT‐3/TMQ800, and MOLT‐3/VCR1,000, respectively. The relative ratio of MDR1 mRNA measured by the competitive RT‐PCR among three sublines was similar to that of MDR1 transcript determined by Northern analysis (1:4:12) and to that of P‐gp measured by flow cytometry (FCM) analysis. In mononuclear cells from patients with leukemia, MDR1 mRNA could be sufficiently quantified by the competitive RT‐PCR established, while FCM assay could scarcely detect P‐gp. This study demonstrated that the competitive RT‐PCR assay using heterologous competitor RNA is a rapid, reliable, and non‐radioactive procedure and is acceptable for the evaluation of MDR1 expression in clinical samples. J. Clin. Lab. Anal. 11:258–266, 1997.

Expression of variant dihydrofolate reductase with decreased binding affinity to antifolates in MOLT-3 human leukemia cell lines resistant to trimetrexate.

Various alterations of the dihydrofolate reductase (DHFR) gene are involved in resistance. In order to understand the mechanism that induce such gene alterations in human leukemia cells, we studied the expression products of DHFR gene in trimetrexate (TMQ)- and/or methotrexate (MTX)-resistant sublines derived from a MOLT-3 human leukemia cell line. A 200-fold TMQ-resistant subline (MOLT-3/TMQ200) expressed the mutated DHFR mRNA, with a base change (T-->C) at the second position of codon 31, as well as the wild type gene. A MTX-resistant subline derived from MOLT-3/TMQ200 (MOLT-3/TMQ200-MTX500) showed a further increase in the expression of the mutated DHFR mRNA, compared to MOLT-3/TMQ200, with a marked decrease of expression of the wild type DHFR mRNA, which is confirmation of amplification of the mutated DHFR gene. By contrast, a 10,000-fold MTX-resistant subline (MOLT-3/MTX10,000) over-expressed the wild type DHFR mRNA, which is confirmation of amplification of the wild type gene. Increased levels of the DHFR enzyme in these sublines were proportional to expression levels of the DHFR mRNA. The DHFR enzyme expressed in MOLT-3/TMQ200-MTX500 cells showed a 40-fold increase in the Ki values for both MTX and TMQ, compared with values for the wild type DHFR expressed in both MOLT-3/MTX10,000 and its parent cell line. These findings suggest that the altered DHFR gene, which was introduced in MOLT-3 cells by exposure to TMQ, gave rise to a variant enzyme with reduced affinity to antifolates, and that complex DHFR alterations confer drug-resistant phenotypes in antifolate-resistance. Structural difference between the antifolates could be important in the introduction of the differential DHFR gene alterations in the antifolate resistance.