Thymidylate synthase mRNA level in adenocarcinoma of the stomach: a predictor for primary tumor response and overall survival

PURPOSE

We tested the hypothesis that polymerase chain reaction (PCR) quantitation of the enzyme thymidylate synthase (TS) within a primary adenocarcinoma of the stomach, has an inverse relationship to response and survival for patients who receive fluorouracil (5FU)-based chemotherapy.

PATIENTS AND METHODS

Before systemic chemotherapy, the genetic expression of TS (TSmRNA level) was determined using a PCR method. Gene expression was calculated by determining the ratio between the amount of radiolabeled PCR product with the linear amplification range of the TS gene and the beta-actin gene. Chemotherapy consisted of two cycles of protracted infusion (PI) 5FU 200 mg/m2/d administered for 3 weeks with leucovorin 20 mg/m2/w. Cisplatin 100 mg/m2 was administered on day 1.

RESULTS

Sixty-five patients with primary gastric cancer had a median TS mRNA level of 4.6 x 10(-3) (range, 0.9 to 20.1 x 10(-3)). Thirty-five percent of patients had measurable responses in their primary tumors. The mean gastric cancer TSmRNA level in responding and resistant patients is statistically significant (P < .001). The median survival time was 43+ months for treated patients with TSmRNA levels less than the median and 6 months for those with TS m-RNA levels greater than the median (P = .003).

CONCLUSION

The genetic expression of TS (TSmRNA level) influences response to 5FU-based chemotherapy and survival for a cohort of patients with primary gastric cancer. Confirmation of these data could lead to therapeutic decisions based on specific molecular properties within a tumor.

Identification of mutations by RNA conformational polymorphism "bar code" analysis

DNA single-strand conformational polymorphism (SSCP) analysis is widely used for detection of point mutations in clinical specimens. Performing SSCP analysis with cRNA instead of DNA has been shown to improve mutation detection frequency. RNA can exist in numerous meta-stable conformations, which appear as patterns of bands on nondenaturing electrophoresis gels. Single base mutations can cause not only mobility shifts of major bands, but also loss of some conformations and appearance of new conformations. Unique RNA SSCP patterns associated with specific base sequences in many cases allow visual identification of point mutations. However, in some cases, the RNA SSCP pattern of a single base change in a sequence is not sufficiently different for a positive identification of the mutation. Improvement in the detection capability of RNA SSCP was obtained by adding 3'-deoxynucleotides to the transcription reaction. The presence of chain-terminating nucleotides in the transcription reaction formed numerous new RNA fragments, thereby generating complex band patterns ("bar codes") unique to each RNA sequence. This method was applied to analyzing p53 mutations in patients with colon cancer.

Thymidylate synthase gene and protein expression correlate and are associated with response to 5-fluorouracil in human colorectal and gastric tumors

Thymidylate synthase (TS) is the target enzyme for 5-fluorouracil (5-FU). We have correlated TS protein and gene expression with the response in patients with colorectal (n = 9) and gastric cancer (n = 12) treated with infusional 5-FU plus leucovorin (LV) or infusional 5-FU/LV and cisplatin, respectively. TS protein expression was analyzed by Western blot using TS106 monoclonal antibody and densitometry scanning. TS gene expression was measured by PCR analysis using beta-actin as an internal standard and expressed as a TS:beta-actin mRNA ratio. A close linear relationship was noted between TS protein expression and TS gene expression (r2 = 0.60) for the 21 tumor samples analyzed. TS immunohistochemical staining on 15 of the 21 samples revealed that the TS staining intensity correlated closely with TS protein and mRNA expression. In two biopsy samples, TS protein levels and TS gene expression did not correlate; however, one of these exhibited a focal TS staining pattern. Both the TS protein level and TS gene expression were significantly associated with response to 5-FU-based therapy. Patients with responsive disease had a mean TS protein level of 0.17 +/- 0.03 arbitrary units (range, 0.05 to 0.38), whereas in patients whose tumors did not respond, the mean TS protein level was significantly higher 0.60 +/- 0.09 (range, 0.06 to 1.01; P < 0.01). A similar pattern was noted with TS gene expression. In patients with responsive disease, the mean TS:beta-actin gene ratio was 1.36 +/- 0.3 (range, 0.5-3.3 x 10(-3). In contrast, biopsies from patients with unresponsive disease had a mean TS:beta-actin gene ratio of 15.4 +/- 2.6 x 10(-3) (range, 2.7-35.9; P < 0.01). TS protein and TS mRNA expression are highly correlated, and each predict for response to 5-FU/LV-based chemotherapy in patients with colorectal and gastric cancer.

Gene amplification and multidrug resistance induced by the phosphatase-inhibitory tumor promoter, okadaic acid

The mechanism by which tumor promoters contribute to cellular transformation and tumorigenesis is not completely understood. To investigate further the molecular events involved in these processes, we used okadaic acid, a non-phorbol ester type tumor promoter that specifically inhibits certain protein phosphatases. We describe here that the continuous treatment of murine NIH 3T3 fibroblast cell cultures with okadaci acid resulted in a 50-fold amplification of two genes, mdr-1a and mdr-1b, that conferred multidrug resistance. As a consequence, the cells became cross-resistant to the cytotoxic effects of adriamycin, an antineoplastic drug used in the treatment of human tumors. Since genetic changes have been correlated with cell transformation and tumorigenesis, our results suggest that these processes may constitute an additional factor contributing to tumor promotion by okadaic acid.

Incorporation of 5-fluorouracil into U2 and U6 snRNA inhibits mRNA precursor splicing

The splicing activities of 5-fluorouracil (FUra)-substituted U2 and U6 small nuclear RNAs (snRNAs) were examined in an in vitro splicing system. Yeast splicing extracts were specifically depleted of endogenous U2 and U6 snRNAs by antisense oligonucleotide-directed RNase H hydrolysis. Splicing activity was recovered when the extracts were reconstituted with synthetic U2 and U6 snRNAs. However, U2 snRNA with all uracils substituted with FUra (FU2) did not restore any splicing activity. Nondenaturing gel electrophoresis showed that FU2 failed to promote the assembly of spliceosome complexes. The ability of U2 snRNA to restore splicing in U2-depleted extracts increased as FUra content decreased but was still only 60% of control activity at 25% substitution of uracils with FUra. Addition of FU2 to nondepleted extracts caused strong inhibition of splicing accompanied by increased degradation of the pre-mRNA, suggesting that FU2 forms an inactive complex with a protein splicing factor that normally binds to the pre-mRNA. FU6 restored full splicing activity to U6-depleted extracts, but at a 5-fold higher concentration than U6 snRNA. These results demonstrate that the incorporation of FUra can impair the functions of catalytic RNA molecules.

Quantitative determination of the ratio of mutated to normal ras genes in the blood of leukemia patients by allele-specific PCR

By combining allele-specific PCR amplification and a PCR-based quantitation approach, a method has been developed to estimate the mutated K-ras gene content in the blood of AML patients as a percentage of total K-ras. One PCR primer set was designed not to discriminate between mutant K-ras and wild-type K-ras and thus amplified the total K-ras gene. The other PCR primer set was designed to be allele-specific for K-ras gene containing a G to C mutation at codon 12. This primer set could discriminate the mutant and wild-type genes when the proportion of the mutated sequence was 0.2% of the total K-ras gene. To test the method on biological specimens, genomic DNA samples were analyzed from the peripheral blood of a patient who had secondary AML with the same codon 12 K-ras mutation. Two samples taken from this patient 2 months apart during follow-up had myeloblast cell contents of 67 and 80%. However, the percentage of mutated K-ras was 50% in both samples, suggesting that this patient may be inherently heterozygotic in this particular mutation. This ratio of mutated to normal K-ras in the patient's cells was confirmed by RNA-SSCP analysis and RNA sequencing. This quantitation method can provide a sensitive and specific estimation of the content of mutated K-ras alleles in patient samples.

Effect of 5-fluoro- and 5-bromouracil substitution on the translation of human thymidylate synthase mRNA

The synthesis of thymidylate synthase (TS) from 5-fluorouracil (FUra)- and 5-bromouracil (BrUra)-substituted mRNAs was examined to investigate the effect of incorporation of uracil (Ura) analogs on translation. Human TS cDNA was transcribed in the presence of Ura-, FUra-, or BrUTP to obtain 100% substituted mRNA. The mRNAs were translated in a rabbit reticulocyte lysate system. The TS protein that was formed from each of the templates reacted identically with TS antibody in Western blots. Time courses of TS formation revealed a characteristic peak which occurred at 45 min for the Ura- and FUra-RNAs and at 2 h for the BrUra-RNA. Substitution of Ura with FUra did not alter the rate of translation, while substitution of BrU for Ura decreased the rate of translation. Substitution of Ura with FUra or BrUra enhanced the stability of the mRNAs in the rabbit reticulocyte lysate by 3- and 10-fold, respectively. Incorporation of BrUra influenced the binding and catalysis on the ribosome, resulting in a 3.5-fold greater rate of activation (Kact) and 6-fold lower Vmax than the equivalent values for the Ura- and FUra-substituted mRNAs. Nondenaturing gel electrophoresis revealed that different conformations exist among the mRNAs. These data show that translation can be influenced by the incorporation of fraudulent bases into mRNA and those bases that stabilize RNA secondary structure will have the greatest inhibitory effect on translation.

Molecular detection and characterization of clonal cell populations in acute lymphocytic leukemia by analysis of conformational polymorphisms of cRNA molecules of rearranged T-cell-receptor-gamma and immunoglobulin heavy-chain genes

Junctional regions of rearranged T-cell-receptor-gamma (TCR) and immunoglobulin heavy-chain (IgH) genes represent an idiotypic DNA sequence for an individual lymphocytic cell, and any clone or clonal disease developing from this cell. In this study, a novel methodology for detection and characterization of clone specific DNA sequences in patients with acute lymphocytic leukemia (ALL) was developed. Junctional regions of rearranged TCR-gamma IgH genes in specimens of bone marrow aspirates of patients with ALL (precursor-B-ALL ten, T-ALL two, null-ALL one; ALL not classified one), of a patient with lymphoid blast crisis of chronic myeloid leukemia, of a B-cell chronic lymphocytic leukemia, and in DNA from peripheral blood mononuclear cells of ten healthy volunteers were amplified by polymerase chain reaction (PCR). The PCR products were transcribed into complementary RNA (cRNA). Conformational polymorphisms of cRNA molecules were analyzed by non-denaturing polyacrylamide gel electrophoresis. A specific cRNA banding pattern for rearranged TCR-gamma or IgH genes was observed in all patients with lymphocytic leukemia. In contrast, analysis of DNA from healthy volunteers yielded a smear of confluent polymorphisms representing multiple different cRNA molecules. In two patients with precursor-B-ALL, cRNA banding patterns of junctional regions of rearranged TCR-gamma genes were analyzed in sequential bone marrow aspirates. The banding patterns disappeared after chemotherapy and achievement of blast clearance. This novel and rapid molecular assay offers several advantages as compared to Southern blot analyses and previous PCR based methodologies for the detection of clonal lymphocytic populations. With this methodology, studies on the clonal evolution of lymphoproliferative disorders (e.g. the prognostic significance of the emergence of additional clones) can be performed more easily than with any other traditional molecular method.

Detection of clonal T-cell populations in gastrointestinal lymphomas by analysis of cRNA conformational polymorphisms of rearranged T-cell-receptor-gamma genes

Analysis of complementary RNA molecules of junctional regions of rearranged T-cell-receptor-gamma genes show a pattern of conformational polymorphisms which is specific for an individual lymphocytic clone. In a blinded study we analysed formalin-fixed, paraffin-embedded histological specimens from gastrointestinal lymphomas and control tissues (lymphomas: pleomorphic T-cell 10, anaplastic large cell [Ki1+] 9, centroblastic 5, immunocytoma 1, B-CLL 2, Hodgkin's 2, centroblastic-centrocytic 1, MALT [mucosa associated lymphoid tissue] 1, T-cell acute lymphoblastic leukaemia 1, non-lymphoid or polyclonal lymphoid tissues 5). Junctional regions of rearranged TCR-gamma genes were amplified by the polymerase chain reaction and the products were transcribed into cRNA. Conformational patterns of cRNA molecules were analysed by polyacrylamide gel electrophoresis. 13/20 T-lineage lymphomas and the T-cell acute lymphocytic leukaemia displayed a distinct cRNA band pattern, all B-lineage lymphomas and the non-lymphoid control tissues were negative. Only one case of nasopharyngeal (lymphoepithelial, Schmincke-Regaud) carcinoma showed a faint cRNA banding pattern. This novel and non-radioactive assay allows for the rapid detection and molecular characterization of clonal lymphoid populations in minute histological biopsy specimens.

Differential S100 beta expression in choroidal and skin melanomas: quantitation by the polymerase chain reaction

PURPOSE

S100 beta, a member of a calcium-binding protein family (S100s), is an important clinical marker for skin melanoma. In contrast, uveal melanomas appeared to express S100 beta protein less frequently and to a lesser degree. This study was performed to verify and extend this finding to the mRNA level.

METHODS

A quantitative polymerase chain reaction (PCR)-based method was used. A ratio, comparing the S100 beta PCR fragment to that of beta-actin (an internal reference gene), was generated to compare S100 beta mRNA expression among samples.

RESULTS

The ratios for skin melanomas (1.2 to 3.9; three tissues and two cell lines) were significantly higher than that for choroidal melanomas (0.1 to 0.63; seven of eight primary tumors and four of four cell lines). Only one choroidal melanoma biopsy had a ratio greater than 1. The PCR products from choroidal melanoma were identical in size and sequence to the S100 beta, as determined by gel electrophoresis and RNA conformational polymorphism. Because the ratios were also low in choroidal melanoma cell lines, the S100 beta phenotype appears to be genetically stable.

CONCLUSION

S100 beta is differentially expressed at the RNA and protein levels by skin and choroidal melanomas, which are derived from distinct populations of melanocytes. However, choroidal melanomas expressing little or no S100 beta were significantly stained by antiserum specific for the S100 protein family. Taken together, these data suggest that choroidal melanocytes express another, perhaps even novel, S100 protein(s).

Detection of point mutations in human DNA by analysis of RNA conformation polymorphism(s)

RNA molecules were found to separate into numerous metastable conformational forms upon non-denaturing gel electrophoresis. The equilibration of the conformations was accelerated by heating or mild denaturing conditions. Single-base substitutions in the sequence of the RNAs caused changes in the conformational patterns, including mobility shifts of major and minor conformations, appearance of new conformations and loss of other conformations. This sequence-dependent RNA conformational polymorphism was used to detect point mutations in p53 and, dihydrofolate reductase genes. Sense and anti-sense RNA strands corresponding to the same segment of the p53 gene gave entirely different conformational patterns. To generate the RNA, short regions of the target genes (up to about 250 bp) were amplified by the polymerase chain reaction and the resulting DNA segments transcribed to RNA by T7 RNA polymerase. The method is rapid, simple, amenable to non-radioactive visualization and was successful in several cases when DNA single-strand conformational polymorphism analysis (Orita et al. (1989) Genomics 5, 874-879) failed to detect the point mutation.

NAD(P)H:quinone oxidoreductase gene expression in human colon carcinoma cells: characterization of a mutation which modulates DT-diaphorase activity and mitomycin sensitivity

NAD(P)H:quinone oxidoreductase (DT-diaphorase; DTD) is an obligate two-electron reductase which may play a role in the bioactivation of antitumor quinones such as mitomycin C (MMC). We studied 10 colon carcinoma cell lines showing different levels of DTD activity (range, 0-3447 nmol/min/mg protein), as measured by the reduction of dichlorophenolindophenol. Expression of the NAD(P)H:quinone reductase gene (NQO1), which codes for the DTD enzyme, as measured by a polymerase chain reaction amplification technique was then correlated with enzymatic activity in all cell lines. HT-29 cells, which have intermediate DTD activity (769 +/- 144 nmol/min/mg protein, mean +/- SD) and are sensitive to MMC, showed high NQO1 expression relative to beta-actin (taken as 100% here for comparative purposes). BE cells which have no detectable DTD activity and are resistant to MMC showed moderate NQO1 expression (91% of HT-29). RNA single-strand conformational polymorphism analysis and subsequent sequencing of BE complementary DNA revealed a C to T mutation in the NQO1 complementary DNA. This confers a proline to serine substitution in the amino acid sequence of the protein. Additionally, HCT-116 cells showed both moderate DTD activity (390 +/- 41 nmol/min/mg protein) and NQO1 expression (41% of HT-29), while resistant subclones of these cells, exposed to MMC during 11 and 44 weeks, showed low gene expression (5 and 9% of HT-29 respectively) and enzymatic activity (11 +/- 6 and 36 +/- 16 nmol/min/mg protein). These results support the ideas that reductive activation of MMC by DTD may be important in the cytotoxicity of MMC and that polymerase chain reaction may be a useful technique for quantitating the relative expression of genes in human tumors.

Inactivation of Tetrahymena rRNA self-splicing by cis-platin proceeds through dissociable complexes

The anti-cancer drug cis-diamminedichloroplatinum (II) (cis-DDP) reacted with Tetrahymena self-splicing rRNA ribozyme, causing loss of self-splicing activity and formation of a number of platinated RNA species. The formation of one distinct platinated product, migrating at an apparent size of 2400 nt, was closely associated with ribozyme inactivation. This platinated RNA was resistant to T1 ribonuclease digestion, suggesting the presence of inter-strand Pt cross-links. The reaction rate of cis-DDP with the ribozyme followed first order kinetics and showed a saturation effect with increasing cis-DDP concentration, characteristic of an affinity-label type of interaction rather than bimolecular collision. The apparent KI for binding of cis-DDP to the ribozyme was 62 microM. Ribozyme treated with urea was not inactivated by cis-DDP, indicating that the native structure of the RNA is required for reaction with cis-DDP. Mg++, which binds to the ribozyme and causes conformational changes in the molecule, protected the ribozyme from inactivation by cis-DDP and also prevented the formation of platinated RNA. These results suggest that binding of cis-DDP to sites formed by certain secondary or tertiary structural elements of the RNA enhance the rate and the specificity of reaction of the reagent with the ribozyme.

Effect of 5-fluorouracil substitution on the self-splicing activity of Tetrahymena ribosomal RNA

Recent studies from several laboratories have suggested that the anticancer drug-5-fluorouracil (FUra) promotes abnormal splicing of precursor RNA molecules. In order to determine the effects of FUra on the chemistry of RNA splicing, we studied the splicing reaction of FUra-containing Tetrahymena rRNA [(FUra) RNA], a Group I self-splicing system having one intron [intervening sequence (IVS)] and two exons. When subjected to splicing conditions, the (FUra) precursor RNA gave all of the normal splicing products, ligated exons, IVS, circulation IVS (C-IVS), and the hydrolyzed circle (L-19 IVS) as well as other hydrolysis side products. No abnormal products indicative of missplicing were observed at pH 7.5. However, the presence of FUra in the RNA decreased the rates and extents of formation of all of the product species. At pH 7.5, the rate of ligated exon formation of (FUra) RNA was inhibited 3-fold and the maximum yield of ligated exons was 50% of normal. Substitution with FUra inhibited the rate of formation of C-IVS about 2-fold, while the extent of formation of this product was decreased by more than 3-fold compared to uracil-containing RNA [(Ura) RNA]. The circularization of (Ura) IVS remained constant to pH 9 and then increased, while that of (FUra) IVS declined abruptly after pH 7.3, indicating that ionization of the FUra residues of (FUra) RNA abolishes its catalytic activity. A temperature dependence experiment showed that the circularization activity of (FUra) IVS was lost at a temperature 15 degrees C lower than that of the (Ura) IVS. The labile phosphodiester bond of the (FUra) C-IVS was more stable to hydrolysis than was that of the (Ura) C-IVS at all pH values. The data suggest that a major effect of FUra substitution is to destabilize the active conformation of RNA because of weaker base pairing between FUra and adenine owing to partial ionization of the FUra residues.

Deoxyuridylate effects on thymidylate synthase-5-fluorodeoxyuridylate-folate ternary complex formation

The competitive basis and specificity of deoxyuridylate (dUMP)-mediated decreases in thymidylate synthase-5'-fluorodeoxyuridylate-folate (TS-FdUMP-folate) ternary complex formation at low concentrations of folates were investigated using charcoal isolation of protein-bound [3H]FuUMP ligand. Reaction conditions used 0.02 microM TS (Lactobacillus casei) and 0.10 microM [3H]FdUMP incubated for 10 min at 37 degrees and pH 7.4. Decreases in counts below control (C) values in dUMP-added samples (S) were expressed as C/S ratios. At CH2--H4PteGlu1 or H4PteGlu1 concentrations below 10 microM, highly linear relationships were found to exist between C/S value and dUMP concentrations, expressed as dUMP/FdUMP ratios. For H4PteGlu1, maximal C/S values for dUMP interference occurred at the lowest H4PteGlu1 concentrations, approaching the value of the TS-FdUMP binary complex. The efficiency of ternary complex formation by H4PteGlu1 was 28 +/- 5% of CH2--H4PteGlu1 values at concentrations below 1.0 microM. The protective effect of increasing H4PteGlu1 against dUMP interference resulted in a linear relationship between the logarithm of H4PteGlu1 concentration and the slope of dUMP interference (C/S vs dUMP/FdUMP). In contrast, the results with CH2--H4PteGlu1 were biphasic. At concentrations of CH2--H4PteGlu1 lower than 0.5 microM, C/S values were greater than those for binary complex alone, a result related to CH2--H4PteGlu1 consumption based on [5-3H]dUMP tritium-release studies. At concentrations of CH2--H4PteGlu1 above 1.0 microM, however, dUMP interference was nearly abolished. Kinetic analysis of the data suggests that this effect of the 5,10-methylene moiety may result in part from positive allosteric effects of first site TS-FdUMP-CH2--H4PteGlu1 ternary complex binding on acceleration of second site binding, in addition to slowed rates of dissociation. Other folylmonoglutamates showed relatively poor TS-[3H]FdUMP-folate complex formation: at 500 microM folate, as a percentage of CH2--H4PteGlu1 values, these were 29.6% for dihydrofolate, 7.5% for 5-CH3--H4PteGlu1, 3.0% for CH = H4PteGlu1, 1.6% for folic acid, 1.1% for 5-CHO--H4PteGlu1 (leucovorin) and 0.9% for 10-CHO--H4PteGlu1. Inhibitory effects by dUMP were consistent with binary complex effects alone for these folates. Study of methotrexate, as the monoglutamate and the hexaglutamate, suggested that ternary complexes with dUMP are favored over those with FdUMP at high concentrations of the antifolate. Our results indicate that activation of leucovorin to over 0.5 microM in intracellular CH2--H4PteGlu1 equivalents may be a requirement for achieving complete TS inhibition by FdUMP in the presence of excess conce

Characterization of the mode of binding of substrates to the active site of Tetrahymena self-splicing RNA using 5-fluorouracil-substituted mini-exons

Splice-site selection specificity in Tetrahymena self-splicing RNA is thought to be mediated by a base-paired complex between a CUCUCU sequence on the end of the 5' exon and a GGGAGG guide sequence in the intron. The substitution of uracil (U) in oligonucleotide mini-exons with 5-fluorouracil (UF), an analogue bearing a much more acidic N-3 proton, allowed us to test the role of hydrogen bonding between complementary bases in the splice-site selection process. The affinities of (U) and (UF) mini-exons for the ribozyme active site were similar and several orders of magnitude greater than expected from base pairing alone. In contrast to CUCU, the CUFCUF mini-exon lost substrate activity with increasing pH, presumably due to ionization of the UF residues. However, the apparent pK values of these residues were several pK units above that of free UF, indicating that the mini-exon is shielded from the solvent by an active site of low polarity. Loss of the pyrimidine N-3 hydrogen bond by selective ionization of the UF residues decreased the binding of CUFCUF to the ribozyme only 3-fold but did prevent its ligation to the 3' exon. Temperature dependence of substrate activity was identical for both (U) and (UF) mini-exons, whereas the UF-substituted ribozyme lost activity at a considerably lower temperature than did the natural (U) ribozyme. These observations indicate that hydrogen-bonded base pairs involving the U residues contribute little to the total binding energy of the 5' splice site with the active site of the ribozyme, but probably help to align the splice sites properly for ligation.

Activity of thymidylate synthetase and its inhibition by 5-fluorouracil in highly enzyme-overproducing cells resistant to 10-propargyl-5,8-dideazafolate

Mouse FM3A mammary adenocarcinoma cells exposed to the specific thymidylate synthetase (TS) inhibitor 10-propargyl-5,8-dideazafolate (PDF) responded by overproducing TS up to 200-fold. In the absence of inhibitor, the elevation of TS levels decayed with a half-life of about 4 weeks. Southern blot analysis of restricted DNA from the PDF-resistant cells using a TS-specific probe showed that the TS gene was amplified to the same extent as enzyme levels. The PDF-resistant cells showed moderate cross-resistance to growth inhibition by 5-fluoro-2'-deoxyuridine, which increased with TS overproduction, but cross-resistance to 5-fluorouracil (FUra) was less (2- to 3-fold) and did not change with increased TS levels. TS activity, measured as release of tritium from [5-3H]2'-deoxyuridine, was no higher in the intact PDF-resistant cells than in wild-type cells. Inhibition of TS activity by FUra in the wild-type cells was accompanied by a proportional decrease in the amount of free TS, presumably due to formation of the tight binding complex of TS with 5-fluoro-2'-deoxyuridylate and 5,10'-methylenetetrahydrofolate. However, in the PDF-resistant cells, most the TS was still in the free form even though TS activity was substantially (85-90%) inhibited. Addition of folinic acid did not change either the sensitivity of the cells to FUra or the rates of tritium release in the cells having overproduced TS. These results are consistent with compartmentalization of TS, possibly in a multienzyme complex.

Concentration and time dependence of the toxicity of fluorinated pyrimidines to HT 29 colorectal carcinoma cells

To determine the optimal concentration time factors for the fluoropyrimidines 5-fluorouracil (FU), 5-fluorouridine (FUR), and 5-fluoro-2'-deoxyuridine (FUdR) in regional chemotherapy, we tested these drugs against the colorectal carcinoma cell line HT 29 at various dosages and exposure times. The measure of cytotoxicity used was the degree of inhibition of colony formation in soft agar after drug treatment compared with untreated control cells. Colonies were visible after 6 days of growth in soft agar, so the initial evaluation of toxicity was done at this time. Additional colonies were found 10 and 16 days after the first evaluation, so the dishes containing the treated cells were also evaluated for this delayed growth phenomenon ("regrowth"), which we considered to be due to a cell growth inhibition effect of the drugs rather than a cytocidal effect. Exposure times of the cells to the drugs ranged from 5 min to 24 h and the doses, between 0.01 and 1000 micrograms/ml. The toxicity of FUdR was concentration-dependent, but its time dependence ceased after a relatively short exposure time. There was a cell population that was not susceptible to FUdR regardless of dose and exposure time; consequently, FUdR treatment was always accompanied by substantial regrowth of colonies. With FU and FUR, conditions could be achieved that resulted in complete cell death (no regrowth), but high concentrations and long exposure times were required with FU. With FUR, on the other hand, both cytostasis and cytotoxicity could be achieved with substantially lower doses and shorter exposure times than with FU. These results indicate that FUR has the potential to be an effective drug in chemotherapy protocols not involving systemic administration.

Synthesis and interaction with uridine phosphorylase of 5'-deoxy-4',5-difluorouridine, a new prodrug of 5-fluorouracil

5'-Deoxy-4',5-difluorouridine (4'-F-5'-dFUrd) (10) has been synthesized on the basis of the rationale that the labilization of the glycosidic linkage caused by the 4'-fluoro substituent might allow this compound to be a better prodrug form of the anticancer drug 5-fluorouracil (FUra) than is the widely studied fluoropyrimidine 5'-deoxy-5-fluorouridine (5'-dFUrd). The rate of solvolytic hydrolysis of the glycosidic linkage of 4'-F-5'-dFUrd at pH 1 was about 500-fold faster than that of 5'-dFUrd. Since uridine phosphorylase is thought to be the enzyme that causes degradation of 5'-dFUrd in vivo to generate FUra, we compared the substrate interactions of 5'-dFUrd and 4'-F-5'-dUrd with this enzyme. The Vmax for hydrolysis of 4'-F-5'-dFUrd to FUra by uridine phosphorylase was about 5-fold greater than that of 5'-dFUrd, whereas the Km value of 4'-F-5'-dFUrd was 10-fold lower. The combination of these two factors results in 4'-F-5'-dFUrd having a 50-fold higher value of V/K than does 5'-dFUrd. Against L1210 cells in culture, the IC50 value for growth inhibition by 4'-F-5'-dFUrd was 3 X 10(-7) compared to 3 X 10(-6) for 5'-dFUrd.

Methotrexate analogues. 29. Effect of gamma-aminobutyric acid spacers between the pteroyl and glutamate moieties on enzyme binding and cell growth inhibition

A series of "stretched" methotrexate (MTX) analogues containing up to five 4-aminobutyryl (Gab) spacers between the 4-amino-4-deoxy-N10-methylpteroyl (MeAPA) moiety and the glutamate (Glu) side chain was prepared. Interest in these compounds stemmed from their relationship to MTX gamma-polyglutamates, from which they differ only in lacking "internal" alpha-carboxyl groups. The ability of the MeAPA-Gabn-Glu derivatives to inhibit dihydrofolate reductase (DHFR) and thymidylate synthase (TS) in vitro and to inhibit the growth of tumor cells in culture was evaluated. The IC50 for DHFR inhibition increased progressively from 0.082 to 0.84 microM as the number of Gab spacers was varied from one to five. At the same time the introduction of Gab spacers was found to produce substantial TS inhibition (Ki 0.1-0.4 microM) similar to that reported for MTX polyglutamates. Despite the activity of the MeAPA-Gabn-Glu derivatives as combined inhibitors of TS and DHFR, there was a steep loss of cell growth inhibitory potency as the number of Gab spacers was increased. This most likely reflects low cell uptake and the fact that when n greater than 1 there is almost total abolition of substrate activity for folylpolyglutamate synthetase, which had previously been observed with n = 1.

Kinetics and mechanism of interaction of 10-propargyl-5,8-dideazafolate with thymidylate synthase

The interaction of Lactobacillus casei thymidylate synthase (TS) with 10-propargyl-5,8-dideazafolate (NPQ) in the presence of 2'-deoxyuridylate (dUMP) has been investigated. After formation of a rapidly reversible dUMP-NPQ-enzyme complex, a slow isomerization occurs to provide a ternary complex that can be isolated on nitrocellulose membranes or by gel filtration. Unusual features of the isolable complex are the slow rate by which it is formed (t1/2 = 0.88 h) and the slow rate at which it dissociates (t1/2 = 26.5 h). The ternary complexes contain 2 mol of dUMP and 2 mol of NPQ bound per mol of dimeric enzyme. Ultraviolet difference spectra of the dUMP-NPQ-TS complex shows a high wavelength maximum that has been attributed to perturbations of the enzyme and/or ligand chromophores that occur upon binding. Data are presented that suggest that the formation of the isolable ternary complex involves nucleophilic attack by a catalytic thiol group of the enzyme to the 6-position of dUMP. Evidence for this is as follows: first, there is a decrease in the absorbance of the pyrimidine chromophore at 265 nm that occurs at the same rate as the formation of the isolable complex; second, using [6-3H]dUMP there is a large, inverse alpha-secondary kinetic isotope effect (kappa H/kappa T = 0.83) upon formation of the complex that is in accord with sp2 to sp3 rehybridization of the 6-carbon of the heterocycle. Treatment of the complex with sodium dodecyl sulfate (NaDodSO4) results in the dissociation of both ligands in an unmodified form, which is consistent with proposed structure of the complex. Isolable ternary complexes are also formed when the enzyme is incubated with 5-fluoro-2'-deoxyuridylate (FdUMP) and NPQ. Interestingly, the dissociation of FdUMP from these complexes is biphasic, with one-half of the bound nucleotide dissociating at an exceedingly slow rate (t1/2 congruent to 100 h). The findings are discussed with relationship to the possible use of NPQ as an anticancer agent.