Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
[CANCER RESEARCH 48, 2426-2431. May I, 1988] Role of Folylpolyglutamate Synthetase in the Regulation of Methotrexate Polyglutamate Formation in H35 Hepatoma Cells1 Thomas B. Johnson, M. G. Nair, and John Galivan2 Wadswonh Center for Laboratories and Research, New York State Department of Health, Albany, New York [T. B. J., J. G.]; and Department of Biochemistry, University of South Alabama, Mobile, Alabama /M. G. N.J ABSTRACT The effect of culture conditions on the glutamylation of methotrexate by intact H35 hepatoma cells and folylpolyglutantate synthetase (FPGS) activity in the corresponding crude extracts has been examined. The rate of cellular glutamylation of methotrexate observed in rapidly dividing cultures was 4-fold higher than confluent cultures, and was accompanied by an increase in extract FPGS activity (2.2-fold). The depletion of cellular folates produced comparable increases in both cellular metho trexate glutamylation and extract FPGS activity (approximately 1.8fold). Near-quantitative reductions in cellular methotrexate glutamylation were caused by media additions of reduced folates and methotrexate to confluent cultures of wild-type and folate-depleted H35 cells. However, these produced relatively modest reductions in FPGS activity in the correspondingcrudeextracts (approximately 50%). Methionine exclusion resulted in a greater than 50% decrease in FPGS activity in crude extracts of these cells compared to extracts of control cultures. The combination of methionine exclusion and folinic acid addition lowered the FPGS activity to less than 25% that of control. The data suggest that the changes in the glutamylation rate of methotrexate in whole cells due to culture conditions such as folate restriction, reduced folate addition, methionine exclusion, and growth state are at least in part a consequence of alterations in FPGS activity. This conclusion is consistent with the proposition that the metabolism of slow-acting substrates for FPGS (such as 4-amino antifolates and their corresponding polyglutamates) may be sensitive to changes in enzyme levels or activity (Cook et al., Biochemistry, 26: 530-539, 1987). Analysis of the products formed by FPGS from extracts using methotrexate as the substrate revealed no significant amounts of polyglutamate species higher than 4-MI.-1 (M 'II,PteGluj. In contrast, when using the thymidylate synthase inhibitor V"propargyl-5,8-dideazafolic acid as the starting substrate under identical assay conditions, FPGS from extracts catalyzed the formationof predom inantly long chain polyglutamate derivatives (Gliu and higher). These results reflect the relative efficacy of methotrexate and W'-propargyl5,8-dideazafolic acid, as well as their polyglutamate derivatives, as sub strates for FPGS. INTRODUCTION FPGS3 (EC 6.3.2.17) catalyzes the conversion of folates and folate analogues into 7-polyglutamate derivatives. The polyglutamylation of folates (1-5) serves several physiological func tions. Folylpolyglutamates appear to be the preferred substrates for many of the folate-requiring enzymes of one-carbon metab olism (1,6), and have been shown to be more efficiently "chan neled" than monoglutamates from one active site to another in folate-requiring multienzyme complexes (7). Polyglutamate de rivatives of folates and methotrexate are more avidly retained in cells than their monoglutamate counterparts, and thus the Received 11/3/87; revised 2/1/88; accepted 2/8/88. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1This work was supported by NIH Grants CA25933, CA32687, and CA34314 by the National Cancer Institute, USPHS/Department of Health and Human Services. 1 To whom requests for reprints should be addressed. 3 The abbreviations and trivial names used are: FPGS, folylpolyglutamate synthetase; 4-NH2-10-CH3-PteGlu, methotrexate (MTX) or 4-amino- 10-methylpteroyl glutamic acid; PDDF, /V'°-propargyl-5,8-dideazafolic acid, N-\4¡(2-amino-4-hydroxy-6-quinazolinyl)methyl]prop-2-ynyl amino|benzoyl-L-glutamic acid, also known as CB3717. FPGS-catalyzed synthesis of methotrexate polyglutamates is a major determinant in the cytotoxicity of this antifolate (1,814). Methotrexate polyglutamates have been shown to be at least as effective as methotrexate against the target enzyme, dihydrofolate reducÃ-ase (8, 9, 11, 12, 15, 16), and therefore their prolonged cellular retention affords them greater cytotoxic potential than the parent drug. Mammalian FPGS has been partially purified from rat (17), mouse (18), and beef liver (19), and has only recently been purified to homogeneity from hog liver (20). While studies utilizing these preparations have been extremely useful in elu cidating the chemical and physical properties of the enzyme, a limited amount of data are available as to its mode of activity in cells and its role in regulating the balance of cellular polyglu tamates of methotrexate. Furthermore, relatively few investi gations on FPGS have used tumor cells as an enzyme source. Finally, there appear to be marked differences in the ability of FPGS to glutamylate methotrexate in whole cells versus exper iments with isolated enzymes, especially with regard to the distribution of polyglutamate products (13, 14, 21-25). Previous studies in our laboratory (26, 27) have shown that the ability of H35 hepatoma cells to glutamylate methotrexate can be altered by changes in growth state and culture conditions. In an effort to determine if these alterations are a consequence of changes in FPGS activity, as well as to gain additional insight into the process of glutamylation in intact cells, we have quantitated the FPGS activity in extracts of H35 hepatoma cells grown under those conditions which caused the greatest differ ences in cellular glutamylation of methotrexate. We have also attempted to understand what factors account for the differ ences in polyglutamate product distribution with isolated en zyme compared with that seen in cultured tumor cells by evaluating the glutamylation of methotrexate and the thymi dylate synthase inhibitor PDDF (28-30). It is envisioned that a better understanding of the role of FPGS in the glutamylation of these antifolates will contribute to an enhanced understand ing of their activities as chemotherapeutic agents. MATERIALS AND METHODS Materials. Swims' medium S-77, folie acid-free Swims' medium, methionine-free Swims' medium, fetal calf serum (dialyzed and undialyzed), and horse serum (dialyzed and undialyzed) were purchased from Grand Island Biological Company. L-[2,3-3H]Glutamic acid (NET-395, 25 Ci/mmol) was obtained from New England Nuclear (Boston, MA). Cellulose power (CF 11) and DEAE-cellulose (DE 52) were purchased from Whatman. Methotrexate, [3',5',7-3H]methotrexate, and 4-NH2-10-CH3-Pte-[G-3H]Glu2 were purchased from Moravek Biochemicals (La Brea, CA). Methotrexate polyglutamate standards were the kind gift of the National Cancer Institute. PDDF, PDDF polyglutamates, and [14C]PDDF labeled with [G-MC]glutamate were synthesized by the method of Nair et al. (31). Thymidine, hypoxanthine, and ATP were purchased from Sigma. Folinic acid was supplied by ICN Pharmaceuticals, and 5-methyltetrahydrofolic acid was prepared as previously described (32). Methotrexate and all folates were purified by DEAE-cellulose chromatography prior to use (32). All other chem icals were reagent grade. 2426 Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1988 American Association for Cancer Research. REGULATION BY FOLYLPOLYGLUTAMATE SYNTH ETASE Cell Culture. H-11-E-C3 cells derived from the Reuber H35 rat hepatoma (referred to as H35 cells) were grown on 60- or 100-mm Falcon dishes in a 5% CO.- atmosphere and were subcultured weekly. The culture medium was Swims' medium S-77 supplemented with 5% fetal calf serum, 20% horse serum, and 4 HIMglutamine, which was routinely changed at 72 and 120 h after plating. The cells were released from the dishes with 0.05% trypsin, counted with a Model ZBI Coulter counter and diluted to a seeding density of 5 x 10' cells/ml. Folatedepleted H3S cells were generated as before (26) and cultured in a similar fashion, but in folate-free Swims' medium S-77 supplemented with 50 /¡Mthymidine, 50 /IM hypoxanthine, 5% dialyzed fetal calf serum, 20% dialyzed horse serum, and 4 m\i glutamine. Preparation of Crude Extracts. Wild-type or folate-deplete H35 cells were seeded in 100-mm Falcon dishes at a density of 6 x 10s cells/ dish, and unless otherwise noted, were grown for 96 h prior to changing to the desired culture medium for experiments, which was accomplished by washing each 100-mm plate with 5 ml of the desired medium lacking serum and replenishing with 5 ml of the same medium. After the indicated time, the cultures were cooled on ice, the medium was removed, and each dish was washed with 3- x 4-ml ice-cold phosphate buffered saline (pii 7.6). The cells were scraped into ice-cold 0.5 M Tris Cl (pH 8.85), 0.2 M 2-mercaptoethanol (approximately 1 ml buffer/15 mg total cell protein) and lysed by freeze-thawing (3x) in dryice/ethanol. The debris was removed by centrifugation (4"C) at 27,000 x g for 30 min. Following removal of an aliquot for protein determi nation by the Lowry method (33), the supernatant was stored at — 70"C. The FPGS activity of the crude extracts stored in this manner did not change for 6 weeks. The activity of a particular extract was typically determined on the day following preparation. Ammonium Sulfate Fractionation. Solid ammonium sulfate was added to the crude supernatant over 20 min at 0°Cwith stirring until 35% saturation was reached. The solution was stirred at the same tempera ture for l h and centrifuged (4"( ') for l h at 27,000 x g. The supernatant was discarded, and the pellet was suspended in a minimum volume of 20 HIMpotassium phosphate buffer (pH 7.5), 50 HIM2-mercaptoethanol, 20% glycerol. Following protein determination (33) the sample was divided into 1.5-ml portions and stored at —70°C. FPGS activity was determined without dialysis. Enzyme Assays. Folylpolyglutamate synthetase was assayed accord ing to the method of McGuire et al. (17). Methotrexate was used routinely at saturation (250 /IM) as the substrate for FPGS activity determinations (21). The stock solution of glutamate (50 in M) was prepared by combining 50 mM sodium glutamate (pH 7.5), 150 HIM sodium glutamate (pH 9.0), and L-[2,3-3HJglutamate (NET-395) from New England Nuclear in a 3:1:2 volume ratio. The specific radioactivity of the glutamate stock solution was 12-14 x 10" dpm/Vmol. Each assay solution contained, in a final volume of 0.250 ml at pH 8.4, 100 HIMTris-Cl (pH 8.85), 10 mM ATP, 20 mM MgO2, 20 mM KC1, 100 HIM 2-mercaptoethanol, 4 mM [3H]glutamate, and methotrexate or PDDF. Following termination of the reaction with addition of l-ml ice-cold 5 HIMNa-glutamate (pH 7.5), 25 mM 2-mercaptoethanol, the unreacted [3H]glutamate was separated from the polyglutamates by DEAE-cellulose column chromatography (column buffer 10 mM TrisCl (pH 7.5), 110 mM NaCl, 25 mM 2-mercaptoethanol). The DEAE minicolumns were then washed with 2 ml water prior to elution of the radioactive polyglutamates with 3 ml 0.1 N HC1. The FPGS activity for a particular extract was determined by assaying the extracts in duplicate for l h at 400 ;<gof added protein and for 2 h at 200 ¿tgof added protein. One-third of the acid eluant was used for counting and the remainder was lyophilized for product analysis. -y-GIutamyl hydrolase (conjugase) activity was monitored according to the method of Samuels et al. (34). Methotrexate and PDDF Glutamylation in Intact H35 Hepatoma Cells. Wild-type H35 hepatoma cells were seeded on 60-mm Falcon dishes at a density of 2 x IO5cells/dish and grown in culture for 96 h. The medium was changed to folate-free Swims' medium S-77 with with 4- x 4-ml ice-cold isotonic saline containing 10 mM potassium phosphate, pH 7.4. Four dishes were used for each experimental point. Two were washed separately with l N NaOH to determine the total intracellular polyglutamates and cellular protein. The two remaining dishes were pooled in 2 ml distilled water, heated in a boiling water bath for 10 min, and lyophilized for product analysis. Product Analysis. Methotrexate polyglutamate product distribution was determined by high-pressure liquid chromatography as previously described (27). The same system was employed for the analysis of the polyglutamates of PDDF except that a linear gradient of 4 to 24% acetonitrile in 0.1 M sodium acetate over 40 min was used. Charcoal Controls. A crude extract of folate-deplete H35 hepatoma cells which had been exposed to 20 //\i folinic acid for 24 h was divided into two 0.5-ml portions and placed on ice. Sufficient ice-cold 2.5% dextran-treated charcoal (35) to bind 10 nmol of folate was added to one extract, while an equal volume of ice-cold water was added to the control. After gentle mixing both extracts were centrifuged (4°C)for 10 min at 12,000 rpm (Beckman Microfuge B), and the supernatants were each assayed for FPGS activity. RESULTS Conditions for Determination of FPGS Activity. In order to determine FPGS activity in extracts of H35 cells, duplicate incubations of 400-/xg extract protein for l h and 200 /¿g protein for 2 h were conducted. Under several conditions shown in Table 1, this combination of assays exhibited equivalent product formation with 250 J¿M methotrexate as substrate. These results are consistent with linearity with respect to time and protein concentration. However, extension of assays to longer time periods with either amount of protein caused a loss of linearity. Shorter incubation times could be utilized, but these could compromise the sensitivity of the assay, especially with less active extracts. Thus, we routinely used the assay format de scribed in Table 1 for measuring the activity of FPGS in cell extracts. The linearity of the assay suggested that 7-glutamyl hydrolase (conjugase) was not interfering with the measurement of meth otrexate polyglutamate formation in assays of crude extracts. To determine the relative activity of 7-glutamyl hydrolase under the conditions of the FPGS assay, the hydrolysis of 200 pmol of 4-NH2-10-CH3-Pte-[G-3H]Glu2 (a typical amount formed in an assay) was measured by incubation with 400 ¿tgof extract for l h or 200 n%of extract for 2 h at pH 8.4. High-performance liquid chromatography analysis (34) showed that in each case less than 0.3% of the substrate had been converted to the monoglutamate. Thus 7-glutamyl hydrolase, which has a rela tively low activity in these cells (36) did not interfere with the measurement of FPGS under these conditions. Methotrexate was compared to tetrahydrofolate as a sub strate for FPGS from crude extracts under the standard assay Table I Product formation from ['HJglutamate and methotrexate with FPGS from extracts of wild-type H35 hepatoma cells with respect to time and protein concentration The results are the average of identical duplicate experiments. Confluent and log phase cells were harvested after 96 and 48 h in culture medium, respectively. Extracts of cells lacking serum in culture were harvested after exposure of cells grown as described in "Materials and Methods" to Swims' S-77 medium + insulin (10 mU/ml) * 4 mM glutamine for 24 h (total culture time, 120 h). in Culture conditionConfluentConfluent—Serum—SerumLog time(min)601206012060120Protein added (^g)400200400200400200[3H]Glutamate corporated (pmol)260251285320520621 insulin (10 mU/ml), and after 24 h the cells were incubated for the indicated time period with 10 /IM [3H]methotrexate or 10 /¿M(14C]PDDF at a specific radioactivity of 4 x IO4 dpm/nmol. Polyglutamate formation was terminated by cooling the cultures on ice and washing 2427 phaseLog phaseIncubation Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1988 American Association for Cancer Research. REGULATION BY FOLYLPOLYGLUTAMATE SYNTHETASE conditions. At saturating concentrations (250 and 35 /IM for methotrexate and tetrahydrofolate, respectively) methotrexate typically incorporated approximately 1.8-fold as much gluta mate as did tetrahydrofolate (data not shown). Alterations in Cellular MIX Glutamylation and Extract FPGS Activity as a Function of Growth State and Culture Condition. A comparison of the rate of glutamylation of methotrexate by intact H35 hepatoma cells and FPGS activity in crude extracts of the same cultures is shown in Table. 2. The greater rate of cellular glutamylation of methotrexate observed in rapidly di viding cultures (>4-fold higher than confluent cultures) was accompanied by an increase in extract FPGS activity (2.2-fold). The depletion of cellular folates produced comparable increases in both cellular methotrexate glutamylation and extract FPGS activity (approximately 1.8-fold). The addition of exogenous reduced folates or methotrexate to confluent cultures of wildtype and folate-depleted H35 cells, which in some cases caused near-quantitative reductions in cellular methotrexate glutamy lation, produced relatively modest reductions in FPGS activity in the corresponding crude extracts (approximately 50%). To determine if the reduction in FPGS activity in the cytosol of the folate-treated cells represented substrate competition be tween the added folate in the cell extracts and the substrate, methotrexate, an extract of folate-deplete cells which had been exposed to 20 /¿M folinic acid for 24 h was treated with 2.5% dextran-treated charcoal and reassayed for FPGS activity. No significant change compared to a charcoal-untreated extract was observed, indicating that an increase in cellular folates following exposure to folinic acid was not altering the apparent FPGS activity in extracts. The relationship between extract FPGS activity and growth state of wild-type H35 cells was further investigated by meas uring the enzyme activity in extracts of cells harvested at various points in the normal growth cycle. The results of this study are depicted graphically in Fig. 1. Extract FPGS activity was shown to gradually increase for the first 48 h and then peak sometime between 48 and 54 h before returning to the initial level in extracts of confluent cells (500-600 pmol/h/mg). Alterations in FPGS Activity in Extracts of Folate-Deplete 1135 Hepatoma Cells Deprived of Methionine. Earlier studies 1500 1000 co=)40 ÃŽ_jUJü£ co O 1I•-8060cctu 0- 500 -r1• 20 ._'.j—Õ1'"X,,\1 24 48 54 72 144 CULTURE (hrs) Fig. 1. The relationship between extract FPGS activity (H) and growth state (O) of wild-type H35 cells. The cells were seeded at a density of 5 x IO4cells/ml (2 X 10s cells/60-mm dish), cultured for the indicated time in Swims' S-77 medium, 5% fetal calf serum, 20% horse serum, and 4 mM glutamine. Cell numbers were determined by visual counting using a Zeiss model E invertoscope equipped with a reticle. Preparation of extracts of these cultures and measurement of FPGS activity were accomplished as described in "Materials and Methods." The results are the average of three experiments except for the 24-h assay for FPGS which is the mean of two observations. Table 3 Effect ofmethionine exclusion and folate addition on FPGS activity in extracts of folate-deplete H35 hepatoma cells Data are presented as the mean of at least four separate measurements under linear assay conditions with standard deviation. Folate-depleted cells were grown in culture for 96 h. Media was removed and the cultures were rinsed with 4 ml sterile Hanks' solution, and then exposed to folate-free Swims' S-77 medium + glutamine (4 mM), insulin (10 mU/ml), ±methion ine (100 /IM), + folinic acid (20 MM),or 5-methyltetrahydrofolate (20 /¿M) for 18 h. Rinsing and exposure was then repeated and the cells were harvested 24 h later. FPGS activity (pmol/h/mg) Control -Methionine —Methionine + folinic acid -Methionine + 5-methyltetrahydrofolate 1103 ±93 476 ±96 245 ±60 515° " Two determinations. Table 2 Alterations in the glutamylation of methotrexate by intact H35 hepatoma cells and FPGS from crude extracts by growth state and culture conditions Log phase cells were harvested after 48 h in culture medium. Cells at confluence were exposed to folinic acid (20 /IM), 5-methyltetrahydrofolate (20 pM), or methotrexate (20 MMin whole cell experiments, 10 pM for cells to be made into extracts) for 24 h prior to incubation with [3H]methotrexate or harvesting. In whole cell experiments, the indicated additions were removed 30 min before the 4-h incubation with labeled methotrexate. For all the conditions tested, intact H35 cells synthesize methotrexate polyglutamates linearly with respect to time for 4 h at a media concentration of 10 MM(26). The results for the activity of FPGS in extracts are the average of two separate duplicate assays performed as described under "Materials and Methods." The average standard deviation when comparing extracts of cells grown under identical conditions is 15% (n = 55). Wild-type cellsLog H35 phaseControl (confluent)+ acid+ Folinic MethotrexateFolate-deplete cellsControl H35 (confluent)+ acid+ Folinic 5-Methyltetrahydrofolate-tMethotrexateWhole * Data from Ref. 26. from this laboratory have shown that the exclusion of methionine from the culture media caused nearly a 70% reduction in the glutamylation of methotrexate by intact folate-deplete H35 hepatoma cells (14). As can be seen in Table 3, methionine exclusion resulted in a greater than 50% decrease in FPGS activity in crude extracts of these cells compared to extracts of control cultures. The combination of methionine exclusion and folinic acid addition lowered the FPGS activity to less than 25% that of control. The combination ofmethionine exclusion and 5-methyltetrahydrofolate addition reduced the activity to approximately half that seen with extracts of control cultures. cellMTX Analysis of Products Formed from Methotrexate and PDDF gluta FPGSactivity(pmol/h/mg)14266383463621198775616675 mylation*(pmol/4 in Intact r135Hepatoma Cells and by FPGS from Crude Extracts h/mg)211470.212837215Extract or Resuspended 0-35% Ammonium Sulfate Pellets. Previous investigations with liver FPGS indicated a limited ability of the enzyme to form longer chain length methotrexate polygluta mates (Gluj and higher) (21, 22, 24, 25), as was the case with cultured hepatocytes during short term incubations (14). Since the rat hepatoma cells readily formed longer chain derivatives under similar conditions (13,26), an evaluation of the products formed with hepatoma cell extracts was conducted to determine if the enzyme from this source could more readily produce 4NH2-10-CH3-Pte-Glu4and Glu5. Methotrexate at an extracel2428 Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1988 American Association for Cancer Research. REGULATION BY FOLYLPOLYGLUTAMATE SYNTHETASE lular concentration of 10 /¿-M was extensively glu tam ylaied by intact wild-type H35 hepatoma cells in 6 h (Table 4), with the predominating products (73%) containing four or more gluta mate residues. In contrast, FPGS from crude extracts synthe sized polyglutamates from methotrexate only up to the triglutamate level (Table 5). Partial purification of the extracts by ammonium sulfate fractionation did not result in a preparation which could catalyze the formation of significant amounts of the tetraglutamate (Table 6). This partially purified extract was capable of converting 4-NH2-10-CH3-PteGlu3 to 4-NH2-10CH.i-Pte-Glu4, and the amount formed was highly dependent on the amount of substrate (Table 6). The same reaction oc curred in crude extracts, but at a 50-70% reduction amount (data not shown). The monoglutamate of the thymidylate synthase inhibitor PDDF (10 MM)was also glutamylated by intact H3S hepatoma cells (Table 4), with the glutamylated products consisting nearly exclusively of the tetra- or higher polyglutamate derivatives. A radiolabeled peak near the expected retention time for the pentaglutamate derivative of PDDF was detected during the high-performance liquid chromatography product analysis, which increased when the incubation time was extended from 6 to 24 h. Lack of a pentaglutamate standard prevented positive identification. In contrast to methotrexate, PDDF was readily Table 4 Metabolism of methotrexate and PDDF in H35 hepatoma cells H35 cells were grown as described in "Materials and Methods" for 96 h. The culture medium was changed to folate-free Swims' S-77 medium plus insulin (10 mU/ml), and 24 h later the cells were incubated with [3H]methotrexate or ['V| PDDF for the indicated time period. Polyglutamate formation was monitored as described in "Materials and Methods." Data are presented as the mean of at least two separate experiments. Glu, is indicative of all species of Glu«and greater. Cellular polyglutamates Compound Concentration OIM) Incubation (h) Concentration (nmol/g) % Distribution Glu2 (.lu, Glu« able to form substantial amounts of the tetra- and higher polyglutamate derivatives when using crude extracts of folatedeplete H35 hepatoma cells as an enzyme source (Table 5). Using a resuspended 0-35% ammonium sulfate pellet from crude extracts as a source for FPGS (Table 6), the product distribution formed after 6 h from PDDF at initial concentra tions lower than 1 ^M was virtually identical to that seen in the experiments with intact cells (i.e., near-exclusive formation of the tetra- and higher polyglutamate derivatives). DISCUSSION Several detailed investigations using FPGS from normal mammalian tissues have appeared in the recent literature (1720), but relatively few studies have used rapidly dividing or transformed cells as an enzyme source. Taylor and Ilamia showed that the FPGS activity of suspension-cultured Chinese hamster ovary cells was not altered by the growth state of the cells or the addition of several metabolites (glycine, adenosine, thymidine, folates, and methionine) to the media (37), suggest ing that changes in growth state and culture conditions did not result in altered activities or amounts of the enzyme. However, the finding that the ability of H35 hepatoma cells to glutamylate methotrexate could be markedly altered by changes in growth state, cellular folate levels, and the exclusion of methionine from the media (14, 26) suggested the possibility that changes in FPGS activity in cells could be involved in regulating the glutamylation of methotrexate. This possibility has been tested in the present study by measuring FPGS activity in extracts of H35 hepatoma cells grown under those culture conditions which caused the greatest alterations in whole cell glutamyla tion. Using crude extracts of H35 cells as a source for FPGS, linear incorporation of [3H]glutamate into 250 MMmethotrexate took place for up to 1 h with 400 /¿gof added protein, and up to 2 h with 200 Mgof added protein (Table 1). These data, along with the finding that -y-glutamyl hydrolase [optimum pH 7.2 (34)] was virtually inactive under these conditions (pH 8.4), Table 5 Analysis of products formed by FPGS from crude extracts offolateindicated that a reliable comparison of FPGS activity could be depleted H35 hepatoma cells using methotrexate and PDDF as substrates Standard assay procedure as described in "Materials and Methods" was em made by assaying the extracts in duplicate for 1 h at 400 /¿g of ployed using the indicated concentration of methotrexate or PDDF, a 6-h incu added protein and for 2 h at 200 ¿tgof added protein. At bation time, and 0.8 mg of added protein. A crude extract of confluent folate saturating substrate concentrations, the 1.8-fold higher gluta depleted H3S cells was used as an enzyme source. mate incorporation for methotrexate versus tetrahydrofolate % Polyglutamate compared favorably with the results of similar experiments by SubstrateMethotrexateMethotrexateMethotrexatePDDFPDDFPDDF(MM)1010.21010.2(pmol)8831113180817942Glu26358783210Glu,364222565839Glu«»100414151 McGuire et al. using partially purified rat liver FPGS, which showed an approximate 1.6-fold greater glutamate incorpora tion with methotrexate (21). The results of these experiments suggest that the alterations in the cellular glutamylation of methotrexate are in part a consequence of changes in the activity of FPGS (Table 2). The reduction of cellular folates (i.e., comparing control cultures of Table 6 Analysis of products formed by FPGS from resuspended 0-35% wild-type versus folate-deplete H35 cells), caused comparable ammonium sulfate pellet of crude extracts of 1135 hepatoma cells increases in both whole cell glutamylation and extract FPGS Standard assay procedure as described in "Materials and Methods" was em activity (approximately 1.8-fold), indicating that alterations in ployed using a 6-h incubation time and 400 Mgof added protein. the enzyme itself may be largely responsible for this effect. Polyglutamate1Glui655859500800Glu,324241129059120Glu«*30083911003388100 However, media additions of folinic acid and 5-methyltetrahyConcentration incorporatec OIM)4-NHj-lO-CHj-PteGlu,4-NHj-lO-CHj-PteGlu,4-NH2-10-CH,-PteGlu,4-NH2-10-CH,-PteGlu,4-NHj-10-CH,-PteGlu,4-NH2-10-CH,-PteGlu,PDDF Substrate (pmol)10111381824140364316734% drofolate to wild-type and folate-deplete cultures for 24 h, which caused a near-quantitative reduction in cellular methotrexate glutamylation [Table 2, (26)], resulted in only a 50% decrease in extract FPGS activity, demonstrating that these changes in enzyme activity are probably not large enough to totally account for alterations in glutamylation in intact cells. Thus, other (Glu),PDDF (Glu),PDDF factors which could alter cellular glutamylation of the drug, (Glu),1010.21001011010.2Glutamate such as the competition between reduced folates and methoMethotrexatePDDFPDDF1010106624113.011.122.240023207398100 2429 Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1988 American Association for Cancer Research. REGULATION BY FOLYLPOLYGLUTAMATE SYNTHETASE trexate for transport (38) and FPGS (21, 22, 39), can also import an additional effect. The increased FPGS activity in extracts of H35 cells har vested in log phase (Table 2 and Fig. 1) indicates that changes in enzyme activity are also involved in the increased ability of dividing cells to glutamylate methotrexate. These data are the first of their kind and are in contrast to the results of similar experiments by Taylor and Hanna (37), who reported that FPGS activity of suspension-cultured Chinese hamster ovary cells was not altered by growth state. A report of increased FPGS activity in dividing LSI78Y lymphoma cells has recently appeared in abbreviated form (40). Methionine has been shown to alter both folylpolyglutamate product distribution in Chinese hamster ovary cells (41) and the rate of polyglutamylation in hepatic cells. In the latter case, glutamylation was reduced by the presence of methionine in normal hepatocytes and increased by the presence of methio nine in folate-restricted hepatoma cells (14). FPGS activity in extracts of H3S cells deprived of methionine was decreased 57% compared to extracts of control cultures (Table 3), while the corresponding decrease in whole cell glutamylation was 70% (14). The combination of folinic acid addition and methi onine exclusion caused the largest reduction in extract FPGS activity measured during this study (>7S%). The mechanism by which methionine alters FPGS activity is not understood, but the effect is particularly interesting because of its magnitude and divergent nature in hepatocytes and hepatoma cells. Ex periments in which FPGS activity was measured in extracts of hepatocytes indicated that the reduction in glutamylation is not associated with changes in enzyme activity. One possible additional mechanism for alterations in gluta mylation is an alteration in the steady state intracellular levels of methotrexate as a result of different conditions. Several positive effectors of the rate of MTX glutamylation which include insulin, methionine, and folate-Iacking medium (26) caused no discernible alteration in the steady state levels of methotrexate during the period of polyglutamate formation in intact cells.4 Similarly the other substrates for FPGS (ATP and glutamate) are not altered in amount by the changes in growth conditions. However dividing hepatoma cells accumulated more intracellular methotrexate than confluent cultures when incu bated in the presence of a concentration of the drug (10 /IM) that is saturating for transport and glutamylation. Under these conditions confluent cells had 8.8 ±3.7 nmol/g and dividing cells had 14.2 ±6.7 nmol/g methotrexate (N = 6). Thus, this 1.6-fold increase in cellular MTX may also contribute to the enhanced rate of glutamylation observed in dividing cells. Shane and his coworkers have conducted a series of elegant studies in which purified FPGS from hog liver was used to model folate and folate analogue metabolism, and concluded that changes in cellular folylpolyglutamate synthetic rates and distributions can be explained by substrate specificities and affinities for the enzyme (25). Furthermore, Shane has pointed out that an exception to autoregulatory control by substrate specificity would exist in the case of substrates such as metho trexate (and its polyglutamates), whose metabolism is relatively slow. Under these conditions, modest changes in the level or activity of FPGS would play an important role in the rate of conversion to polyglutamates (25). Thus, the present finding that extracts of H35 hepatoma cells with differential capacities for glutamylation have correspondingly altered FPGS activities suggests that the cellular levels or activity of this enzyme may 4 T. B. Johnson and J. Galivan, unpublished results. play an important role in the regulation of methotrexate me tabolism in transformed mammalian tissue. The activity of 7glutamyl hydrolase, the enzyme which catalyzes the hydrolysis of folyl and antifolyl polyglutamates, does not appear to be a major factor in altering glutamylation in intact H35 cells as a result of growth state, folate, and methionine content, as its activity in extracts of cells grown under these conditions is not significantly changed.4 Long chain methotrexate polyglutamates (those with four or more glutamate residues) can be formed in hepatoma cells as early as 2 h after exposure to methotrexate and can be the major species within 4 h (13, 26), while virtually no methotrex ate species above the level of the triglutamate are formed in extended (>36 h), enzyme-supplemented incubations with par tially purified FPGS from rat and mouse liver (22, 24). The latter is consistent with the reduced capacity of cultured hepa tocytes to form longer chain length polyglutamates during short-term incubations (<6 h). In this investigation, FPGS from crude extracts synthesized polyglutamates from methotrexate only up to the triglutamate level (Table 5), in contrast to observations with intact cells (13, 26). The absence of higher polyglutamate products was not a consequence of 7-glutamyl hydrolase activity, since ammonium sulfate fractionation [which greatly reduced the hydrolase activity in the extracts (23)] failed to yield a preparation which could form significant amounts of 4-NH2-10-CH3-Pte-Glu4 from methotrexate (Table 6). Additionally, the ammonium sulfate fraction was able to form 4-NH2-10-CH3-Pte-Glu4 from 4-NH2-10-CH3-Pte-Glu3, indicating that this conversion could take place provided the triglutamate was present at effective concentrations. Thus in spite of the fact that the hepatoma cells have a greater capacity to form long chain polyglutamates of methotrexate relative to hepatocytes, the FPGS extracted from both sources has a limited capacity to convert the drug to its high polyglutamates. In contrast to the result with methotrexate, both intact H35 hepatoma cells and FPGS from extracts metabolized the nionoglutamate of the thymidylate synthase inhibitor PDDF to long chain polyglutamates (Tables 4 and 5). Furthermore, a similar product distribution for PDDF as was seen in whole cell exper iments was noted when using FPGS from crude extracts or resuspended 0-35% ammonium sulfate pellets (Table 6). These results suggest that the inability of methotrexate to form long chain polyglutamates outside of the cellular environment is due to the triglutamate being formed in quantities insufficient to overcome its poor kinetic properties as a substrate for further glutamylation. Shane et al. have shown that the diglutamate was the highest chain length formed for several 4-amino unti folates tested with the purified hog liver enzyme (25). Decreased substrate activity of methotrexate polyglutamates as the gluta mate chain length is increased has been reported by Schoo et al., who showed that 4-NH2-10-CH3-Pte-Glu3 had about 3-5% of the substrate activity of methotrexate for the beef liver enzyme (22), and by Clarke and Waxman, who found that the relative Vm^/Km ratios of 4-NH2-10-CH3-Pte-Glu3.5 for the human liver enzyme were each less than 20% that of metho trexate (42). It does not appear that the catalytic ability of FPGS is severely compromised outside of the cell, since the use of other substrates such as reduced folates (4, 17, 19, 23, 25) and PDDF (Tables 5 and 6) can result in the formation of long chain polyglutamates. Other possible factors contributing to the limited glutamylation of methotrexate in experiments with isolated enzyme may be the inability to achieve the effective cellular concentration of FPGS, and the absence of cellular proteins which may alter glutamylation. 2430 Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1988 American Association for Cancer Research. REGULATION BY FOLYLPOLYGLUTAMATE SYNTHETASE The results of this study show that the activity of FPGS may be an important parameter in the regulation of methotrexate polyglutamate formation in H35 hepatoma cells, and that the decreased ability to form long chain methotrexate polyglutamates in assays with isolated enzyme may be a consequence of the decreased substrate activity of the polyglutamates of this antifolate, rather than solely an impairment of the enzyme catalytic ability. The interaction of methotrexate and PDDF with intact cells and FPGS may take on added importance since it has recently been shown that this combination and others like it can exhibit synergistic cytotoxicity against hepatoma cells in vitro (43). 20. 21. 22. 23. 24. 25. ACKNOWLEDGMENTS The authors gratefully acknowledge Dr. John J. McGuire for help and advice with the FPGS assay and Zenia Nimec for her excellent technical assistance. 27. REFERENCES 28. 1. McGuire, J. J., and Berlino, J. R. Enzymatic synthesis and function of folylpolyglutamates. Mol. Cell. Biochem., 38:19-48, 1981. 2. Cichowicz, D. J., Foo, S. K., and Shane, B. Folylpoly-y-glutamale synthesis by bacteria and mammalian cells. Mol. Cell. Biochem., 39: 209-228, 1981. 3. Kisliuk, R. L. Pteroylpolyglutamates. Mol. Cell. Biochem., 39: 331-346, 1981. 4. McGuire, J. J., and Coward, J. K. Pteroylpolyglutamates: biosynthesis, degradation, and function. In: R. L. Blakley, and S. J. Benovic (eds.), Folates and Pterins, pp. 135-190. New York: John Wiley and Sons, 1984. 5. Shane, B., and Siokstad. E. L. R. The interrelationships among folate, vitamin B12 and methionine metabolism. Ann. Rev. Nutr., 5:115-141,1985. 6. Blakley, R. L., Crane, A., Cocco, L., and Baugh, C. M. Molecular basis for the interaction of folie acid and its analogs with dihydrofolate reducÃ-ase,/n: l. D. Goldman, B. A. Chabner, and J. R. Berlino (eds.), Folyl and Antifolyl Polyglutamates, pp. 1-18. New York: Plenum Publishing, 1983. 7. Mackenzie, R. E., and Baugh, C. M. Interaction of tetrahydropteroylpolyglutamates with two folate-dependent multifunctional enzymes. In: I. D. Goldman, B. A. Chabner, and J. R. Bertino (eds.), Folyl and Anlifolyl Polyglutamates, pp. 19-34. New York: Plenum Publishing, 1983. 8. Galivan, J. Transport and metabolism of methotrexate in normal and resist ant cultured rat hepatoma cells. Cancer Res., 39:735-743, 1979. 9. Galivan, J. Evidence for the cytotoxic activity of polyglutamate derivatives of methotrexate. Mol. Pharmacol., 17: 105-110,1980. 10. Balinska, M., Gali van, J., and Coward, J. K. Efflux of methotrexate and its polyglutamate derivatives from hepatic cells in vitro. Cancer Res., 41: 27512756, 1981. 11. Fry, D. W., Yalowich, J. C., and Goldman, I. D. Rapid formation of poly--yglutamyl derivatives of methotrexate and their association with dihydrofolate reducÃ-aseas assessed by high pressure liquid chromatography in the Ehrlich Asches tumor cell in vitro, i. Biol. Chem., 257:1890-1896, 1982. 12. Jolivet, J., Schilsky, R., Bailey, B., Drake, J. C.. and Chabner, B. A. Synthesis. retention, and biological activity of methotrexate polyglutamates in cultured human breast cancer cells. J. Clin. Invest., 70: 351-360, 1982. 13. Balinska, M., Nimec, Z., and Galivan, J. Characteristics of metholrexate polyglutamate formation in cultured hepatic cells. Arch. Biochem. Biophys., 276:466-476, 1982. 14. Galivan, J., Pupons, A., and Rhee, M. S. Hepatic parenchymal cell glutamylation of methotrexate studied in monolayer culture. Cancer Res., 46: 670-675, 1986. 15. Jacobs, S. A., Adamson, R. II., Chabner, B. A., Derr, C. J., and Johns, D. G. Stoichiometric inhibition of mammalian dihydrofolate reducÃ-aseby the 7-glutamyl metabolite of methotrexate, 4-amino-4-deoxyWV"'-methylpteroylglutamyl-7-glutamate. Biochem. Biophys. Res. Commun., 63: 692-698, 1975. 16. Whitehead, V. M. Synthesis of methotrexate polyglutamates in L1210 mu rine leukemia cells. Cancer Res., 37: 408-412, 1977. 17. McGuire, J. J., Hsieh, P., Coward, J. K., and Bertino, J. R. Enzymalic synthesis of folylpolyglulamales. Characterization of the reaction and its products. J. Biol. Chem., 255: 5776-5788, 1980. 18. Moran, R. G., and Colman, P. D. Mammalian folyl polyglutamate synthetase: partial purification and properties of the mouse liver enzyme. Biochem istry, 23:4580-4589, 1984. 19. Pristupa. Z. B., Vickers, P. J., Sephton, G. B., and Scrimgeour, K. G. 26. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. Folylpolyglutamate synlhelase from beef liver: assay, slabilizalion, and char acterization. Can. J. Biochem. Cell. Biol., 62:495-506, 1984. Cichowicz, D. J., and Shane, B. Mammalian folyl--y-polyglutamate synlhelase. 1. Purificalion and general properlies of the hog liver enzyme. Biochem istry, 26: 504-512, 1987. McGuire, J. J., Hsieh, P., Coward, J. K., and Bertino, J. R. In vitro methol re vate polyglutamate synthesis by rat liver folylpolyglutamate synthetase and inhibition by bromosulfophthalein. In: I. D. Goldman, B. A. Chab ner, and J. R. Bertino (eds.), Folyl and Anlifolyl Polyglulamates, pp. 199214. New York: Plenum Publishing, 1983. Schoo, M. M. J., Pristupa, P. B., Vickers, P. J., and Scrimgeour, K. G. Folate analogues as substrates of mammalian folylpolyglutamate synthetase. Cancer Res., 45:3034-3041, 1985. Scrimgeour, K. G. Biosynthesis of polyglutamates of folates. Biochem. Cell Biol., «.-667-674, 1986. McGuire, J. J., and Coward, J. K. D.L-Threo-4-fluoroglutamic acid. J. Biol. Chem., 260; 6747-6754, 1985. Cook, J. D., Cichowicz, D. J., George, S., Lawler, A., and Shane, B. Mammalian tolvlpoK -, glutamate synthetase. 4. ¡nvitro and in vivo metab olism of folates and analogues and regulation of folate homeostasis. Bio chemistry, 26:530-539, 1987. Nimec, Z., and Galivan, J. Regulatory aspects of the glulamylalion of methotrexate in cultured hepatoma cells. Arch. Biochem. Biophys., 226: 671-680, 1983. Galivan, J. Hormonal alterai ion of methotrexate and folate polyglutamate formation in cultured hepatoma cells. Arch. Biochem. Biophys., 230: 355362, 1984. Jones, T. R., Calvert, A. H., Jackman, A. L., Brown, S. J., and I larrup. K. R. A polenl antitumor quinazoline inhibitor of thymidylate synthetase: synthesis, biological properties, and Iherapeulic results in mice. Eur. J. Cancer, / 7:11 -23, 1981. Jackson, R. C., Jackman, A. L., and Calvert, H. Biochemical effects of a quinazoline inhibitor of Ihymidylale synthetase, A (4 ((2 amino 4 hydroxy6-quinazolinyl)methyl)prop-2-ynyl amino)benzoyl-L-glutamic acid (CB3717) on human lymphoblastoid cells. Biochem. Pharmacol., 32:3783-3790,1983. Cheng, Y-C, Dutschman, G. E., Starnes, M. C., Fisher, M. H., Nanavathi, N. T., and Nair, M. G. Activity of the new antifolate JV'°-propargyI-5,8dideazafolate and its polyglutamates against human dihydrofolate reducÃ-ase, human thymidylate synthetase, and KB cells containing different levels of dihydrofolate reducÃ-ase.Cancer Res., 45: 598-600, 1985. Nair, M. G., Nanavati, N., Nair, I. G., Kisliuk, R. L., Gaumont, Y., Hsiao, M. C., and Kaiman, T. I. Folate analogues 26. Synthesis and anlifolate activity of 10•substitutedderivatives of 5,8-dideazafolic acid and of Ihe poly 7-glutamyl metabolites of ¿V10-propargyl-5,8-dideazafolicacid (PDDF). J. Med. Chem., 29:1754-1760, 1986. Galivan, J. H., Maley, G. F., and Maley, F. Factors affecting substrate binding in Lactobacillus casei thymidylate synthelase as sludied by equilib rium dialysis. Biochemislry, IS: 356-361, 1976. Lowry, O. H., Rosebrough, N. J., Fair, A. L., and Randall, R. J. Protein measurement with the folin phenol reagent. J. Biol. Chem., 193: 265-275, 1951. Samuels, L. L., Goûtas,L. J., Priest, D. G., Piper, J. R., and Sirotnak, F. M. Hydrolytic cleavage of methotrexate -, pol>glutamates by folylpolyglutamyl hydrolase from various tumors and normal lissues of Ihe mouse. Cancer Res., «.-2230-2235, 1986. Kamen, B. A., Takach, P. L., Valev, R., and Caston, J. D. A rapid, radiochemical-ligand binding assay for metholrexate. Anal. Biochem., 70:54-63,1976. Galivan, J., Johnson, T., Rhee, M., McGuire, J. J., Priesl, D., and Kesevan, V. The role of folylpolyglulamale synthetase and -y-glutamyl hydrolase in altering cellular folyl- and antifolylpolyglulamates. Adv. Enzyme. Régulât., 26":147-155, 1987. Taylor, R. T., and Hanna, M. L. Folate-dependent enzymes in cultured Chinese hamster cells: folylpolyglutamale synlhelase and ils absence in mutants auxotrophic for glycine + adenosine + Ihymidine. Arch. Biochem. Biophys., 181: 331-344, 1977. Galivan, J. 5-Melhylletrahydrofolate transport by hepaloma cells and resist ant sublines in culture. Cancer Res., 41: 1757-1762,1981. Moran, R. G., Colman, P. D., Rosowsky, A., Forsch, R. A., and Chan, K. K. Structural features of 4-amino anlifolales required for substrati- aclivily with mammalian folylpolyglutamate synthelase. Mol. Pharmacol., 27: 156-166, 1985. Cole, D. E., Jolivet, J, Allegra, C. J., and Poplack, D. G. Metholrexate (MTX) resistance of non-dividing tumor cells: role of MTX polyglutamates (MTXPG). Proc. Am. Assoc. Cancer Res., 27: 256, 1986. Foo, S. K., and Shane, B. Regulation of folylpoly-f-glutamate synthesis in mammalian cells. J. Biol. Chem., 257: 13587-13592, 1982. Clarke, L., and Waxman, D. J. Human liver folylpolyglulamale synthelase: biochemical characterization and interaction with folates and folate antago nists. Arch. Biochem. Biophys., 256:585-596, 1987. Galivan, J., Nimec, Z., and Rhee, M. Synergislic growth inhibilion of ral hepaloma cells exposed in vitro in A"°-propargyl-5,8-dideazafolate with meth otrexate or the lipophilic antifolates trimelrexale and meloprine. Cancer Res., 47: 5256-5260, 1987. 2431 Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1988 American Association for Cancer Research. Role of Folylpolyglutamate Synthetase in the Regulation of Methotrexate Polyglutamate Formation in H35 Hepatoma Cells Thomas B. Johnson, M. G. Nair and John Galivan Cancer Res 1988;48:2426-2431. Updated version E-mail alerts Reprints and Subscriptions Permissions Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/48/9/2426 Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at [email protected]. To request permission to re-use all or part of this article, contact the AACR Publications Department at [email protected]. Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1988 American Association for Cancer Research.