Identification, cloning, and sequencing of a cDNA coding for rat gamma-glutamyl hydrolase.

Purified gamma-glutamyl hydrolase secreted from rat H35 hepatoma cells has been characterized as a diffuse band of 55 kDa on SDS-polyacrylamide gel electrophoresis that is converted to bands of 35 and 33 kDa after enzymatic removal of N-linked carbohydrate. Polyclonal antibodies against 55-kDa gamma-glutamyl hydrolase captured the enzyme activity and recognized the glycosylated and both deglycosylated forms of gamma-glutamyl hydrolase. A complete cDNA sequence of gamma-glutamyl hydrolase was obtained using degenerate oligonucleotides derived from peptide sequences, screening of a rat hepatoma cDNA library, and reverse transcription polymerase chain reaction. Based upon the deduced amino acid sequence the peptide component of gamma-glutamyl hydrolase had a molecular weight of 33,400. The results of amino acid analysis of the purified protein agreed with the deduced amino acid sequence in which there are seven potential asparagine-containing glycosylation sites.

The properties of the secreted gamma-glutamyl hydrolases from H35 hepatoma cells.

gamma-Glutamyl hydrolase has been partially purified and characterized from conditioned culture medium of H35 hepatoma cells. Evidence for heterogeneity of the enzyme is derived from its elution as three distinct peaks of enzymatic activity when the enzyme is purified by TSK-butyl-Sepharose column chromatography. These three enzyme fractions appear to have identical catalytic properties but, as yet, the basis for their resolution is not understood. A rapid, sensitive and simple assay based on reverse-phase HPLC fluorescent detection with pre-column derivatization using o-phthalaldehyde (OPA) was developed to separate OPA-derivatives of poly-gamma-glutamates and glutamic acid. Using this assay and the standard HPLC assay for pteroylpolyglutamates, the enzyme appears to be an endopeptidase with respect to pteroylpenta-gamma-glutamate (PteGlu5), methotrexate penta-gamma-glutamate (4-NH2-10-CH3PteGlu5) and p-aminobenzoyl-penta-gamma-glutamate (pABAGlu5). The initial products are PteGlu1 (or 4-NH2-10-CH3PteGlu1 or pABAGlu1) and intact tetra-gamma-glutamate, which is subsequently degraded to glutamic acid. When penta-gamma-glutamate is the substrate, the cleavage of the gamma-bonds by the enzyme is less ordered, with the early appearance of mono-, di-, tri- and tetraglutamate. Poly-alpha-glutamate is not a substrate nor are pABA-gamma-Glu5 or penta-gamma-glutamate covalently linked to albumin. 4-NH2-10-CH3PteGlu2 or Glu5 bound to dihydrofolate reductase is not a substrate for the enzyme, offering further evidence that protein-associated poly-gamma-glutamates are poor substrates for gamma-glutamyl hydrolase from H35 hepatoma cells.

The properties and function of gamma-glutamyl hydrolase and poly-gamma-glutamate.

gamma-Glutamyl hydrolase is a ubiquitous enzyme that has the capacity to cleave gamma-glutamyl bonds of cellular folyl- and antifolylpoly-gamma-glutamates. This study has revealed that the enzyme is secreted by primary cultures of rat hepatocytes and by H35 hepatoma cells. It was found that more than 99% of the total enzyme from H35 cells accumulated in the medium after 48 hr incubation with the serum-free medium. The cells were shown to remain intact during the secretion period since lactate dehydrogenase, dihydrofolate reductase and lysosomal hydrolases other than gamma-glutamyl hydrolase were retained within the cell. When PteGlu5 (folylGlu4) is used as a substrate the initial product is PteGlu (folate), and there is no appearance of intermediate chain length pteroyl polyglutamates. Therefore, the secreted and cellular gamma-glutamyl hydrolase from hepatoma cells appears to be an endopeptidase. Polyclonal antibodies to the poly-gamma-glutamate substrates of the enzyme were prepared and characterized. The antibodies recognize the structural differences between alpha- and gamma-glutamyl linkages but appear equally active with PteGlu5 and its analogs such as 4-NH2-10-CH3PteGlu5 and pABAGlu5. The affinity of the antibodies is related to the gamma-glutamyl structure since L-glutamic acid, folate or p-aminobenzoic acid are inactive with the antibodies. Furthermore, poly-gamma-glutamate has lower affinity for the antibodies than the poly-gamma-glutamate derivatives of PteGlu, 4-NH2-10-CH3PteGlu or pABA.

Secretion of gamma-glutamyl hydrolase in vitro.

gamma-Glutamyl hydrolase (also known as conjugase) is a ubiquitous enzyme that has the capacity to cleave folyl- and antifolylpolyglutamates. This study has revealed that the enzyme is secreted by primary cultures of rat hepatocytes and by H35 hepatoma cells. H35 cells have lower cellular levels of gamma-glutamyl hydrolase than do hepatocytes but secrete a greater proportion of gamma-glutamyl hydrolase. More than 99% of the total enzyme from H35 cells accumulated in the medium after 48 h. The cells were shown to remain intact during the secretion period since lactate dehydrogenase, dihydrofolate reductase, and lysosomal hydrolases other than gamma-glutamyl hydrolase were retained within the cell. Using the substrate 4-amino-10-methyl-pteroyldiglutamate (4-NH2-10-CH3-Pte-Glu2), the intracellular and secreted enzyme form(s) from H35 cells were found to have the following properties (a) Km values of 24.3 +/- 3.7 microM and 34.8 +/- 8.6 microM, respectively, and (b) maximal activity at pH 5 to 7 and apparent molecular weights of 120,000 by gel filtration. Both the cellular and secreted enzymes convert 4-NH2-10-CH3-PteGlu4 and pteroylpentaglutamate acid, to the corresponding monoglutamates with little or no appearance of intermediate chain length polyglutamates. This suggests that both act primarily as endopeptidases. Thus far, the cellular and secreted enzymes cannot be differentiated although the current studies do not establish this point unequivocally. Alterations in the cellular and secreted H35 cell gamma-glutamyl hydrolase levels in response to changes in culture conditions revealed that glutamine enhances activity while insulin diminishes it. Other transformed cells found to secrete this protein are Hep-G2 human hepatoma, JAR human choriocarcinoma, HeLa, and rat glioma. gamma-Glutamyl hydrolase could not be detected in medium conditioned by human MCF-7 breast cancer cells, and relatively low activities were found in the medium from CCRF-CEM or K562 leukemia cells. These studies directly establish for the first time the secretion of gamma-glutamyl hydrolase in vitro.

Folate analogues. 34. Synthesis and antitumor activity of non-polyglutamylatable inhibitors of dihydrofolate reductase.

Five analogues of methotrextate (MTX), 10-deazaaminopterin (10-DAM), and 10-ethyl-10-deazaaminopterin (10-EDAM) in which the glutamate moiety was replaced by either a gamma-methyleneglutamate or beta-hydroxyglutamate were synthesized and evaluated for their antifolate activity. These analogous are 4-amino-4-deoxy-N10-methylpteroyl-beta-hydroxyglutamic acid (1), 4-amino-4-deoxy-10-deazapteroyl-beta-hydroxyglutamic acid (2), 4-amino-4-deoxy-N10-methylpteroyl-gamma-methyleneglutamic acid (3, MMTX), 4-amino-4-deoxy-10-deazapteroyl-gamma-methyleneglutamic acid (4, MDAM), and 4-amino-4-deoxy-10-ethyl-10-deazapteroyl-gamma-methyleneglutamic acid (5, MEDAM). None of these compounds were metabolized to the respective polyglutamate derivative as judged by their inability to serve as substrates for CCRF-CEM human leukemia cell folylpolyglutamate synthetase (FPGS) in vitro. All compounds inhibited recombinant human-dihydrofolate reductase (DHFR) at nearly equivalent magnitude as MTX. Growth-inhibition studies with H35 hepatoma, Manca human lymphoma, and CCRF-CEM human leukemia cells established greater cytotoxic effects with compounds 3-5 than with compounds 1 and 2. gamma-Methyleneglutamate derivatives 3-5 were transported to H35 hepatoma cells better than MTX or beta-hydroxyglutamate derivatives 1 and 2. Compound 3 was 2.5 times better than MTX in competing with folinic acid transport in H35 hepatoma cells. Compound 1 did not have a significant inhibitory effect on folinic acid transport even at 50 microM under identical conditions. The IC50 for compound 1 against H35-hepatoma cell growth was 8.5-fold higher than MTX. Compounds with the gamma-methyleneglutamate moiety (3-5) exhibited almost equal or lower IC50 values than MTX against the growth of CCRF-CEM human leukemia cells. These studies show that on continuous exposure, the non-polyglutamylatable inhibitors DHFR (3-5) can exhibit superior antifolate activity compared to the polyglutamylatable methotrexate, presumably due to their enhanced transport to these cell lines. Compounds 3-5 appear to be excellent models to study the role of polyglutamylation of antifolates in antitumor activity and host toxicity.

Synthesis of N-[N-(4-deoxy-4-amino-10-methylpteroyl)-4-fluoroglutamyl]- gamma-glutamate, an unusual substrate for folylpoly-gamma-glutamate synthetase and gamma-glutamyl hydrolase.

N-[N-(4-Deoxy-4-amino-10-methylpteroyl)-4-fluoroglutamyl]-ga mma-glutamate has been synthesized and its ability to serve as a substrate for folylpolyglutamate synthetase and gamma-glutamyl hydrolase has been investigated. It was anticipated that this compound would be a substrate for both of these enzymes. Although the title compound proved to be a good substrate for folylpolyglutamate synthetase, hydrolysis catalyzed by gamma-glutamyl hydrolase was unexpectedly slow. These results suggest the use of fluoroglutamate-containing peptides as hydrolase-resistant folates or antifols in a variety of chemotherapeutic regimens.

The Chromophore and Polypeptide Composition of Aplysia Ink.

The composition of the ink of the sea hare, Aplysia, was studied in regard to its tetrapyrrole and polypeptide content. The ink was separated into three pigment components by both thin-layer and gel filtration chromatography. These three pigments have distinctive visible absorption spectra, and--by comparison with known tetrapyrroles--we have demonstrated that they are derived from the three tetrapyrrole chromophores (bilins) found on the biliproteins of certain red algae, which constitute a portion of the Aplysia diet. The red component is phycourobilin; the purple is phycoerythrobilin; and the blue is phycocyanobilin. Sodium dodecyl sulfate gel electrophoresis experiments were also performed. The results of these experiments showed several polypeptides, and major bands at 78,000 and 61,000 molecular weight were noted. Biliproteins, at most, would be minor components of the ink.

Biochemical and growth inhibitory effects of the erythro and threo isomers of gamma-fluoromethotrexate, a methotrexate analogue defective in polyglutamylation.

We previously reported (J. Galivan et al., Proc. Natl. Acad. Sci. USA, 82: 2598-2602, 1985) the synthesis and characterization of DL-erythro,threo-gamma-fluoromethotrexate (FMTX). The individual diastereomers, DL-erythro-FMTX (eFMTX) and DL-threo-FMTX (tFMTX), and their radiolabeled counterparts have now been prepared and characterized. Transport of eFMTX (Km = 9.3 microM; Vmax = 7.5 pmol/min/10(7) cells) was similar to that of methotrexate (MTX: Km = 6.6-9.9 microM; Vmax = 11.4-14.2 pmol/min/10(7) cells), while tFMTX (Km = 65.1 microM; Vmax = 8.4 pmol/min/10(7) cells) was transported less efficiently. Both isomers were able to saturate intracellular dihydrofolate reductase and accumulate further as unbound intracellular drug. Based on competition experiments and studies with MTX transport-defective cell lines, both isomers utilized the reduced folate/MTX transport system. Efflux half-times for the isomers were similar to those of MTX. Each isomer was equivalent to MTX in its ability to inhibit dihydrofolate reductase activity and bind to intracellular dihydrofolate reductase when the intracellular drug concentration was limiting. Both isomers had drastically diminished capacity to be metabolized to poly(gamma-glutamyl) metabolites by isolated folylpolyglutamate synthetase and in whole cells; tFMTX was metabolized to a slightly lesser extent than eFMTX. Using the CCRF-CEM human leukemia and H35 rat hepatoma cell lines, the growth-inhibitory effects of eFMTX were almost the same as those of MTX during continuous exposure, while tFMTX was slightly less potent. This difference in growth-inhibitory potency of the two isomers correlated with their ability to inhibit de novo thymidylate synthesis in the H35 cell line. These results indicate that both diastereomers of FMTX are similar in their properties to MTX, except that both are incapable of being readily converted to polyglutamate derivatives. As a result of these properties, both isomers could be used under appropriate conditions in comparative studies with MTX to define the roles of MTX polyglutamates.

The characteristics and consequences of folate depletion in hepatoma cells in vitro by inhibition of dihydrofolate reductase.

Growth of rat hepatoma cells in subtoxic concentrations of the DHFR inhibitor metoprine caused a marked time and concentration dependent reduction in cellular folates. As much as 75% total cellular folates can be lost without impairing growth. Increasing the concentration of metoprine into a range that causes inhibition of growth results in no further reduction in cellular folates. This effect is presumably mediated through inhibition of DHFR and several mechanisms are discussed which may account for these results. Cells grown in medium in which the concentration of folate is changed from 4 microM to 20 nM had intracellular folate levels that were reduced 85%. This is nearly the same reduction caused by treating cells grown in normal medium (4 microM folate) with continuous, subtoxic levels of metoprine. The reduction in cellular folates caused by growth in nM folic acid caused enhanced growth inhibitory activity of several antifolates. On a concentration basis metoprine was 12-fold more active under these conditions, PDDF was 37-fold more active and DDATHF was 44-fold more active. The reason for the enhanced sensitivity to PDDF and DDATHF may also be analogous to the reason for their synergism with the low concentration of metoprine and trimetrexate (12).

Antifolate drug interactions: enhancement of growth inhibition due to the antipurine 5,10-dideazatetrahydrofolic acid by the lipophilic dihydrofolate reductase inhibitors metoprine and trimetrexate.

The presence of low concentrations of the lipophilic dihydrofolate reductase inhibitors metoprine or trimetrexate, which cause little inhibition in the growth of cultured hepatoma cells in combination with weakly inhibiting concentrations of 5,10-dideazatetrahydrofolate, exhibit greater activity than would be predicted by the activity of the individual components. Growth inhibition by this inhibitor of glycineaminoribonucleotide transferase alone or in the presence of the reductase inhibitors is prevented by hypoxanthine indicating that the combination of drugs is enhancing the activity of 5,10-dideazatetrahydrofolate against purine biosynthesis. H35 hepatoma cells resistant to methotrexate (100-fold) as a result of a transport defect are 40-fold resistant to 5,10-dideazatetrahydrofolate suggesting that this analogue enters hepatoma cells at least in part by the reduced folate coenzyme-methotrexate transport system. The transport-resistant cells are also susceptible to enhanced inhibition of cell growth by low levels of reductase inhibitors in combination with 5,10-dideazatetrahydrofolate. These results have a corollary in an earlier study showing that the same concentrations of metoprine and trimetrexate could enhance the growth inhibition and cytotoxicity of the folate-based inhibitor of thymidylate synthase, 10-propargyl-5,8-dideazafolic acid (Galivan et al., Cancer Res., 47: 5256-5260, 1987). Combinations of 5,10-dideazatetrahydrofolic acid and 10-propargyl-5,8-dideazafolic acid are less growth inhibitory than that predicted by each of the folate analogues alone. It is possible that the effects of all these combinations are related to distortions in the folate pools caused by the folate analogues being used in combination. Two methods of analysis, one graphical and one mathematical, were used to analyze the drug interactions described in this presentation. The enhancement effect seen with the lipophilic dihydrofolate reductase inhibitors and 5,10-dideazatetrahydrofolate clearly represents a supraadditive or a synergistic drug interaction. In contrast the combination of the folate-based inhibitors of purine (5,10-dideazatetrahydrofolic acid) and thymidylate biosynthesis (N10-propargyl-5,8-dideazafolate) exhibit frank antagonism under certain conditions.

Synergistic growth inhibition of rat hepatoma cells exposed in vitro to N10-propargyl-5,8-dideazafolate with methotrexate or the lipophilic antifolates trimetrexate or metoprine.

The growth inhibitory effects of combinations of antifolates on hepatoma cells in culture have been examined. In these studies methotrexate or the lipophilic inhibitors of dihydrofolate reductase were used with the thymidylate synthase inhibitor N10-propargyl-5,8-dideazafolate (PDDF). Under certain conditions partial growth inhibition by methotrexate and trimetrexate is reduced by noninhibitory to slightly inhibitory concentrations (less than 1 microM) of PDDF. At somewhat higher concentrations (1.6-4 microM) of PDDF, synergy is observed with methotrexate, trimetrexate, or metoprine. Trimetrexate exerted greater synergistic effects than methotrexate. A noninhibitory concentration of trimetrexate (2 nM) in combination with a partially inhibitory concentration of PDDF reduced growth by 93%. Metoprine was capable of replacing trimetrexate and exhibits slightly greater inhibitory activity in combination than trimetrexate. Both metoprine and trimetrexate in combination with PDDF caused synergistic inhibition of the de novo synthesis of thymidylate in intact cells as measured by tritium release from [5-3H]deoxyuridine. Clonal assays were used to demonstrate synergy between trimetrexate or metoprine and PDDF, attesting to the cytotoxic properties of this combination. Thymidine alone can protect against both the synergistic combination of trimetrexate or metoprine and PDDF and PDDF alone, but has only a weak protective effect on toxic concentrations of trimetrexate and metoprine. These observations suggest that growth inhibition is mediated by the activity of N10-propargyl-5,8-dideazafolate on thymidylate synthase. These results are discussed with regard to the mechanism of inhibition of thymidylate synthase by the 5,8-dideazafolates and the possibility of enhancing the inhibitory activity of this class of compounds by using them with inhibitors of dihydrofolate reductase.

gamma-Fluoromethotrexate: synthesis and biological activity of a potent inhibitor of dihydrofolate reductase with greatly diminished ability to form poly-gamma-glutamates.

A methotrexate (MTX) analog containing fluorine at the gamma-carbon of the glutamate moiety, gamma-fluoromethotrexate (FMTX), has been synthesized and evaluated for its biochemical and pharmacological properties. FMTX inhibition of dihydrofolate reductase from several sources is nearly equivalent to that shown by MTX. Most important, FMTX is an exceedingly poor substrate for folylpoly (gamma-glutamate) synthetase, the enzyme that catalyzes the biosynthesis of the highly-retained, cytotoxic MTX polyglutamates. Uptake experiments in H35 hepatoma cells show that FMTX accumulates to approximately the same extent as MTX at steady state. The rapid efflux of both derivatives is also very similar. The major difference detected in cells between the two compounds is the meager glutamylation of FMTX, due to the electronegative properties of the fluorine adjacent to the potential amide-forming carboxyl group. Exposure of dividing cells to 50 microM MTX for 2 and 6 hr results in the formation of 55 and 130 nmol, respectively, of the polyglutamates (more than two glutamate residues)/g of cell protein. With FMTX these values were reduced by 98% and 93%, respectively. Growth inhibition studies show that MTX is only 12-fold more toxic than FMTX when the cells are exposed to each derivative continuously for 72 hr. When the exposure time is reduced, a greater disparity between the inhibitory effects is observed; with a 2-hr pulse, MTX is 2300-fold more effective than FMTX. These data correlate with the effects of pulses of FMTX and MIX on de novo thymidylate biosynthesis in intact cells. The results indicate that of the parameters examined, the vastly reduced toxicity of FMTX after its removal from the culture medium is best correlated with impaired glutamylation. The data strongly suggest that prolonged toxicity of MTX is a result of metabolic conversion to MTX polyglutamates and that these effects are far more dramatic in short-term than in long-term exposure to the antifolates.

Glutamylation of methotrexate in hepatoma cells in vitro: regulation and the development of specific inhibitors.

Methotrexate is glutamylated in cultured hepatoma cells to derivatives that contain a total of 2 to 5 gamma-glutamyl residues. The rate of polyglutamate formation and extent of accumulation are saturable with respect to both medium concentration of methotrexate and time. Maximal rates of glutamylation and accumulation of methotrexate polyglutamates at steady state occur at approximately 10 microM extracellular methotrexate. Inclusion of physiologic concentrations of insulin or removal of folate from the medium each cause a doubling of the rate of glutamylation, and these effects are additive. Insulin and folate restriction also enhance the accumulation of methotrexate polyglutamates. In combination they result in a doubling in the intracellular methotrexate polyglutamate pool at steady state and a shift in the polyglutamate distribution to longer-chain-length species. The importance of the longer-chain-length polyglutamates is apparent from the 6-hr retention of the polyglutamate species: Glu2, 15%; Glu3, 21%; Glu4, 50%; and Glu5, 83%. In probing the glutamylation reaction, a new series of inhibitors have been initiated. These are based upon replacing the incoming glutamate with 4-fluoroglutamate or synthesizing methotrexate with the glutamate replaced by 4-fluoroglutamate. The 4-fluoroglutamyl analogs of methotrexate are effective inhibitors of dihydrofolate reductase but cannot be glutamylated. They will be utilized to probe the role of glutamylation in antifolate activity and folate metabolism.

Regulatory aspects of the glutamylation of methotrexate in cultured hepatoma cells.

The glutamylation of methotrexate has been evaluated in H35 hepatoma cells in vitro as a function of the conditions of culture. Glutamylation yields methotrexate polyglutamate with two to five additional glutamate residues and is a saturable process. The rate of glutamylation increases little above 10 microM extracellular methotrexate which corresponds to an intracellular concentration of approximately 4 microM. The rate of glutamylation measured over a 6-h period was stimulated by a reduction in cellular folates and prior incubation of the cells with insulin. Glutamylation was also more rapid in dividing cultures than in confluent cells. The combination of insulin inclusion and folate reduction, which was additive, caused approximately a fourfold increase in the rate of glutamylation over control cells under the conditions tested. The maximal rate of methotrexate glutamylation, which was 100 nmol/g/h, occurred in folate-depleted, insulin-supplemented cells. Supplementing folate-depleted cells with reduced folate coenzymes caused the glutamylation to be reduced by more than 90%. The turnover of methotrexate polyglutamates in cells saturated with these derivatives occurred at approximately one-half the rate of net synthesis and was stimulated to nearly the same extent by folate depletion and insulin. In addition to showing that folates can modify the rates of methotrexate polyglutamate formation, data are presented suggesting that methotrexate polyglutamates can regulate their own synthesis. The consequences of the formation of these retained forms of methotrexate in H35 hepatoma cells (M. Balinska, J. Galivan, and J.K. Coward (1981) Cancer Res. 41,2751-2756) and the effects of potential regulators of this process are discussed in terms of the glutamylation of folates in the cells and the chemotherapeutic effects of antifolates.

Effects of folinic acid on hepatoma cells containing methotrexate polyglutamates.

The effects of folinic acid on a toxic pulse exposure of cultured hepatoma cells to methotrexate (4-amino-10-methylpteroylglutamic acid) is reported. Inclusion of folinic acid (5-formyl-5,6,7,8-tetrahydropteroylglutamic acid) (10 micro M) with the 2-hr pulse of methotrexate (10 micro M) nearly completely prevents the uptake and gamma-glutamylation of methotrexate and prevents toxicity. Addition of folinic acid after methotrexate results in a partial rescue that is time and concentration dependent. Restoration of cell growth in the presence of increasing amounts of folinic acid is accompanied by a concentration-dependent elevation in tritium release from [5-3H]deoxyuridine. In the absence of folinic acid, the release of tritium from [5-3H]deoxyuridine remains inhibited for three days after exposure to methotrexate, which can be related to the cellular formation and retention of methotrexate polyglutamates. Following the 2-hr pulse of methotrexate, the cellular pool consists of 70% polyglutamates of which the predominant species has three glutamate residues (4-NH2-10-CH3PteGlu3). When methotrexate is removed from medium, following the pulse, unmetabolized methotrexate rapidly leaves the cells, and 4-NH2-10-CH3PteGlu3 is converted to methotrexate polyglutamates containing four to six glutamate residues. Addition of folinic acid after the methotrexate pulse prevents the conversion of 4-NH2-10-CH3PteGlu3 to the higher-chain-length derivatives and causes a reduction in the total methotrexate cell pools over the next 48 hr. These results suggest that the effects of folinic acid on methotrexate polyglutamates may play a role in the rescue of cells containing these derivatives.

Factors controlling the concentrations of methotrexate in cultured hepatic cells.

The polyglutamate metabolites of methotrexate are as inhibitory to the target enzyme dihydrofolate reductase as is methotrexate. Because of their greater retention they have a longer half-life within the cells and thus a greater potential for cytotoxicity. These metabolites have been found in numerous cells and tissues and are extensively synthesized in cultured hepatic cells. Uptake of methotrexate by primary cultures of rat hepatocytes occurs by a pathway which is independent of the folate coenzymes but appears to be related in some way to cholic acid and organic anion uptake. The evidence for the commonality of these pathways is (a) an instability of both uptake systems in the absence of hormones in the culture medium, (b) nearly equal inhibition of uptake by PCMS and NEM, and (c) cross competition of cholic acid and methotrexate for entry into the cells. Cholic acid and BSP can also selectively inhibit methotrexate polyglutamate formation in hepatocytes. Methotrexate entry into H35 hepatoma cells is mediated by the transport system which is shared by folate coenzymes and is not inhibited by cholic acid, BSP or sulfhydryl reagents. At concentrations of cholic acid or BSP which inhibit methotrexate polyglutamate formation in hepatocytes there is little or no loss of polyglutamate formation in H35 cells, possibly because BSP and cholic acid are taken up less by H35 cells than by hepatocytes.