Abstract
Prior studies from this laboratory documented the prevalence of methotrexate (MTX) transport activity with a low pH optimum in human solid tumor cell lines. In HeLa cells, this low pH activity has high affinity for pemetrexed [PMX (Alimta)] and is reduced folate carrier (RFC)-independent because it is not diminished in a RFC-null subline (R5). R5 cells also have residual transport activity, with high specificity for PMX, at neutral pH. In the current study, a R5 subline, R1, was selected under MTX selective pressure at a modest reduction in pH. There was markedly decreased MTX and PMX transport at both pH 5.5 and pH 7.4. When MTX was removed, there was a slow return of transport activity, and when MTX was added back, there was loss of transport at both pH values within 8 weeks. In R1 cells, there was a marked decrease in accumulation of PMX, MTX, and folic acid along with a decrease in growth inhibition by these and other antifolates that require a facilitative process to gain entry into cells. These data demonstrate that (i) RFC-independent transport in HeLa cells at low and neutral pH contributes to antifolate activity (in particular, to PMX activity) and can be diminished by antifolate selective pressure and (ii) the loss of these activities results in marked resistance to PMX, an agent for which there is little or no loss of activity when transport mediated by RFC is abolished. These observations suggest that transport activity in RFC-null HeLa R5 cells at neutral and low pH may reflect the same carrier-mediated process.
INTRODUCTION
Membrane transport is an important determinant of the chemotherapeutic efficacy of antifolates and an essential step in the utilization of natural folates that are one-carbon donors for key biosynthetic processes within cells. There are two major, highly specific routes for the transport of folates and antifolates: (i) The reduced folate carrier (RFC), an anion exchanger and member of the superfamily of facilitative carriers (1 , 2) ; and (ii) two membrane folate receptors that deliver folates into cells by an endocytotic process (1) . RFC has a physiologic pH optimum in leukemia cells (3) ; the optimum pH for folate receptor-mediated transport is lower (4) . Beyond this, it has been well recognized that there are other folate transport activities with low pH optima (4, 5, 6, 7, 8, 9, 10) , but beyond a role in intestinal transport, which has been attributed to RFC (11, 12, 13) , there has been little information regarding their characteristics, underlying mechanisms, and physiologic and pharmacological importance.
The functional importance of low pH transport activities for folates and antifolates may be linked to the low pH noted under some physiologic and pathological conditions. Hence, there is a low pH at the surface of villi in the upper segment of the small intestine (14) , which provides a suitable environment for low pH transporter(s) that play a role in intestinal absorption of folates, a process with a low pH optimum (15) . Likewise, there is a low pH in the interstitium of solid tumors so that low pH transporters could play a role in the delivery of antifolates to tumor cells under these conditions (16, 17, 18, 19) .
Recently, this laboratory demonstrated, in a diverse spectrum of human solid tumor cell lines, the prevalence of transport activity with a low pH optimum that, in most cases, was equal to or greater than RFC-mediated transport at physiologic pH (19) . This low pH transport activity was shown (i) to be entirely independent of RFC in HeLa cells because it was not decreased in a cell line (R5) in which RFC was deleted from the genome and (ii) to have the characteristics of a carrier-mediated mechanism with high affinity for natural folates and some antifolates. The highest affinity (Kt = 45 nmol/L at pH 5.5) was noted for a new generation antifolate, pemetrexed [PMX (Alimta); ref. 20] . There was also substantial residual transport activity for PMX at physiologic pH in RFC-null R5 cells (21) . This neutral pH activity was shown to be saturable with an affinity for PMX (Kt = 12 μmol/L) at pH 7.4 that is lower than the affinity of RFC for this agent but much higher than the affinities for several other antifolates such as ZD1694, ZD9331, and PT523 (21) .
To further explore the pharmacological and physiological importance of the low pH folate transport (LPFT) activity in HeLa cells, studies were undertaken to determine whether further methotrexate (MTX) selective pressure applied to the RFC-deficient R5 cell line under mildly acidic conditions would result in alterations in the low pH transport route and/or the residual transport activity observed at physiologic pH in these cells.
MATERIALS AND METHODS
Chemicals.
[3′,5′,7,9-3H]Folic acid (69 Ci/mmol) and [3′,5′,7-3H]MTX (28 Ci/mmol) were purchased from Moravek Biochemicals, Inc. (Brea, CA). [3H]PMX (50 or 1 Ci/mmol) and unlabeled PMX were provided by the Eli Lilly Company (Indianapolis, IN). ZD1694 and ZD9331 were provided by AstraZeneca, and AG331 was provided by Agouron Pharmaceuticals Inc. (La Jolla, CA). All radiochemicals were purified by high-performance liquid chromatography before use (22) . The stability of radiochemicals was verified by high-performance liquid chromatography on a regular basis, and materials were repurified as necessary.
Cell Culture Conditions.
R5 cells obtained from HeLa cells under MTX selective pressure lack RFC expression due to a gene deletion (19) . These cells were maintained in pH 6.9 RPMI 1640, unless otherwise specified, supplemented with 10% fetal bovine serum (Gemini Bio-Products, Calabasas, CA), 2 mmol/L glutamine, 20 μmol/L 2-mercaptoethanol, penicillin (100 units/mL), and streptomycin (100 μg/mL) at 37°C in a humidified atmosphere of 5% CO2. The pH of this medium was obtained by decreasing the concentration of sodium bicarbonate from 24 to 7.2 mmol/L; osmolality was maintained by increasing the sodium chloride concentration by 17 mmol/L (19) . The pH of the regular RPMI 1640 with the same supplementation was 7.3. Mycoplasma adherent to mesothelioma cells produce a folate transport activity with very high affinity for PMX (23 , 24) . This is also the case for HeLa cells, but to a much lesser extent (24) .1 Therefore, cell cultures were monitored regularly with a Mycoplasma detection kit (American Type Culture Collection, Manassas, VA) and shown to be free of this microorganism.
Section of HeLa Cells Deficient for the Low pH Transport Activity.
A genome-wide insertion mutagenesis technique, gene trapping, was used in an attempt to inactivate low-ph folate transporter activity and thus to identify genes coding for low-ph folate transporter. This method was initially used to inactivate genes in mouse embryonic cells and has also been used to identify radiation-sensitive and cytokine-responsive genes (25, 26, 27) . The gene trapping vector ROSAβGeo (28) , provided by Dr. Philippe Soriano (Fred Hutchinson Cancer Research Center, Seattle, WA), includes a splice acceptor sequence immediately upstream from a promoterless LacZ-neo fusion gene. This splice acceptor is not bypassed. Hence, after insertion of the vector in the genome and subsequent splicing, a gene trapped loses its function but drives expression of both the lacZ and neomycin resistance genes. The gene-trapped clones are selected with G418, and the trapped genes can be identified by rapid amplification of cDNA ends polymerase chain reaction.
R5 cells grown in pH 6.9 RPMI 1640 were transfected with the ROSAβGeo trapping vector using LipofectAMINE Plus Reagent (Invitrogen, Carlsbad, CA). Two days later, cells were exposed for 2 weeks to 500 nmol/L MTX, which inhibits the growth of R5 cells in this medium, and 600 μg/mL G418. Four clones survived under these conditions, but only one (R1) exhibited a decrease in MTX influx at pH 5.5. As illustrated in Fig. 1⇓ , the clones were maintained in pH 6.9 RPMI 1640 for 2 months in the presence or absence of 500 nmol/L MTX, 600 μg/mL G418, or both. R1 cells grown with MTX had low LPFT activity, whereas R1 cells grown without MTX regained LPFT activity, regardless of the presence or absence of G418. Based on this observation, R1 cells were maintained in pH 6.9 medium with 500 nmol/L MTX (R1+MTX cells) or without MTX (R1-MTX cells). R1+ MTX cells grew at a rate ∼10% slower than R1-MTX cells.
Schema of the method of generation of R1+MTX and R1-MTX cells. R5 cells were transfected with RosaβGeo trapping vector, and selection with 500 nmol/L MTX and 600 μg/mL G418 was started 2 days later. After 2 weeks, the surviving clone, R1, which exhibited decreased LPFT activity, was grown in the presence or absence 500 nmol/L MTX, 600 μg/mL G418, or both for 2 months. R1-MTX and R1+MTX along with R5 cells were used in the current study.
Growth Inhibition by Antifolates.
Cells grown in pH 6.9 RPMI 1640 were transferred to 96-well plates (1,000 cells per well for R5 and R1-MTX cells; 2,000 cells per well for R1+MTX cells) in either pH 6.9 or pH 7.3 RPMI 1640 and exposed continuously to a spectrum of antifolate concentrations for 6 days. MTX (500 nmol/L), which was used to maintain the stability of R1+MTX cells, was absent in growth inhibition assays. Cell growth rates were quantified by sulforhodamine B staining (29) .
Folic Acid, MTX, or PMX Accumulation.
For measurement of folic acid accumulation, cells grown in pH 6.9 RPMI 1640 were transferred to either pH 6.9 or pH 7.3 folic acid-free RPMI 1640 containing 2 μmol/L [3H]folic acid and grown for 7 to 9 days. Cells were reseeded once to keep them subconfluent and to minimize acidification of the medium. For measurement of MTX and PMX accumulation, cells grown in pH 6.9 RPMI 1640 were plated and grown for 3 days under subconfluent conditions in pH 6.9 or pH 7.3 RPMI 1640 containing 50 nmol/L [3H]MTX or [3H]PMX in addition to 200 μmol/L glycine, 100 μmol/L adenine, and 10 μmol/L thymidine. Intracellular tritium was determined as in the transport studies described below.
Transport Studies.
Analysis of transport followed a protocol designed for rapid uptake determinations on cells growing in monolayer cultures (30) . Cells (4 × 105 R5 and R1-MTX cells and 5 × 105 R1+MTX cells) were seeded in 20-mL Low Background glass vials (Research Products International Corp., Prospect, IL) and grown for 3 days in drug-free medium to reach early confluence. The cells were washed twice with HBS [20 mmol/L HEPES, 140 mmol/L NaCl, 5 mmol/L KCl, 2 mmol/L MgCl2, and 5 mmol/L glucose (pH 7.4)] or MBS [20 mmol/L 4-morpholinepropanesulfonic acid, 140 mmol/L NaCl, 5 mmol/L KCl, 2 mmol/L MgCl2, and 5 mmol/L glucose (pH 5.5)] and incubated in this buffer at 37°C for 20 minutes. After removal of the incubation buffer, uptake was initiated by the addition of 0.5 mL of HBS or MBS buffer at 37°C containing radiolabeled folates at the desired concentrations. Uptake was terminated by injection of 5 mL of ice-cold HBS into the vials, after which the adherent cells were washed three times with 5 mL of ice-cold acid buffer [10 mmol/L NaAc and 150 mmol/L NaCl (pH 3.5)]; at this pH, any folate bound to folate receptor at the cell surface is extracted by the wash procedure. The cells were then dissolved by incubation in 0.2 mol/L NaOH (0.5 mL) at 65°C for 45 minutes. Radioactivity in 0.4 mL of the lysate was determined, and 10 μL of lysate were processed for protein determination (BCA; Pierce, Rockford, IL). Cellular uptake is expressed in pmol/mg protein.
For trans-stimulation studies, cells were incubated in MBS buffer (pH 5.5) containing 100 μmol/L 5-formyltetrahydrofolate (5-CHO-THF) for 20 minutes, after which 5 mL of ice-cold HBS or MBS buffer were added, and the cells were washed three times with the same buffer. Influx of PMX was initiated by adding 0.5 mL of prewarmed (37°C) HBS containing 2 μmol/L [3H]PMX and 20 μmol/L unlabeled folic acid or MBS containing 0.1 μmol/L [3H]PMX, and the cells were processed as described above. Control cells underwent the same procedure, but in the absence of 5-CHO-THF.
RESULTS
Selection of RFC-Null R5 Cells for Loss of LPFT Activity.
Studies were undertaken to determine whether MTX selective pressure at a reduced pH might alter the expression of LPFT activity using an approach that might identify the genes coding for LPFT activity. To accomplish this, R5 cells were first transfected with a gene trapping vector, followed by continuous exposure to 500 nmol/L MTX and G418 at a pH of 6.9. Four MTX resistance clones were obtained, but only one (R1) exhibited a marked decrease in LPFT activity. R1 cells grown in the presence of 500 nmol/L MTX for 2 months maintained suppression of the LPFT activity whether or not G418 was present, and LPFT activity returned to R5 levels in cells grown in the absence of MTX whether or not G418 was present (Fig. 1)⇓ . Hence, the suppression of LPFT activity appeared to be largely reversible but did not appear to be related to inactivation of the LPFT gene by the trapping process.
To facilitate description of the various experimental conditions, R1 cells that were grown continuously in medium containing MTX, in which LPFT activity was suppressed, are indicated as R1+MTX. R1 cells grown in MTX-free medium in which LPFT activity was present are indicated as R1-MTX (Fig. 1)⇓ .
Transport of MTX, PMX, and Folic Acid at pH 5.5 in LPFT-Deficient R1 Cells.
Influx of [3H]MTX (0.5 μmol/L), [3H]PMX (0.1 μmol/L), and [3H]folic acid (0.5 μmol/L) was assessed in R1+MTX, R1-MTX, and R5 cells at pH 5.5. As shown in Fig. 2⇓ , influx of all three folates at pH 5.5 was markedly suppressed (5–8-fold) in the R1+MTX cells as compared with R5 and R1-MTX cells. Influx of folic acid in the R1-MTX cells was equivalent to that of R5 cells. Influx of PMX and MTX in R1-MTX cells was increased to levels only slightly less than that of R5 cells, consistent with reversion of the R1 transport phenotype to that of R5 cells in the absence of MTX. Of note is that whereas the concentration of PMX was one-fifth that of MTX or folic acid, the rates of transport in R5 and R1-MTX cells were similar, consistent with a higher affinity of the LPFT process for PMX than the other folates (20) .
Influx of folates into R5, R1-MTX, and R1+MTX cells at pH 5.5. R5 (▪), R1-MTX, (▴), and R1+MTX (•) cells were maintained in pH 6.9 medium, washed, and incubated with MBS for 20 minutes at 37°C. Influx was initiated by addition of 0.5 mL of prewarmed MBS containing 0.5 μmol/L [3H]MTX (top panel), 0.1 μmol/L [3H]PMX (middle panel), or 0.5 μmol/L [3H]folic acid (bottom panel). Data are the mean ± SE from four independent experiments for MTX and PMX and three experiments for folic acid.
Transport of PMX and MTX in LPFT-Deficient Cells at pH 7.4.
As reported previously, there is a secondary transport pathway for antifolates at pH 7.4 in RFC-null R5 cells with higher affinity for PMX than MTX (21) . Studies were conducted to determine whether this pathway is altered in cells in which the LPFT activity is diminished. As indicated in the top panel of Fig. 3⇓ , PMX influx was markedly decreased in R1+MTX cells at pH 7.4, but the rate was similar to that of the R5 line when cells were grown in the absence of MTX (R1-MTX). This reduction in influx was comparable with the decrease observed at pH 5.5 (Fig. 2)⇓ . MTX influx was also deceased, but to a lesser extent (by a factor of ∼3), in R1+MTX as compared with R5 cells (Fig. 3⇓ , bottom panel). Influx of MTX in R1-MTX cells was only slightly less than that in R5 cells.
Influx of PMX and MTX in R5 (▪), R1-MTX (▴), and R1+MTX (•) cells at pH 7.4. The extracellular concentration of [3H]PMX (top panel) and [3H]MTX (bottom panel) was 2 μmol/L. Folic acid (20 μmol/L) was present during PMX influx determinations to decrease the ordinate intercept, which represents surface binding, without alteration of initial uptake rates in R5 cells at pH 7.4 (21) . Folic acid (50 nmol/L) was present during MTX influx determinations to block any possible contribution by folate receptors. Data are the mean ± SE from three independent experiments.
Accumulation of Folic Acid and Antifolates under Growth Conditions.
The impact of the loss of LPFT on accumulation of these folates was assessed under growth conditions in pH 6.9 and pH 7.3 RPMI 1640 containing 50 nmol/L [3H]MTX, 50 nmol/L [3H]PMX, or 2 μmol/L [3H]folic acid (Table 1)⇓ . Under these conditions, essentially all radiolabel represents polyglutamate derivatives (21) . MTX and folic acid accumulation was higher at pH 6.9 than at pH 7.3, consistent with the transport component mediated by LPFT. In contrast, PMX accumulation was somewhat less at pH 6.9 than at pH 7.3. At both pH values, PMX accumulation was 3- to 6-fold greater than MTX accumulation, consistent with more rapid polyglutamation of the former drug. At pH 6.9 (Table 1)⇓ , there was marked (∼84–89%) suppression of the accumulation of all folates in R1+MTX cells that lack LPFT activity. There was a full return to R5 levels in R1-MTX cells. The pattern was similar when cells were grown in pH 7.3 RPMI 1640, except for slightly lower accumulation levels in R1-MTX cells than in R5 cells. The levels of folic acid accumulation in R1+MTX cells, although markedly decreased, were apparently sufficient to support cell growth at near usual rates (see Materials and Methods).
Accumulation of cell tritium after growth in the presence of 50 nmol/L [3H]MTX, 50 nmol/L [3H]PMX, or 2 μmol/L [3H]folic acid in R5 cells, R1-MTX cells with LPFT activity, and R1+MTX cells with markedly decreased LPFT activity
Growth Inhibition by Antifolates.
The impact of dihydrofolate reductase (DHFR) and thymidylate synthase (TS) inhibitors was assessed in pH 6.9 and pH 7.3 RPMI 1640 in the various cell lines. Trimetrexate (TMQ), a DHFR inhibitor, enters cells by passive diffusion and does not form polyglutamate derivatives. AG331 has similar properties but is an inhibitor of TS. ZD9331 is a TS inhibitor transported into cells by RFC but does not form polyglutamate derivatives (2) . The data are summarized in Table 2⇓ . In pH 6.9 medium (Table 2)⇓ , there were negligible differences between the IC50 in R1-MTX and R5 cells with all antifolates tested; in each case, sensitivity was largely restored when R1 cells were maintained in the absence of MTX. R1+MTX cells were ∼3-fold resistant to MTX but ∼ 10-fold collaterally sensitive to TMQ, as compared with R1-MTX and R5. The latter is consistent with the marked contraction of folate cofactor pools in these cells that results in enhanced suppression by TMQ of DHFR (Table 2)⇓ . Under all conditions, cells were equally sensitive to AG331, indicative of a lack of change at the level of TS and a lack of effect of cell folate pools on the inhibitory effect of this drug at this enzyme site. R1+MTX cells were 6-fold resistant to ZD9331 and 4-fold resistant to PMX, consistent with the loss of transport. On the other hand, the decrease in activity of ZD1694 was smaller, with a ∼2-fold increase in IC50. The growth inhibition by these antifolates at pH 7.3 was similar, except resistance to PMX was increased, and resistance to ZD9331 was decreased in R1+MTX cells (Table 2)⇓ . Hence the decrease in influx activity and drug accumulation at both pH values resulted in increased resistance to all of the antifolates that require facilitative transport to gain entry into cells. However, the magnitude of the change varied among the antifolates.
Inhibition of growth of R5, R1-MTX, and R1+MTX cells by antifolates
Time Course of Reversibility of LPFT Suppression and the Relationship between LPFT Activity and RFC-Independent Transport Pathway at Physiologic pH.
Studies were designed to assess the time course of (i) resumption of LPFT activity when MTX was removed from the growth medium of R1+MTX cells and (ii) loss of LPFT activity when 500 nmol/L MTX was added to the medium of R1-MTX cells. As indicted in the top panel of Fig. 4⇓ , resumption of LPFT activity, as assessed by measurement of [3H]folic acid influx at pH 5.5, was gradual after removal of MTX. By the 8th week, LPFT activity had reached ∼ 65% the level of R1-MTX cells. With time (several months), this ultimately reached the level of R1-MTX cells (data not shown). On the other hand, on addition of MTX, LPFT activity in R1-MTX cells decreased more rapidly and was at the level of R1+MTX cells within 6 weeks.
The time course of changes in folic acid and PMX transport activity in R1 cells with the addition or withdrawal of MTX. Top panel, analyses at pH 5.5. [3H]Folic acid influx (0.5 μmol/L) in R1+MTX cells maintained continuously with MTX (▪) is compared with cells in which MTX was removed at week 0 (○). R1-MTX cells (□) are compared with cells to which MTX was added at week 0 (•). Bottom panel, analysis at pH 7.4. [3H]PMX influx (2 μmol/L) in the presence of 20 μmol/L folic acid is compared in R1-MTX cells (▪), R1+MTX cells in which MTX was removed 8 weeks earlier (▴), and R1-MTX cells to which MTX was added 8 weeks earlier (♦). Data are the mean ± SE from three independent experiments.
There was a similar response to the presence or absence of MTX in R1 cells at neutral pH, as indicated in the bottom panel of Fig. 4⇓ . In R1-MTX cells, PMX influx at pH 7.4 was markedly decreased within 8 weeks after addition of MTX. In R1+MTX cells, PMX influx increased to ∼60% that of R1-MTX cells within 8 weeks after MTX was removed, similar to what was observed at pH 5.5 (see above). Hence, the increase or decrease in LPFT activity at pH 5.5 correlated with the increase or decrease in PMX influx activity at pH 7.4.
Trans-Stimulation of PMX Influx at Both Low and Physiologic pH.
One important characteristic of carrier-mediated processes is trans-stimulation: augmentation of the unidirectional flux of a substrate into cells by the opposite flux of the same or another substrate out of cells via the same carrier (31 , 32) . This has been demonstrated at physiologic pH for RFC-mediated transport (33) . Trans-stimulation of MTX influx was also observed in R5 cells at pH 5.5 (20) . As expected, PMX influx at pH 5.5 was doubled in R5 cells loaded with 5-CHO-THF (Fig. 5⇓ , top panel). PMX influx at pH 7.4 was increased to the same extent as at pH 5.5 when cells were preloaded with 5-CHO-THF (Fig. 5⇓ , bottom panel). This is strong evidence that the low and neutral pH transport activities in R5 cells are carrier-mediated.
Trans-stimulation of PMX influx in cells loaded with 5-CHO-THF. R5 cells were incubated with 100 μmol/L 5-CHO-THF (▪) or buffer (□) for 20 minutes at 37°C and pH 5.5, after which the cells were washed with ice-cold MBS or HBS. Influx was initiated by adding prewarmed MBS containing 0.1 μmol/L [3H]PMX (top panel) or HBS containing 2 μmol/L [3H]PMX and 20 μmol/L unlabeled folic acid (bottom panel). Data are the mean ± SE from three independent experiments.
DISCUSSION
Recently, this laboratory demonstrated the presence of very prominent LPFT activity in more than 30 human solid tumor cell lines (19) . In HeLa cells, this was shown to be RFC- independent, to contribute to the delivery of MTX at pH levels relevant to the interstitium of human solid tumors, and to be unrelated to the expression of folate receptors (19) . Additional studies indicated that this LPFT activity is mediated by a Na+-independent facilitative carrier with very high affinity for PMX (Kt = 45 nmol/L) and lesser but still substantial (≤1 μmol/L) affinity for 5-CHO-THF, 5-methyltetrahydrofolate, folic acid, and MTX at pH 5.5 (20) . PT523 and PT632, better substrates for RFC than MTX, are very poor substrates for the LPFT activity at low pH (20) .
The present study demonstrates for the first time that the LPFT activity in RFC-null R5 cells can be markedly diminished in response to MTX selective pressure and that this is accompanied by a decrease in transport at neutral pH. The data suggest that the low and neutral pH activities may be related. First, in R1 cells that have lost LPFT activity, there was a comparable loss of PMX transport at physiologic pH. Furthermore, the resumption of LPFT activity in R1 cells grown in the absence of MTX was accompanied by a similar change in PMX transport at neutral pH. The relationship between LPFT activity and residual transport activity at pH 7.4 was also supported by their preferences for different folates. For example, PMX is the preferred substrate at both pH values, whereas MTX, folic acid and ZD1694 have comparable Ki values that are ∼8× greater than for PMX at pH 7.4 (21) and at pH 5.5 (20) . Finally, PMX influx was trans-stimulated to the same extent at both pH 5.5 and pH 7.4 when cells were loaded with 5-CHO-THF, consistent with a facilitative carrier-mediated process under both conditions. However, it is also possible that these activities at pH 5.5 and pH 7.4 are not molecularly related and that changes in MTX selective pressure may produce changes in two distinct processes that diminish drug transport, one with a neutral pH optimum and the other with a low pH optimum.
The folate transport activities at acidic and neutral pH, irrespective of their origins, may be of pharmacological importance; this may be especially significant for PMX. PMX, in combination of cisplatin, was recently approved for the treatment of pleural mesothelioma (34) . In another phase III trial, PMX efficacy was comparable with that of docetaxel as a second-line therapy of non–small-cell lung cancer but with significantly fewer side effects (35) . This agent has activity in a variety of other solid tumors, as well (36) . PMX activity is not decreased by the loss of RFC in HeLa cells grown with 5-CHO-THF (21) . In fact, whereas deletion of RFC resulted in high-level resistance to ZD1694 (40-fold) and PT523 (250-fold) in the R5 cell line, there was 2-fold collateral sensitivity to PMX (21) . This must be related, in part, to the preservation of RFC-independent transport at low and physiologic pH along with the contraction of cell folate cofactor pools (see below). The finding in the current study that loss of transport activity at pH 5.5 and pH 7.4 was associated with high-level resistance to PMX is direct evidence for the important role this transport plays as a determinant of PMX activity. Because transport activity at low pH was shown to be equal to or greater than transport at pH 7.4 in the majority of human solid tumor cell lines (19) , RFC-independent PMX transport may be an important route for the delivery of this drug within the acidic environment of solid tumors (19) . These observations raise the possibility that loss of RFC activity, alone, is unlikely to be an important basis for acquired resistance to PMX in vitro or in the clinical setting.
The activity of a variety of antifolates (in particular, PMX) is highly dependent on the level of intracellular folate cofactors (2) . An increase in cellular folates, achieved by increasing folate in the growth medium, decreases accumulation of antifolate polyglutamates and diminishes growth inhibition (37) . This was also observed in a murine leukemia cell line selected for resistance to 5,10-dideazatetrahydrofolate, in which the affinity of RFC for folic acid was increased, and cellular folate pools were expanded (38) . Likewise, expansion of folate pools due to loss of MRP1 exporter expression was associated with resistance to a spectrum of antifolates (39, 40, 41) . On the other hand, a decrease in cellular folate cofactors achieved by decreasing folate in the growth medium increases antifolate activity (37) . There is generally a good correlation between retention of PMX activity under these conditions and collateral sensitivity to TMQ. This is due to decreased competition between the drug and dihydrofolate at the level of DHFR because the lower levels of reduced folates are interconverted and oxidized when this enzyme is suppressed (42) . For example, RFC-deficient cells are fully sensitive to TMQ but resistant to PMX in growth medium containing folic acid. However, these cells are collaterally sensitive to TMQ and much less resistant or not resistant at all to PMX when grown in medium containing 5-CHO-THF (21 , 43) . In the latter case, cellular folate pools are contracted due to diminished transport of 5-CHO-THF, which, unlike folic acid, depends primarily on this carrier for transport into cells. For agents such as PMX that require polyglutamation for activity, decreased constitutive feedback inhibition of antifolate polyglutamation by the lower level of intracellular folates compensates in part for the transport defect (44) . In the current study, at both pH 6.9 and pH 7.3, LPFT-deficient cells were ∼7-fold collaterally sensitive to TMQ; however, these cells were 6- to 7-fold resistant to PMX due to the loss of its alternative transport pathway(s). Hence, PMX influx in LPFT-deficient cells was so low that the marked accompanying contraction of cellular folates (see Tables 1⇓ and 2⇓ ) could no longer compensate for the transport defect.
These studies shed some light on the basis for the difference in activity between ZD1694 and PMX. Both agents have comparable affinity for folylpoly-γ-glutamate synthetase (45) , and their polyglutamate derivatives have comparable inhibitor constants at the level of TS (46 , 47) . Yet, despite a secondary inhibitory effect of PMX at the level of glycinamide ribonuleotide formyltransferase (47 , 48) , the IC50 for ZD1694 (1.5 nmol/L) is one-fifteenth that of PMX (22 nmol/L) in HeLa cells grown in folic acid medium (21) . This pattern is reversed when RFC activity is lost in R5 cells. Under these conditions, the ZD1694 IC50 increased to a level ∼4-fold higher than the PMX IC50 (Table 2⇓ ; pH 7.3 medium), suggesting a much greater dependence of ZD1694 on RFC activity. However, when the RFC-independent transport activity was lost in the R1+MTX cells, the IC50 values of PMX and ZD1694 were increased by 7- and 2-fold, respectively, as compared with R5 cells, suggesting a higher dependence of PMX activity on the RFC-independent transport pathway. Furthermore, under the latter conditions, the IC50 for PMX and ZD1694 was the same; hence, membrane transport is a major determinant of the differences in activities of these drugs.
Although R5 cells were transfected with a gene trapping agent, the LPFT gene did not appear to be trapped in the R1 cell line, based on the observation that LPFT activity appeared to be related to MTX selective pressure, but not to G418 selective pressure. It is of interest that resumption of LPFT activity in the absence of MTX or marked diminution of LPFT activity with the addition of MTX required a long interval. This suggests that MTX does not play a direct role in regulation of the LPFT activity but rather results in the selection of a population of transport-deficient cells. Hence, LPFT-deficient cells regain activity in the absence of MTX by a random, slow process, and the resulting clones with LPFT activity overgrow LPFT-deficient cells because they have a growth advantage likely related to their much higher cellular folate levels. On re-exposure to MTX, cells with LPFT activity are killed, and the LPFT-deficient clone reemerges. Studies are currently underway to identify the basis for the LPFT activity at the molecular level.
Footnotes
-
Grant support: National Institutes of Health grant CA-82621 and a grant from the Eli Lilly Company, which also provided tritiated PMX and other nonlabeled PMX reagents for these studies.
-
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.
-
Requests for reprints: I. David Goldman, Departments of Medicine and Molecular Pharmacology and the Albert Einstein Cancer Research Center, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, NY 10803.
-
↵1 Unpublished observations.
- Received May 11, 2004.
- Revision received August 31, 2004.
- Accepted September 8, 2004.
References
Volume 10, Issue 24
