Loss of Folic Acid Exporter Function with Markedly Augmented Folate Accumulation in Lipophilic Antifolate-resistant Mammalian Cells*

(Received for publication, April 29, 1997)

Yehuda G. Assaraf Dagger and I. David Goldman §

From the § Departments of Medicine and Molecular Pharmacology, the Albert Einstein College of Medicine Cancer Center, Bronx, New York 10461 and Dagger  Department of Biology, The Technion, Israel Institute of Technology, Haifa 32000, Israel

ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENT
REFERENCES


ABSTRACT

We previously described a 1,000-fold pyrimethamine-resistant Chinese hamster ovary cell line (PyrR100) which retains parental dihydrofolate reductase activity and methotrexate (MTX) sensitivity. This study characterizes the basis for the 14-fold decrease in folic acid and leucovorin concentrations required for clonogenic growth of PyrR100 cells relative to parental AA8 cells. Under conditions in which folic acid reduction was blocked by trimetrexate, PyrR100 cells displayed relative to parental AA8 cells a: 1) 17- and 5-fold increase in the net transport of folic acid and MTX, respectively; 2) 23- and 5-fold decrease in the efflux rate constant for folic acid and MTX, respectively; and 3) 2-fold increase in folic acid influx with no significant change in MTX influx. The markedly increased net folic acid transport in PyrR100 cells could not be explained by cellular folic acid binding, mitochondrial sequestration, polyglutamylation, nor by a decreased membrane potential.

The effect of energy deprivation on folic acid and MTX transport in both cell lines was quite different. Glucose and pyruvate deprivation nearly abolished the increase in net folic acid transport in PyrR100 cells. In contrast, energy deprivation increased net MTX transport in AA8 cells, whereas no change was seen with PyrR100 cells. Furthermore, while folic acid influx in PyrR100 and AA8 cells was markedly reduced with energy deprivation, MTX influx was not affected. Provision of glucose and pyruvate to energy-deprived cells resulted in a rapid onset of MTX efflux from parental AA8 cells but not from PyrR100 cells. Taken together these results indicate that the markedly enhanced net transport of folic acid and MTX in PyrR100 cells is largely due to the complete loss of exit pump activity. Furthermore, the energy source that sustains the augmented levels of folic acid appears linked to the influx process and is distinct from the energy source that sustains MTX gradients under these conditions. We conclude that the loss of folic acid efflux is an efficient means of augmenting cellular uptake of folate cofactors and subsequent survival on picomolar folate concentrations. This constitutes the first demonstration of the loss of folic acid exporter activity in mammalian cells as a response to lipophilic antifolate selective pressure.


INTRODUCTION

Folic acid cofactors play a key role in one-carbon metabolism and are essential for the biosynthesis of purine and pyrimidine precursors of nucleic acids, for the metabolism of certain amino acids, as well as for initiation of macromolecular synthesis in mitochondria (1, 2). However, mammalian cells cannot synthesize folates and therefore must rely on their uptake from exogenous sources. Several plasma membrane routes have been described in mammalian cells that can accommodate transport of folates and their 4-amino analogs including methotrexate (MTX).1 1) The reduced folate carrier (RFC), a major folate uptake route which is a bidirectional anion exchanger (3, 4) with a high affinity (Km = 0.3-5 µM) for reduced folates and MTX, and low affinity (Km = 200-400 µM) for folic acid (3-5). 2) Folate receptors, glycosylphosphatidylinositol membrane-anchored proteins that mediate the unidirectional uptake of folates into mammalian cells with a high affinity for folic acid and 5-methyltetrahydrofolate (KD = 1-10 nM) but lower affinity (KD = 10-300 nM) for other reduced folates and MTX (6-9). 3) An apparently separate transport system with optimal activity at low pH which recognizes folic acid, reduced folates, and MTX with comparable affinities (1-5 µM) (10-12).

However, although folates can be taken up efficiently by mammalian cells via these systems (and possibly by additional yet unidentified transport routes), they can be rapidly lost by efflux through RFC as well as ATP-dependent efflux pumps (13) such as the multispecific organic anion transporters (MOATs) (14-17). Thus, intracellular levels of free monoglutamate reflect the net contributions of these multiple bidirectional processes. Furthermore, another element of cellular folate retention is the extent of folylpolyglutamylation (18, 19), as well as compartmentation in organelles including mitochondria (20).

In this study we characterized the mechanism(s) underlying the extraordinary decrease in the folate growth requirement of Chinese hamster ovary (CHO) PyrR100 cells that display high level resistance to lipophilic antifolates including 2,4-diaminopyrimidines (e.g. pyrimethamine and metoprine), trimetrexate (TMQ), and piritrexim (21). We find that these cells have markedly augmented folic acid accumulation and provide evidence that this is due predominantly to the loss of folic acid exporter activity. This constitutes the first demonstration of the loss of folic acid efflux activity in mammalian cells as a mechanism of adaptation to dihydrofolate reductase (DHFR) inhibition with a lipophilic antifolate.


EXPERIMENTAL PROCEDURES

Chemicals

[3',5',7',9-3H]folic acid, [carboxyl-14C]inulin, and [3H]tetraphenylphosphonium (TPP+) were obtained from Amersham Corp., and [3',5',7'-3H]MTX was purchased from Moravek Biochemicals (Brea, CA). Radiolabeled and unlabeled folic acid and MTX were purified prior to use by high performance liquid chromatography (19). Trimetrexate was a generous gift from Dr. D. Fry (Warner-Lambert, Parke-Davis, Ann Arbor, MI). Folic acid, leucovorin (calcium salt), MTX, and pyrimethamine were obtained from Sigma.

Cell Lines and Tissue Cultures

CHO PyrR100 cells were established by stepwise selection of parental AA8 cells in gradually increasing concentrations of pyrimethamine (21). The multiple step selection initiated at 100 nM pyrimethamine (the LD50) was terminated at 100 µM, and the 1000-fold pyrimethamine-resistant cells were therefore termed PyrR100. Parental AA8 cells and their PyrR100 subline were maintained in monolayer or suspension culture conditions in RPMI 1640 containing 2.3 µM folic acid (Life Technologies, Inc.), supplemented with 5% dialyzed fetal bovine serum (Gemini Bio-Laboratories Inc.), 1 mM sodium pyruvic acid (Mediatech), 2 mM glutamine, 100 units/ml penicillin, and 100 µg/ml streptomycin. Unless otherwise stated, the growth medium of PyrR100 cells was supplemented with 100 µM pyrimethamine. Monolayer cells were passaged biweekly by a standard trypsinization protocol.

Folate Cofactor Growth Requirement

Monolayers of AA8 and PyrR100 cells grown for at least five doublings in pyrimethamine-free medium were washed with phosphate-buffered saline and detached by trypsinization. Cells were then seeded at 500-1,000 cells/60-mm Petri dish in 5 ml of folic acid-free RPMI 1640 medium containing a range of folic acid or leucovorin concentrations. After 7-12 days of incubation at 37 °C when colonies (>50 cells/colony) became visible, cells were washed with phosphate-buffered saline, fixed with methanol, stained with crystal violet, and counted. The control clonogenic growth was determined in medium containing 2.3 µM folic acid or 100 nM leucovorin.

Folic Acid and MTX Transport

Influx measurements were performed according to a previously described method (22) with some modifications. Exponentially growing cells from suspension cultures in pyrimethamine-free growth medium were collected by centrifugation (750 × g for 2 min), and washed three times with ice-cold HBS (20 mM HEPES pH 7.4, 140 mM NaCl, 5 mM KCl, 2 mM MgCl2, and 5 mM glucose). Cells were resuspended to a density of 2-6 × 107 cells/ml in 1 mM pyruvate-containing HBS with 5 µM TMQ (only for folic acid transport measurements) to ensure complete blockade of folic acid reduction by DHFR (23). Following 20 min of incubation at 37 °C, uptake was initiated by addition of [3H]folic acid or [3H]MTX and 0.5-1-ml portions of the cell suspension were drawn at given times, and transport was terminated by injection into 10 ml of ice-cold HBS. After centrifugation (750 × g for 2 min) at 4 °C, cells were resuspended in ice-cold HBS, washed twice in the same buffer, and processed for determination of intracellular radiolabel with normalization to dry weight (22). Intracellular water of AA8 and PyrR100 cells was determined using the membrane-impermeable [carboxyl-14C]inulin as the extracellular marker as described elsewhere (24). Efflux measurements were performed by first loading cells with [3H]folic acid (in the presence of TMQ) or [3H]MTX for 30 min at 37 °C. Then, a portion was removed for determination of the intracellular level prior to efflux. The remaining radiolabeled cells were sedimented by centrifugation and resuspended in a large volume of prewarmed HBS containing 1 mM pyruvate and 5 µM TMQ. Portions of cell suspension were drawn at given times and processed as detailed above for cellular radiolabel. The DHFR binding capacity was the level of [3H]MTX remaining in cells after efflux for 30-60 min, an incubation time sufficient to allow for the loss of all free drug.

Measurement of Plasma Membrane Potential

We have employed the method of equilibrium distribution of the lipid-soluble cation TPP+ across the plasma membrane to determine the membrane potential of AA8 and PyrR100 cells (25). The uptake of TPP+ was measured at 2 µM and reached a steady-state after 10 min of incubation at 37 °C. Total cellular TPP+ was corrected for a bound component that was retained in cells after efflux into a buffer free of TPP+. The Nernst equation was employed to calculate the membrane potential using the cell volume determination (see Table II) and intracellular/extracellular TPP+ concentrations.

Table II. Comparison of folic acid transport parameters in AA8 and PyrR100 cells

Data presented are the means ± S.E. of 4-13 independent experiments.
AA8       PyrR100 PyrR100/AA8

Initial uptake rate (nmol/min/g dry wt) 0.058  ± 0.006 0.108  ± 0.014 1.9
Efflux rate constant (min-1) 0.084  ± 0.012 0.0036  ± 0.0011 0.043
Total cell folic acid (nmol/g dry wt) 0.51  ± 0.03 7.52  ± 1.17 14.8
H2Oi/dry wt (µl/mg) 4.98 4.43 0.89
Observed [folic acid]i (µM) 0.102 1.70 16.7
Predicted [folic acid]i (µM)a 0.07 0.07 1
Observed/Predicted 1.46 24.3 16.6
Membrane potential (mV)  -35  -35 1.0

a Based upon the Nernst equation using the measured membrane potential at an extracellular folic acid level of 1 µM.

Analysis of the Polyglutamate Derivatives of Folic Acid and MTX

[3H]folic acid, [3H]MTX, and their polyglutamate derivatives were analyzed by HPLC (19) using two continuous linear gradients of 0-10% and 10-15% acetonitrile over 35 and 15 min, respectively. Unlabeleld MTX-Glu(1-6) standards (Schircks Laboratories, Switzerland) served as internal markers and were monitored by UV absorbance at 280 nm. Typically, MTX-Glu(6-3) eluted during the 0-10% acetonitrile gradient whereas MTX-Glu(2-1) resolved during the 10-15% acetonitrile elution.

Analysis of Intracellular Folic Acid Compartmentation and Binding

Exponentially growing AA8 and PyrR100 cells (2.5-5 × 108 cells) in pyrimethamine-free suspension cultures were harvested by centrifugation and washed three times with ice-cold HBS. Cells were resuspended in prewarmed HBS containing 1 mM pyruvate and 5 µM TMQ, incubated for 20 min at 37 °C, and then loaded for 1 h with 1 µM [3H]folic acid in the same buffer. A portion of radiolabeled cells was removed for determination of cell-associated folic acid. The remaining cells were washed three times with ice-cold HBS, resuspended in 10 volumes of ice-cold hypotonic buffer A (10 mM Tris-HCl, pH 7.8, 5 mM MgCl2, 2 mM CaCl2), and allowed to swell for 20-30 min at 0 °C (26). Following light microscopy verification of >95% cell swelling, beta -mercaptoethanol was added to a final concentration of 1%, and cells were disrupted with 50 strokes in a tight-fitting Dounce homogenizer (Wheaton, pestle A); >99% cell lysis was achieved in both parental AA8 and PyrR100 cells. Sucrose (Ultrapure, Pharmacia Biotech Inc.) was added to a final concentration of 0.25 M, and the homogenate was layered over 5 ml of 25% sucrose and centrifuged at 600 × g for 10 min at 4 °C to sediment intact cells, nuclei, and cell debris. The supernatant was then collected and centrifuged at 10,000 × g for 10 min at 4 °C to separate mitochondrial and cytosolic fractions. Alternatively, to isolate mitochondria under osmoprotective conditions to minimize potential [3H]folic acid leakage, cells loaded with [3H]folic acid were suspended in 10 volumes of isotonic buffer B (10 mM Tris, pH 7.8, 233 mM sucrose, 5 mM MgCl2, 2 mM CaCl2). Under these conditions 230-300 Dounce homogenizer strokes, and 50-85 strokes were necessary to disrupt PyrR100 and AA8 cells, respectively.


RESULTS

Parental AA8 cells and their PyrR100 subline with high level resistance to lipophilic antifolates were examined for their clonogenic growth requirements for folic acid and leucovorin. Whereas parental AA8 cells maintained a 50% clonogenic viability at 70 nM folic acid and 0.8 nM leucovorin, PyrR100 cells required 14-fold less folate cofactors and displayed a 50% clonogenic viability at leucovorin concentrations as low as 60 pM (Table I). Consequently, PyrR100 cells could grow solely on the residual (~75 pM) folates present in a folic acid-free medium supplemented with 5% dialyzed fetal bovine serum (data not shown). Thus, PyrR100 cells possess a markedly decreased folate growth requirement and picomolar concentrations of leucovorin are sufficient to support clonogenic growth.

Table I. Folic acid and leucovorin clonogenic growth requirement for parental AA8 and PyrR100 cells

Data presented for the 50% clonogenic viability (CV50) values are the means of two experiments.
CV50
AA8 PyrR100 AA8/PyrR100

Folic acid 70 nM 5 nM 14
Leucovorin 800 pM 60 pM 13

To explore the role of membrane transport in the markedly decreased folate growth dependence of PyrR100 cells, [3H]folic acid transport was measured in the presence of 5 µM TMQ to abolish DHFR activity and subsequent folic acid reduction (23). Net uptake of folic acid into AA8 and PyrR100 cells reached steady-state at 30 min (Fig. 1, inset) and 60 min (Fig. 1) with folic acid levels at 0.51 ± 0.03 and 7.52 ± 1.17 nmol per g of dry weight, respectively (Table II). Based on the measured intracellular water, AA8 and PyrR100 cells achieved free intracellular folic acid concentrations of 0.1 and 1.7 µM, respectively (Table II). Therefore, PyrR100 cells achieved a 17-fold higher intracellular folic acid concentration than parental AA8 cells.


Fig. 1. Time course of folic acid uptake in PyrR100 and AA8 cells. Exponentially growing suspension cultures of PyrR100 (squares) and AA8 cells (triangles) were harvested by centrifugation, washed three times with ice-cold HBS, and suspended to 5 × 107 cells/ml in 5 µM TMQ-containing HBS supplemented with 1 mM pyruvic acid. After 20 min of incubation at 37 °C, transport was initiated by adding [3H]folic acid to a final concentration of 1 µM. Portions (0.5-1 ml) of the cell suspension were removed at designated times into ice-cold HBS following which cells were processed for determination of intracellular radiolabel as described under "Experimental Procedures." Inset, expanded time course of [3H]folic acid accumulation in AA8 cells.
[View Larger Version of this Image (18K GIF file)]

Net uptake of [3H]MTX (Fig. 2) was also higher in PyrR100 than in AA8 cells (7.63 ± 0.64 versus 3.40 ± 0.28 nmol MTX/g of dry weight, respectively, Table III). To assess free MTX, cells loaded with [3H]MTX were resuspended in MTX-free buffer allowing for discrimination between the free and bound components. While the tightly bound MTX fraction was only slightly increased in PyrR100 cells, the free intracellular MTX concentration of 1.1 µM was increased 4.8-fold in this cell line as compared with 0.23 µM in AA8 cells (Table III). Hence, net accumulation of free folic acid in PyrR100 cells was much greater than for MTX.


Fig. 2. Time course of MTX accumulation in PyrR100 and AA8 cells. Following 20 min incubation of PyrR100 (solid squares) and AA8 (solid triangles) cell suspensions (3 × 107/ml) in HBS containing 1 mM pyruvate, transport was initiated by addition of [3H]MTX to a final concentration of 1 µM. Following 30 min of loading with 1 µM MTX, the cells were resuspended into a large volume of drug-free buffer for determination of the loss of free drug (indicated by the double-headed arrows) and the tightly bound component (open symbols).
[View Larger Version of this Image (13K GIF file)]

Table III. Comparison of MTX transport parameters between AA8 and PyrR100 cells

Data presented are the means ± SEM of 4-10 independent experiments.
AA8 PyrR100 PyrR100/AA8

Initial uptake rate (nmol/min/g dry wt) 0.366  ± 0.036 0.421  ± 0.030 1.15a
Efflux rate constant (min-1) 0.22  ± 0.06 0.041  ± 0.012 0.186
Total cell MTX (nmol/g dry wt) 3.396  ± 0.276 7.63  ± 0.64 2.25
Bound MTX (nmol/g dry wt) 2.25  ± 0.17 2.70  ± 0.24 1.20
Free MTX (nmol/g dry wt) 1.14 4.93 4.30
Observed [MTX]i 0.23 1.11 4.83
Predicted [MTX]i (µM)b 0.07 0.07 1
Observed/Predicted 3.29 15.9 4.8

a The difference is not statistically significant (p = 0.1805).
b Based upon the Nernst equation using the measured membrane potential of -35 mV and an extracellular MTX level of 1 µM.

To assess the role that changes in influx and efflux parameters contribute to the alterations in steady-state (anti)folate levels, bidirectional fluxes were measured. While folic acid influx was increased 2-fold in PyrR100 cells (Table II), no significant difference in the influx of MTX was observed between PyrR100 and AA8 cells (Table III). Analysis of efflux revealed that parental AA8 cells had a very rapid loss of folic acid (t1/2 = 4 min), whereas the efflux t1/2 in PyrR100 cells was 80 min (Fig. 3A). While folic acid efflux in PyrR100 cells was not complete even after 3 h of incubation at 37 °C, the process could be described by a single exponential in both cell lines (Fig. 3B). The folic acid efflux rate constant (k') was decreased by a factor of 23 in PyrR100 relative to AA8 cells (k' = 0.0036 min-1 and 0.084 min-1, respectively, Table II). MTX efflux was also decreased in PyrR100 cells (Fig. 4A), albeit to a lesser extent than observed for folic acid with a 5-fold fall in the MTX efflux rate constant in PyrR100 relative to AA8 cells (0.041 min-1 and 0.22 min-1), respectively (Fig. 4B and Table III). Hence, the large increase in the steady-state folic acid levels in PyrR100 cells is associated with a two fold increase in influx but more than an order of magnitude greater decline in efflux. The smaller increase in MTX uptake in PyrR100 cells is due entirely to a decline in efflux and to a much smaller extent than observed for folic acid in PyrR100 cells. A variety of studies were undertaken to confirm that the markedly diminished efflux of folic acid and MTX in PyrR100 cells was due entirely to a change in membrane transport and not due to: 1) increased (anti)folylpolyglutamylation, 2) increased compartmentation of folate in mitochondria, 3) sequestration by a highly overexpressed (anti)folate-binding protein, or 4) decreased plasma membrane potential. After loading cells with 1 µM [3H]MTX for 1 h, cell extracts were analyzed by HPLC for MTX and its polyglutamate derivatives. PyrR100 and AA8 cells both contained~ 85% of the radiolabel as the MTX monoglutamate (data not shown). The remaining 15% of radiolabel was associated with decreasing amounts of diglutamate, triglutamate, and tetraglutamate in both cell lines. Thus, the vast majority of radiolabeled MTX in both PyrR100 and AA8 cells remains in the monoglutamate form after 1 h incubation with [3H]MTX. HPLC analysis of cell extracts following 1 h incubation with 1 µM [3H]folic acid in the presence of 5 µM TMQ revealed that greater than 95% of the radiolabel in both PyrR100 and AA8 cells represented folic acid monoglutamate. Hence, there was no metabolism of folic acid under these conditions.


Fig. 3. Time course of folic acid efflux in PyrR100 and AA8 cells. Following 30 min of loading of PyrR100 (squares) and AA8 cells (triangles) with 1 µM [3H]folic acid in 10 µM TMQ-containing HBS at 37 °C, cells were rapidly sedimented by centrifugation and resuspended in TMQ-containing folic acid-free HBS at 37 °C, and efflux was followed for up to 3 h (A). To determine the folic acid efflux rate constants, the log value of the percentage of initial free folic acid level at each time point was plotted as a function of time (B).
[View Larger Version of this Image (15K GIF file)]


Fig. 4. Time course of MTX efflux in PyrR100 and AA8 cells. After 30 min loading of PyrR100 (squares) and AA8 cells (triangles) with [3H]MTX at 1 µM and 5 µM (to compensate for the small exchangeable pool of MTX in AA8 cells, see Table III), respectively, cells were rapidly sedimented by centrifugation and suspended in drug-free HBS (A). To determine the MTX efflux rate constants, the log value of the percentage of the initial free MTX (corrected for tightly bound MTX, see Table III) was plotted as a function of time (B).
[View Larger Version of this Image (15K GIF file)]

To examine the levels of compartmentation of folic acid in mitochondria, cells were loaded with 1 µM [3H]folic acid in the presence of 5 µM TMQ for 1 h, following which mitochondria were isolated in isotonic or hypotonic buffer. Under these conditions, mitochondria-associated folic acid in both PyrR100 and AA8 cells did not exceed 0.5% of cellular radiolabel. To explore the possibility of intracellular folic acid sequestration by a potentially overexpressed folate binding protein, cells were loaded with 1 µM [3H]folic acid for 1 h after which G-25 gel filtration of total cellular lysate was performed. All the radiolabel (>99.9%) was retained on the column indicating that all the [3H]folic acid was free and not protein-associated.

The possibility of a decline in the membrane potential in PyrR100 cells that would result in an increase in the the levels of folates was explored by the equilibrium distribution of the lipid-soluble cation TPP+ across the plasma membrane. Based on the Nernst equation the mean membrane potential value for both PyrR100 and AA8 cells was calculated to be -35 mV (Table II). Hence, a decline in the membrane potential did not play a role in the marked accumulation of folic acid in PyrR100 cells.

Since the data indicated that changes in net folic acid and MTX transport were associated with a large increase in the transmembrane electrochemical potential difference for these folates in PyrR100 cells, studies were focused on the bioenergetic characteristics of these processes in the two cell lines. It has been well established that at least one component of MTX efflux in various mammalian cells is mediated by energy requiring exit pump(s) (14-17), while uphill transport into cells appears to be mediated by RFC, through an anion exchange mechanism with intracellular organic phosphates (3, 4). Hence studies were undertaken to assess the energy requirement for (anti)folate transport in PyrR100 and AA8 cells. Energy depletion achieved by incubating cells for 30 min in glucose- and pyruvate-deficient transport buffer (27) decreased the augmented net folic acid transport in PyrR100 cells by 85%, while net folic acid uptake in AA8 cells at steady-state was essentially unchanged (Fig. 5). In contrast, glucose and pyruvate deprivation increased net MTX transport in AA8 (Fig. 6A) but not in PyrR100 cells (Fig. 6B). Provision of glucose and pyruvate after energy deprivation resulted in an immediate fall in the free MTX in AA8 (Fig. 6A, arrow) but no change in PyrR100 cells (Fig. 6B, arrow). Folic acid influx in PyrR100 and AA8 cells was decreased by 80 and 60%, respectively, in the absence of glucose and pyruvate and under these conditions was comparable in the two cell lines (Fig. 7). Energy deprivation did not alter MTX influx in either AA8 or PyrR100 cells (based upon four experiments).


Fig. 5. Effect of energy deprivation on folic acid transport in PyrR100 and AA8 cells. Exponentially growing cells in medium containing 5 mM glucose and 1 mM pyruvate were divided into two portions one of which was washed three times in ice-cold HBS containing 5 mM glucose and 1 mM pyruvate (energy replete), and the other was washed with ice-cold HBS lacking glucose and pyruvate. PyrR100 (solid squares) and AA8 cells (solid triangles) from the energy replete group, as well as PyrR100 (open squares) and AA8 cells (open triangles) from the energy deplete group were resuspended in their respective buffers with the addition of 5 µM TMQ. After incubation for 30 min at 37 °C (to deplete intracellular ATP), uptake was initiated with the addition of 1 µM [3H]folic acid.
[View Larger Version of this Image (14K GIF file)]


Fig. 6. Effect of energy deprivation on MTX transport in PyrR100 and AA8 cells. The experimental details were those described in Fig. 5 legend except for the omission of TMQ here. Transport into AA8 cells under energy replete (solid triangles) or deplete conditions (open triangles), as well as into PyrR100 cells under energy replete (solid squares) or deplete conditions (open squares) was initiated by adding 1 µM [3H]MTX. Following energy depletion, 5 mM glucose and 1 mM pyruvate were added simultaneously (see arrows) to energy deplete AA8 (solid triangles) and PyrR100 (solid squares) cells and the effects on MTX levels assessed.
[View Larger Version of this Image (16K GIF file)]


Fig. 7. Effect of energy depletion on the initial rates of folic acid uptake in PyrR100 and AA8 cells. The experimental conditions and symbols are those described in the legend to Fig. 5.
[View Larger Version of this Image (15K GIF file)]


DISCUSSION

This study demonstrates that the underlying basis for the markedly decreased folate growth requirement of PyrR100 cells is due to a major decline in folic acid efflux and a small increase in folic acid influx. This resulted in a markedly augmented folic acid accumulation. Net uptake of MTX was also increased in PyrR100 cells but to a lesser extent; this was due entirely to a fall in efflux. Studies excluded the possibility that augmented folic acid uptake was related to: 1) increased expression of a cellular folate-binding protein, 2) increased sequestration of folates in mitochondria, 3) increased polyglutamylation, or 4) a fall in the membrane potential of PyrR100 cells. This constitutes the first demonstration of the loss of folic acid exporter activity in mammalian cells selected for lipophilic antifolate resistance.

The marked decrease in folic acid and MTX export activities in PyrR100 cells resulted in a marked increase in the transmembrane gradients for these folates. Hence, the chemical gradients for folic acid and MTX increased by factors of 17 and 5 in PyrR100 as compared with AA8 cells. When the membrane potential, measured at -35 mV in both cell lines is considered, the expected ratio of intracellular to extracellular concentrations of both folates should be 0.07 when the extracellular level is 1 µM, if transport were passive or equilibrating, i.e. no energy consumed in the transport process. In fact, the intracellular level of folic acid is only 50% greater, and MTX is three times higher than this predicted value in AA8 cells, representing only a small electrochemical potential difference across the cell membrane. On the other hand, net transport is markedly increased in PyrR100 cells, the folic acid level exceeds the predicted equilibrium value by a factor of 24, and MTX exceeds this value by a factor of 16. The maintenance of these high electrochemical potential differences across the cell membrane must require a substantial energy source.

Uphill transport of MTX into mammalian cells (most thoroughly characterized in mouse L1210 leukemia cells) is dependent upon an RFC-mediated anion exchange with intracellular organic anions. In this model, the downhill flow of these organic anions out of the cell via the carrier results in the uphill flow of MTX into the cell by the same mechanism. This process is opposed by independent MTX exit pump(s) that are tightly coupled to ATP hydrolysis (14-16). When the exit pump is inhibited under conditions of energy deprivation, influx of MTX is unchanged or only slightly increased, efflux is markedly decreased, and the net uptake of MTX is consequently rapidly and substantially increased.

This was the behavior observed for MTX in AA8 cells. There was essentially no change in influx with deprivation of energy substrates, and net uptake was increased, consistent with a reduction in efflux due to cessation of efflux pump activity. When energy substrates were added back to the cells, there was an immediate and rapid net loss of MTX from the cells consistent with the energization of the efflux pump, the pattern that has been described for L1210 mouse leukemia cells (13, 14, 28-30). The pattern in PyrR100 cells was fundamentally different; while net MTX uptake was markedly greater than in AA8 cells this process was not affected by the absence of energy substrates, consistent with a loss of exporter function but no change in energy-coupling to the RFC system.

The energetics of folic acid uptake in AA8 cells was different from that observed with MTX. Influx was decreased, but net transport essentially unchanged with energy deprivation. The affinity of folic acid for RFC is one or two orders of magnitude lower than that for MTX (3-5). However, the export pump(s) have a potent effect on folic acid efflux. Hence the exit pump dominates, and only very low steady-state levels of folic acid are achieved in AA8 cells. In PyrR100 cells the lack of energy substrates markedly decreased both influx and net transport of folic acid. Thus, folic acid influx in these cells appears to be linked to metabolic energy in a process that accounts for sustained uphill transport under conditions in which the exporter is not functional. This appears to be a mechanism that is specific for folic acid with little or no impact on the transport of MTX.

Of note is that the MTX efflux rate constant (k' = 0.22 min-1) obtained for CHO AA8 cells was identical to that observed by the Sirotnak group (28, 29) with mouse leukemia L1210 cells (k' = 0.21 min-1). Moreover, the latter group reported that under energy deprivation this MTX efflux rate constant was decreased by 83% (k' = 0.036 min-1); this value is again almost identical to the MTX efflux rate constant obtained with PyrR100 cells (k' = 0.041 min-1). In contrast, the folic acid efflux rate constant in parental AA8 cells was 0.084 min-1 but was profoundly decreased in PyrR100 cells (k' = 0.0036 min-1). Taken together these data suggest that, as with mouse L1210 cells (30), ~80% of the MTX efflux appears to be mediated via a route distinct from RFC, such as the MOATs (14-16), most likely through MOAT 3 (16), whereas the residual MTX efflux ostensibly occurs via RFC (30). In contrast, since RFC has a very much lower affinity for folic acid than MTX, the vast majority of the efflux of this folate in CHO AA8 cells presumably exits via a MOAT. Hence, when transport by the MOAT is abolished, as presumably occurs in PyrR100, efflux is markedly reduced, and high levels of folic acid accumulate.

Recent studies have shown that some MOATs particularly MOAT 3 and to some degree MOAT 4 can mediate an ATP-mediated efflux of variety of organic anions including MTX (16). A canalicular rat liver organic anion transporter cDNA (termed cMOAT) has been recently cloned which mediates the ATP-driven hepatocellular excretion of numerous toxic organic anions (31). This cMOAT which contains the ATP-binding domains (Walker A and B) was shown to be a member of the well established ABC superfamily of transporters (32). It is possible that like MTX (14, 15), the efflux of folic acid is also mediated by a related ATP-driven exit pump.

PyrR100 cells exhibit a genetically stable resistance to lipid-soluble antifolates that is fully retained even after long term growth (1,300 cell doublings) in pyrimethamine-free medium (21, 27). Furthermore, the markedly augmented folic acid accumulation as well as the loss of folate efflux activity are stable in PyrR100 cells grown in the continuous presence of pyrimethamine and in its absence even after 1,300 cell doublings. These characteristics strongly suggest that the putative genetic alteration responsible for the loss of folic acid efflux activity in PyrR100 cells is stable. It is of interest that the in vivo loss of organic anion transporter activity in canalicular liver cells was associated with a single nucleotide deletion in the cMOAT gene leading to congenital jaundice in these rats (31).

PyrR100 cells display over 1,000-fold resistance to pyrimethamine and a 30-fold cross-resistance to trimetrexate and piritrexim, all of which are lipid-soluble inhibitors of mammalian DHFR (21, 23). The ability of PyrR100 cells to accumulate and retain high levels of folic acid is a very useful adaptation to lipophilic antifolates in particular to 2,4-diaminopyrimidines including pyrimethamine which have relatively low affinity for DHFR (21). Augmented net transport and free intracellular folic acid levels results in increased formation of reduced folates and retention of their polyglutamyl derivatives.2 This, in turn, results in the generation of high dihydrofolate levels as antifolate associates with DHFR, reduced folates are oxidized, and dihydrofolate then effectively competes with the antifolate for the small percentage of enzyme binding sites (<5%) that are sufficient to sustain tetrahydrofolate cofactor pools (33).

The potential loss of folic acid and MTX exporter activity in human tumors due to treatment with, for example, lipophilic antifolates may have potentially important implications for selective chemotherapy. The transporter that mediates the efflux of MTX and folic acid has a broad specificity for various organic anions some of which are cytotoxic including cholate and taurocholate (14-16). Thus, PyrR100 cells that have lost folic acid efflux activity are likely to accumulate high levels of such toxic organic anions and possess a prominent hypersensitivity to these compounds. Hence, it may be possible to exploit this property to selectively kill these tumor cells with low concentrations of cytotoxic agents that are converted intracellularly to amphiphilic anion conjugates. Thus, the loss of folic acid exporter activity may provide a novel approach for selective elimination of drug-resistant tumor cells with this phenotype.


FOOTNOTES

*   This work was supported by National Cancer Institute Grant CA-39807.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
   To whom correspondence should be addressed: Dept. of Medicine and Molecular Pharmacology, Albert Einstein Cancer Center, 1300 Morris Park Ave., Bronx, NY 10461. Tel.: 718-430-2302; Fax: 718-430-8550.
1   The abbreviations used are: MTX, methotrexate; RFC, reduced folate carrier; DHFR, dihydrofolate reductase; TMQ, trimetrexate; MOAT, multispecific organic anion transporter; CHO, Chinese hamster ovary; TPP+, tetraphenylphosphonium; HBS, Hepes-buffered saline; HPLC, high performance liquid chromatography.
2   Y. G. Assaraf, H. M. Barr, and G. Jansen, unpublished data.

ACKNOWLEDGEMENT

We thank Pi-Jun Wang for expert technical assistance.


REFERENCES

  1. Blakely, R. D., and Benkovics, S. J. (1985) Folates and Pterins: Chemistry and Biochemistry of Pterins, Vol. 2, John Wiley & Sons, New York
  2. Blakely, R. D., and Whitehead, V. M. (1986) Folates and Pterins: Nutritional, Pharmacological, and Physiological Aspects, Vol. 3, John Wiley & Sons, New York
  3. Goldman, I. D. (1971) Ann. N. Y. Acad. Sci. 186, 400-422 [Medline] [Order article via Infotrieve]
  4. Henderson, G. B., and Zevely, E. M. (1981) Biochem. Biophys. Res. Commun. 99, 163-169 [Medline] [Order article via Infotrieve]
  5. Sirotnak, F. M. (1987) NCI Monogr. 5, 27-35 [Medline] [Order article via Infotrieve]
  6. Henderson, G. B. (1990) Annu. Rev. Nutr. 10, 319-335 [CrossRef][Medline] [Order article via Infotrieve]
  7. Wang, X., Shen, F., Freisheim, J. H., Gentry, L. E., and Ratnam, M. (1992) Biochem. Pharmacol. 44, 1898-1901 [Medline] [Order article via Infotrieve]
  8. Antony, A. C. (1992) Blood 79, 2807-2820 [Medline] [Order article via Infotrieve]
  9. Brigle, K. E., Spinella, M. J., Westin, E. H., and Goldman, I. D. (1994) Biochem. Pharmacol. 47, 337-345 [CrossRef][Medline] [Order article via Infotrieve]
  10. Henderson, G. B., and Strauss, B. P. (1990) Cancer Res. 50, 1709-1714 [Abstract]
  11. Sierra, E. E., Brigle, K. E., Spinella, M. J., and Goldman, I. D. (1997) Biochem. Pharmacol. 53, 223-231 [CrossRef][Medline] [Order article via Infotrieve]
  12. Kumar, C. K., Moyer, M. P., Dudeja, P. K., and Said, H. M. (1997) J. Biol. Chem. 272, 6226-6231 [Abstract/Free Full Text]
  13. Goldman, I. D. (1969) J. Biol. Chem. 244, 3779-3785 [Medline] [Order article via Infotrieve]
  14. Saxena, M., and Henderson, G. B. (1995) J. Biol. Chem. 270, 5312-5319 [Abstract/Free Full Text]
  15. Saxena, M., and Henderson, G. B. (1996) Biochem. Pharmacol. 51, 975-982 [CrossRef]
  16. Saxena, M., and Henderson, G. B. (1996) Biochem. J. 320, 273-281 [Medline] [Order article via Infotrieve]
  17. Saito, H., Masuda, S., and Inui, K. (1996) J. Biol. Chem. 271, 20719-20725 [Abstract/Free Full Text]
  18. Rosenblatt, D. S., Whitehead, V. M., Vera, N., Pottier, A., Dupont, M., and Vuchich, M. J. (1978) Mol. Pharmacol. 14, 1143-1147 [Abstract]
  19. Fry, D. W., Yallowich, J. C., and Goldman, I. D. (1982) J. Biol. Chem. 257, 1890-1896 [Free Full Text]
  20. Horne, D. W., Holloway, R. S., and Said, H. M. (1992) J. Nutr. 122, 2204-2209 [Medline] [Order article via Infotrieve]
  21. Assaraf, Y. G., and Slotky, J. I. (1993) J. Biol. Chem. 268, 4556-4566 [Abstract/Free Full Text]
  22. Brigle, K. E., Spinella, M. J., Sierra, E. E., and Goldman, I. D. (1995) J. Biol. Chem. 270, 22974-22979 [Abstract/Free Full Text]
  23. Assaraf, Y. G., Molina, A., and Schimke, R. T. (1989) J. Biol. Chem 264, 18326-18334 [Abstract/Free Full Text]
  24. Bobzien, W. F., III, and Goldman, I. D. (1972) J. Clin. Invest. 51, 1688-1696 [Medline] [Order article via Infotrieve]
  25. Lichtenstein, D., Dunlop, K., Kaback, H. R., and Blume, A. J. (1979) Proc. Natl. Acad. Sci. U. S. A. 76, 2580-2584 [Abstract]
  26. Trent, D. F., Seither, R. L., and Goldman, I. D. (1991) Biochem. Pharmacol. 42, 1015-1019 [Medline] [Order article via Infotrieve]
  27. Sprecher, H., Barr, H. M., Slotky, J. I., Tzukerman, M., Eytan, G. D., and Assaraf, Y. G. (1995) J. Biol. Chem. 270, 20668-20676 [Abstract/Free Full Text]
  28. Dembo, M., Sirotnak, F. M., and Moccio, D. M. (1984) J. Membr. Biol. 78, 9-17 [Medline] [Order article via Infotrieve]
  29. Sirotnak, F. M., Moccio, D. M., and Yang, C.-H. (1984) J. Biol. Chem. 259, 13139-13144 [Abstract/Free Full Text]
  30. Henderson, G. B., and Zevely, E. M. (1984) J. Biol. Chem. 259, 1526-1531 [Abstract/Free Full Text]
  31. Paulusma, C. C., Bosma, P. J., Zaman, G. J. R., Bakker, C. T. M., Otter, M., Scheffer, G. L., Schepper, R. J., Borst, P., and Oude Elferink, R. P. J. (1996) Science 271, 1126-1128 [Abstract]
  32. Higgins, C. F. (1992) Annu. Rev. Cell Biol. 8, 67-72 [CrossRef]
  33. Seither, R. L., Trent, D. F., Mikulecky, D. C., Rape, T. J., and Goldman, I. D. (1989) J. Biol. Chem. 264, 17016-17023 [Abstract/Free Full Text]

©1997 by The American Society for Biochemistry and Molecular Biology, Inc.