ARCHIVES

OF BIOCHEMISTRY

AND

BIOPHYSICS

Vol. 298, No. 1, October, pp. 121-128, 1992

Transport of Methotrexate Vesicles from Rat Liver’ Donald

W. Horne*‘tz2

and Kathleen

in Basolateral Membrane A. Reed?

*Biochemistry Research Laboratory (151), Department of Veterans Affairs Medical Center, Nashville, Tennessee 37212-2637, and tDepartment of Biochemistry, Vanderbilt University, School of Medicine, Nashville, Tennessee 37232

Received April 27, 1992, and in revised form May 22, 1992

Transport of the antifolate cancer drug methotrexate was studied in vesicles isolated from the basolateral membrane of rat liver. Transport of methotrexate by basolateral membrane vesicles (BLMVs) was mostly via uptake into an osmotically active intravesicular space, with some binding (approximately 9%), as shown by initial uptake studies and by varying medium osmolarity with increasing concentrations of sucrose. Methotrexate transport was linear for the first 20 s of incubation. Transport was not affected by imposition of a Naf gradient across the vesicular membrane. Transport of methotrexate displayed a broad pH optimum: at an intravesicular pH of 7.5, the initial rate of uptake was not significantly different at extravesicular pH values ranging from 5.5 to 7.5, but uptake was less at extravesicular pH of 5.0 or 8.0. Methotrexate transport was saturable: K, = 0.15 -+ 0.05 pun and V,,, = 11.4 f 1.1 pmol 10 s-l mg-’ protein. Methotrexate uptake into BLMVs was not inhibited by 5-methyltetrahydrofolate nor by B-formyltetrahydrofolate but was weakly inhibited by folic acid in a concentration-dependent manner. Uptake was also inhibited by the anion-exchange inhibitor 4,4’-diisothiocyanostilbene-2,2’-disulfonic acid (DIDS), and by the structurally unrelated anions ATP, ADP, Cl-, SO:-, and oxalate’-. Adenosine (no negative charge) had no effect on transport. When vesicles were preloaded with anions (ADP, SO:-, oxalate2-) such that an anion gradient existed from the intra- to the extravesicular compartment, and methotrexate uptake was measured, no stimulation of uptake was seen. Methotrexate uptake into rat liver BLMVs was electrogenic as shown by stimulation of the initial rate of uptake by a valinomycin-imposed K+ diffusion potential across the vesicular membrane. These results suggest that methotrexate is transported into the hepatocyte across the basolateral membrane by an electrogenic, multispecific anion carrier system. 0 1992Academic Press,

Inc.

0003.9861/92

$5.00 1992 by of reproduction

Copyright 0 All

rights

Folic acid and its derivatives play a crucial role in metabolism. These coenzymes are involved in metabolism of amino acids and in the synthesis of purines and thymidine for RNA and DNA, and provide the formyl group for initiation of protein synthesis in mitochondria (1). Folate metabolism has been implicated in several diseases: megaloblastic anemias due to folate and vitamin B12 deficiencies (l), carcinogenesis (2), and mental retardation in fragile-X syndrome (3). The central role of folates in nucleic acid synthesis has lead to the use of “antifols,” e.g., methotrexate and trimetrexate, in the treatment of cancer. Both the folates and the antifolates are involved in an enterohepatic circulation (4-6). Since methotrexate is toxic for both the liver and the intestine (7,8), the enterohepatic circulation exposes these organs repeatedly to this toxic substance during therapy. For these reasons, our studies have been aimed at understanding the mechanism(s) whereby folates and antifolates are transported into the liver and are excreted into bile. Previous studies in our laboratory and that of Goldman have shown, by the different effects of inhibitors, that isolated hepatocytes transport 5-methyltetrahydrofolate by a carrier-mediated process different from that responsible for methotrexate transport (9-13). This characteristic is not seen in most other mammalian cell types as these have similar systems for transport of reduced folates and methotrexate (14, 15). Our recent findings further highlight the differences in 5-methyltetrahydrofolate and methotrexate transport in rat liver cells (16). We showed that 5-methyltetrahydrofolate uptake was dependent on a transmembrane proton gradient: transport increased as the extracellular pH was ’ This study was supported by the Medical Research Service of the Department of Veterans Affairs and by NIH Grant DK-32189. * To whom correspondence should be addressed at the Research Service (151), VA Medical Center, Nashville, TN 37212-2637.

121 Academic Press, Inc. in any

form

reserved.

122

HORNE

AND

lowered. Analysis of the data by Hill plots indicated a stoichiometry of 1:l for coupling of Smethyltetrahydrofolate transport with H+ ions. On the other hand, transport of methotrexate was maximal at an extracellular pH of about 7, wherein the transmembrane pH gradient was essentially zero. The isolated hepatocyte retains the basolateral and the canalicular membranes; therefore, we cannot know for certain whether transport studies reflect uptake across one or both membranes. To circumvent this problem, as well as the complication of metabolism and intracellular protein binding, we have utilized membrane vesicles isolated from the basolateral membrane of rat liver for our studies of Smethyltetrahydrofolate transport (17). These studies have confirmed the findings, discussed above, regarding Smethyltetrahydrofolate transport into liver cells. In the present study we report that transport of methotrexate into the basolateral membrane vesicles is a saturable process; is weakly inhibited by the structural analog folic acid but not by Smethyltetrahydrofolate nor 5-formyltetrahydrofolate; is independent of imposed Na+ or H+ gradients; is electrogenic in nature; and is inhibited by a number of structurally dissimilar anions. MATERIALS

REED

at 37’C and were started by adding 20 ~1 of BLMV suspension to 80 al of incubation medium [0.26 M sucrose, 0.04 M Hepes, pH 7.5 (Mes replaced Hepes in experiments at pH 6 or less)] which contained labeled and unlabeled substrate and/or other substituents as stated. Incubations were terminated at the desired time by adding 2 ml of ice-cold incubation buffer. The mixture was filtered and the filters washed twice with 2 ml of ice-cold incubation buffer. Uptake of methotrexate was estimated by counting the dried filters using a TM Analytic Betatrac Model 6895 liquid scintillation spectrometer. Adsorption of labeled methotrexate to the filters was assessed using a blank which contained all reagents except vesicles; this value was subtracted from all raw counts before calculating uptake.

RESULTS Effect of medium osmolurity on methotrexate uptake into BLMVs. In Figure 1, uptake of 0.25 PM methotrexate was assessed at 60 min of incubation and the results are plotted as a function of the reciprocal of medium osmolarity. Osmolarity was increased by increasing the extravesicular concentration of sucrose. In theory, at infinite osmolarity, there is zero intravesicular space and any measured “uptake” represents binding of the substrate to the membrane. The results show that methotrexate uptake was a linear function of l/osmolarity. Extrapolating the line to zero (infinite osmolarity) indicated that

AND METHODS

Materials. [3’,5’,7,9-3H]Methotrexate was purchased from Moravek Biochemicals, Inc. (Brea, CA). Isotopic purity was determined using our HPLC method for folates (18) (methotrexate eluted from the column in a baseline-separated peak after 5-methyltetrahydrofolate). [HMethotrexate was purified by the same HPLC procedure (if necessary) such that in no instance was substrate of 198% radiochemical purity used in these studies. Cellulose nitrate membrane filters (0.45-pm pore size) were purchased from Millipore. Unlabeled methotrexate, 5formyltetrahydrofolate, folic acid, 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid (Hepes)3 4-morpholineethanesulfonic acid (Mes), 4,4’-diisothiocyanostilbene-2,2’-disulfonic acid (DIDS), were from Sigma Chemical Company. (GS)-5-Methyltetrahydrofolate was synthesized according to Horne et al. (9). All other chemicals were reagent grade and were purchased from commercial sources. Basolateral membrane isolation. Basolateral membrane vesicles (BLMVs) were isolated from rat liver by our modification (17, 19) of the procedure of Blitzer and Donovan (20), which utilizes Percoll density gradient centrifugation. The isolated vesicles were resuspended in =3 ml of suspension buffer (0.3 M sucrose, 0.01 M Hepes-KOH, pH 7.5) at 3-5 mg of protein/ml by three passages through a 23-gauge needle. The vesicles were aliquoted and frozen at -70°C until needed. The frozen vesicles were thawed in a room-temperature water bath, allowed to equilibrate with transport buffer (as stated) for 2 h at room temperature, and resuspended by three passages through a 23-gauge needle prior to transport experiments. Protein was assayed according to Bradford (21). Transport assays. Methotrexate uptake into BLMVs was performed as described previously (17, 19). In brief, incubations were performed

’ Abbreviations used: PteGlu, pteroylglutamic acid, folic acid; H,PteGlu, 5,6,7,8,-tetrahydrofolic acid; 5HCO-H,PteGlu, 5-formyltetrahydrofolic acid; 5-CH,-H,PteGlu, 5-methyltetrahydrofolic acid; Hepes, 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid; Mes, 4morpholineethanesulfonic acid, DIDS, 4,4’-diisothiocyanostilbene-2,2’disulfonic acid; BLMV, basolateral membrane vesicle.

0.5

1.0

1.5

Z.0

2.5

3.0

3.5

1/Osmolarity FIG. 1. Effect of medium osmolarity on methotrexate transport into rat liver basolateral membrane vesicles. Vesicles were preincubated with 300 mM sucrose and 10 mM Hepes, pH 7.5, for 2 h at room temperature. Incubation was performed for 60 min at 37°C in an incubation buffer of 270 mM sucrose and 40 mM Hepes at pH 6.5 and sufficient additional amounts of sucrose to give the indicated osmolarities. Methotrexate (0.25 hM) was added to the incubation medium at the onset of incubation. Results are means + SE for triplicate determinations from at least three vesicle preparations.

METHOTREXATE r

I

f 0

/I

I

, 2w

TRANSPORT

//

, 400

I

‘, 600

/3ooo

Time (s) FIG. 2. Transport of methotrexate into BLMVs as a function of time and pH. Vesicles were preloaded with a buffer of 300 mM sucrose and 10 mM Hepes, pH 7.5 (0, V), or 300 mM sucrose, 10 mM Mes, pH 5.0 (0) at room temperature for 2 h. Uptake of 0.25 pM [sH]methotrexate was estimated at 37°C in medium of 270 mM sucrose, 40 mM Hepes, pH 7.5 (0), or 270 mM sucrose, 40 mM Mes, pH 5.0 (0, V). Results are means f SE for triplicate determinations from at least two vesicle preparations.

only about 9% of uptake could be accounted for by binding; therefore, about 90% of uptake represents transport into an osmotically active, intravesicular space. The possiMetabolism of methotrexate by BLMVs. bility that BLMVs might metabolize methotrexate was examined by incubating 0.25 PM methotrexate with vesicles for 60 min at 37°C. The vesicles were separated, and washed free of medium by filtration, and sonicated, the label taken up was analyzed by HPLC as described under Materials and Methods. Greater than 95% of the radioactivity eluted in the position of authentic methotrexate. This experiment shows that the vesicles do not significantly metabolize methotrexate. pH gradient dependence of methotrexate transport into BLMVs. Figure 2 illustrates experiments designed to determine the effect of a pH gradient, imposed across the vesicle membrane, on the uptake of methotrexate into rat liver BLMVs. As can be seen, the initial rate of uptake was similar under all three conditions (pH,, = pHi, = 7.5; pH,,, = 5.0, pHi” = 7.5; and pH,,t = pH, = 5.0). At times later than 2 min, uptake at pHo = 5.0, pHi = 7.5 and at pH, = pHi = 5.0 was similar, but in both these cases uptake was lower than uptake at pHo = pHi = 7.5.

IN LIVER

MEMBRANE

123

VESICLES

In separate experiments (data not shown) the initial rate of uptake was determined at an initial intravesicular pH of 7.5 and extravesicular pH from 5.0 to 8.0. The results showed that methotrexate uptake demonstrated a broad pH optimum: uptake was essentially the same at extravesicular pH ranging from 5.5 to 7.5, but uptake was somewhat less at extravesicular pH of 5.0 and 8.0. Na+gradient dependence of methotrexate transport into BLMVs. Previous reports have suggested that a Na+ gradient may energize uphill transport of methotrexate into the hepatocyte (11, 12). Unambiguous results may be obtained using isolated BLMVs wherein metabolism, protein binding, uptake through the canalicular membrane, or effects on membrane electrical gradients is not in question. Figure 3 shows that uptake of methotrexate into BLMVs was similar in the presence of imposed Nat and K+ gradients and no “overshoot” (an indication that transport could be energized by the imposed gradient) was seen. Concentration dependence of methotrexate transport into BLMVs. Figure 4 illustrates that methotrexate uptake into BLMVs was biphasic, characterized by two components. One was saturable (K, = 0.15 f 0.05 PM, V,,, = 11.4 + 1.1 pmol 10 s-l mg-’ protein) and the other, apparently nonsaturable with a slope of 1.31 f 0.12 pmol 10 s-l mg-’ protein PM-‘.

F-----+11 50

100

150

200

Time

250

300

4000

(s)

FIG. 3. Transport of methotrexate into BLMVs as a function of time in the presence and absence of an imposed Na+ gradient. Vesicles were preloaded for 2 h at room temperature with a solution of 300 mM sucrose, 10 mM Hepes, pH 7.5. Uptake of 0.25 pM [3H]methotrexate was estimated in incubation medium composed of 70 mM sucrose, 100 mM sodium gluconate, 40 mM Hepes, pH 6.5 (V), or 70 mM sucrose, 100 mM potassium gluconate, 40 mM Hepes, pH 6.5 (v). The results are means * SE for triplicate determinations from three vesicle preparations.

HORNE

124

[Methotrexote]

AND

(#A)

FIG. 4. Transport of methotrexate into BLMVs as a function of concentration. Rat liver BLMVs were preloaded at room temperature for 2 h with a buffer of 300 mM sucrose, 10 mM Hepes, pH 7.5. [3H]Methotrexate along with the appropriate amount of unlabeled methotrexate (as shown) was added at the onset of incubation. Uptake was estimated after 10 s of incubation in medium composed of 270 mM sucrose, 40 mM Hepes, pH 6.5. The curve was fitted to the data using SigmaPlot 4.1 (Jandel Scientific) employing the equation u = k*s + (( V,,,.ps)/(K~ + s)), where u, s, K,,,, and V,, have their usual meanings and k is the slope of the linear component of uptake. Results are means f SE of three or four determinations from three vesicle preparations.

Effect of structural analogs on methotrexate transport into BLMVs. As shown in Table I, uptake of methotrexate was not inhibited by the naturally occurring folate coenzymes 5methyltetrahydrofolate and 5-formyltetrahydrofolate. These results confirm our previous findings (12) and those of Goldman and associates (11) that methotrexate transport is not inhibited by these reduced, folate coenzymes in hepatocytes. Table I also shows that folic acid was somewhat inhibitory for methotrexate transport into the vesicles. Effect of different anions on methotrexate transport into BLMVs. It has recently been shown that liver has a multispecific, anion-exchange carrier localized to the basolateral membrane (22). Uptake of sulfate by this system was inhibited by various anions: SO;-, acetate, oxalate, etc. Figure 5 illustrates that methotrexate uptake into BLMVs was inhibited by several structurally unrelated anions. ADP and ATP were inhibitory, whereas adenosine showed no inhibition. Methotrexate uptake was also inhibited by Cl-, SO;, and by oxalate2-. Inhibition by acetate was not statistically significant.

REED

We tested the ability of several of the above anions to stimulate methotrexate uptake when an outwardly directed gradient of these anions existed across the vesicle membrane. Vesicles were preincubated for 2 h with 4 mM Na oxalate, Na acetate, or K2S04 or 50 PM ADP and uptake was determined over 20 s such that the extravesicular concentrations were 10 @M for ADP and 0.8 mM for the other anions. Figure 6 shows that there was no stimulation of methotrexate uptake by an imposed anion gradient (in > out). DIDS is an inhibitor of the red cell band 3 anion exchanger (23, 24). Its inhibition of anion transport has been used to imply that transport systems may function by anion exchange. Table II shows that DIDS inhibited methotrexate uptake in a concentration-dependent manner. However, the lack of transstimulation by outwardly directed anion gradients reported above (Fig. 6) argues against an anion-exchange mechanism for methotrexate uptake into BLMVs. Effect of an imposed electrical gradient on methotrexate uptake into BLMVs. An electrical potential may be generated across the vesicular membrane using valinomycininduced K+ diffusion (17, 19, 25). In our case an inside positive potential was generated by incubating vesicles (no intravesicular Kf ions) in medium containing 75 mM potassium gluconate. In the control (no net potential), vesicles were preincubated with 75 mM potassium gluconate and valinomycin and incubated in medium of the same Kf concentration. Uptake of methotrexate under these conditions is shown in Fig. 7. It can be seen that methotrexate uptake was much faster at early time periods as a result of the imposed inside-positive electrical gradient. Indeed, the electrical gradient caused a transient TABLE

I

Effect of Structural Analogs on Transport of Methotrexate in Rat Liver BLMVs

Inhibitor concentration (PM) 0 5.0 10.0 25.0 50.0 100.0

Methotrexate uptake (pm01 10 s-i mg-’ protein) 5-CH,-H,PteGlu 5.4 f 0.7 5.2 f 0.9 5.8 + 0.6 5.1 + 0.5 5.3 + 0.9 ND

5-HCO-HIPteGlu 5.4 f 0.7 5.5 It 0.2 5.6 k 0.4 5.8 2 0.7 5.6 f 0.4 ND

PteGlu 5.1 + 0.6 4.4 AI 0.4 3.7 5 0.4 ND” 2.0 f 0.3 1.3 +_0.6

Note. The results are the averages + SE for triplicate determinations from three different vesicle preparations. Vesicles were preincubated for 2 h at room temperature in 300 mM sucrose and 10 mM Hepes at pH 7.5. Transport of 0.25 PM [3H]methotrexate was measured at 37°C in medium composed of 270 mM sucrose, 40 mM Mea, pH 6.5, and unlabeled inhibitors at the concentrations indicated. ’ Not determined.

METHOTREXATE

TRANSPORT

IN LIVER

MEMBRANE

125

VESICLES

Inhibitors

c

Anions

FIG. 5. Effect of structurally unrelated anions on methotrexate transport into BLMVs. Vesicles were preincubated for 2 h at room temperature in a medium of 300 mM sucrose, 10 mM Hepes, pH 7.5. Uptake of 0.25 PM methotrexate was estimated at 37°C after 10 s of incubation in medium, at pH 6.5, which contained 10 pM adenosine, ADP, or ATP; 40 mM KCl; or 4 mM sodium acetate, potassium sulfate, or sodium oxalate. Isoosmolarity was maintained by adjusting the concentration of sucrose accordingly. Results are means + SE of triplicate determinations from at least three vesicle preparations. Columns marked by asterisks are significantly different from control, P < 0.05.

FIG. 6. Lack of transstimulation of methotrexate transport into BLMVs by various anions. Vesicles were preincubated for 2 h at room temperature in medium (20 ~1) composed of 300 mM sucrose, 10 mM Hepes, pH 7.5 (control); or the same medium plus 50 PM ADP; or 288 mM sucrose with either 4 mM K2S04 or 4 mM sodium oxalate, pH 7.5; or 292 mM sucrose, 4 mM sodium acetate, pH 7.5. Uptake of 0.25 PM [3H]methotrexate was estimated at 37°C after 20 s incubation by the addition of 80 gl of medium (270 mM sucrose, 40 mM Hepes, pH 6.5). Results are means + SE for three or four determinations from two vesicle preparations.

uptake of methotrexate above the steady-state level seen at later times (300 and 600 s). These results clearly indicate that methotrexate uptake is electrogenic in rat liver BLMVs.

strates against their chemical gradients (the so-called “overshoot” phenomenon) in response to an inwardly imposed Na+ gradient as reported (19, 20, 27). These characteristics demonstrate that our BLMV preparation is suitable for assessing transport of methotrexate into the liver cell.

DISCUSSION

In the present study we have used membrane vesicles isolated from the basolateral (sinusoidal) face of rat liver to study the events responsible for transport of methotrexate into liver cells from the blood. These vesicles were isolated by Percoll density-gradient centrifugation by our modification (17,19) of the procedure of Blitzer and Donovan (20). We have shown (17,19) that this vesicle preparation is suitable for transport studies by several means. The marker for the basolateral membrane (Na+/K+ ATPase) was enriched over 17-fold while markers for contaminating organelles were either decreased or showed only slight enrichment when compared to homogenate activities. The vesicles were found to be mostly (-70%) right-side-out by freeze-fracture electron microscopy and ouabain binding to the vesicles. The vesicles transported taurocholate (26, 27) and biotin (19) by a Na+-gradient dependent process, transiently accumulating these sub-

TABLE Effect (DIDS)

II

of 4,4’-Diisothiocyanostilbene-2,2’-disulfonic on Transport of Methotrexate in Rat Liver

DIDS concentration (mM) 0 0.1 0.5

Acid BLMVs

Methotrexate uptake (pm01 10 s-i mg-i protein) 4.25 f 0.53 2.89 + 0.26 1.17 + 0.18

Note. The results are the averages + SE for triplicate determinations from three different vesicle preparations. Vesicles were preincubated for 2 h at room temperature in 300 mM sucrose and 10 mM Hepes at pH 7.5. Transport of 0.25 PM [3H]methotrexate was measured at 37°C in medium composed of 270 mM sucrose, 40 mM Mes, pH 6.5, and the concentrations of DIDS indicated.

HORNE

, -

i 0

100

200

300

400

500

800

Time (s) FIG. 7. Effect of electrical gradient on BLMV transport of methotrexate. Vesicles were preincubated at room temperature for 115 min in medium either without potassium (0,300 mM sucrose, 10 mM Hepes, pH 7.5) or with potassium (0, 150 mM sucrose, 75 mM potassium gluconate, 10 mM Hepes, pH 7.5). Valinomycin (10 pg/mg protein) was added and incubation was continued for an additional 5 min. [3H]Methotrexate (0.25 PM) was added in medium such that in both incubations the extravesicular concentrations were potassium gluconate, 75 mM; sucrose, 150 mM; Mes, 10 mM, pH 6.5, and uptake was estimated at the times indicated. Inside positive electrical gradient (0) ([K+ out @ in). No electrical gradient (0) ([K+] out = in). Results are means + SE for triplicate determinations from three vesicle preparations. Asterisks indicate significant difference at P < 0.05.

Our results (Fig. 1) showed that methotrexate transport into BLMVs was mostly (=91%) transport into an osmotically active, intravesicular space and that binding to the membrane was only about 9% of total uptake. Studies were also performed to determine whether any metabolism of methotrexate occurred in our BLMV preparation. Labeled methotrexate was incubated with vesicles for 60 min and the accumulated label was subjected to HPLC analysis. Approximately 95% of the radioactivity coeluted with authentic methotrexate, indicating that there was no significant metabolism of the substrate. We have previously shown that Ei-methyltetrahydrofolate is transported into freshly isolated hepatocytes (16) and into rat liver BLMVs (17) by cotransport with H+ ions. In BLMVs, transport of 5methyltetrahydrofolate displayed the transient “overshoot” phenomenon (uptake at 3-5 min of incubation was twofold higher than at the steady state at 60 min) in response to an imposed H+ gradient (pH,,, = 5.0, pHi, = 7.5). Uptake at pH 5.0 (both

AND

REED

in and out, no pH gradient) was faster than at pH 7.5 (again, no gradient), but in neither of these cases was an overshoot seen. Figure 2 shows a time course of methotrexate uptake into BLMVs under different pH conditions. The initial rates of methotrexate uptake were similar under all three pH gradient conditions. At times later than 2 min, uptake was greater at pH 7.5 (both intra- and extravesicular) than at pH 5.0 (whether a initial pH gradient existed or not). The reason for the latter finding in not known but could result from differences in intravesicular volume or membrane binding at different pH. In any case, the important finding was that there was no evidence for an H+ gradient-dependent overshoot (Fig. 2, pH, = 5.0, pHi = 7.5) for methotrexate transport, indicating that this compound is not cotransported with H+ ions. Methotrexate uptake into BLMVs with intravesicular pH of 7.5 was determined at extravesicular pH ranging from 5.0 to 8.0 (i.e., the pH gradient across the vesicular membrane was systematically varied). Methotrexate uptake showed little or no dependence on the magnitude of the pH gradient; indeed the results only demonstrate a broad pH optimum for methotrexate transport. Again, this is in contrast with uptake of 5-methyltetrahydrofolate into the vesicles in which case the initial rate of uptake was proportional to the magnitude of the imposed H+ gradient (17). Our group (12) and that of Goldman (11) have reported that methotrexate uptake into freshly isolated rat liver cells is dependent on extracellular sodium ions. In the present study using BLMVs we found no evidence for Na+ gradient dependence of methotrexate transport (see Fig. 3). Uptake of methotrexate was similar in the presence of imposed gradients (out > in) of 100 mM Na or K gluconate and no overshoot was apparent. The reason for the lack of sodium dependence in the BLMVs and the sodium dependence seen in isolated hepatocytes is not known. However, the isolated hepatocyte is a much more complex system than the isolated BLMVs. The hepatocyte has both the basolateral membrane, through which compounds are transported from and into the circulation, and a canalicular membrane, through which compounds are secreted into bile. Therefore, transport characteristics in the isolated hepatocyte could reflect a combination of the properties of both membrane transport systems. Indeed, it has been shown that a Na+ gradient-dependent system for uptake of glutamic acid exists in the bile canalicular membrane (28). A similar partitioning could explain the apparently contradictory results for methotrexate transport in BLMVs and in isolated hepatocytes. An alternative interpretation rests on the fact that methotrexate transport into BLMVs is an electrogenic process (see below). Isolated hepatocytes normally maintain an electrical potential of about 32 mV (cytosol negative with respect to extracellular medium) and ouabain results in depolarization (29, 30) of this potential. Since ouabain

METHOTREXATE

TRANSPORT

inhibits the Na+/K+ ATPase, replacement of sodium in the medium by choline should also lead to depolarization (less negative potential) and, therefore, increased uptake of the anionic methotrexate. Transport of methotrexate into BLMVs showed the characteristics of a carrier-mediated system. The initial uptake rate was a saturable function of extravesicular concentration of methotrexate as shown in Fig. 4. Transport was composed of a high-affinity saturable component with apparent K, = 0.15 PM and V,,, = 11.4 pmol 10 s-l mg-‘. A second, and apparently nonsaturable, component was evident at extravesicular concentrations up to 15 PM. As shown in Table I, methotrexate uptake was inhibited by the structural analog folic acid, but not by &methyltetrahydrofolate nor by 5formyltetrahydrofolate. This lack of inhibition of the latter two analogs is consistent with observations in isolated hepatocytes reported by us (12) and by Goldman and co-workers (11). A recent report by Hugentobler and Meier (22) has characterized a multispecific anion exchange system in BLMVs. Since methotrexate is a dianion at physiological pH, experiments were performed to ascertain if this exchange system might be responsible for methotrexate transport in BLMVs. Figure 5 shows that a number of anions-ADP, ATP, chloride, sulfate, and oxalate-exhibited cis-inhibition of methotrexate uptake into BLMVs. Further studies were carried out to determine if these anions, when present inside the vesicle, might be exchanged for extravesicular methotrexate. The vesicles were preincubated with a high concentration of the appropriate anion. Uptake of methotrexate was measured by diluting the vesicles into medium without the anion such that an outward gradient (concentration inside was fivefold greater than outside) was established. The results (Fig. 6) showed that there was no transstimulation of methotrexate uptake by any of the anions tested, suggesting that methotrexate was not entering the vesicles in exchange for these anions. The electrical nature of methotrexate transport was determined by creating a relatively positive electrical potential across the membrane using valinomycin-induced K+ diffusion into the intravesicular space. Under this condition, not only was the initial rate of methotrexate uptake faster in the presence of a K+ diffusion potential, but uptake showed a transient “overshoot” in response to the imposed positive electrical potential (Fig. 7). This “overshoot” is attributed to enhanced uptake of methotrexate while an inside positive potential is maintained. As time progresses and the magnitude of the potential declines due to equilibration of potassium across the membrane, the accumulated methotrexate will efflux from the vesicles and reach a steady state. These results clearly demonstrate that uptake of the negatively charged methotrexate is electrogenic in nature; i.e., it is the anionic species that crosses the vesicular membrane. This obser-

IN LIVER

MEMBRANE

127

VESICLES

vation could help explain the increased efflux of methotrexate caused by dibutyryl cyclic AMP in hepatocytes as reported by Gewirtz et al. (31). It has been shown that cyclic AMP causes a slow (over 30 min) hyperpolarization (15-20 mV more negative) of the hepatocyte membrane [see Ref. (29) for review]. Thus, the increased negative charge across the hepatocyte membrane could result in increased efflux of methotrexate. It is interesting to compare the properties of methotrexate uptake into liver BLMVs with those reported for other systems. In L1210 mouse leukemia cells (15) uptake of methotrexate and the reduced folates (e.g., 5-methyltetrahydrofolate and 5-formyltetrahydrofolate) appear to share the same system. Further studies by Henderson and Zevely (32) and by Sirotnak and colleagues (33) indicated that transport was via anion exchange. In the intestine, uptake of folic acid, methotrexate, and reduced folates appears to be via the same system which is energized by cotransport with hydrogen ions (14, 34, 35). In contrast, methotrexate transport in BLMVs was inhibited by anions but there was no evidence of an anion-exchange mechanism, uptake was not inhibited by the reduced folates (whereas folic acid did inhibit), and there was no evidence for hydrogen ion cotransport. In conclusion, methotrexate transport into rat liver basolateral membrane vesicles is a carrier-mediated process, is independent of imposed Na+ or H+ gradients, is electrogenic in nature, and is inhibited by various structurally unrelated anions and the analog folic acid but not 5methyltetrahydrofolate nor 5-formyltetrahydrofolate. REFERENCES 1. Blakley, R. L. (1969) The Biochemistry of Folic Acid and Related Pteridines, North-Holland, Amsterdam/London. 2. Mikol, Y. B., Hoover, K. L., Creasia, D., and Poirier, Carcinogenesis 4, 1619-1629. 3. Krumdieck, C. L., and Howard-Peebles, Genet. 16,23-28.

L. A. (1983)

P. N. (1983) Am. J. Med.

4. Strum, W. B., and Liem, H. H. (1977) Biochem. Pharmacol. 12351240.

26,

5. Strum, W. B., and Liem, H. H. (1980) Res. Commun. Chem. Pathol. Pharmacol. 30, 493-507. 6. Hillman, R. S., and Steinberg, 345-354.

S. E. (1982) Annu. Reu. Med. 33,

7. Trier, J. S. (1962) Gastroenterology

42, 295-305.

8. Dahl, M. G. C., Gregory, M. N., and Scheuer, P. J. (1971) Br. Med.

J. 1,625-630. 9. Horne, D. W., Briggs, W. T., and Wagner, C. (1978) J. Biol. Chem. 253,3529-3535. 10. Horne, D. W. (1986) in Chemistry and Biology of Pteridines (Cooper, B. A., and Whitehead, V. M., Eds.), pp. 559-564, Walter de Gruyter, Berlin/New York. 11. Gewirtz, D. A., White, J. C., Randolph, J. K., and Goldman, I. D. (1980) Cancer Res. 40, 573-578. 12. Horne, D. W., Briggs, W. T., and Wagner, C. (1976) Biochem. Biophys. Res. Commun. 68,70-76.

128

HORNE

13. Horne, D. W., Briggs, W. T., and Wagner, C. (1979) Arch. Biochem. Biophys. 196,557-565. 14. Said, H. M., and Strum, W. B. (1983) J. Phurmacol. Exp. Z’her. 226,95-99. 15. Goldman, I. D. (1971) Ann. N.Y. Acad. Sci. l&X6,400-422. 16. Horne, D. W. (1990) Biochim. Biophys. Acta 1023,47-55. 17. Horne, D. W., Reed, K. A., and Said, H. M. (1991) Am. J. Physiol. 262, G150-G158. 18. Horne, D. W., Patterson, Biophys. 270, 729-733.

D., and Cook, R. J. (1989) Arch. Biochem.

19. Said, H. M., Korchid, S., Horne, D. W., and Howard, M. (1990) Am. J. Physiol. 259, G865-G872.

20. Blitzer, B. L., and Donovan, C. B. (1984) J. Biol. Chem. 259,92959301.

21. Bradford, M. (1976) Anal. Biochem. 72, 248-254. 22. Hugentobler, G., and Meier, P. J. (1986) Am. J. Physiol. 251, G656G664. 23. Cabantchik, Z. I., and Rothstein, A. (1972) J. Membr. Biol. 10,311330. 24. Cabantchik, Z. I., and Rothstein, A. (1974) J. Membr. Biol. 15,207226.

AND

REED

25. Meier, P. J., St. Meire-Abt, A., Barrett, J. Biol. Chem. 259, 10614-10622.

C., and Boyer, J. L. (1984)

26. Deleted in proof. 27. Deleted in proof. 28. Ballatori,

N., Moseley, R. H., and Boyer, J. L. (1986) J. Biol. Chem.

261,6216-6221. 29. Moule, S. K., and McGivan, J. D. (1990) Biochim. Rev. Biomembr. 103 1, 383-397. 30. Hoek, J. B., Nichols, D. G., and Williamson, Chem. 255,1458-1464.

Biophys. Actu

J. R. (1980) J. Biol.

31. Gewirtz, D. A., Randolph, J. K., and Goldman, I. D. (1981) Biochem. Biophys. Res. Commun. 101, 366-374. 32. Henderson,

G. B., and Zevely, E. M. (1983) Arch. Biochem. Biophys.

221,438-446. 33. Yang, C.-H., Sirotnak,

F. M., and Dembo, M. (1984) J. Membr. Biol.

79,285-292. 34. Said, H. M., Ghishan, F. K., and Redha, R. (1987) Am. J. Physiol. 252,G229-G236. 35. Schron, C. H., Washington, Inuest. 76,2030-2033.

C., Jr., and Blitzer, B. L. (1985) J. Clin.

Transport of methotrexate in basolateral membrane vesicles from rat liver.

Transport of the antifolate cancer drug methotrexate was studied in vesicles isolated from the basolateral membrane of rat liver. Transport of methotr...
865KB Sizes 0 Downloads 0 Views