Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Cyclosporine-A (CsA), a P-glycoprotein (P-gp) inhibitor, produces marked changes in the handling of L-3,4-dihydroxyphenylalanine (L-dopa), the precursor of natriuretic dopamine. In the rat, chronic administration of CsA is accompanied by antinatriuresis, hypertension, and a reduction in daily urinary excretion of dopamine, which appears to result from a reduction in the amount of L-dopa made available to the kidney (Pestana et al., 1995). On the other hand, when CsA-treated rats were given exogenous L-dopa, an enhanced accumulation of newly formed dopamine in kidney was observed (Pestana et al., 1995), suggesting that CsA treatment favored the intracellular accumulation of L-dopa to be decarboxylated to dopamine. This view was confirmed in subsequent studies in the renal epithelial cell line LLC-PK₁ (Soares-da-Silva et al., 1998b). In LLC-PK₁ cells, the uptake of L-dopa is a facilitated, saturable, and energy-dependent process, and short-term exposure (30 min) to CsA increased in the intracellular availability of L-dopa (Soares-da-Silva et al., 1998b). By contrast, long-term exposure (14 h) to CsA was found to produce opposite effects on the intracellular availability of L-dopa (Soares-da-Silva et al., 1998b). The short-term effects of CsA are apparently related to a decrease in the ability of LLC-PK₁ cells to extrude L-dopa, whereas the long-term effects of CsA appear to be due to an increase in the cellular extrusion of L-dopa (Soares-da-Silva et al., 1998b). These effects correlated well with the effects of CsA on P-gp activity, suggesting that cell extrusion of L-dopa could be promoted by this transporter. CsA is known to acutely reverse P-gp-mediated multidrug resistance (Keller et al., 1992) and to effectively inhibit P-gp-mediated transfer mechanisms (Leveque and Jehl, 1995). On the other hand, chronic exposure to CsA stimulates P-gp, and this is believed to represent a mechanism of cellular detoxification (Garcia del Moral et al., 1995). P-gp is a 170-kDa plasma membrane protein encoded by the mammalian multidrug resistant (MDR1) gene (Endicott and Ling, 1989), which functions as an ATP-driven active efflux pump of a wide variety of drugs (Ford and Hait, 1990) and also acts as an efflux pump to expel hydrophobic substances from the cells (Rao, 1995).

LLC-PK₁ cells express proximal tubule cell-like properties in vitro (Hull et al., 1976) and have been used for the purpose of study dopamine receptors and the renal actions of the amine. These cells have been shown to contain high levels of aromatic L-amino acid decarboxylase (AADC) and convert L-dopa to dopamine in a saturable fashion (Dawson and Phillips, 1990; Grenader and Healy, 1991; Soares-da-Silva et al., 1997). Newly formed dopamine in LLC-PK₁ was demonstrated to stimulate cAMP accumulation, an effect attenuated by an equimolar concentration of carbidopa or blocked by the D₁ antagonist Sch 23390 (Grenader and Healy, 1991); this suggests that locally formed dopamine in LLC-PK₁ cells, as in epithelial cells of proximal tubules, can act as an autocrine/paracrine substance, and these cells constitute a good model for the in vitro study of the renal dopaminergic system. LLC-GA5 Col300 cells, which overexpress human P-gp on the apical membranes, were derived from a clone of LLC PK₁ cells stable transfected with a cDNA encoding the human P-gp (Saeki et al., 1993).

The present work was aimed at studying the role of P-gp on the basal-to-apical uptake and flux of L-dopa in LLC-PK₁ and LLC-GA5 Col300 cells and examining the effect of short-term (30 min) and long-term (3 h) exposure to UIC2, a monoclonal anti-P-gp antibody (Mechetner and Roninson, 1992), and verapamil on the L-dopa apical extrusion and P-gp activity. It is reported that P-gp activity in LLC-GA5 Col300, as evidenced by their ability to extrude rhodamine 123 (Rh123) or calcein, was twice that observed in LLC-PK₁ cells, and both the native P-gp in LLC-PK₁ and the human P-gp LLC-GA5 Col300 cells may promote the apical outward transfer of L-dopa. Other inhibitors or substrates of P-gp (vinblastine, Rh123, quinidine, and daunomycin) were also found to inhibit basal-to-apical flux of L-dopa in both cell lines, although more potently in LLC-PK₁ cells.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cell Culture. LLC-PK₁ cells, a porcine-derived proximal renal tubule epithelial cell line that retains several properties of proximal tubular epithelial cells in culture (Hull et al., 1976), were obtained from the American Type Culture Collection (Rockville, MD). LLC-GA5 Col300 cells, which were obtained from the Riken Cell Bank, derive from a clone of LLC PK₁ cells stable transfected with a cDNA encoding the human P-gp (Saeki et al., 1993). Both cell lines were maintained in a humidified atmosphere of 5% CO₂, 95% air at 37°C. LLC-PK₁ cells (ATCC CRL 1392; passages 210-215) were grown in Medium 199 (Sigma Chemical Co., St. Louis, MO) supplemented with 100 U/ml penicillin G, 0.25 µg/ml amphotericin B, 100 µg/ml streptomycin (Sigma), 3% fetal bovine serum (Sigma), and 25 mM HEPES (Sigma). The culture medium used to grow LLC-GA5 Col300 cells (passages 16-25) was similar to that described earlier with the exception that it contained 300 ng/ml colchicine. For subculturing, the cells were dissociated with 0.05% trypsin-EDTA, split 1:4, and subcultured in Costar flasks with 75- or 162-cm² growth areas (Costar, Badhoevedorp, the Netherlands). For uptake studies, the cells were seeded onto collagen-treated 0.2-µm polycarbonate filter supports (internal diameter, 12 mm; Transwell; Costar) at a density of 13,000 cells per well (2.0 × 10⁴ cells/cm²). For studies on P-gp activity, the cells were seeded onto collagen-treated glass coverslips (10 mm diameter) at a density of 30,000 cells per coverslip. The cell medium was changed every 2 days, and the cells reached confluence after 3 to 5 days of incubation. For 24 h before each experiment, the cell medium was free of fetal bovine serum. Experiments were generally performed 2 to 3 days after cells reached confluence and 6 to 8 days after the initial seeding, and each square centimeter contained about 100 µg of cell protein.

P-gp Activity. In the first set of experiments, P-gp activity was measured according to the procedures previously described (Mechetner and Roninson, 1992; Fardel et al., 1996), with minor modifications. In brief, LLC-PK₁ cells cultured in collagen-treated plastic 24 wells were incubated for 30 min in the presence of UIC2 (3 µg/ml) and then incubated with 100 µM Rh123 for an additional 15 min. UIC2 is a mouse monoclonal antibody that recognizes an extracellular epitope of the human P-gp (Mechetner and Roninson, 1992). Uptake was terminated by the rapid removal of the medium containing Rh123; thereafter, the cells were washed with 2 ml of ice-cold Hanks' medium and 1.5 ml of 0.1% (v/v) Triton X-100 (dissolved in 5 mM HCl, pH 7.4) was added to each 2-cm² well to solubilize the cells. Rh123 was measured by spectrophotometry at 499 nm. Triton X-100 (0.1 v/v) was found not to alter Rh123 measurements.

Transport Studies. On the day of the experiment, the growth medium was aspirated, and the cells were washed with Hanks' medium; thereafter, the cell monolayers were preincubated for 15 min in Hanks' medium at 37°C. The upper and lower chambers contained 600 and 200 µl, respectively. The Hanks' medium used in these experiments also contained benserazide (50 µM) and tolcapone (1 µM) to inhibit the enzymes AADC and catechol-O-methyltransferase, respectively. During preincubation and incubation, the cells were continuously shaken and maintained at 37°C. Uptake was initiated by the addition of 600 µl of Hanks' medium with a given concentration of L-dopa to the lower chamber only. In experiments designed to study the effect of UIC2 (3 µg/ml), verapamil (25 µM), vinblastine (10 µM), Rh123 (10 µM), quinidine (50 µM), and daunomycin (10 µM), cells were incubated with 2.5 µM L-dopa applied from the basal cell border, and uptake (accumulation in the cell monolayer) and flux (transfer to opposite chamber) were measured over a 6-min period. Inhibitors of P-gp were all applied from the apical side only and were present during the preincubation and incubation periods. [³H]Sorbitol (0.4 µM) was used to estimate paracellular fluxes and extracellular trapping of L-dopa during L-dopa uptake studies. At the end of incubation, cells were placed on ice, and the medium bathing the apical cell border was collected, acidified with perchloric acid, and stored at 4°C until assay for L-dopa. The cells were washed with ice-cold Hanks' medium, and 0.2 mM perchloric acid was added (100 and 500 µl in the upper and lower chambers, respectively); the acidified samples were stored at 4°C before injection into the high-pressure liquid chromatograph for the assay of L-dopa, as previously described (Soares-da-Silva et al., 1998b). The lower limits for detection of L-dopa ranged from 350 to 500 fmol.

Protein Assay. The protein content of monolayers of LLC-PK₁ and LLC-GA5 Col300 cells was determined according to the method of Bradford (1976), with human serum albumin as a standard.

Cell Viability. Cells cultured in collagen-treated plastic supports were preincubated for 30 min or 3 h at 37°C in the absence or the presence of test drugs and then incubated in the absence or the presence of L-dopa for an additional 6 min. Subsequently, the cells were incubated at 37°C for 2 min with trypan blue (0.2% w/v) in phosphate buffer. Incubation was stopped by rinsing the cells twice with Hanks' medium, and the cells were examined using a Leica microscope. Under these conditions, more than 95% of the cells excluded the dye.

Data Analysis. P-gp activity was determined by the slope of the accumulation of calcein (in picomoles per milligram of protein) measured by linear regression analysis (Neame and Richards, 1972). K_m and V_max values for the uptake of L-dopa, as determined in saturation experiments, were calculated from nonlinear regression analysis using the GraphPad Prism statistics software package (GraphPad Software, San Diego, CA). Fractional outflow (apical or basal) was calculated using the expression L-dopa^fluid/(L-dopa^fluid + L-dopa^cell), where L-dopa^fluid indicates the amount of L-dopa (in nanomoles per milligram of protein) that reached the apical or the basal chamber and L-dopa^cell (in nanomoles per milligram of protein) indicates the amount of L-dopa accumulated in the cell monolayer. Arithmetic mean values are given with S.E. values. Statistical analysis was performed by one-way ANOVA, followed by Newman-Keuls test for multiple comparisons. A P value of less than .05 was assumed to denote a significant difference.

Drugs. Calcein and calcein-AM were obtained from Molecular Probes (Eugene, OR). Daunomycin, L-dopa, quinidine, verapamil hydrochloride, and vinblastine were purchased from Sigma Chemical Co. Tolcapone was kindly donated by the late Prof. Mosé Da Prada (Hoffman-La Roche, Basel, Switzerland). UIC2 was obtained from Immunotech (Marseille, France).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

The first set of experiments were aimed to provide functional evidence on the presence of P-gp in both LLC-PK1 and LLC-GA5 Col300 cells and to identify possible differences in the sensitivity of these cells lines to P-gp inhibitors. As shown in Fig. 1A, LLC-GA5 Col300 cells accumulate less Rh123 (2.9 ± 0.1 nmol/mg protein) than do LLC-PK₁ cells (8.1 ± 0.6 nmol/mg protein). In agreement with this result is the finding that 30-min treatment with the anti-P-gp monoclonal antibody UIC2 (3 µg/ml) inhibited P-gp in LLC-PK₁ cells, resulting in a 15% increase (P < .05) in Rh123 accumulation, whereas in LLC-GA5 Col300 cells, it failed to change Rh123 accumulation. Figure 2 shows one transmitted light image and several consecutive fluorescent images in single LLC-PK₁ (Fig. 2A) and LLC-GA5 Col300 (Fig. 2B) cells loaded with calcein-AM (0.5 µM) and then treated with verapamil (25 µM); total exposure time was 50 min, and verapamil was added at t = 10 min. As shown in the figure, only the LLC-PK₁ cell was found to accumulate calcein, this being particularly evident from 30 min on. Figure 2C shows the results on the intracellular accumulation of calcein in single LLC-PK₁ and LLC-GA5 Col300 cells using the protocol described earlier in single cells in five independent experiments. Cells were loaded with calcein-AM (0.5 µM) and then treated with verapamil (25 µM); total exposure time was also 50 min, and verapamil was added at t = 10 min. Intracellular calcein steadily rose for 15 min until reaching a plateau. The latency between the addition of verapamil and the rise in calcein levels was 10 to 12 min.

View larger version (13K): [in this window] [in a new window]   Fig. 1.   A, accumulation of Rh123 in LLC-PK1 cells (open columns) and LLC-GA5 Col300 cells (filled columns). B, effect of 30-min incubation with UIC2 (3 µg/ml) on the accumulation of Rh123 in LLC-PK1 cells (open columns) and LLC-GA5 Col300 cells (filled columns). Columns represent mean values of six experiments per group, and vertical lines indicate S.E. Significantly different from values for LLC-Col 300 cells (*P < .05) and corresponding control values (#P < .05).

View larger version (30K): [in this window] [in a new window]   Fig. 2.   Transmitted light and consecutive fluorescent images in single LLC-PK1 cells (A) and LLC-GA5 Col300 cells (B) loaded with calcein-AM (0.5 µM) and then treated with verapamil (25 µM); total exposure time was 50 min, and verapamil was added at t = 10 min. The calibration bar on the right indicates levels of calcein in nanomolar. C, graph shows the time course of the intracellular accumulation of calcein in single LLC-PK1 and LLC-GA5 Col300 cells using the protocol described above in single cells in five independent experiments. Cells were loaded with calcein-AM (0.5 µM) and then treated with verapamil (25 µM); total exposure time also was 50 min, and verapamil was added at t = 10 min.

Because LLC-GA5 Col300 cells accumulate less Rh123 and calcein and UIC2 and verapamil were less effective in enhancing the accumulation of these two substrates of P-gp, it was hypothesized that these effects could be related to the overexpression of P-gp in LLC-GA5 Col300 cells. It was decided, therefore, to increase the incubation period with verapamil from 30 min to 3 h, with the aim of enhancing the inhibition of P-gp. Figure 3 shows the accumulation of calcein in LLC-PK₁ and LLC-GA5 Col300 cells in control conditions during a 20-min incubation period and after exposure to verapamil (25 µM) for 30 min or 3 h. In control conditions, LLC-GA5 Col300 cells accumulate less calcein (7.5 ± 0.3 pmol/mg protein/s) than LLC-PK₁ cells (11.2 ± 0.3 pmol/mg protein/s), as evidenced by the slope of the regression lines. In LLC-PK₁ cells, pretreatment with verapamil (25 µM) for 30 min increased the rate of accumulation of calcein by 5-fold, whereas in LLC-GA5 Col300 cells, no significant change in the rate of accumulation of calcein was observed (see also Table 1). Exposure for 3 h to verapamil (25 µM) was found to increase the rate of accumulation of calcein by 2.5-fold in LLC-PK₁ cells and by 3.7-fold in LLC-GA5 Col300 cells (see also Table 1).

View larger version (18K): [in this window] [in a new window]   Fig. 3.   Accumulation of calcein (0.5 µM) for 20 min in LLC-PK1 cells (A) and LLC-GA5 Col300 cells (B) in control conditions and after short-term (30 min) and long-term (3 h) exposure to verapamil (25 µM). Values are percentage of calcein levels at t = 0 min. Symbols represent mean values of five to seven experiments per group; S.E. values were 5 to 10% of corresponding mean values. , control; , 30 min verapamil; , 3 h verapamil.

Previous studies from our laboratory (Soares-da-Silva et al., 1998a,b) have shown that the accumulation of L-dopa, in time course experiments, increased linearly with time for several minutes until attaining equilibrium at 12 min. Therefore, in experiments on the accumulation and flux of L-dopa across cell monolayers, the cells were incubated for 6 min with a nonsaturating concentration of L-dopa (2.5 µM). The K_m value for the accumulation of L-dopa applied from the basal cell side was found to be similar to that from the apical side (Soares-da-Silva et al., 1998a,b). As shown in Fig. 4, the accumulation of L-dopa (2.5 µM) applied from the basal cell border in LLC-PK₁ cells was similar to that observed in LLC-GA5 Col300 cells, but the basal-to-apical flux of L-dopa in LLC-GA5 Col300 cells was 1.5-fold that in LLC-PK₁ cells. By contrast, when the substrate was applied from the apical cell side, the accumulation of L-dopa (2.5 µM) in LLC-GA5 Col300 cells was less than half that observed in LLC-PK₁ cells and the apical-to-basal flux of L-dopa did not differ between the two cell lines. The accumulation of L-dopa applied from the basal cell border in both cell lines was significantly greater (P < .05) than that observed when the substrate was applied from the apical cell side. Paracellular leakage measured by the basal-to-apical flux of [³H]sorbitol was minimal in both cell lines and represented 0.1% of the amount applied at the cell surface.

View larger version (23K): [in this window] [in a new window]   Fig. 4.   Accumulation and basal-to-apical or apical-to-basal flux of L-dopa in LLC-PK1 cells (open columns) and LLC-GA5 Col300 cells (filled columns) incubated for 6 min at 37°C. Columns represent mean values of four experiments per group, and vertical lines indicate S.E. Significantly different from corresponding values (*P < .05) in LLC-PK1 cells. Significantly different from corresponding vlaues (#P < .05) when applied from the basal cell border.

Because LLC-GA5 Col300 cells required a more prolonged exposure to verapamil (3 h instead of 30 min in LLC-PK₁) to demonstrate inhibition of P-gp (enhanced accumulation of calcein), we decided to apply the same methodology in L-dopa transport studies. As shown in Fig. 5, a 30-min exposure to UIC2 (3 µg/ml) or verapamil (25 µM) increased L-dopa accumulation in LLC-PK₁ cells by 27 ± 4 and 88 ± 14% and reduced L-dopa apical extrusion by 29 ± 4 and 23 ± 1%, respectively. On the other hand, exposure to UIC2 (3 µg/ml) or verapamil (25 µM) for 3 h was found to produce less marked increases in the accumulation of L-dopa in the cell monolayer and less marked reductions in the apical extrusion of L-dopa. In contrast to that observed in LLC-PK₁ cells, exposure of LLC-GA5 Col300 cells to UIC2 (3 µg/ml) or verapamil (25 µM) for 30 min produced no significant changes in cell accumulation and apical extrusion of L-dopa (Fig. 6). A more prolonged exposure (3 h) to UIC2 (3 µg/ml) or verapamil (25 µM) resulted in a marked increase in L-dopa accumulation in the cell (105 ± 13 and 146 ± 24% increase) and a pronounced decrease (91 ± 1 and 92 ± 1% reduction) in the apical extrusion of L-dopa (Fig. 6).

View larger version (23K): [in this window] [in a new window]   Fig. 5.   Effect of UIC2 (3 µg/ml) and verapamil (25 µM) on the apical fractional outflow (A) and cellular uptake of L-dopa (B) in LLC-PK1 cells. Cells were preincubated for 30 min (hatched columns) or 3 h (filled columns) with UIC2 or verapamil and then incubated for 6 min at 37°C with 2.5 µM L-dopa, added from the basolateral cell border. Columns represent mean values of six experiments per group, and vertical lines indicate S.E.M. Significantly different (*P < .05) from controls (open columns).

View larger version (21K): [in this window] [in a new window]   Fig. 6.   Effect of UIC2 (3 µg/ml) and verapamil (25 µM) on the apical fractional outflow (A) and cellular uptake of L-dopa (B) in LLC-GA5 Col300 cells. Cells were preincubated for 30 min (hatched columns) or 3 h (filled columns) with UIC2 or verapamil and then incubated for 6 min at 37°C with 2.5 µM L-dopa, added from the basolateral cell border. Columns represent means of six experiments per group and vertical lines show S.E. Significantly different (*P < .05) from controls (open columns).

Figure 7 shows the effect of several compounds that are substrates or inhibitors of P-gp on the basal-to-apical flux of L-dopa; test compounds were applied from the apical cell side only, whereas L-dopa was applied from the basolateral cell side. As shown in the figure, a 30-min exposure to these substances produced a more pronounced inhibition of the apical flux of L-dopa in LLC-PK₁ than in LLC-GA5 Col300 cells. This was particularly evident, with the difference attaining statistical significance (P < .05), for vinblastine (10 µM), Rh123 (10 µM), and quinidine (50 µM). In another set of experiments designed to evaluate the directional nature of the phenomenon we described, vinblastine (10 µM) and Rh123 (10 µM) were applied from the basolateral cell side and L-dopa was applied from the apical cell side. Both drugs failed to affect the apical-to-basolateral flux of L-dopa (Fig. 8).

View larger version (19K): [in this window] [in a new window]   Fig. 7.   Effect of vinblastine (Vin; 10 µM), Rh123 (Rh; 10 µM), quinidine (Quin; 50 µM), and daunomycin (Dau; 10 µM) on the basal-to-apical flux of L-dopa in LLC-PK1 (open columns) and LLC-GA5 Col300 (filled columns) cells. Cells were preincubated for 30 min with test compounds and then incubated for 6 min at 37°C with 2.5 µM L-dopa, added from the basolateral cell border. Columns represent mean values of six experiments per group, and vertical lines indicate S.E. Significantly different from corresponding controls (*P < .05) or from values obtained in LLC-GA5 Col300 cells (#P < .05).

View larger version (26K): [in this window] [in a new window]   Fig. 8.   Effect of vinblastine (Vin; 10 µM) and Rh123 (Rh; 10 µM) on the apical-to-basal flux of L-dopa in LLC-PK1 (open columns) and LLC-GA5 Col300 (filled columns) cells. Cells were preincubated for 30 min with test compounds and then incubated for 6 min at 37°C with 2.5 µM L-dopa, added from the basolateral cell border. Columns represent mean values of four experiments per group, and vertical lines show the S.E.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The results presented here clearly show that intracellular L-dopa can be extruded from the cell over the apical cell border, with the expression of this phenomenon being more pronounced in LLC-GA5 Col300 cells than in LLC-PK₁ cells. This is demonstrated by greater basal-to-apical flux of L-dopa in LLC-GA5 Col300 cells than in LLC-PK₁ cells. In agreement with this view is the finding that cell accumulation of L-dopa in LLC-GA5 Col300 cells was significantly lower than that in LLC-PK₁ cells when the substrate was applied from the apical cell border. Both sets of results suggest that intracellular L-dopa is actively extruded out of the cell through the apical cell border, most probably through P-gp. This is in agreement with the finding that LLC-GA5 Col300 cells are endowed with greater P-gp activity than LLC-PK₁ cells, which fits well with the fact that the former cell line was engineered to overexpress the human P-gp at the apical cell border (Saeki et al., 1993). In this respect, it is interesting to observe that the accumulation of L-dopa from the basal side and the apical-to-basal flux of L-dopa was identical in the two cell lines.

The first point worth discussing concerns the presence of P-gp activity in LLC-PK₁ cells, given the discrepant reports in the literature. Although the majority of functional studies reported in the literature favor the view that LLC-PK₁ cells are endowed with P-gp (Tanigawara et al., 1992; Ito et al., 1993a,b,c; Decorti et al., 1998), there are reports suggesting the absence of P-gp in this cell line (Cvetkovic et al., 1999; Matsuzaki et al., 1999). Undoubtedly, P-gp activity should be markedly higher in cells stable transfected with a cDNA encoding the human P-gp than in the original cell clone (Saeki et al., 1993; Cvetkovic et al., 1999; Matsuzaki et al., 1999). This, however, does not exclude the presence of native P-gp in LCC-PK₁ cells and its functional importance in handling P-gp substrates. The evidence gathered in the present study favors the presence of this transporter in LLC-PK₁ cells. These cells were found to accumulate Rh123, with this being sensitive to the anti-P-gp monoclonal antibody. More importantly, these cells were also found to accumulate calcein, one of the best markers for the functional assay of P-gp activity (Holló et al., 1994), with this being sensitive to verapamil. In the present study, the rate of accumulation of calcein and its enhancement by verapamil was determined in confluent cell monolayers at considerably low concentrations of the fluorophore (0.5 µM). Furthermore, taking advantage of microscope fluorescence techniques, we were also able to demonstrate in single LLC-PK₁ cells the rate of P-gp activity, as a function of calcein accumulation after the addition of verapamil. The finding that calcein accumulation in these particular experiments reached a plateau within 10 min suggests that this transporter may play an important role in this particular cell line.

The finding that short-term (30 min) exposure to UIC2 and verapamil reduced L-dopa apical extrusion and increased L-dopa accumulation in LLC-PK₁ cells strongly suggests that L-dopa in these cells is extruded through P-gp. On the other hand, the observation that short-term (30 min) exposure of LLC-GA5 Col300 cells to UIC2 and verapamil failed to affect the accumulation and apical extrusion of L-dopa may call into question the suggestion that L-dopa transfer at the apical border is promoted by P-gp. To answer this question, one also must consider the results obtained for the effect of verapamil on the rate of accumulation of calcein. Despite the fact that LLC-GA5 Col300 cells were found to accumulate calcein at a low rate, suggesting the presence of high levels of P-gp, the short-term (30 min) exposure to verapamil failed to increase the rate of calcein accumulation. This contrasts with the results obtained in LLC-PK₁ cells in which verapamil enhanced by 6-fold the rate of accumulation of calcein. Because LLC-GA5 Col300 cells extrude more calcein (low rate of calcein accumulation) and verapamil was less effective in enhancing the accumulation of calcein, it was hypothesized that these effects could be related to overexpression of P-gp. It was decided, therefore, to increase the incubation period with UIC2 and verapamil from 30 min to 3 h, with the aim of enhancing the inhibition of P-gp. This strategy was found to be successful, and the long-term exposure of LLC-GA5 Col300 cells to UIC2 or verapamil produced a marked increase in the accumulation of both L-dopa and calcein and drastically reduced the apical extrusion of L-dopa. These results agree with the view that both cell lines express P-gp, and differences in sensitivity to UIC2 and verapamil may be explained by the fact that LLC-PK₁ cells exhibit lower levels of P-gp activity than LLC-GA5 Col300 cells. This fits well with the finding that basal-to-apical flux of L-dopa in LLC-PK₁ cells is more sensitive to inhibition by other P-gp inhibitors (vinblastine, Rh123, and quinide) than LLC-Col300 cells. However, transporters other than P-gp may be involved in the outward transfer of L-dopa. In fact, drugs used to interfere with P-gp activity are not selective for this multispecific transporter and may interact with other transporters rather than P-gp for the outward transfer of L-dopa. On the other hand, the effect of these drugs (especially in LLC-PK₁ cells) was far from being a complete inhibition of basal-to-apical flux of L-dopa.

The long-term exposure of LLC-PK₁ cells to UIC2 and verapamil resulted in a less pronounced increase in cell accumulation of L-dopa and less pronounced reduction in the apical extrusion of L-dopa compared with the short-term exposure. One possible explanation for this result might be the presence of an increased number of P-gp units after prolonged exposure to UIC2 and verapamil, as a rebound phenomenon. We have no direct evidence to support this hypothesis, except the observation that prolonged exposure to verapamil results in a less pronounced increase in calcein accumulation compared with that observed after a short-term exposure to verapamil. In fact, the rate of calcein accumulation after short-term exposure to verapamil was twice that observed after a long-term exposure to verapamil (see Table 1). Other evidence favoring this view is the finding that chronic exposure to CsA (another inhibitor of P-gp) stimulated P-gp activity, as a mechanism of detoxification (Garcia del Moral et al., 1995; Soares-da-Silva et al., 1998b).

The findings described here in cultured renal cells suggest that P-gp may act as an extrusive mechanism for intracellular L-dopa and that as a result of its activation, it would reduce the availability of L-dopa to be decarboxylated to dopamine. This has important implications for the understanding of effects of drugs that interfere with P-gp activity in the kidney, such as CsA and several anticancer drugs, and the role of renal protective effects of dopamine. Favoring this view is the following evidence. The treatment of rats with the P-gp inhibitor CsA increases the accumulation of L-dopa in isolated renal tubules and enhances the synthesis of dopamine in kidney slices loaded with L-dopa, without affecting AADC activity (Pestana et al., 1995). In addition, the short-term exposure of LLC-PK₁ cells to CsA results in a decrease in the apical outward transfer of intracellular L-dopa. Taken together, these results are in agreement with the view that while inhibiting P-gp, CsA decreases the extrusion of intracellular L-dopa and facilitates its decarboxylation to dopamine. Then it is possible that the increased formation of dopamine in CsA-treated rats, resulting from the inhibition of P-gp, may counteract the indirect antinatriuretic effects of CsA (Sturrock and Struthers, 1994; Pestana et al., 1995). This is in agreement with our recent observation of a gradual increase in the daily urinary excretion of DOPAC, the deaminated metabolite of dopamine, in human kidney transplant recipients who underwent therapy with CsA (Pestana et al., 1997).

In conclusion, the data presented here show that LLC-PK₁ cells are endowed with P-gp and that the outward transfer of L-dopa at the apical cell border in both LLC-PK₁ and LLC-GA5 Col300 cells is in part promoted through this transporter.