The Multidrug Resistance Protein 1: A Functionally Important Activation Marker for Murine Th1 Cells¹

Murine CD4⁺ T lymphocytes can be subdivided mainly into two phenotypically distinct subpopulations termed Th1 and Th2 cells (1, 2). The two subpopulations differ with respect to the cytokine pattern they produce: Th1 cells secrete IL-2, IFN-γ, and TNF-β, whereas the Th2 cytokine repertoire consists of IL-4, IL-5, IL-6, IL-9, IL-10, and IL-13 (3). Th1 and Th2 cells act as regulators and activators of the immune response and trigger the immune system toward a macrophage-dominated type 1 or a B cell-dominated type 2 response, respectively. Several in vivo models have been generated for studying the differences between Th1- and Th2-mediated immunity. One of them, experimental cutaneous murine leishmaniasis, caused by s.c. infection with the protozoan parasite Leishmania major, leads to a Th1 response in resistant C57BL/6 mice and to a Th2 response in susceptible BALB/c mice defined by the respective cytokine profile (4, 5, 6, 7).

Such in vivo models have prompted the search for specific surface molecules which would facilitate identification of the two CD4⁺ T cell subsets. In this regard, it was recently shown that the IL-18 receptor is selectively expressed on the surface of Th1 cells (8). On the other hand, Th2 cells could be detected with the help of the cell surface molecule ST2L (9) as well as by high expression of CD30 (10). Similarly, chemokine receptors were shown to be useful markers for the discrimination of T cell subsets. For instance, the chemokine receptors CXCR3 and CCR5 were reported to be mainly expressed on the surface of Th1 cells, although this finding was only partially confirmed by others (11, 12, 13). Abs to the eotaxin receptor CCR3 were used for isolation of Th2, but not Th1 cells from human peripheral blood (14). The highest preference for Th2 lymphocytes was shown by the chemokine receptor CCR4 (11, 13).

Recently, we have described (15) another candidate for Th1 and Th2 discrimination, namely, a transmembrane transporter molecule. The protein exports the fluorescent anion Fluo-3 out of the cytoplasm of T cells (16) and is differentially expressed in resting Th1 and Th2 cells: resting Th2 cells transport Fluo-3 whereas Th1 cells are unable to do so. Stimulation of Th1 cells with Ag and APCs results in an up-regulation of the anion transporter comparable to resting Th2 cells. Northern blotting and dye extrusion inhibition studies in resting Th1 and Th2 cells raised the possibility that this pump might be murine multidrug resistance protein 1 (mrp1),³ the ortholog of the human MRP1. Indeed, in one Th1 cell clone, mrp1 protein was up-regulated after antigenic restimulation like the Fluo-3 pump, as revealed by immunoblot analysis.

MRP1 is a member of the family of ATP-binding cassette transporters that has been shown to mediate multidrug resistance in tumor cells. It is also expressed in normal tissues, including several types of epithelia, muscle cells, and macrophages (17, 18). Low levels of MRP mRNA are expressed in normal peripheral blood CD4⁺ T cells (19).

In the present investigation, we further studied the relation of mrp1 and the Fluo-3 pump and their expression in CD4⁺ T cells. We directly show that mrp1 exports Fluo-3, the cleaved intracellular fluorescent anion of the acetomethyl ester fluo-3-acetoxymethyl ester (Fluo-3-AM). In addition, we demonstrate at the single-cell level that not only antigenic activation, but also cytokine stimulation, enhances Fluo-3 extrusion as well as mrp1 expression in Th1 cells. Finally, we present evidence that mrp1 activity is a prerequisite for activation of Th1 cells by Ag or cytokines.

Fluo-3-AM, Triton-X-100, paraformaldehyde, sodium citrate, FCS, Staphylococcus aureus enterotoxin B (SEB), and saponin were purchased from Sigma (Deisenhofen, Germany). Cyclosporin A (CsA) was purchased from Calbiochem (La Jolla, CA). MK571 was purchased from Biomol (Hamburg, Germany). PSC 833 was purchased from Novartis (Nuremberg, Germany). The cytokines TNF-α, IL-2, and IL-12 were purchased from PharMingen (Hamburg, Germany). For ELISA studies, commercial Ab pairs from PharMingen were used.

Female BALB/c mice at the age of 4–6 wk were purchased from Charles River Breeding Laboratories (Sulzfeld, Germany). They were injected with 10 μg of SEB in 50 μl of PBS into the hind footpad. Control mice were injected with 50 μl of PBS only. After 12 h, the draining popliteal lymph nodes were removed and a single cell suspension was prepared for flow cytometry.

The Th1 cell clones used were B10BI derived from a C57BL/6 mouse and specific for bovine insulin, and LNC-2 derived from a BALB/c mouse and specific for Mycobacterium tuberculosis-purified protein derivative (PPD) (20, 21). The Th2 cell clone used was L1/1, derived from a BALB/c mouse, with specificity for LmAg, a L. major Ag preparation (20). The in vitro propagation of these T cell clones has been described (21). Briefly, 5 × 10⁵ cells/well were restimulated every 4 wk in 12-well tissue culture plates (Costar, Cambridge, MA) with syngeneic irradiated (25 Gy) spleen cells (5 × 10⁶/well) and Ag in a total volume of 3 ml of Clicks’ RPMI 1640 medium (Biochrom, Berlin, Germany) supplement as described (22). Forty-eight hours after restimulation, the contents of the wells were divided into two, and culture supernatant (SN, 4%) of the cell line X63Ag8-653/IL-2 (23) was added as a source of recombinant murine (rm) IL-2. Thereafter, rmIL-2 was added once per week. Cells were routinely used for the experiments 4 wk after their final antigenic restimulation and 2 wk after the final addition of rmIL-2.

For in vitro restimulation, 5 × 10⁵ T cells were cultured in 24-well plates (Nunc, Wiesbaden, Germany) either in the presence of Ag and 2 × 10⁶ syngeneic spleen cells or in the presence of rm cytokines, as detailed in the legends of Figs. 1–9. After 48 h, the T cells were harvested and processed for the analysis of Fluo-3 extrusion and mrp1 expression. Cytokine production in the SNs of the stimulated T cells was determined in triplicate using commercial ELISA kits (PharMingen).

For the labeling of T cells with Fluo-3-AM and secretion of the Fluo-3 anion, a procedure was used that is described in detail elsewhere (16). However, in the present study dye extrusion during labeling of cells with Fluo-3-AM in HBSS containing 5% FCS was inhibited by the presence of CsA (25 μmol) instead of probenecid. After washing in HBSS/5% FCS, the cells were incubated for 10 min in the presence of CsA at 37°C to allow for complete export of the fluorogenic anion Fluo-3 after cleavage cytoplasmic esterases. Thereafter, CsA was washed out of the medium, and the cells were allowed to export Fluo-3 for 60 min at 37°C. Fluo-3 accumulation in the cells was analyzed using a FACSCalibur (Becton Dickinson, Heidelberg, Germany) and the CellQuest software (Becton Dickson). For determination of Fluo-3 extrusion in mrp1-transfected HEK 293 cells, a simplified procedure was followed. Thus, cells were labeled for 60 min with Fluo-3-AM in the absence of CsA to permit simultaneous uptake of Fluo-3-AM and secretion of Fluo-3. Steady-state Fluo-3 levels were determined immediately afterward.

For detection of an intracellular epitope of murine mrp1 at the single-cell level, 1 × 10⁶ cells were washed twice with PBS and incubated in PBS containing 2% paraformaldehyde at room temperature for 20 min. Cells were then permeabilized using 0.1% Triton-X-100 and 0.1% sodium citrate in H₂O at 37°C for 10 min. After washing, the cells were incubated with mAb MRPr1 (1 μg/ml) (24) in PBS containing 0.5% saponin and 2% FCS. In parallel, cells were labeled with a rat IgG2a isotype control Ab (PharMingen). Binding of mAb was revealed with PE-conjugated goat anti-rat (10 μg/ml; Jackson ImmunoResearch, West Grove, PA) or indodicarbocyanine-conjugated donkey anti-rat Ig (1 μg/ml; Jackson ImmunoResearch). Staining intensity of mrp1 was measured by flow cytometry.

In lymph node cells from SEB-injected mice, mrp1 was similarly labeled, and, in addition, the cells were stained with anti-CD4 mAb conjugated to Cy-Chrome (2 μg/ml; PharMingen) and either anti-Vβ8 mAb (10 μg/ml; PharMingen) or anti-Vβ4 mAb (1:50; BioSource International, Camarillo, CA), both conjugated to FITC.

For detection of the surface Ag CD69 after antigenic or cytokine stimulation, T cells were incubated with biotinylated rat anti-mouse CD69 mAb (3 μg/ml; PharMingen) in PBS containing 1% FCS at 4°C for 10 min. Binding of mAb was visualized by streptavidin conjugated to PE (1 μg/ml; PharMingen). Whenever T cells were stimulated with Ag and APCs, flow cytometry analysis allowed for a clear distinction and gating of the respective T cell population from the APCs on the basis of forward-site scatter characteristics. All data shown are gated on the T cell population.

For histochemistry, T cells were prepared and labeled as described above for mrp1 detection by flow cytometry. In an additional step, the fixed and labeled T cells were spun onto glass slides using the Cytospin 2 (Shandon, Frankfurt, Germany). Thereafter, the T cells were scanned and analyzed using the Leica TCS confocal microscope system (Leica, Heidelberg, Germany).

HEK 293 cells (25) were transfected with the episomal expression vector pcEBV7 (26, 27) containing murine mrp1 or human MRP1 cDNA. Control HEK 293 cells were transfected with the parental vector and grown under identical conditions, which are described in detail elsewhere (28).

Previous analysis had established the existence of a protein pump in T cells that extrudes the fluorescent anionic dye Fluo-3 (16). Dye extrusion inhibition studies raised the possibility that the Fluo-3 anion pump was mrp1 (15). In experiments described here, we attempted to formally prove that mrp1 transports the Fluo-3 anion using HEK 293 cells transfected with the episomal expression vector pcEBV7 containing either the human MRP1 or the murine mrp1 cDNA (28). The cells were incubated with the nonfluorescent hydrophobic acetomethyl ester Fluo-3-AM, and accumulation of the cleaved fluorescent anion Fluo-3 was allowed to reach steady state. After 60 min, the Fluo-3 staining intensity indicating the amount of intracellular Fluo-3 anion was analyzed (Fig. 1). Cells transfected with the human or murine MRP1 cDNA were considerably less stained, demonstrating enhanced Fluo-3 transport activity, as compared with control transfected cells. Thus, this experiment demonstrates that Fluo-3 is a substrate for murine mrp1 and human MRP1 and can be used to measure transport function of these proteins.

In our previous studies (15), Fluo-3-extruding activity was found to be enhanced in Th1 cells after in vitro restimulation with Ag and APCs. In this study, we show that, in addition, Fluo-3 extrusion is up-regulated in Th1 cells after stimulation by cytokines. Th1 cells of the two clones B10BI and LNC-2 were incubated with different concentrations of IL-2. After 48 h, T cells were harvested and the Fluo-3 extrusion capacity was tested (Fig. 2). Comparison of the degree of Fluo-3 staining intensity immediately after labeling in the presence of CsA (Fig. 2, shaded histograms) and after 60 min of incubation in the absence of CsA (Fig. 2, open histograms) shows that IL-2 dose-dependently up-regulated the Fluo-3-extruding activity. Dye extrusion could be inhibited by further incubation in CsA (data not shown). The data obtained for the two clones were comparable and are given for only one of them.

IL-12 and TNF-α, which by themselves were almost unable to induce Fluo-3 extrusion, were very potent Fluo-3 pump activators when combined with a limiting concentration of IL-2 (Fig. 3). In these experiments, the combination of rmIL-12 (10 ng/ml; Fig. 3, left) or of rmTNF-α (10 ng/ml; Fig. 3, right) with a low concentration of rmIL-2 (0.06 ng/ml) induced the Fluo-3 pump activity to the same level as the optimal amount of rmIL-2 (10 ng/ml) alone. Similar data were obtained for the Th1 cell clones B10BI (Fig. 3) and LNC-2 (data not shown). In contrast, the cytokines IFN-α, IFN-γ, GM-CSF, IL-1β, IL-3, IL-4, IL-6, IL-9, IL-10, and IL-15 were unable to induce any Fluo-3 pump activity (data not shown).

Next, we analyzed, whether the up-regulation of Fluo-3 extrusion activity, was accompanied by an increase in mrp1 protein at the single-cell level. Resting Th1 and Th2 cells were stimulated with optimal amounts of rmIL-2 (10 ng/ml) for 48 h (Fig. 4,A). Thereafter, the clones were processed for fluorescence staining of an intracellular mrp1 epitope with mAb MRPr1. Resting Th1 cells displayed a weak mrp1 signal compared with resting Th2 cells, confirming our earlier observations (15). After cytokine stimulation, mrp1 expression was unchanged in Th2 cells, but was elevated in Th1 cells to almost the same level as in Th2 cells. Moreover, mrp1 expression in Th1 cells was up-regulated by IL-2 dose dependently (Fig. 4,B), confirming the above reported effects of IL-2 on mrp1 function (Fig. 2).

The ability to detect MRP at the single-cell level permitted the localization of the mrp1 protein in the T cells by confocal laser microscopy. Again, mrp1 could be detected only in resting Th2 cells and IL-2-activated Th1 cells but not in resting Th1 cells. In all cells expressing mrp1, i.e., in resting Th2 cells and activated Th1 cells, mrp1 was localized almost exclusively at the plasma membrane (Fig. 5). No staining was observed with an isotype control mAb (Fig. 5).

So far, our studies have demonstrated that mrp1 is an activation marker for long-term in vitro cultured Th1 cell clones. To verify that our findings also apply for CD4⁺ T cells in vivo, we injected the superantigen SEB into the hind footpad of BALB/c mice. SEB activates T cells expressing the Vβ8 family of the TCR (29, 30). This experimental setup allowed us to study mrp1 expression in TCR-triggered Vβ8⁺ T cells compared with Vβ4⁺ T cells, as a negative control. Twelve hours after injection, the popliteal lymph node cells from SEB-injected and from vehicle-injected control mice were processed for a three-color staining of mrp1, CD4, and Vβ4 or Vβ8. The panels in Fig. 6 depict lymphocytes gated for CD4⁺ expression. Control mice (Fig. 6, A and B) demonstrated almost no mrp1 expression in the Vβ4- or Vβ8-positive cells. In SEB-injected mice, mrp1 expression in the irrelevant Vβ4⁺ TCR subset was comparable to that of control mice (Fig. 6,C, upper right quadrant), whereas the relevant Vβ8⁺ TCR subset was almost entirely mrp1 positive (Fig. 6,D, upper right quadrant). The negative result of the isotype control labeling in the SEB-injected mice (Fig. 6, E and F) confirms the specificity of the mrp1 detection by mAb MRPr1. In a normal mouse, the Vβ8⁺ TCR subset consists mostly of naive T cells; consequently, our findings demonstrate the in vivo reliability of mrp1 as a marker for T lymphocytes that have been triggered for the first time and only for 12 h via their TCR.

In another set of experiments, we examined the physiological relevance of mrp1 in the activation process of Th1 and Th2 cells. Resting Th1 cells were stimulated for 48 h with rmIL-2 (Fig. 7, A–C) or Ag and APCs (Fig. 7, D–F) in the presence or absence of MK571, an inhibitor of MRP (31). A shift in the forward scatter confirmed the activation of the Th1 cells by both stimuli. Importantly, this shift was reduced to the value of resting Th1 cells when MK571 (75 μM) was added (Fig. 7, A and D). Along with blastogenesis, expression of the activation marker CD69 (32) was increased by IL-2 or antigenic stimulation. Like blastogenesis, CD69 expression was markedly blocked by simultaneously added MK571 (75 μM) (Fig. 7, B and E). As a third parameter for Th1 cell activation, the production of IFN-γ was analyzed. In the absence of MK571, both Th1 clones produced IFN-γ in response to IL-2 (Fig. 7,C) or Ag plus APCs (Fig. 7,F). In the presence of MK571, the secretion of IFN-γ was blocked. Similar results were obtained when the Th1 cells were cocultured with PSC 833, a different MRP inhibitor (15), instead of MK571 (data not shown). In addition, the effect of MK571 was not restricted to Th1 cells in that the Ag-induced secretion of IL-4 by Th2 cells was also distinctly blocked by simultaneous incubation with MK571 (Fig. 8,B). In a direct comparison, the MK571 effect on Th2 cells tended to be smaller than the one on Th1 cells (Fig. 8, A and B).

In control experiments, toxic effects of MK571 on activated Th cells could be excluded (Fig. 9). To show this, resting Th1 cells were stimulated with rmIL-2 for 24 h, leading to a shift in CD69 expression (Fig. 9,A). Thereafter, the cells were washed and further incubated either in medium only or in medium containing the MRP-specific inhibitor MK571. After an additional 6- or 12-h incubation, CD69 expression had not changed in the MK571 group compared with the control group (Fig. 9 B). Likewise, IFN-γ production was not reduced after 6 or 12 h in the culture SNs containing MK571 compared with the control samples.

In conclusion, inhibition of MRP limits the Th1 cell response to activation via two different stimuli such as cytokines or Ag plus APCs. To be effective, however, inhibition has to occur during the initiation of T cell activation and not once the T cells are already activated.

In this study, we demonstrate that the fluorescent anion Fluo-3 can be transported by the human MRP1, a known mediator of multidrug resistance in tumor cells and its murine ortholog mrp1. Although this is not unexpected given the known substrate specificity of MRP1/mrp1, this is the first direct demonstration that Fluo-3 can be used to monitor MRP1/mrp1 activity. Identification of MRP1/mrp1-positive cells by Fluo-3-labeling instead of Ab-labeling has the significant advantage that the cells to be tested remain viable, can be isolated using flow cytometry, and can be used for further functional tests or long-term in vitro cultures. So far, the only mAbs available for MRP1/mrp1 detection recognize intracellular epitopes and therefore require fixation and subsequent permeabilization of the cells (24, 33).

We have recently hypothesized that mrp1 is present and functional in nontumorogenic, cloned murine Th cells (15). This hypothesis was based on dye extrusion inhibition studies as well as on a strong correlation between the presence of the Fluo-3 pump and the expression of the mrp1 gene. The data presented here based on studies with MRP1/mrp1-transfected cells corroborate our original hypothesis. We have also previously shown that the Fluo-3 pump was differentially expressed in Th cells in that resting Th2 cells, but not resting Th1 cells, exported the Fluo-3 anion. After restimulation with Ag and APCs, Th1 cells acquired the potential to extrude Fluo-3 to the same extent as resting Th2 cells. This was revealed at the level of Th1 and Th2 cell clones in vitro as well as at the level of freshly purified Th1 and Th2 cells ex vivo. There was no change of the Fluo-3 pump activity in Th2 cells after antigenic restimulation compared with the resting state.

We now report that in addition to antigenic restimulation, several cytokines are able to up-regulate the expression of the Fluo-3 pump in Th1 cells. IL-2 seems to be the most powerful cytokine in this regard. IL-12 and TNF-α were substantially less effective than IL-2, but both cytokines strongly synergized with a limiting concentration of IL-2. It should be mentioned that we were unable to influence the Ag-triggered induction of the Fluo-3 pump with a neutralizing anti-IL-2 Ab (data not shown). We are currently pursuing the hypothesis that a TCR signal may up-regulate the Fluo-3 pump independently of cytokines.

Again, not only the Fluo-3 pump activity was up-regulated after IL-2 stimulation, but also mrp1 protein expression, as revealed by FACS analysis at the single-cell level. Therefore, Fluo-3 pump up-regulation correlated with up-regulation of mrp1 protein, providing further evidence that the Fluo-3 pump is mrp1. Together, our studies clearly implicate mrp1 as an activation marker for Th1 cells in vitro. In addition, mrp1 is constitutively expressed in Th2 cells, at least at the level of T cell clones.

The present study also demonstrates that mrp1 is an activation marker for Th cells in vivo. We had already suspected that this was the case on the basis of the Fluo-3 pump levels expressed during experimental murine leishmaniasis (15). Almost all L. major Ag-specific CD4⁺ T cells were detected within the Fluo-3-extruding cell subset. Here, we extend these earlier findings and report that indeed mrp1 protein is specifically induced in CD4⁺ T cells in vivo after TCR stimulation. Unfortunately, permeabilization of the cells needed for detection of intracellular mrp1 epitopes leads to diffusion of the Fluo-3 dye. This prevented double staining of mrp1 protein and the activity of the Fluo-3 pump in L. major Ag-specific T cells and prevented the use of the Leishmania model, in which it is impossible to directly identify the L. major-specific CD4⁺ T cells. Therefore, we used the model Ag SEB, which binds to MHC II molecules, and in this state directly interacts with any TCR containing a β-chain of the Vβ8 family. Therefore, all Vβ8⁺ T cells, which can be identified by Ab staining, should be recognized and activated by SEB (29). We found that already 12 h after SEB injection into mice, only the relevant TCR Vβ8 subset, but not the Vβ4 subset used as a control, contained elevated levels of the mrp1 protein.

Taken together, we propose mrp1 as a new activation marker for murine T cells, not only in long-term in vitro cultures but also in freshly ex vivo-isolated murine T cells. To our knowledge, this is the first report that characterizes mrp1 as an inducible molecule in untransformed cells rather than tumor cells. So far it was only known that low levels of MRP mRNA are present in normal peripheral blood CD4⁺ T cells (19).

The role of mrp1 in the activation process of Th cells is not known. MRP knockout mice had increased levels of glutathione in blood mononuclear leukocytes as well as in bone marrow cells and erythrocytes, indicating that mrp1 might pump this reducing tripeptide (34). Glutathione conjugates such as leukotriene C₄ (LTC₄) have been demonstrated to be transported by MRP1 in vitro (31, 35, 36). LTC₄ is involved in a number of signal transduction pathways that control vascular permeability and smooth muscle contraction (37). Not surprisingly, therefore, mrp1-deficient mice displayed an impaired response to an inflammatory stimulus as a result of their defective LTC₄ secretion (38). In vitro studies have also shown that oxidized glutathione acts as a substrate for MRP (39). Consequently, it has been suggested that MRP may be involved in protection against oxidative stress and maintenance of the intracellular redox potential (40). In T cells, IL-2 initiates proliferation which results in enhanced metabolic activity, accumulation of oxygen intermediates, and changes in the intracellular redox potential. Our results raise the possibility that one of the responses to this change in redox state is up-regulation of mrp1 expression. On the other hand, IL-12 triggers Th1 differentiation (41). Therefore, the combination of IL-2 and IL-12 triggers two different signal transduction pathways for proliferation and differentiation and leads to strongly enhanced metabolism. This could explain the necessity for synergistic expression of mrp1 in Th1 cells stimulated by IL-2 plus IL-12. TNF-α, which we have shown to synergize with IL-2, has been reported previously to increase MRP expression at the level of mRNA and protein in a colon carcinoma cell line (42).

Alternatively, it is conceivable that mrp1 is directly involved in cytokine secretion. It is known that P-glycoprotein, another member of the multidrug resistance group of proteins, is involved in the transport of IL-2 in PHA-stimulated human peripheral blood leukocytes (43). However, this transporter is not expressed in the herein used murine Th1 clones (15) and its role may be filled by mrp1.

Whatever the role of mrp1 in murine T cells may be, it has to account for the fact that this molecule is constitutively expressed in Th2 cells (15). The significance of this finding for the physiology of Th2 cells is completely unknown and under investigation. Possibly, Th2 cells have an increased level of cell metabolism even in the resting state, which requires that the detoxification machinery is continuously present.

One would expect that for mrp1 to be functional in protection against oxidative stress or in cytokine secretion, the protein should be localized in the plasma membrane. Using confocal laser microscopy, we have confirmed that this is the case in both resting Th2 as well as stimulated Th1 cells. Previous studies examined MRP localization in MRP-overexpressing tumor cells and MRP gene-transfected cells and came to similar conclusions (44, 45).

The idea of MRP as a quick detoxifier of the cell may also explain a discrepant result for Fluo-3 extrusion and mrp1 protein expression in resting Th1 cells. Here, a significant baseline expression of mrp1 protein was detectable by fluorescence staining, whereas no Fluo-3 extrusion occurred. At present, we are unable to exclude that this baseline staining reflects nonspecific binding of the mAb MRPr1. Alternatively, however, the reason for this discrepancy may reside in the fact that the MRP molecule has two nucleotide binding sites (46), both of which have to be occupied by ATP molecules for functionality. Thus, inactive mrp1 may be located at the membrane (and detectable by Ab staining) to keep the cell alert for a critical situation when quick detoxification or cytokine secretion is needed. Such a scenario would enable the cell to react much faster toward changes in the micromilieu, as it could, if de novo mrp1 synthesis was required.

In an attempt to determine whether the elevated expression of mrp1 in activation of Th1 cells has a functional role in the activation process, we stimulated Th1 cells in the presence of the MRP-specific inhibitor MK571 (31). Blastogenesis as well as CD69 (32) expression was reduced in Th1 cells coincubated with MK571. Thus, blocking the mrp1 transport during the initiation of Th1 activation appears to interfere with completion of the activation process. Similarly, IFN-γ production was inhibited by MK571 to the level of resting cells. Similar results were obtained when MRP was blocked by PSC 833 instead of MK571, indicating that the effects of MK571 were not restricted to this compound. In Th2 cells, MK571 interfered with the Ag-induced secretion of IL-4. This set of data indicates that mrp1 transport is essential not only for Th1, but also for Th2 cell-mediated immunity.

MK571 works at the inner leaflet of the plasma membrane, and therefore published data on inhibition of MRP by MK571 used inside-out vesicles and relatively low amounts of the inhibitor (31). In our studies, we tested the effects of MK571 on viable T cells, which required higher concentrations of the inhibitor. To exclude artificial results caused by a putative toxicity of MK571 at the concentrations used, we incubated already activated T cells with MK571 for an additional 12 h. No changes in IFN-γ production, blastogenesis (data not shown) and CD69 expression were observed compared with the non-MK571-treated control. Therefore, a toxic effect of MK571 on T cell function could be excluded. Thus, these results clearly demonstrate that MRP inhibition only influences T cell function when it occurs along with the initiation of T cell activation, but not once T cells are already activated.

Taken together, these data demonstrate that mrp1 function is necessary for efficient Th1 cell activation, suggesting that the protein either secretes molecules such as cytokines that are involved in activation or effluxes intracellular inhibitors of the process. Further studies will try to identify intracellular signaling pathways that are blocked by MK571 and may shed light on the exact molecular nature of MRP function in this context.

We thank Susanne Bischof for excellent and patient assistance throughout these studies and Derek Schulze for FACS analysis of Fluo-3 accumulation in the transfected cells. We also thank Dr. Keppler (Deutsches Krebsforschungszentrum, Heidelberg, Germany) for helpful discussions.