The plasma membrane transport protein P-glycoprotein (P-gp) is expressed by subsets of both CD4+ and CD8+ T cells in mice. The proportion of T cells that express P-gp goes up with age, and the P-gp-expressing subset of the CD4 memory population is hyporesponsive in many in vitro assays. The significance of P-gp expression for T cell function has not been well established, although several reports have suggested that it may promote cytokine export and/or cytotoxic T cell function. To elucidate which T cell functions may require P-gp, we have compared a variety of responses using T cells from wt and P-gp knockout mice. Protein expression and rhodamine-123 efflux studies revealed that peripheral T cells exclusively utilize the mdr1a-encoded isoform rather than the homologous mdr1b or mdr2 isoforms. Comparisons of T cells from mdr1a+/+ and mdr1a−/− mice showed no differences in proliferation or in secretion of IL-2, IL-4, IL-5, IL-10, or IFN-γ in response to polyclonal stimulation. Moreover, mdr1a−/− T cells produced strong allospecific cytotoxic responses comparable to those of wt T cells. Our results show that P-gp is not a necessary component of peripheral T cell functional responses. Further investigation will be needed to determine the significance of P-gp expression in T lymphocytes.

The P-glycoprotein (P-gp)3 family of transporters consists of large integral membrane proteins that couple the hydrolysis of ATP to the extrusion of a broad range of substrates from the cytoplasm to the extracellular environment (1). Since their discovery in drug-resistant tumor cells (2), P-gps have been studied primarily with respect to their ability to facilitate multidrug resistance (mdr) in cancer cells following chemotherapy (3). However, it has been well documented that many diverse cell types ranging from hemopoietic stem cells and T lymphocytes to secretory cells in the liver, kidney, and adrenals also normally express P-gp (4, 5, 6, 7). In addition to cellular detoxification, a number of functional properties have been attributed to P-gp, including transport of steroid hormones (8), cell volume regulation by chloride transport (9), and peptide extrusion (10). Despite these provocative findings, the functional significance of P-gp expression in normal, untransformed cells, including T lymphocytes, is still uncertain.

Mammalian P-gps belong to the ATP-binding cassette superfamily of transporters (11) and can be divided into two subgroups based upon their substrate specificity. One group includes isoforms that can transport hydrophobic drugs and that are associated with mdr (1); these are encoded by the human MDR1 (12), murine mdr1a and mdr1b (13, 14), and hamster pgp1 and pgp2 genes (15). The human MDR1 and mouse mdr1 isoforms, although highly homologous, can be further distinguished based on distinct differences in their substrate preferences and sensitivity to inhibitors (14, 16). The other group includes isoforms that primarily transport phospholipids and that appear to be necessary for normal hepatocyte secretion of phosphatidylcholine into the bile (17, 18, 19); these are encoded by the human MDR3 (20) and murine mdr2 genes (21).

P-gp is expressed on subsets of CD4+ and CD8+ peripheral T cells in both humans and mice (7, 22), and several groups have shown that the proportion of P-gphigh T cells increases with age (23, 24, 25). There is substantial evidence that P-gphigh T cells, isolated from the CD4 memory pools of young or old mice, respond less well than cells of the complementary P-gplow subset in tests for proliferation and cytokine secretion (26, 27), although the published literature includes some conflicting data on this point (28). Presently, there are also a few reports that suggest that the human MDR1-encoded isoform of P-gp is involved in CD8+ T cell cytotoxic effector function (29) and in the release of certain cytokines from PBL (30, 31). In both cases, these proposed functions were inferred from experiments where inhibition of P-gp transporter function, with either anti-P-gp Abs or drugs (chemosensitizers), led to an observed impairment in either cytotoxic activity or release of IL-2, IL-4, and IFN-γ. However, impairment of cytotoxic activity required the use of high concentrations of an anti-P-gp Ab to produce only minor reductions in CTL activity (29). Moreover, concentrations and classes of chemosensitizers that have other cellular effects unrelated to P-gp inhibition were used for examining cytokine release from human PBL, and the presence of active P-gp in those T cells actually releasing IL-2, IL-4, and IFN-γ was not demonstrated (30, 31). Taking these factors into consideration, it is evident that more convincing evidence is needed to establish the identity of cellular functions that actually require direct involvement of P-gp.

In the current study, we have sought to determine whether proliferation, cytokine secretion, or generation of cytotoxicity requires expression of P-gp in T cells from mice. To avoid the ambiguities associated with drug or Ab inhibition of P-gp, we have characterized the responses of T cells from mice that are homozygous for gene disruptions in specific mdr genes, comparing them to T cells from littermates with wild-type (wt) mdr genes. We report that mouse peripheral T cells exclusively utilize the mdr1a-encoded isoform of P-gp and that an absence of mdr1a does not significantly alter the proliferative or cytokine responses of either CD4+ or CD8+ T cells to polyclonal activation. Similarly, we found that T cells from mdr1a knockout (KO) mice are capable of strong allospecific cytotoxic responses equivalent to those of wt T cells. In contrast to the results of studies using drugs or Abs to block P-gp function, our data show that P-gp is not required by peripheral T cells to perform any of the investigated cellular functions.

Strains.

Eight-wk-old and retired breeder (>5 mo old) FVB, FVB/mdr1a KO, and FVB/mdr1a/b KO male mice were purchased from Taconic (Germantown, NY). Retired breeder FVB/mdr2 KO male mice and 8-wk-old male BALB/c and female C57BL/6 mice were purchased from The Jackson Laboratory (Bar Harbor, ME). All mice were housed and maintained under specific pathogen-free conditions at the University of Michigan for at least 1 wk before use.

Breeding and genotyping.

Following an initial C57BL/6 × FVB/mdr1a KO cross, B6 × FVB F1 mice were mated to produce a heterogeneous F2 population. Using genomic DNA isolated from tail snips, each F2 mouse was genotyped at the mdr1a locus by Southern blot (32). The mdr1a gene probe was a generous gift from Dr. Alfred H. Schinkel (The Netherlands Cancer Institute, The Netherlands). The homozygous mdr1a+/+ and mdr1a−/− F2 male mice were retained and allowed to reach an age (typically 7–8 mo) at which substantial numbers of P-gphigh T cells are typically seen in wt mice before being used for experimentation.

Splenic T cells were prepared by negative selection on anti-Ig-coated plates as previously described (26). In some experiments, T cells were further separated into CD4+ and CD8+ subsets by immunomagnetic depletion using a combination of either anti-CD8 (53-6.7) or anti-CD4 (L3T4) ascites, respectively, followed by goat anti-rat Ig coupled to magnetic beads (PerSeptive Biosystems, Framingham, MA). Average purity of subset-depleted CD3+ T cells was >95% for the desired population (CD4+ or CD8+) as determined by flow cytometry.

T cells (1 × 106/sample) were incubated with 1 μM rhodamine-123 (R-123) (Molecular Probes, Eugene, OR) for 10 min, washed three times, and then allowed to extrude the dye for 30 min at 37°C in the absence or presence of 10 μg/ml of the P-gp inhibitor verapamil (Sigma, St. Louis, MO) as previously described (23). Cells were counterstained with either anti-CD4-PE, anti-CD8-PE, or isotype control-PE Abs from PharMingen (San Diego, CA). During the entire procedure, cells were protected from light and, with the exception of the R-123 extrusion step, kept at 4°C until analyzed on a FACScan with Lysis II software (Becton Dickinson, Mountain View, CA). Nonviable lymphocytes were excluded from analysis using a light scatter gate established with propidium iodide-stained control samples. Fluorescence values from 20,000 gated events were recorded for each sample. R-123 histogram profiles from CD4+ or CD8+ gated T cells were generated and analyzed with WinMDI software (Joseph Trotter, Scripps Research Institute, La Jolla, CA).

Crude membrane fractions were prepared from T cells (4 × 106/sample) as previously described (33). Proteins were resolved by 7.5% SDS-polyacrylamide gel electrophoresis and then transferred to polyvinylidene difluoride (PVDF) membrane (Bio-Rad, Hercules, CA) using Towbin buffer with 5% MeOH. Blots were subsequently incubated overnight at 4°C in blocking buffer (PBS/0.1% Tween 20 + 3% BSA), followed by incubation with mdr/Ab-1 (Oncogene Research Products, Cambridge, MA) at 1 μg/ml in blocking buffer. Ab-1 recognizes a conserved peptide sequence (SALDTESEKVVQEALDKAREG) that is found in the C-terminal cytoplasmic domain of all three murine P-gp isoforms. Detection of bound Ab-1 was performed in a sequential manner using a 1:1000 dilution of biotinylated goat anti-rabbit (Zymed, San Francisco, CA), a 1:5000 dilution of streptavidin-conjugated alkaline phosphatase (Zymed), and Vistra enzymatic chemifluorescence (ECF) substrate reagent (Amersham, Piscataway, NJ), respectively. Chemifluorescence images were acquired with a Storm 840 PhosphorImager (Molecular Dynamics, Sunnyvale, CA).

Proliferative capacities of purified CD4+ and CD8+ splenic T cells from wt and mdr1a KO B6 × FVB F2 male mice were determined as previously described (27) with minor modifications. In brief, 96-well round-bottom culture plates were coated with 10 μg/ml goat anti-rat Ig (Sigma) overnight at 4°C. After three washes, designated wells were secondarily coated with 0.3 μg/ml anti-CD3 (145-2C11) ascites ± 1 μg/ml anti-CD28 Ab (PharMingen) for 2 h at 37°C. All wells were washed twice before the addition of 5 × 104 CD4+ or CD8+ purified T cells/well in RP-FC medium (RPMI, 10% FCS, 5 × 10−5 M 2-ME, l-glutamine, and antibiotics). Some samples, where indicated, also included 50 ng/ml PMA and 200 ng/ml ionomycin (Sigma). Cells were incubated at 37°C and 5% CO2 for 64 h, pulsed with 0.5 μCi/well [3H]thymidine (ICN, Costa Mesa, CA) for an additional 8 h, then harvested and processed for beta-scintillation counting. Experiments were always performed using T cells isolated from one wt and one mdr1a KO mouse on the same day.

Purified CD4+ and CD8+ splenic T cells from wt and mdr1a KO B6 × FVB F2 male mice were stimulated as described above, except that supernatants were harvested at either 24 h or 72 h for cytokine analysis. IL-2 levels in 24-h supernatants were measured by bioassay using CTLL-20 indicator cells and are expressed as U/ml based on a recombinant murine IL-2 (Genzyme, Cambridge, MA) standard curve determination. The CTLL-20 cell line was a generous gift from Dr. Keith Bishop (University of Michigan, Ann Arbor, MI). IL-4, IL-5, IL-10, and IFN-γ levels were measured as previously described (34) by two-site ELISA assays using recombinant cytokines as standards.

Allospecific cytotoxicity was assessed using a target cell survival assay as previously described (35) with minor modifications. In brief, equal numbers of mitomycin C-treated BALB/c splenocytes (as stimulators) and purified splenic T cells from either a wt or mdr1a KO B6 × FVB F2 male mouse were combined and cultivated in RP-FC media for 5 days. Resultant effector cells were harvested, counted, and then diluted at defined E:T ratios in 96-well plates, to which were then added 1 × 104 [3H]thymidine-labeled P815 (target) or EL-4 (control) cells. After incubation at 37°C for 1 h, the number of nonkilled targets was estimated by harvesting cells and measuring [3H]thymidine values. The percentage of cell lysis was calculated using these values as follows: % lysis = 100 × [(cpmtarget only − cpmtarget + effector)/cpmtarget only].

Proliferation and cytokine data from wt and mdr1a KO T cell samples were compared statistically using paired t tests (results shown as p values). For cytotoxic responses, lytic units (target cells lysed/106 effectors) were first calculated at the 30% lysis level for wt and mdr1a KO effector responses to target cells, and then used to perform a paired t test (result shown as p value). Unless otherwise noted, all other results are presented in the text as mean values ± SEM.

Although it is well established that human peripheral T cells express the MDR1 isoform of P-gp (7, 22), it is not yet known which of the specific mdr isoforms is responsible for P-gp activity in mouse T cells. Therefore, before investigating the role of P-gp in T cell functions, we first determined which of the three murine P-gp isoforms are active in murine peripheral T cell subsets. Using efflux of the fluorescent P-gp substrate R-123 as a measure of P-gp activity, we compared the abilities of resting splenic T cells from 7-mo-old wt, mdr1a KO, and mdr2 KO FVB mice to extrude R-123 (Fig. 1). Although the absence of the mdr2 isoform did not significantly alter the proportion of CD4+ or CD8+ T cells that were able to extrude R-123 compared with T cells from wt mice (36 ± 0.7% (n = 3) vs 38 ± 0.6% (n = 3) for CD4+ T cells, and 73 ± 1.4% (n = 3) vs 75 ± 0.8% (n = 3) for CD8+ T cells, respectively), neither CD4+ nor CD8+ T cells from mdr1a KO mice were able to extrude R-123 (0.2 ± 0.09% (n = 3) for CD4+, and 0.4 ± 0.03% (n = 3) for CD8+ T cell subsets). In separate experiments (not shown), we found that T cells isolated from other lymphoid organs of mdr1a KO mice, including thymus, lymph nodes, and Peyer’s patches, were also unable to extrude R-123. These data show that both CD4+ and CD8+ peripheral T cells express and utilize the mdr1a isoform of P-gp. We were unable to test T cells from mdr1b KO mice, but the nearly complete absence of efflux activity exhibited by mdr1a KO T cells suggests that the mdr1b isoform does not play a major role in mediating R-123 efflux in peripheral T cells, even when mdr1a function has been genetically obliterated.

FIGURE 1.

P-glycoprotein activity in T cells from wt and mdr KO mice. Splenic T cells isolated from 7-mo-old wt, mdr1a KO, or mdr2 KO FVB mice were stained with R-123 to assay P-gp-mediated dye efflux activity. R-123 fluorescence histograms are shown from CD4+ or CD8+ gated T cells that were allowed to extrude the dye in the absence (shaded profile) or presence (open profile) of verapamil. Markers shown on each histogram were set to exclude >99% of verapamil-treated cells and were used to calculate the percentage (value above marker) of untreated cells that were capable of extruding R-123. Data shown are representative of at least three separate experiments for each mouse strain.

FIGURE 1.

P-glycoprotein activity in T cells from wt and mdr KO mice. Splenic T cells isolated from 7-mo-old wt, mdr1a KO, or mdr2 KO FVB mice were stained with R-123 to assay P-gp-mediated dye efflux activity. R-123 fluorescence histograms are shown from CD4+ or CD8+ gated T cells that were allowed to extrude the dye in the absence (shaded profile) or presence (open profile) of verapamil. Markers shown on each histogram were set to exclude >99% of verapamil-treated cells and were used to calculate the percentage (value above marker) of untreated cells that were capable of extruding R-123. Data shown are representative of at least three separate experiments for each mouse strain.

Close modal

To test the idea that other P-gp isoforms might be present but unable to extrude R-123, we analyzed expression of P-gp in CD4+ and CD8+ T cells from wt and mdr1a KO mice. Using an Ab that recognizes a conserved epitope found in all three P-gp isoforms, immunoblot analysis revealed that P-gp was abundant in CD4+ and even more so in CD8+ T cells from wt mice as expected, but was absent in both T cell types from mdr1a KO mice (Fig. 2). Considered together with the R-123 efflux data, these results indicate mdr1a is exclusively expressed in peripheral T cells.

FIGURE 2.

Immunodetection of P-glycoprotein in T cells from wt and mdr1a KO mice. Crude membrane fractions were prepared from purified CD4+ or CD8+ splenic T cells that were isolated from 6-mo-old wt or mdr1a KO FVB mice. Proteins were resolved by SDS-PAGE (4 × 106 cell equivalents/lane), transferred to polyvinylidene difluoride, and probed with mdr/Ab-1. Arrow denotes expected position of ∼170- to 180-kDa P-gp isoforms.

FIGURE 2.

Immunodetection of P-glycoprotein in T cells from wt and mdr1a KO mice. Crude membrane fractions were prepared from purified CD4+ or CD8+ splenic T cells that were isolated from 6-mo-old wt or mdr1a KO FVB mice. Proteins were resolved by SDS-PAGE (4 × 106 cell equivalents/lane), transferred to polyvinylidene difluoride, and probed with mdr/Ab-1. Arrow denotes expected position of ∼170- to 180-kDa P-gp isoforms.

Close modal

Several previous studies have noted differences between P-gphigh and P-gplow T cells in proliferative responses in vitro, although one of these reported deficient proliferation in the P-gphigh subset (26) whereas the other found the P-gphigh cells to be substantially more responsive than P-gplow cells (28). We therefore compared T cells from wt and mdr1a KO mice to investigate whether P-gp regulates the proliferative response induced by anti-CD3 with our without anti-CD28, or by the combination of PMA and ionomycin. For this set of experiments, and all others except as noted, we used 7- to 8-mo-old homozygous wt and homozygous mdr1a−/− F2 mice from a C57BL/6 × FVB/mdr1a KO cross. In this age range, significant proportions of both CD4+ (≥30%) and CD8+ (≥70%) resting T cells from wt mice exhibit P-gp activity (Fig. 1). Additionally, use of the F2 generation allows functional comparisons to be made on a heterogeneous genetic background, reducing possible confounds due to inbred strain-specific idiosyncrasy. Fig. 3 shows the proliferative responses of purified CD4+ and CD8+ splenic T cells from wt and mdr1a KO B6 × FVB F2 mice after 72 h of stimulation. In all cases, regardless of the subset examined or the stimulus employed, T cells from wt and mdr1a KO mice showed similar levels of proliferation, and there were no significant differences between the wt and KO mice.

FIGURE 3.

Proliferative responses of T cells from wt and mdr1a KO mice. Purified CD4+ or CD8+ splenic T cells isolated from 7-mo-old wt or mdr1a KO B6 × FVB F2 mice were stimulated as indicated for 72 h. Proliferative responses, estimated by [3H]thymidine incorporation, are shown on each graph both as average responses from individual mice (circles) and as mean responses ± SEM for the entire group (bars). Values in the upper right hand corner of each graph indicate the statistical p value results from paired t test analyses.

FIGURE 3.

Proliferative responses of T cells from wt and mdr1a KO mice. Purified CD4+ or CD8+ splenic T cells isolated from 7-mo-old wt or mdr1a KO B6 × FVB F2 mice were stimulated as indicated for 72 h. Proliferative responses, estimated by [3H]thymidine incorporation, are shown on each graph both as average responses from individual mice (circles) and as mean responses ± SEM for the entire group (bars). Values in the upper right hand corner of each graph indicate the statistical p value results from paired t test analyses.

Close modal

To assess the possibility that mitogenic activation of mdr1a KO T cells might induce increased expression of one of the other P-gp isoforms, we monitored P-gp activity by R-123 efflux over a period of 6 days after mitogen exposure and found no increase in the levels of P-gp activity in either CD4+ or CD8+mdr1a KO T cells (data not shown). We also compared the proliferative responses of splenic T cells from mdr1a/b double KO FVB mice, which express neither mdr1a nor mdr1b, to wt and mdr1a KO FVB inbred mice, and found no differences in their proliferative responses (data not shown). These results indicate that P-gp does not significantly alter proliferation in murine T cells.

Two previous publications (30, 31) have suggested that P-gp is involved in the release of specific cytokines, including IL-2, IL-4, and IFN-γ, from human PBL. To see whether P-gp might be involved in cytokine release from murine T cells, we compared production of IL-2, IL-4, IL-5, IL-10, and IFN-γ by T cells from wt and mdr1a KO B6 × FVB F2 mice. Two stimulation conditions were used: anti-CD3 alone, which leads to suboptimal activation, and the combination of anti-CD3 + anti-CD28 at concentrations that lead to maximal cytokine accumulation. Fig. 4 shows IL-2 release from purified CD4+ and CD8+ T cells after 24 h of stimulation. P-gp activity does not appear to be required for IL-2 release from either CD4+ or CD8+ murine T cells; wt and mdr1a KO T cells produced equivalent levels of IL-2 in response to either stimulus. To assess the possibility that compensatory increases in the expression of mdr1b might lead to P-gp function after activation of T cells from mdr1a KO mice, we compared the abilities of splenic T cells from wt, mdr1a KO, and mdr1a/b double KO FVB mice to release IL-2 and observed no differences in a series of three replicate experiments (data not shown).

FIGURE 4.

IL-2 production by T cells from wt and mdr1a KO mice. Purified CD4+ or CD8+ splenic T cells isolated from 7- to 8-mo-old wt or mdr1a KO B6 × FVB F2 mice were stimulated as indicated for 24 h. Bar graphs show the mean levels ± SEM of bioactive IL-2 that were present in culture supernatants following stimulation. Values in parentheses below each graph represent the total number (N) of individual mice tested in replicate experiments for each designated T cell subset and mouse type.

FIGURE 4.

IL-2 production by T cells from wt and mdr1a KO mice. Purified CD4+ or CD8+ splenic T cells isolated from 7- to 8-mo-old wt or mdr1a KO B6 × FVB F2 mice were stimulated as indicated for 24 h. Bar graphs show the mean levels ± SEM of bioactive IL-2 that were present in culture supernatants following stimulation. Values in parentheses below each graph represent the total number (N) of individual mice tested in replicate experiments for each designated T cell subset and mouse type.

Close modal

Secretion of other cytokines (IL-4, IL-5, IL-10, and IFN-γ) from wt and mdr1a KO purified CD4+ and CD8+ splenic T cells was examined after 72 h of stimulation with anti-CD3 or anti-CD3 plus anti-CD28. As shown in Fig. 5, there were no statistically significant differences in the levels of any of the tested cytokines between wt and mdr1a KO T cells in response to either stimulus. CD4+ T cells from the mdr1a KO mice tended to make slightly higher amounts of IL-10, and CD8+ T cells from the KO mice tended to produced less IFN-γ, but neither difference was dramatic nor statistically significant and may have reflected mere chance fluctuations in a series of comparisons. Thus, these data show that P-gp is not required for cytokine secretion by murine T cells.

FIGURE 5.

Th1- and Th2-type cytokine production by T cells from wt and mdr1a KO mice. Purified CD4+ or CD8+ splenic T cells isolated from 7- to 8-mo-old wt or mdr1a KO B6 × FVB F2 mice were stimulated as indicated for 72 h. Bar graphs show mean levels ± SEM of IL-4 (top row), IL-5 (second row), IL-10 (third row), and IFN-γ (bottom row) that were present in culture supernatants following stimulation. Values in parentheses underneath graphs represent the total number (N) of individual mice tested in replicate experiments for each designated T cell subset and mouse type.

FIGURE 5.

Th1- and Th2-type cytokine production by T cells from wt and mdr1a KO mice. Purified CD4+ or CD8+ splenic T cells isolated from 7- to 8-mo-old wt or mdr1a KO B6 × FVB F2 mice were stimulated as indicated for 72 h. Bar graphs show mean levels ± SEM of IL-4 (top row), IL-5 (second row), IL-10 (third row), and IFN-γ (bottom row) that were present in culture supernatants following stimulation. Values in parentheses underneath graphs represent the total number (N) of individual mice tested in replicate experiments for each designated T cell subset and mouse type.

Close modal

Studies of human NK and CD8+ T cell cytotoxic function have suggested that P-gp may be required for cytotoxic effector function (29, 36). To see whether P-gp was involved in generation or expression of cytotoxic function by mouse CD8+ T cells, splenic T cells from wt or mdr1a KO B6 × FVB F2 mice (H-2Db/q) were cultivated with mitomycin C-treated splenocytes from BALB/c mice (H-2Dd) for 5 days and then tested for allospecific cytotoxicity using [3H]thymidine-labeled P815 (H-2Dd) target cells or EL-4 (H-2Db) cells as a control. The results (Fig. 6) show that T cells from both wt and mdr1a KO mice produced equally strong and equally specific alloreactive CTL. The responses of the mdr1a KO T cells were, if anything, slightly higher than those of wt mice, although the difference was not statistically significant (p = 0.45, n = 5). The data thus suggest that P-gp does not play a significant role in murine T cell cytotoxic responses.

FIGURE 6.

Cytotoxic activity of T cells from wt and mdr1a KO mice. The ability of effector T cells, generated in vitro using splenic T cells isolated from 7-mo-old wt or mdr1a KO B6 × FVB F2 mice, to kill allogeneic (P815) or syngeneic (EL-4) target cells was determined. Mean percentage lysis values ± SEM are shown at various E:T ratios for wt (filled symbols) and mdr1a KO (open symbols) effector cells. The p value shown indicates the result of a paired t test comparison between wt and mdr1a KO allospecific lytic values calculated at the 30% lysis level (dotted line).

FIGURE 6.

Cytotoxic activity of T cells from wt and mdr1a KO mice. The ability of effector T cells, generated in vitro using splenic T cells isolated from 7-mo-old wt or mdr1a KO B6 × FVB F2 mice, to kill allogeneic (P815) or syngeneic (EL-4) target cells was determined. Mean percentage lysis values ± SEM are shown at various E:T ratios for wt (filled symbols) and mdr1a KO (open symbols) effector cells. The p value shown indicates the result of a paired t test comparison between wt and mdr1a KO allospecific lytic values calculated at the 30% lysis level (dotted line).

Close modal

Aside from its role in mdr, the role of P-gp in normal physiological processes has proven difficult to elucidate. Numerous in vitro studies using drug-resistant, mdr gene-transfected, or chemosensitizer-treated cells have shown that P-gp may participate in some capacity in a variety of cellular functions that range from acting as volume-sensitive chloride ion channels to specialized exporters (8, 10, 37, 38, 39), but there is little direct evidence to connect these properties of the P-gp pump to key physiological or pathological functions. The recent production of mouse lines deficient in one or more of the mdr genes (17, 32, 40) thus provides useful models for examining the implications of P-gp expression in specific cell types and organ systems. Studies using the mdr2 KO strain have, for example, led to the discovery that the murine mdr2 and by implication the homologous human MDR3 P-gp isoforms function in hepatocytes to export phosphatidylcholine into the bile (17, 19, 41). The physiological role of the drug-transporting isoforms of P-gp are still obscure but could well be clarified by studies using these KO mice.

In the present study, we have compared wt to mdr KO mice to gain greater insight into the normal cellular functions of P-gp in peripheral T cells. Previous studies attempting to resolve this matter have primarily relied on the treatment of human T cells with P-gp-inhibiting drugs and/or Abs to see whether these agents altered T cell function (29, 30, 31). The concern that pleiotropic actions of these agents, independent of their effects on P-gp function, might have contributed to their observed effects on T cell function prompted the current series of experiments using mdr KO mice. Although there is always apprehension about the potential developmental effects one may encounter when using gene knockout animals, we were initially encouraged by previous characterizations of these strains, which reported general development, reproductive capacity, and health to be normal in the mdr1a, mdr1b, and double KO stocks (32, 40). Indeed, mdr1a KO mice display normal numbers and proportions of naive, memory, CD4+, and CD8+ peripheral T cells through 20 mo of age (our unpublished results).

Our data show that peripheral T cells in mice exclusively utilize the mdr1a isoform and do not increase expression of either of the other two isoforms when mdr1a is absent. This was somewhat unexpected, in that Schinkel et al. have reported both mdr1a and mdr1b isoforms to be active in hemopoietic progenitor cells (40), and because certain tissues in mdr1a KO mice, including the liver and kidney, showed increased mdr1b mRNA expression (32). We have, however, found compensatory expression of the mdr1b isoform in the intestinal intraepithelial lymphocyte population of mdr1a KO mice.4

The proportion of peripheral T cells that express P-gp activity increases with age in mice (23, 42). In addition, studies from our lab have shown that P-gphigh memory CD4+ T cells from young mice are significantly impaired in their ability to proliferate and release cytokines in response to a variety of stimulation conditions, although they can be induced to proliferate by the combination of PMA + ionomycin (26, 27). In contrast, Bommhardt et al. have reported that P-gphigh naive CD4+ T cells exhibit increased functional responsiveness to activation (28). Our current results do not address this discrepancy directly but do show clearly that differences among P-gp subsets in functional responsiveness are unlikely to reflect effects of the P-gp transporter itself in the process of T cell activation or maturation. Instead, our data suggest that expression of P-gp is one element of a developmental transition that leads, among other things, to functional differences among the P-gp subsets. It is worth note that studies of TCR-transgenic mice (43) have reported an increase with age in the proportion of transgene-positive naive T cells that express P-gp, in parallel with a decline in the ability of these cells to proliferate and produce cytokines.

Our previous experience with the P-gphigh subset of memory CD4+ T cells, which exhibit impaired release of IL-2, IL-4, IL-5, IL-10, and IFN-γ compared with P-gplow cells (26, 27), also led us to question whether P-gp is actually required for the export of IL-2, IL-4, and IFN-γ from human T cells as reported by Drach et al. (30) and Raghu et al. (31). Our results clearly show that P-gp does not affect the production or secretion of these cytokines or the others tested, at least in mice. It seems possible, though unlikely, that humans and mice may have evolved different mechanisms to export certain cytokines. The reports dealing with human T cells have not addressed the question of whether cytokine production differs between the P-gphigh and P-gplow subsets. The ability of CD4+ T cells from young adult humans and mice to produce high levels of cytokines, despite their relatively low proportion of P-gp+ cells, further suggests that cytokine release from peripheral T cells does not require P-gp activity.

Several groups have also reported that P-gp appears to be involved in the cytotoxic effector function of human NK and CD8+ T cells (29, 36, 44). Both Chong et al. (36) and Klimecki et al. (44) inferred that P-gp participated in human NK cell cytotoxicity based on their observations that effector responses were reduced in the presence of P-gp-inhibiting drugs. However, Schinkel et al. have since reported that NK cells from mdr1a/b double KO mice exhibit cytotoxic responses comparable to those of wt mice (40). These contradictory results probably reflect the ambiguities of using P-gp-inhibiting agents to study P-gp function, in that many of these compounds have effects unrelated to P-gp inhibition that can confound interpretation of experimental results. In an alternate approach using an Ab that inhibits human P-gp function (45), Gupta et al. demonstrated that the cytotoxic responses of human CD8+ T cells could be inhibited in a dose-dependent fashion, and postulated that P-gp might be involved in cytotoxic effector molecule export (29). The inhibitory effect was weak, however, achieving only 30% inhibition at Ab doses of 100 μg/ml. Our current data show that T cells from mice lacking P-gp are not impaired in the development or expression of alloreactive cytotoxic T cell function; our results are thus in good agreement with the similar findings of Schinkel et al. (40) on NK cytotoxicity.

Interpretation of experiments using gene knockout technology is often complicated by the possibility that biological redundancy may make it difficult to determine whether a specific protein plays a role in the processes examined. In the current study, we have established that the other two P-gp isoforms are neither functionally active nor present at significant levels in the absence of mdr1a. In a related study, Schinkel et al. also reported that NK cells, which have even higher levels of P-gp activity than T cells (46, 47), did not show any compensatory increases in the mRNA levels of several other transport proteins when both mdr1a and mdr1b were absent (40). Based on these observations, it is unlikely that other proteins act in place of mdr1a in its absence, thus strengthening the case that mdr1a does not play an essential role in any of the currently investigated cellular functions in peripheral T cells.

We thank Erin Belloli, Luann Linsalata, and Julie Eisenbraun for their assistance in breeding and genotyping the (B6 × FVB/mdr1a KO)F2 mice used in this study. We also thank Dr. Alfred Schinkel (Netherlands Cancer Institute) for his technical advice and materials to genotype the murine mdr1a locus.

1

This work was supported by National Institutes of Health (NIH) Grants AG03878 and AG08808. M.D.E. was supported by NIH Training Grants GM07315, AI07413, and AG00114.

3

Abbreviations used in this paper: P-gp, P-glycoprotein; KO, knockout; mdr, multidrug resistance; R-123, rhodamine-123; wt, wild type.

4

M. D. Eisenbraun, D. H. Teitelbaum, R. L. Mosley, and R. A. Miller. Altered development of intestinal intraepithelial lymphocytes in P-glycoprotein-deficient mice. Submitted for publication.

1
Gottesman, M. M., I. Pastan.
1993
. Biochemistry of multidrug resistance mediated by the multidrug transporter.
Annu. Rev. Biochem.
62
:
385
2
Juliano, R. L., V. Ling.
1976
. A surface glycoprotein modulating drug permeability in Chinese hamster ovary cell mutants.
Biochim. Biophys. Acta
455
:
152
3
Goldstein, L. J., H. Galski, A. Fojo, M. Willingham, S. L. Lai, A. Gazdar, R. Pirker, A. Green, W. Crist, G. M. Brodeur.
1989
. Expression of a multidrug resistance gene in human cancers.
J. Natl. Cancer Inst.
81
:
116
4
Fojo, A. T., K. Ueda, D. J. Slamon, D. G. Poplack, M. M. Gottesman, I. Pastan.
1987
. Expression of a multidrug-resistance gene in human tumors and tissues.
Proc. Natl. Acad. Sci. USA
84
:
265
5
Sugawara, I., I. Kataoka, Y. Morishita, H. Hamada, T. Tsuruo, S. Itoyama, S. Mori.
1988
. Tissue distribution of P-glycoprotein encoded by a multidrug-resistant gene as revealed by a monoclonal antibody, MRK 16.
Cancer Res.
48
:
1926
6
Croop, J. M., M. Raymond, D. Haber, A. Devault, R. J. Arceci, P. Gros, D. E. Housman.
1989
. The three mouse multidrug resistance (mdr) genes are expressed in a tissue-specific manner in normal mouse tissues.
Mol. Cell. Biol.
9
:
1346
7
Neyfakh, A. A., A. S. Serpinskaya, A. V. Chervonsky, S. G. Apasov, A. R. Kazarov.
1989
. Multidrug-resistance phenotype of a subpopulation of T-lymphocytes without drug selection.
Exp. Cell Res.
185
:
496
8
Ueda, K., N. Okamura, M. Hirai, Y. Tanigawara, T. Saeki, N. Kioka, T. Komano, R. Hori.
1992
. Human P-glycoprotein transports cortisol, aldosterone, and dexamethasone, but not progesterone.
J. Biol. Chem.
267
:
24248
9
Valverde, M. A., T. D. Bond, S. P. Hardy, J. C. Taylor, C. F. Higgins, J. Altamirano, F. J. Alvarez-Leefmans.
1996
. The multidrug resistance P-glycoprotein modulates cell regulatory volume decrease.
EMBO J.
15
:
4460
10
Raymond, M., P. Gros, M. Whiteway, D. Y. Thomas.
1992
. Functional complementation of yeast ste6 by a mammalian multidrug resistance mdr gene.
Science
256
:
232
11
Higgins, C. F..
1992
. ABC transporters: from microorganisms to man.
Annu. Rev. Cell Biol.
8
:
67
12
Ueda, K., C. Cardarelli, M. M. Gottesman, I. Pastan.
1987
. Expression of a full-length cDNA for the human “MDR1” gene confers resistance to colchicine, doxorubicin, and vinblastine.
Proc. Natl. Acad. Sci. USA
84
:
3004
13
Gros, P., Y. B. Ben Neriah, J. M. Croop, D. E. Housman.
1986
. Isolation and expression of a complementary DNA that confers multidrug resistance.
Nature
323
:
728
14
Devault, A., P. Gros.
1990
. Two members of the mouse mdr gene family confer multidrug resistance with overlapping but distinct drug specificities.
Mol. Cell. Biol.
10
:
1652
15
Endicott, J. A., F. Sarangi, V. Ling.
1991
. Complete cDNA sequences encoding the Chinese hamster P-glycoprotein gene family.
DNA Seq.
2
:
89
16
Tang-Wai, D. F., S. Kajiji, F. DiCapua, D. de Graaf, I. B. Roninson, P. Gros.
1995
. Human (MDR1) and mouse (mdr1, mdr3) P-glycoproteins can be distinguished by their respective drug resistance profiles and sensitivity to modulators.
Biochemistry
34
:
32
17
Smit, J. J., A. H. Schinkel, R. P. Oude Elferink, A. K. Groen, E. Wagenaar, L. van Deemter, C. A. Mol, R. Ottenhoff, N. M. van der Lugt, M. A. van Roon.
1993
. Homozygous disruption of the murine mdr2 P-glycoprotein gene leads to a complete absence of phospholipid from bile and to liver disease.
Cell
75
:
451
18
van Helvoort, A., A. J. Smith, H. Sprong, I. Fritzsche, A. H. Schinkel, P. Borst, G. van Meer.
1996
. MDR1 P-glycoprotein is a lipid translocase of broad specificity, while MDR3 P-glycoprotein specifically translocates phosphatidylcholine.
Cell
87
:
507
19
Crawford, A. R., A. J. Smith, V. C. Hatch, R. P. Oude Elferink, P. Borst, J. M. Crawford.
1997
. Hepatic secretion of phospholipid vesicles in the mouse critically depends on mdr2 or MDR3 P-glycoprotein expression: visualization by electron microscopy.
J. Clin. Invest.
100
:
2562
20
Roninson, I. B., J. E. Chin, K. G. Choi, P. Gros, D. E. Housman, A. Fojo, D. W. Shen, M. M. Gottesman, I. Pastan.
1986
. Isolation of human mdr DNA sequences amplified in multidrug-resistant KB carcinoma cells.
Proc. Natl. Acad. Sci. USA
83
:
4538
21
Gros, P., M. Raymond, J. Bell, D. Housman.
1988
. Cloning and characterization of a second member of the mouse mdr gene family.
Mol. Cell. Biol.
8
:
2770
22
Chaudhary, P. M., E. B. Mechetner, I. B. Roninson.
1992
. Expression and activity of the multidrug resistance P-glycoprotein in human peripheral blood lymphocytes.
Blood
80
:
2735
23
Witkowski, J. M., R. A. Miller.
1993
. Increased function of P-glycoprotein in T lymphocyte subsets of aging mice.
J. Immunol.
150
:
1296
24
Pilarski, L. M., D. Paine, J. E. McElhaney, C. E. Cass, A. R. Belch.
1995
. Multidrug transporter P-glycoprotein 170 as a differentiation antigen on normal human lymphocytes and thymocytes: modulation with differentiation stage and during aging.
Am. J. Hematol.
49
:
323
25
Aggarwal, S., T. Tsuruo, S. Gupta.
1997
. Altered expression and function of P-glycoprotein (170 kDa), encoded by the MDR 1 gene, in T cell subsets from aging humans.
J. Clin. Immunol.
17
:
448
26
Witkowski, J. M., S. P. Li, G. Gorgas, R. A. Miller.
1994
. Extrusion of the P-glycoprotein substrate rhodamine-123 distinguishes CD4 memory T cell subsets that differ in IL-2- driven IL-4 production.
J. Immunol.
153
:
658
27
Bining, N., R. A. Miller.
1997
. Cytokine production by subsets of CD4 memory T cells differing in P-glycoprotein expression: effects of aging.
J. Gerontol. A. Biol. Sci. Med. Sci.
52
:
B137
28
Bommhardt, U., J. C. Cerottini, H. R. MacDonald.
1994
. Heterogeneity in P-glycoprotein (multidrug resistance) activity among murine peripheral T cells: correlation with surface phenotype and effector function.
Eur. J. Immunol.
24
:
2974
29
Gupta, S., C. H. Kim, T. Tsuruo, S. Gollapudi.
1992
. Preferential expression and activity of multidrug resistance gene 1 product (P-glycoprotein), a functionally active efflux pump, in human CD8+ T cells: a role in cytotoxic effector function.
J. Clin. Immunol.
12
:
451
30
Drach, J., A. Gsur, G. Hamilton, S. Zhao, J. Angerler, M. Fiegl, N. Zojer, M. Raderer, I. Haberl, M. Andreeff, H. Huber.
1996
. Involvement of P-glycoprotein in the transmembrane transport of interleukin-2 (IL-2), IL-4, and interferon-γ in normal human T lymphocytes.
Blood
88
:
1747
31
Raghu, G., S. W. Park, I. B. Roninson, E. B. Mechetner.
1996
. Monoclonal antibodies against P-glycoprotein, an MDR1 gene product, inhibit interleukin-2 release from PHA-activated lymphocytes.
Exp. Hematol.
24
:
1258
32
Schinkel, A. H., J. J. Smit, O. van Tellingen, J. H. Beijnen, E. Wagenaar, L. van Deemter, C. A. Mol, M. A. van der Valk, E. C. Robanus-Maandag, H. P. te Riele.
1994
. Disruption of the mouse mdr1a P-glycoprotein gene leads to a deficiency in the blood-brain barrier and to increased sensitivity to drugs.
Cell
77
:
491
33
Barrand, M. A., P. R. Twentyman.
1992
. Differential recognition of mdr1a and mdr1b gene products in multidrug resistant mouse tumour cell lines by different monoclonal antibodies.
Br. J. Cancer
65
:
239
34
Dozmorov, I. M., R. A. Miller.
1998
. Generation of antigen-specific Th2 cells from unprimed mice in vitro: effects of dexamethasone and anti-IL-10 antibody.
J. Immunol.
160
:
2700
35
Matzinger, P..
1991
. The JAM test: a simple assay for DNA fragmentation and cell death.
J. Immunol. Methods
145
:
185
36
Chong, A. S., P. N. Markham, H. M. Gebel, S. D. Bines, J. S. Coon.
1993
. Diverse multidrug-resistance-modification agents inhibit cytolytic activity of natural killer cells.
Cancer Immunol. Immunother.
36
:
133
37
Valverde, M. A., M. Diaz, F. V. Sepulveda, D. R. Gill, S. C. Hyde, C. F. Higgins.
1992
. Volume-regulated chloride channels associated with the human multidrug-resistance P-glycoprotein.
Nature
355
:
830
38
Warren, L., A. Malarska, J. C. Jardillier.
1995
. The structure of P-glycoprotein and the secretion of lysosomal enzymes in multidrug-resistant cells.
Cancer Chemother. Pharmacol.
35
:
267
39
Debry, P., E. A. Nash, D. W. Neklason, J. E. Metherall.
1997
. Role of multidrug resistance P-glycoproteins in cholesterol esterification.
J. Biol. Chem.
272
:
1026
40
Schinkel, A. H., U. Mayer, E. Wagenaar, C. A. Mol, L. van Deemter, J. J. Smit, M. A. van der Valk, A. C. Voordouw, H. Spits, O. van Tellingen, J. M. Zijlmans, W. E. Fibbe, P. Borst.
1997
. Normal viability and altered pharmacokinetics in mice lacking mdr1-type (drug-transporting) P-glycoproteins.
Proc. Natl. Acad. Sci. USA
94
:
4028
41
Mustacich, D. J., J. Shields, R. A. Horton, M. K. Brown, D. J. Reed.
1998
. Biliary secretion of α-tocopherol and the role of the mdr2 P-glycoprotein in rats and mice.
Arch. Biochem. Biophys.
350
:
183
42
Witkowski, J. M., G. Gorgas, R. A. Miller.
1996
. Reciprocal expression of P-glycoprotein and TAP1 accompanied by higher expression of MHC class I antigens in T cells of old mice.
J. Gerontol. A. Biol. Sci. Med. Sci.
51
:
B76
43
Linton, P. J., L. Haynes, N. R. Klinman, S. L. Swain.
1996
. Antigen-independent changes in naive CD4 T cells with aging.
J. Exp. Med.
184
:
1891
44
Klimecki, W. T., C. W. Taylor, W. S. Dalton.
1995
. Inhibition of cell-mediated cytolysis and P-glycoprotein function in natural killer cells by verapamil isomers and cyclosporine A analogs.
J. Clin. Immunol.
15
:
152
45
Hamada, H., T. Tsuruo.
1986
. Functional role for the 170- to 180-kDa glycoprotein specific to drug-resistant tumor cells as revealed by monoclonal antibodies.
Proc. Natl. Acad. Sci. USA
83
:
7785
46
Drach, D., S. Zhao, J. Drach, R. Mahadevia, C. Gattringer, H. Huber, M. Andreeff.
1992
. Subpopulations of normal peripheral blood and bone marrow cells express a functional multidrug resistant phenotype.
Blood
80
:
2729
47
Klimecki, W. T., B. W. Futscher, T. M. Grogan, W. S. Dalton.
1994
. P-glycoprotein expression and function in circulating blood cells from normal volunteers.
Blood
83
:
2451