Functional detection of MDR1/P170 and MRP/P190-mediated multidrug resistance in tumour cells by flow cytometry.

Multidrug resistance (MDR) in tumour cells is often caused by the overexpression of the plasma membrane drug transporter P-glycoprotein (P-gp) or the recently discovered multidrug resistance-associated protein (MRP). In this study we investigated the specificity and sensitivity of the fluorescent probes rhodamine 123 (R123), daunorubicin (DNR) and calcein acetoxymethyl ester (calcein-AM) in order to detect the function of the drug transporters P-gp and MRP, using flow cytometry. The effects of modulators on the accumulation and retention of these probes were compared in several pairs of sensitive and P-gp- as well as MRP-overexpressing cell lines. R123, in combination with the modulator PSC833, provided the most sensitive test for detecting P-gp-mediated resistance. Moreover, in a 60 min drug accumulation assay R123 can be regarded as a P-gp-specific probe, since R123 is not very efficiently effluxed by MRP. In contrast to R123, a 60 min DNR or calcein-AM accumulation test could be used to detect MRP-mediated resistance. The MRP-specific modulator genistein could be used in combination with DNR, but not with calcein-AM. Vincristine (VCR) can be used to increase the cellular uptake of calcein-AM in MDR cells, but is not specific for MRP. Thus, although the combination of DNR with genistein appeared to be as sensitive as the combination of calcein-AM with VCR, the former may be used to probe specific MRP activity whereas the latter provides a combined (P-gp + MRP) functional MDR parameter. With these functional assays the role and relative importance of P-gp and MRP can be studied in, for example, haematological malignancies.

acquired. is a major cause of failure in the curative treatment of haematological malignancies. Among the most active anticancer agents used in the treatment of haematological malignancies are some natural toxin-derived drugs. such as the anthracycline daunorubicin (DNR). the epipodophy llotoxin etoposide (VP-16) and the yinca alkaloid vincristine (VCR). Development of cross-resistance to these structurally and functionally unrelated drugs is called multidrug resistance (MDR). One type of MDR is caused by the overexpression of a plasma membrane protein. P-glycoprotein (P-gp). the product of the .DR1 gene (Endicott and Ling. 1989). P-gp functions as an ATP-dependent drug efflux pump. which results in lower intracellular drug concentrations and, hence. in drug resistance (Bradley et al.. 1988: Broxterman andPinedo, 1991). In some haematological malignancies, such as acute myeloid leukaemias. non-Hodgkin's lymphomas and multiple myelomas. P-gp expression appears to be a poor prognostic factor which might predict refractonrness (Grogan et al.. 1993). In vitro. MDR can be circumvented by many relatively non-cvtotoxic agents. such as calcium channel blockers and cyclosporins (Nooter et al.. 1990: Miller et al.. 1991. Therefore. the possibility of reversing MDR in the clinic has attracted considerable interest.
In addition to P-gp-mediated MDR. a number of in vitro drug-selected tumour cell lines which do not contain P-gp but nevertheless show the MDR phenotype have been described (Kuiper et al.. 1990: Versantsoort et al.. 1992. Several, but not all. of these non-P-gp MDR cell lines show overexpression of the gene encoding the multidrug resistance-associated protein (MRRP). which was recently cloned bv Cole et al. (1992) from the non-P-gp MDR small-cell lung cancer cell line H69AR. MRP is a 180-195 kDa membrane glvcoprotein which, like P-gp. belongs to the ATP-binding cassette (ABC) superfamilv of membrane transport proteins (Higgins. 1992 Gene transfection studies have demonstrated that MRP acts as a plasma membrane drug efflux pump and that MfRP overexpression increases resistance to a w-ide spectrum of natural product drugs (Grant et al.. 1994: Broxterman and Versantvoort. 1995. Detection and characterisation of resistant cells in clinical tumour samples could become an important diagnostic parameter in clinical practice. Detection of MDR is mostly performed by using gene-specific probes to detect mRNA (overexpression or monoclonal antibodies to detect the corresponding proteins (Fojo et al.. 1987: Van der Valk et al.. 1990). However. since MDR is caused by increased drug transport out of the cell. the best wsay of measuring would be to determine the free intracellular drug concentration. By measuring the drug accumulation in the cells. interpretation problems resulting from differences in post-translational modifications of the pump proteins (Center. 1985) and differences in passive drug uptake through the plasma membrane (Spoelstra et al.. 1992) can be avoided. An additional important advantage of a functional assay is that it can determine if the resistance can be modulated by resistance modifiers. In the case of modulation it may be considered to administer modifiers in conjunction swith chemotherapy to the patient.
Flow cytometry is a readily applicable technique to assay MDR in leukaemias because several highly fluorescent molecules appear to be substrates for P-gp. The (Burzer et al.. 1994: Schuurhuis et al.. 1995. it swill be important to studv the functional effect of this protein in leukaemias. Therefore we investigated the specificity. and sensitivity of fluorescent probes for detection of P-gp and MRP-mediated MDR. We compared different fluorescent probes by measuring the effect of modulators on their accumulation and retention in several drug-sensitive and P-gpas well as MRPmediated resistant cell lines using flow cytometry. Cell lines Pigpand MP detcbn by fiw cyluiby N Feller et al with a low resistance factor are included to test the sensitivity of the detection method, since it is likely that low resistance is important in clinical material (Schuurhuis et al., 1995). In this report we present the results of experiments which compare daunorubicin. rhodamine and calcein-AM in P-gp and MRP MDR cells.

Flow cvtometrv
Fluorescence was analysed with a FACScan flow cytometer (Becton Dickinson Medical Systems. Sharon, MA. USA). which was equipped with an argon laser. The fluorescence of 10 000 events was logarithmically measured at a laser excitation wavelength of 488 mm. The fluorescence of rhodamine 123, fluorescein isothiocyanate and calcein was collected through a 530 nm band-pass filter and fluorescence of daunorubicin was collected through a 575 nm filter. The logarithmically amplified signals were converted into values on a linear scale and expressed as relative fluorescence units (FU). from which the mean fluorescence was calculated.
Drug accumulation and retention Cells (0.3 -0.7 x 106) were incubated for 60 min at 37°C in 2 giM DNR (Sigma. St Louis, MO. USA) or 200 ng ml-' R123 (Sigma) or for 10 mmn in 0.5 jLM calcein-AM (Molecular Probes, Eugene, OR, USA) with or without a modulator in medium A (growth medium without phenol red and bicarbonate buffer, but with 20 mm Hepes) + 10% FCS. PSC833 (Sandoz, Basle, Switzerland) at 2 A&M, 200 liM genistein (Sigma) or 100 jim vincristine (Sigma) was used as a modulator. After the incubation the cells were washed in ice-cold medium A + 10% FCS. Non-specific binding of the drugs to the cells was measured by adding ice-cold drugcontaining medium to the cells and washing them immediately.
In order to measure DNR and R123 retention, the cells were loaded with the drug for 60 mmn. and after washing the cells were incubated at 37°C in drug-free medium A + 10% FCS with or without the relevant modulator for 60 min. The efflux was stopped by centrifuging the cells and adding icecold medium. P-gp and .RP detection In order to detect P-gp. viable cells were incubated for 1 h at room temperature with mouse monoclonal antibody MRK-16 (10 jig ml -') (kindly provided by Dr T Tsuruo) or a non-relevant mouse IgG2a (10 fig ml-') (Becton Dickinson).
Cells were washed and for 30 min incubated in the dark with rabbit anti-mouse fluorescein isothiocyanate (1:100) (Dakopatts. Copenhagen. Denmark). In order to detect MRP. cells were permeabilised in 10% (v v) Lysing solution G (Becton Dickinson) for O min and then incubated for 1 h at room temperature with rat monoclonal antibody MRPrl (1.7 jig ml-') or a non-specific rat monoclonal antibody (provided by M Flens and G Zaman: Flens et al.. 1994). Antibody binding was detected with rabbit anti-rat fluorescein isothiocyanate (1:100) (Dakopatts). The mean of the fluorescence of the MRK16or MRP-labelled cells was divided by the mean of the fluorescence of the non-relevant antibody-labelled cells.

Flow-through sYstem
The flow-through sy stem has been described in detail previously ). Basically. a monolayer of approximately 10 cells attached to a glass chamber (surface 50 cm2. height 0.1 mm) was perfused with 2 ILM DNR in medium A. The perfusion medium was passed over the cells at a constant flow of 200 jlI minat 37°C until a steady-state level was reached. Then a series of pulse injectionsk of modulators was introduced into the flowing perfusion medium. causing inhibition of the DNR efflux. which was measured at the outlet of the flow-through system as a decrease in the medium fluorescence of DNR. The fluorescence of DNR was measured by a fluorescence detector (type 821-FP. Jasco. Haschioji City. Japan) at excitation emission wavelengths of 480 nm 560 nm.

Results
P-gp functional assays Standard fluorescent probes for measuring the function of P-gp by flow cytometry are R123 and DNR (Lampidis et al., 1985;Van Acker et al., 1993). A more recently investigated probe for P-gp is calcein-AM (Hollo et al., 1994;Homolya et al.. 1993). In this study we compared these probes in order to assess their sensitivity and specificity for P-gp and MRP detection in accumulation and retention assays. First, we determined the accumulation ratio (the ratio of the fluorescence of drugs accumulated in the resistant and sensitive cell line) for these probes and the effect of the P-gp inhibitor PSC833 on their accumulation in the KB cell lines. Solubilisation of the cells with SDS after accumulation of the dyes confirmed that the effect of modulator on calcein and R123 fluorescence was indeed caused by an increased dye uptake. This excludes the possibility that the higher calcein or R123 fluorescence induced by the modulators is due to an altered intracellular drug distribution, resulting in a higher fluorescence of cells. In particular, the analysis of a slightly resistant cell line, such as KB8 (see last column of Table I for P-gp expression), is cn'tical in these assays, since this low level of resistance appears to be relevant for leukaemias (Schuurhuis et al.. 1995). Table I shows that a reduced DNR accumulation can be detected in KB8 and KB8-5 cells compared with the sensitive parental KB3-1 cells. R123 and calcein accumulation were reduced in KB8-5, but not in the KB8 cell line. However, in the KB8 cell line a distinct effect of the P-gp modulator PSC833 on the R123 accumulation could be detected, and this effect of PSC833 was greater than the effect on DNR accumulation. In contrast, P-gp function in KB8 could not be detected by PSC833 modulation of calcein-AM uptake. Thus. the combination of PSC833 with PSp9u -ddcia by flw cSmy N Feler et a 545 independent experiments. Differences were tested for significance using a Student's t-test: ap<0.05; bP<O.Ol. The acumulation ratio is the basal fluorescence of resistant cells divided by the fluorescence of sensitive cells after drug accumulation. The effect of PSC833 (2 jM) is the ratio of drug accumulation with PSC833 divided by drug accumulation without PSC833. The amount of P-gp is calulated by dividing the mean of the fluorescence of the MRK16-labelled cells by the mean of the fluorescence of the non-relevant antibody-labelled cells (a ratio of 1.0 means that P-gp is undetectable). R123 appeared to be the most sensitive P-gp functional assay in these cell lines. It has been suggested that the sensitivity of a functional assay can be enhanced by measuring drug retention instead of accumulation (Ross et al., 1993). We therefore compared these different experimental set-ups with R123 in the KB cell lines. Figure 1 shows the time course of the R123 accumulation and retention with and without PSC833 in the KB8 cell line. The relative effect of PSC833 on the accumulation and retention was somewhat larger when the accumulation and efflux time respectively were longer (not shown). However, in order to compare the sensitivity of an accumulation and retention assay, time points of 60 min were chosen for practical reasons. Table II shows that no substantial increase in the sensitivity of the assays was obtained by measuring the effect of PSC833 on R123 retention, when cells were loaded without PSC833. However, when cells were loaded with PSC833 and subsequently effiuxed, the effect of the modulator was even much smaller than in the accumulation assay in KB8-5 cells, and in KB8 cells it was hardly detectable. In contrast, when cells were loaded with 20 pM verapamil as modulator instead of 2 IsM PSC833, no difference was observed between the effect of modulator on R123 accumulation and retention (data not shown).
In order to find an explanation for this apparent discrepancy between the effect of the clinically important modulator PSC833 on R123 accumulation and retention, KB8-5 cells were loaded with 2 lM DNR in a flow-through system until steady-state accumulation was reached. Then a pulse of 25 IM verapamil or 2 pM PSC833 was injected into the DNR-containing medium and the dynamic effect of cellular DNR influx by inhibition of P-gp (and DNR efflux after passage of the modulator) was recorded by measuring the fluorescence of DNR in the medium at the outlet of the flow-through system. DNR influx and efflux by cells were measured by the decrease and increase of the DNR fluorescence in the medium respectively ( Figure 2). This experiment showed that the efflux of DNR is much more rapidly restored when verapamil instead of PSC833 is used as modulator. Apparently, PSC833 remains much longer in the cells than verapamil, which explains the slight modulator effects on the DNR efflux (see middle column of Table II). Therefore, PSC833 is not suitable for loading cells with drugs intended for subsequent drug efflux studies.

Modulation of MRP-mediated transport
The previous experiments showed that measuring the modulation of R123 accumulation by PSC833 is a convenient   , 1990). The trace shows the decrease and increase in DNR fluorescence in the medium as a result of increased (influx) and decreased (efflux) cellular DNR uptake respectively. and sensitive functional assay for detecting P-gp activity. In order to establish whether PSC833 is also an efficient modulator for MRP, we first compared the effect of PSC833 with genistein, a compound which has been shown to inhibit MRP-but not P-gp-mediated DNR efflux (Versantvoort et al.. 1993). Table III shows that the effectiveness of PSC833 as a modulator of DNR accumulation varied among the different MRP-overexpressing cell lines. PSC833 was an effective modulator in the MRP-overexpressing HL60/ADR cells, but had no effect or only a slight effect on the accumulation of DNR in the other MRP-overexpressing cells. In conclusion. PSC833 appeared not to be strictly specific for P-gp, since in certain tumour cells it may also interact with DNR transport by MRP.
Since PSC833 was not a potent modulator in most MRP MDR cells. we investigated the use of genistein as a modulator in order to compare its effect on transport of DNR, R123 and calcein-AM. However, in preliminary experiments we found that genistein decreased the fluorescence of calcein as well as of R123 in sensitive cells. Because of this interaction. which was apparently unrelated to the presence of any drug efflux pump, genistein could not be used in combination with R123 or calcein-AM. Therefore, for R123 and calcein-AM we decided to use VCR as a modulator. since it has been shown previously that it modulates P-gpas well as MRP-mediated drug efflux (Milder et al., 1994).
overexpressing HL60/ADR cell line, as shown before (Table  III), we tested the combination of R123 and PSC833 in these cells. However, PSC833 had only a slight effect on R123 accumulation in this cell line. We also tested VCR as a modulator of R123 accumulation in the transfectant SI (MRP) and in the sensitive cell line SW-1573, but again a decrease in the fluorescence of R123 was seen in both cell lines (data not shown), which suggests a non-specific interaction among these drugs, unrelated to the function of the MRP efflux pump. In conclusion, R123 appeared to be an 1200 - Start efflux R123 to probe MRP function The results of the R123 accumulation are summarised in Table IV. It appears that R123 is not a sensitive probe for detecting MRP-resistant cells, including the MRP transfectant. In fact, in a 60 min accumulation assay no consistently decreased R123 accumulation was observed. The GLC4-ADR150 and the HL60 ADR sublines accumulated even more R123 than the related parent cell lines. Nor was there an enhanced efflux of R123 in the GLC4-ADR,50 compared with the GLC4 parent cell line, as illustrated in Figure 3. Only in the SW-1573/2R120 and the Sl(MRP) sublines could a small R123 accumulation defect be detected. Since PSC833 was an effective modulator of DNR transport in the MRP- Fgre 3 Uptake and retention of R123 in the GLC4 and GLC4-ADR150 cells. GLC4 (-----) and GLC4-ADRI% (--A--) cells were loaded for 60 min with 200 ng ml-R123. Retention of R123 was measured after suspending the GLC4 ( -) and GLC4-ADRi,5 (-A-) cells in R123-free medium.   (Table IV). In these cells the accumulation defect of calcein and the effect of the modulator on calcein accumulation were greater than the effect on DNR. Since genistein is a specific modulator of MRP-mediated transport and cannot be used together with calcein (see above), the preferred combination for specifically detecting the MRP function is DNR and gemnstein.

Sensitivity of MRP detection
In order to analyse the sensitivity of the preferred combination of DNR and genistein for detecting the MRP function, DNR accumulation and retention were tested in the GLC4 series of cell lines with increasing MRP expression. Table V shows that DNR accumulation was progressively decreased in the cell lines with increasing MRP expression. Also, the effect of genistein on DNR accumulation and retention was greater in cells with increasing resistance and MRP expression. However, the retention assay was not more sensitive than the accumulation assay. Nor did the method of loading the cells before efflux. with or without genistein. alter the effect of genistein on DNR retention (data not shown). From these data it can be inferred that this assay was not sensitive enough to distinguish between GLC4 and GLC4-ADR, cells, which are 1.4-fold resistant to DNR (Versantvoort et al., 1995a). The GLC4-ADRp, and GLC4-ADR,50 sublines which are 6.3-and 58-fold resistant to daunorubicin respectively. could easily be detected. Notably in the 3-fold resistant SI(MRP) cell line, MRP-related drug resistance could be detected in the DNR accumulation assay. as shown in Table  IV.

Dicussion
Many fluorescent probes (DNR. DOX, R123. Hoechst 33342. BCECF-AM. calcein-AM, FURA-2-AM, etc.) are available to detect P-gp-mediated MDR. These probes have been used to detect P-gp by flow cytometry and laser scan microscopy (Schuurhuis et al., 1991;Homolya et al., 1993;Van Acker et al.. 1993), and the function of P-gp has been characterised in many human tumour cell lines. However. until now systematic studies into the sensitivity and specificity of these probes for detecting P-gp activity have not yet been reported. Moreover, this issue has become more complicated since it is now recognised that at least one other drug pump, namely MRP. may be involved in MDR. Both P-gp-and MRP-overexpressions appear to be important resistance mechanisms in haematological malignancies (Burger et al., 1994;Kuss et al.. 1994;Schuurhuis et al.. 1995).
In order to gain more insight into the specificity and sensitivity of methods of detecting MDR by flow cytometry. we studied three fluorescent probes, R123, DNR and calcein-AM, combined with three different modulators. PSC833, genistein and VCR, in P-gp-and MRP-overexpressing cells. Each of these combinations has its own merits and drawbacks, as discussed below. The fluorescent probes were selected for the following reasons. DNR is a cytotoxic drug with favourable fluorescent properties used in the treatment of acute myeloid leukaemia (AML) (Schuurhuis et al.. 1995). Calcein-AM has been reported to be a good probe for P-gp because it is a hydrophobic non-fluorescent molecule which rapidly permeates the plasma membrane of cells. By cleavage of the ester bonds by intracellular esterases it is transformed to calcein. which is a highly fluorescent, non-membrane-permeable free acid. Moreover. since calcein. in contrast to calcein-AM, is not transported by P-gp. it is well retained by cells, with a fluorescence which is essentially insensitive to changes in pH. as well as to changes in Ca-+ or Mg" concentration (Homolya et al., 1993). R123 is also a sensitive probe for P-gp, but the accumulation of R123 will be dependent on pH and mitochondrial storage capacity. R123 is not well retained by cells and some leakage of the probe occurs upon the transfer of accumulated cells into dye-free medium at O°C (Holl6 et al.. 1994).
In this study we compared these different fluorescent dyes by measuring the effect of modulators on their accumulation in P-gp as well as MRP MDR cells. PSC833. a cyclosporin analogue, was chosen as a modulator for P-gp because it has acceptable toxicity when given to patients in concentrations known to reverse P-gp-mediated drug resistance in vitro.
Interestingly, this study provides evidence that the action of PSC833 on P-gp-mediated DNR pumping is much longer lasting than the action of verapamil (Figure 2). presumably because PSC833 remains longer at the P-gp binding site in the cells. This might also have relevance for clinical modulation studies. Genistein was chosen as a modulator to detect MRP function in cells because it has been shown to inhibit specifically MRP-mediated DNR transport. However. genistein in the presently used concentration is too toxic for clinical applications (Versantvoort et al., 1993).
We first found that the R123 PSC833 combination provided the most sensitive test for detecting P-gp-mediated MDR. This P-gp functional assay is very sensitive since the slightly resistant KB8 cell line (RFDNR 2 2) could be detected by PSC833 modulation. This is despite the fact that the KB8 cell line had no accumulation defect for R123. One possible explanation is that the number of mitochondria is somewhat increased in this cell line. Remarkably. no modulation of calcein accumulation could be detected in the KB8 cell line. whereas in the KB8-5 cell line calcein appeared to be a more sensitive probe than DNR. The reason for this discrepancy is not known. For the time being. when clinical samples are being tested for P-gp function. we would recommend that R123 be used in combination with 2tgM PSC833.
Next, we tested these probes in MRP-mediated drugresistant cell lines. A remarkable result was that R123. the most sensitive probe for P-gp. was not a suitable probe for sensitive detection of MRP-mediated drug resistance in cell lines. Although MRP-overexpressing cells are resistant to R123  and the highly MRP-overexpressing COR-L23 R cell line showed increased efflux of R123 (Twentyman et al.. 1994). no accumulation defect was found in the resistant cell lines GLC4-ADR,1, and HL60 ADR after 60 min incubation with R123. It might be surprising that we found that R123 accumulation after 60 mn is higher in some MRP-overexpressing cell lines than in the parental sensitive cell lines. This may be due to R123 being trapped in cytosolic vesicles containing MRP, as suggested by Cole et al. (1992). In addition. in contrast to P-gp. an accumulation defect for R123 can only be found after a longer incubation time (Twentyman et al.. 1994). Since there is evidence that the function of drug pumps in clinical samples, such as in AML cells. may deteriorate after storage. we prefer short-term (60 mmn) drug accumulation assays. Under these conditions R123 cannot be used as a probe for detecting MRP-mediated resistance and therefore R123 can be regarded as P-gp specific.
Calcein-AM appeared to be an effective probe for MRPmediated resistance but the specific MRP modulator genistein decreased calcein fluorescence in sensitive cells and could not be used as a modulator. For this reason VCR. a modulator of P-gp-and MRP-mediated resistance (Mfilder et al.. 1994). was used to modulate the calcein accumulation in MRP-resistant cells. A disadvantage of the combination of calcein-AM and VCR is that MRP-related resistance cannot be distinguished from P-gp-related resistance. Since the combination of DNR and genistein appeared to be as sensitive as the combination of calcein-AM and VCR. and since DNR modulation by genistein is specific for MRP. we prefer the %% N Feller et al 548 combination of DNR and genistein for testing specific MRPmediated resistance. Some data suggested a different spectrum of MRP activity in the human leukaemia HL60 ADR cells since. in contrast to the other MRP-overexpressing cells. PSC833 is a potent modulator of the DNR accumulation, and it also appeared that calcein-AM is a more effective probe than DNR when this cell line is compared with the other MRP MDR cell lines. For this reason. calcein-AM with VCR as modulator may yet be a very useful alternative for testing leukaemias.
Recent studies have suggested that anti-cancer drugs may be conjugated in cells and that these conjugates may be the substrate for the MRP transporter (Jedlitschky et al.. 1994). Resistance would then depend on the conjugating enzymes as well as on MRP expression. However, it is important to note here that all the MRP-overexpressing cells tested had an accumulation defect for calcein. It has been reported that P-gp extrudes calcein-AM directly from the plasma cell membrane without entering the cellular cytoplasm and that calcein is not a probe for P-gp (Homolya et al.. 1993;Hol1o et al., 1994). If MRP extrudes calcein-AM by a similar mechanism to P-gp. then our data would suggest that MRP may extrude hydrophobic drugs in a non-conjugated form from the plasma membrane compartment. Elucidation of the mechanism of extrusion by MRP of different classes of drugs needs more research (Versantvoort et al.. 1995b).
In the GLC4 series the degree of MRP expression was in accordance with drug resistance (Versantvoort et al.. 1995a) as well as with the DNR accumulation defect. However, the SW-1 573 sensitive cell line had a relatively high level of MRP. as detected with the MRPrl antibody. but genistein had no effect on DNR accumulation. One possible explanation for this apparent discrepancy is that cells had to be permeabilised before being stained with MRPrl, since MRPrl is a monoclonal antibody which recognises an internal epitope . Therefore, no distinction could be made between functional MRP in the plasma cell membrane and MRP in structures such as the Golgi system . Since MRP staining of the plasma membrane SW-1573 cells was undetectable by immunocytochemistry . it could well be that this cell line contains more non-functional MRP than other sensitive cell lines. In future, monoclonal antibodies directed against outer epitopes of MRP will clarify this point. Alternatively, a role of intracellular MRP might be inferred from data on the MRP-overexpressing cell line H69 AR, which showed an altered intracellular distribution of doxorubicin but no doxorubicin accumulation defect (Cole et al.. 1992). For this reason we also explore another approach to detect P-gpas well as MRP-mediated MDR. by using laser scan microscopy to detect the effect of modulators on the amount of nuclear anthracycline fluorescence in relation to P-gp and MRP expression (Broxterman et al.. 1994;Schuurhuis et al.. 1991).
Preliminary data from these studies show that in AML P-gpas well as MRP-mediated MDR is at the level of KB8 resistance (Schuurhuis et al.. 1995).
In conclusion, the present study shows that R123 accumulation in combination with PSC833 is a specific and sensitive test for detecting low levels of P-gp-mediated resistance, which could be of importance in clinical practice. For detection of MRP-mediated resistance. R123 did not appear to be a sensitive probe in a 60 min accumulation assay, whereas the combination of DNR with genistein was more sensitive and selective. The use of calcein-AM with an appropriate modulator (VCR) may be an alternative. The finding that the P-gp-inhibitory action of PSC833 lasts longer than that of verapamil may pertain to the use of PSC833 for modulating MDR in clinical studies.