Effect of the Multidrug Inhibitor GG918 on Drug Sensitivity of Human Leukemic Cells D-C Zhou, G Simonin, A-M Faussat, R Zittoun and J-P Marie
Total Page:16
File Type:pdf, Size:1020Kb
Leukemia (1997) 11, 1516–1522 1997 Stockton Press All rights reserved 0887-6924/97 $12.00 Effect of the multidrug inhibitor GG918 on drug sensitivity of human leukemic cells D-C Zhou, G Simonin, A-M Faussat, R Zittoun and J-P Marie Laboratoire de Cine´tique et Cultures Cellulaires, Universite´ Paris VI, and Formation de Recherche Associe´e Claude Bernard, Hoˆtel Dieu de Paris, France The drug GG918 has been specifically developed for overcom- could be achieved with acceptable toxicity.8–10 However, this ing MDR phenotype and is now in use in clinical trials. In this modulator significantly increased the area under the curve of study, the effects of GG918 on leukemic cell were investigated the co-administrated cytostatic drugs,9–11 and therefore using a 3 day MTT assay. Results showed that, in a highly 10 resistant P-gp(1) leukemic cell line, 0.1 mM of GG918 gives rise increased their toxicity. to a 40-fold sensitization to daunorubicin (DNR) (residual resist- An acridone carboxamide derivative, GG918 (N-{4-[2- ance: 2.1), a 57-fold sensitization to mitoxantrone (residual (1,2,3,4-tetrahydro-6,7-dimethoxy-2-isoquinolinyl)-ethyl]-phenyl} resistance: 1.5), and a 3.3-fold sensitization to idarubicin -9,10-dihydro-5-methoxy-9-oxo-4-acridine carboxamide), (residual resistance: 2.9). When human AB serum was added could sensitize several P-gp(+) cell lines to cytotoxic drugs in to the incubation medium, 1 mM of GG918 was needed to observe the full P-gp modulation potency described above. The a dose-dependent manner through binding to the P-glyco- effect of 1 mM of GG918 was tested on 27 samples of poor prog- protein, inhibiting the efflux, and increasing drug accumu- nosis acute leukemia (25 AML, two ALL). DNR sensitization lation in these cells.12 This derivative was tested on the sensi- (using the MTT assay) and modulation of rhodamine 123 uptake tive myeloma cell line RPMI 8226, and on MDR1(+) sublines were monitored and used as criteria for comparing the in vitro resistant to increasing concentrations of doxorubicin modulation potency of this new compound to the potency of (8226/Dox1, 8226/Dox4, 8226/Dox6 and 8226/40), and on 10 mM of verapamil, which was used as reference. A good corre- lation (r 5 0.8, P 5 0.001) was observed between the results of 27 AML clinical samples. Results showed a potent reversing the two tests. Eleven out of the 26 cases tested were MDR1(1) effect of GG918 (GF120918) for Dox resistance in the 13 (42%), and showed a higher IC50 for DNR than the negative 8226/Dox cell lines, and in 11/27 of the AML samples. In cases (861 6 1284 nM vs 187 1 246 nM, P 5 0.05). GG918 was mice implanted with P-gp(+) P388/Dox leukemic cell line, a able to modulate the in vitro resistance to DNR in eight cases single dose of GG918 restored sensitivity of the tumor to 1 2 (seven MDR1( ), no MDR1( ), one non-tested). Verapamil did doxorubicin administrated soon after, without any toxicity.12 not increase DNR toxicity in four of these eight cases, but was more efficient in one other MDR1(1) case. In conclusion, the More interestingly, GG918 had little effect on the blood and DNR sensitivity of the majority of the fresh AML samples tissue levels of doxorubicin when both drugs were co-admin- 12 expressing P-gp could be modulated in vitro by 1 mM of GG918. istrated in mice. For these reasons, we decided to test this Keywords: multidrug resistance; modulator; acute myelogenous modulator on leukemic human cell lines (MDR1 and/or MRP) leukemia and fresh samples of leukemia. Introduction Materials and methods Multidrug resistant (MDR) phenotype, present from the outset Cells lines of leukemia or arising during chemotherapy, is associated with treatment failure in acute leukemia.1,2 Typical MDR is HL60, a human promyelocytic leukemia cell line; HL60/DNR, related to the overexpression of the MDR1 gene, which enco- a P-gp(+) subline of HL60 resistant to daunorubicin (gifts from des a 170 kDa P-glycoprotein (P-gp) that acts as a transmem- F Lacombe, Bordeaux, France); K562, a human erythroleuke- brane drug efflux pump.3 This efflux can be inhibited by many mia cell line, and K562/Adr, a MDR subline of K562 resistant non-cytotoxic compounds, but until now the ability to reverse to adriamycin (gifts from BI Sikic, Stanford, CA, USA) were this drug resistance induced by the MDR phenotype has not used as positive and negative controls for MDR1 phenotype. been demonstrated in vivo. Another drug efflux pump, the Two K562 sublines resistant to different concentrations of multidrug resistance protein (MRP), was more recently homoharringtonine (HHT): K562/HHT100 (MRP+, MDR1+), described by Cole et al4 in a lung cancer cell line, and is also and K562/HHT200 (MRP+, MDR1++) developed in our lab- able to mimic the MDR phenotype (induced by P-gp) after oratory,14 and A549, a lung adenocarcinoma cell line express- cDNA transfection into sensitive cells.5 ing spontaneously high levels of MRP but no P-gp (gift from Verapamil, the first P-gp modulator described in 1981,6 was S Chevillard, Institut Curie, Paris, France), were used for test- extensively tested in association with doxorubicin to treat ing the effect of GG918 on MRP phenotype. All cell lines were clinically refractory cancers, but phase I and II trials demon- tested for in vitro sensitivity during exponential growth, and strated that the cardio-vascular toxicity is dose-limiting at were routinely cultivated in RPMI medium and 10% fetal calf serum concentrations of 1–2 mM, while in most in vitro models serum (‘growth medium’) (± cytostatic). for MDR, concentrations of 6–10 mM are required for full inhi- bition of P-gp.7 Phases I–II with cyclosporin A (CsA), a potent P-gp modulator in vitro, demonstrated that serum levels of 2– Fresh leukemic cells from patients 4 mM, sufficient in vitro to modulate the MDR phenotype, We investigated fresh leukemic cells from 27 patients with acute leukemia. Twenty-five were acute myeloid leukemias Correspondence: J-P Marie, Service d’He´matologie Biologique, Hoˆ tel- (AML): three AML0, three AML1, six AML2, two AML3, one Dieu, 75181 Paris Cedex 04, France AML4, five AML5 according to FAB criteria, five secondary Received 11 April 1997; accepted 6 May 1997 (II) leukemias (three blast crisis of myeloproliferative syn- MDR reversion by GG918 in vitro in AML D-C Zhou et al 1517 dromes, two preceded by myelodysplasia). Two others were mRNA MDR1 represented at least 30% of the resistant cell B acute lymphoblastic leukemias (one Ph+ at diagnosis, one line K562/HHT100.15 refractory to induction treatment). Fourteen were newly diag- nosed, four relapsed and nine were refractory to at least one cytotoxic treatment. Blood mononuclear cells were isolated Rhodamine and DNR uptake by gradient density, and leukemic cells represented at least 80% of the cells when tested. We described this technique extensively elsewhere.16 Briefly, 106 cells were incubated at 37°C in presence of 0.2 mg/ml of rhodamine 1,2,3 (rho) ± GG918 (0.1–1 mM)or10mMof vera- Growth inhibition assay pamil as reference. After 1 h, cells were washed twice and passed through a cytometer (Facsort; Becton Dickinson, Pont Cell growth inhibition was determined by plating 2 × 105 cells de Clay, France), the signal of FL1 was selected, and the uptake variation was calculated by the variation of the mean in 200 ml growth medium containing serial dilutions (10 nM of fluorescence (rho + modulator/rho). to 30 mM) of the cytostatic drug in 96-well microtiter plates. On some occasions, DNR was used instead of rhodamine. Each concentration of the drug (daunorubicin, mitoxantrone, − Cells were incubated for 1 h with 10 6 M DNR ± modulators idarubicin) was repeated in six wells. The experiences were ° repeated three times, and the mean ± standard deviation was at 37 C, then washed and passed through the cytometer. The FL2 channel displayed the histogram of control cells (without calculated. GG918 (1 mM) (gift from F Hyafil, Glaxo France DNR) for autofluorescence, and the signal given by the spon- Research Lab, Les Ulis, France) or 10 mM of verapamil (Knoll taneous fluorescence of DNR taken up by the cells. The effect France, Levallois-Perret, France) were added into the culture of modulators was calculated as for rhodamine (DNR + medium. After incubation for 3 days at 37°C with 5% CO , 2 modulator/DNR). cell viability was determined using the MTT [3-(4,5-dimethyl- thiazol-2-yl)-2,5-diphenyltetrazolium bromide] assay. Briefly, 20 ml of MTT (5 mg/ml in PBS) was added to each well fol- lowed by 5 h of incubation. The medium and MTT were then Statistics removed from the wells, and formazan crystals were dissolved in 200 ml of DMSO. The absorbance was recorded in a Student’s t-test was used for comparison of IC50/DNR. For microplate reader at the wavelength of 550 nm. The effect of modulation of drug sensitivity and rhodamine uptake, a sim- drug on growth inhibition could be assessed as: ple regression curve was drawn. For evaluation of the effect of modulators on rhodamine uptake, the Kosmogorov–Smir- % of growth inhibition nov test allowed us to set the threshold for significant modu- lation at 1.5 for the ratio of mean fluorescence of absorbance of drug-treated cells = 1 − × 100 dye + modulator/dye alone: above this value, the Kosmogo- absorbance of untreated cells rov–Smirnov test was significant (P , 0.05).