Acute promyelocytic leukemia (APL) cells express a considerable level of CD33, which is a target of gemtuzumab ozogamicin (GO), and a significantly lower level of P-glycoprotein (P-gp). In this study, we examined whether GO was effective on all-trans retinoic acid (ATRA)- or arsenic trioxide (ATO)-resistant APL cells. Cells used were an APL cell line in which P-gp was undetectable (NB4), ATRA-resistant NB4 (NB4/RA), NB4 and NB4/RA that had been transfected with MDR-1 cDNA (NB4/MDR and NB4/RA/MDR, respectively), ATO-resistant NB4 (NB4/As) and blast cells from eight patients with clinically ATRA-resistant APL including two patients with ATRA- and ATO-resistant APL. The efficacy of GO was analyzed by 3H-thymidine incorporation, the dye exclusion test and cell cycle distribution. GO suppressed the growth of NB4, NB4/RA and NB4/As cells in a dose-dependent manner. GO increased the percentage of hypodiploid cells significantly in NB4, NB4/RA and NB4/As cells, and by a limited degree in NB4/MDR and NB4/RA/MDR cells. Similar results were obtained using blast cells from the patients with APL. GO is effective against ATRA- or ATO-resistant APL cells that do not express P-gp, and the mechanism of resistance to GO is not related to the mechanism of resistance to ATRA or ATO in APL cells.
Recently, gemtuzumab ozogamicin (GO, Mylotarg™), a calicheamicin-conjugated humanized anti-CD33 monoclonal antibody (mAb), has been introduced for the treatment of acute myeloid leukemia (AML).1 However, the clinical outcome after the treatment with GO was negatively associated with P-glycoprotein (P-gp) function in AML.2 In our previous studies, we found that resistance to GO was mainly mediated by P-gp.3, 4 Acute promyelocytic leukemia (APL) cells express a considerable level of CD33 antigen and a significantly lower level of P-gp compared with other types of AML.5 Therefore, GO may become established as a successful treatment for APL. In fact, GO has been introduced with the clinical efficacy in the treatment of APL.6, 7 However, in vitro efficacy of GO on APL cells as well as all-trans retinoic acid (ATRA)- and arsenic trioxide (ATO)-resistant ones has not been studied well. Moreover, drug interaction among ATRA, ATO and GO, and the mechanisms of resistance of APL cells to them remain unclear.8
Materials and methods
The cell lines used were a human APL cell line, NB4, which was kindly provided by Dr M Lanotte (Hospital Saint-Louis, Paris, France);9 ATRA-resistant NB4 (NB4/RA) cells; NB4 and NB4/RA cells transfected with MDR-1 cDNA (NB4/MDR and NB4/RA/MDR, respectively); and ATO-resistant NB4 (NB4/As) cells. The NB4/RA and NB4/As cells were obtained by culturing NB4 cells with gradually increasing concentrations of ATRA and ATO, respectively.10 mdr-1 messenger RNA (mRNA) was not detectable in NB4, NB4/RA or NB4/As cells by RT-PCR.11, 12 NB4/MDR and NB4/RA/MDR cells had detectable mdr-1 mRNA, but did not have detectable MDR-related protein (MRP) mRNA or lung-resistant protein (LRP) mRNA. Blasts were collected from four patients with APL at diagnosis, six patients with clinically ATRA-resistant but ATO-sensitive APL and two patients with clinically ATRA- and ATO-resistant APL.
Flow cytometric analysis for CD33 and Pgp expression
For evaluation of CD33 expression, cells were stained with phycoerythrin (PE)-conjugated anti-CD33 mAb (Becton Dickinson Immuno-cytometry Systems, San Jose, CA, USA), according to the manufacturer's instructions. For P-gp analysis, cells were incubated with biotinylated MRK16 (Fab′) mouse mAb or a subclass-matched control mAb, and stained with streptavidine-Per CP (Becton Dickinson Immuno-cytometry Systems) as previously described.11 Over 10 000 events were analyzed with the Epics XL flow cytometer (Beckman Coulter, Fullerton, CA, USA). APL cells obtained from the patients were also gated by CD-45 staining.4
Humanized anti-CD33 mAb and GO
GO consists of three essential parts: an antibody a cytotoxic agent, and a linker. The antibody, humanized IgG4 (hP67.6), targets the CD33 antigen. The cytotoxic agent is N-acetyl (NAc)-γ calicheamicin dimethyl hydrazide (DMH), a derivative of calicheamicin antitumor antibiotics.1, 2 GO, humanized nonconjugated anti-CD33 mAb (hP67.6) and free NAc-γ calicheamicin DMH were kindly provided by the Wyeth Research Division of Wyeth Pharmaceuticals Inc. (Philadelphia, PA, USA). The amount of GO used in an experiment was determined based on the concentration of NAc-γ calicheamicin DMH bound to the antibody. One microgram of GO contains 27.1 ng of NAc-γ calicheamicin DMH, and approximately 97% of a GO molecule is composed of the linker and antibody.
3H-Thymidine (3H-TdR) incorporation analysis for assessment of cell proliferation
Cells were plated in a 96-well microplate (BD Biosciences, Billerica, MA, USA) at 2 × 105 cells per well in the presence or absence of GO containing 5, 10 or 100 ng/ml NAc-γ calicheamicin DMH or the respective concentration of hP67.6, in 100 μl of RPMI 1640 medium containing 10% fetal calf serum (FCS) and 1 μCi of 3H-TdR. The detailed method was described in our previous papers.3 The level of 3H-TdR incorporation upon incubation with GO was compared with that upon incubation with hP67.6. The analysis was repeated five times.
Dye exclusion test with propidium iodide staining
After incubation of cells with GO or hP67.6. for the indicated period of time, cells were stained with 0.2 μg/ml propidium iodide (PI) (Sigma, Saint Louis, MO, USA) solution and counted. The numbers of dye-stained (dead) and unstained (living) cells both decreased and the amount of debris rapidly increased, making it difficult to evaluate the cell viability properly. Therefore, viable cells were evaluated. The viable cell count (/ml) after incubation with GO was compared with that after incubation with hP67.6. The analysis was repeated five times.
Cell cycle distributions
The cell cycle distribution was analyzed by flow cytometry with PI staining. The detailed method was described in our previous papers.3, 4 Cell cycle distribution could be analyzed after incubation with 10 or 100 ng/ml of GO for 24 or 48 h. GO temporarily arrests NB4 cells at the G2/M phase, and increases the percentage of hypodiploid cells, by which we evaluated the effect of GO.3, 4 Then, the GO-sensitive cells rapidly change to debris. The analysis was performed in triplicate.
Flow cytometric analysis of CD33 and P-gp expression on NB4 cells and its sublines
The amount of CD33 expressed on the cells did not significantly differ among NB4, NB4/RA, NB4/MDR, NB4/RA/MDR and NB4/As cells. P-gp was not expressed on NB4, NB4/RA or NB4/As cells, in agreement with previous reports.11, 12 Equivalent levels of P-gp were expressed on NB4/MDR and NB4/RA/MDR cells.
3H-TdR incorporation into NB4 cells and its sublines
Upon 72-h incubation with GO containing 5, 10 or 100 ng/ml of NAc-γ calicheamicin DMH, the level of 3H-TdR incorporation into NB4 cells and its sublines decreased in a dose-dependent manner (Figure 1a). In each cell line, the level of 3H-TdR incorporation was lower than that in the same cell line that had been incubated with the corresponding concentration of hP67.6. Upon incubation with GO containing 100 ng/ml NAc-γ calicheamicin DMH, there were significant differences in the level of 3H-TdR incorporation between NB4 and NB4/MDR cells at 48 and 72 h (P<0.01 each), and between NB4/RA and NB4/RA/MDR cells at 48 and 72 h (P<0.01 each) (Figure 1b). There were no significant differences in the level of 3H-TdR incorporation between NB4 and NB4/RA cells, or between NB4/MDR and NB4/RA/MDR cells at any time point. 3H-TdR incorporation in NB4 cells upon 72-h incubation with GO containing 10 ng/ml NAc-γ calicheamicin DMH corresponded with that with 1000 ng/ml free NAc-γ calicheamicin DMH, which corresponded to a concentration of NAc-γ calicheamicin DMH that was approximately 100 times greater than that in GO. The rate did not change between P-gp-negative and positive NB4 cells.
Viable cell count analysis by flow cytometry
By the incubation with GO containing 5, 10 or 100 ng/ml of NAc-γ calicheamicin DMH for 72 h, the viable cell counts of NB4 cells and its sublines decreased in a dose-dependent manner (Figure 2a). Figure 2c shows their cell counts upon 24-, 48- or 72-h incubation of these cell lines with GO containing 100 ng/ml NAc-γ calicheamicin DMH or the respective concentration of hP67.6. GO similarly decreased the number of NB4 and NB4/RA cells. GO also decreased the count of NB4/As cells. GO slightly reduced the counts of NB4/MDR and NB4/RA/MDR cells. The counts of NB4 cells and its sublines upon 72-h incubation with GO containing 10 ng/ml NAc-γ calicheamicin DMH corresponded with those with 1000 ng/ml free NAc-γ calicheamicin DMH, respectively.
The combination of ATRA and GO reduced the count of NB4 cells by a greater degree than GO or ATRA alone (P=0.019 and P<0.01, respectively), but did not reduce the count of NB4/RA cells by a significantly greater degree than GO or ATRA alone (Figure 2b). Upon incubation with GO and ATO, the counts of NB4 and NB4/RA cells were less than those upon incubation with GO (P<0.01 each) or ATO alone (P<0.01 each).
Cell cycle distribution
The increase percentage in the number of hypodiploid cells on cell cycle distribution upon incubation with GO containing 10 or 100 ng/ml of NAc-γ calicheamicin DMH for 48 h is summarized in Figure 3. The percentage of hypodiploid cells in the NB4, NB4/RA and NB4/As cells was increased upon 12-h incubation with GO, and it was the highest upon 48-h incubation. Beyond 48 h, it was difficult to evaluate the proportion of hypodiploid cells accurately because these GO-sensitive cells transformed into debris, as reported previously (Figure 4).3 Upon incubation with GO, the percentage of hypodiploid cells in the NB4/MDR and NB4/RA/MDR cells were significantly less than those in the NB4 and NB4/RA cells, respectively (P<0.01 each). The addition of ATRA to GO further increased the percentage of hypodiploid cells in NB4 cells (P=0.043), but not in NB4/RA cells (P=0.97). Upon incubation with GO and ATO, the percentage of hypodiploid cells in NB4 and NB4/RA cells were slightly higher than those upon incubation with GO alone (P=0.091 and 0.082, respectively).
Similar results were obtained using blast cells derived from the patients with APL (Table 1). Upon incubation with GO for 48 or 60 h, the hypodiploid portion considerably increased in APL cells that had been obtained from not only the four cases at diagnosis, but also five ATRA-resistant and two ATRA- and ATO-resistant cases. Two patients, whose APL relapsed after achieving complete remission (CR) by ATRA and receiving postremission chemotherapy, were treated according to the Japanese phase 1 and 2 study of GO. They were resistant to re-induction therapy by ATRA, but achieved CR and CR without platelets recovery (CRp) after treatment with GO, respectively.
Recently, GO with or without ATRA was introduced for the treatment of APL, and the clinical efficacy of these therapies has been reported in newly diagnosed or relapsed patients with APL.6, 7 There are two basic reasons that support the clinical application of GO on APL. One is that large amounts of CD33 are commonly expressed on the surface of APL cells. Therefore, several different anti-CD33 mAbs have been used for the treatment of APL, and notable results have been reported, especially in the use of GO.6, 7 Another reason is that a low level of P-gp is expressed on APL cells.5 GO is sometimes not effective in some other subtypes of AML because the detached calicheamicin derivative is pumped out by P-gp.3, 4 Therefore, this mechanism of resistance to GO is not theoretically applicable to APL cells.
In this study, GO showed antiproliferative and cytocidal effects on ATRA-resistant NB4 cells as well as NB4 cells. We previously demonstrated that MDR modifiers, PSC833 and MS209, had no effect on ATRA-resistance in APL, which indicated that P-gp has a limited role in ATRA-resistance.11 Intracellular ATRA concentration was not influenced by P-gp.11 Clinical evidence, including our reports, also supported the independence of P-gp and ATRA-resistance.13 Taking these data into consideration, GO is predictably effective on ATRA-resistant APL unless P-gp is expressed.
In the previous report, the combination of GO and ATRA was given to some patients with APL. In a study conducted in the US, GO was administered with ATRA to 19 patients with untreated APL.6 The CR rate was 16/19 (84%), and 14 became PCR-negative. In relapsed APL, Lo-Coco et al7 reported 14 cases of patients who achieved molecular remission after treatment with GO among 16 relapsed APL cases. However, there has been no in vitro study to explain these clinical efficacies. We performed the present study using NB4 and its drug-resistant sublines in an attempt to elucidate the mechanisms of GO. In APL, the drug resistance, which has been studied and discussed previously, might be built up by multiple causes and procedures.8, 13 Further studies on APL should be performed from many directions. It is also important to determine the optimal dosage of these drugs as well as the optimal timing of their administration.
GO also showed efficacy on ATO-resistant NB4 cells, which do not express P-gp. The cellular glutathione and MRP levels are reported with their relationship to ATO-resistance.10, 14 Walter et al15 reported that MRP1 might attenuate the susceptibility to GO, although by a smaller degree than P-gp. We could not find an obvious relationship between MRP1 and GO-resistance.16 Our data suggest that the MRP and the cellular glutathione levels play limited roles, while P-gp plays a major role in GO-resistance.
GO showed antiproliferative and cytocidal effects on APLs that do not express P-gp (Figure 4). GO increased the percentage of hypodiploid cells (Figure 3) while it inhibited cell proliferation in the early phase (Figures 1 and 2). After undergoing these changes, GO-sensitive cells rapidly collapsed into debris. The time-lag and variation of the effect of GO on APL cells might be explained by differences in the level of CD33 expression on the cells, and the length of time required for binding, and internalization of GO and detachment of calicheamicin from GO. Alternatively, GO could have various different actions against cells. Apoptosis, which is one of the main mechanisms of GO, did not explain all of the observed morphological changes of GO-treated cells in our previous study using videomicroscopy.3 However, analysis of the changes in cell cycle distribution could be a valuable test for analyzing the susceptibility of AML cells to GO. It has a high degree of usability for samples derived from cases that contain different phenotypes.
We confirmed the antileukemia effect of GO on APL in an in vitro study using an APL cell line and its ATRA- and ATO-resistant sublines. GO seems to be promising for the treatment of not only untreated but also relapsed APL. A larger clinical study of GO for the treatment of relapsed and refractory APL is needed. The results of such study may suggest how GO should be integrated into the management of APL.
Sievers EL, Larson RA, Stadtmauer EA, Estey E, Lowenberg B, Dombret H, et al., Mylotarg Study Group. Efficacy and safety of gemtuzumab ozogamicin in patients with CD33-positive acute myeloid leukemia in first relapse. J Clin Oncol 2001; 19: 3244–3254.
Linenberger ML, Hong T, Flowers D, Sievers EL, Gooley TA, Bennett JM et al. Multidrug-resistance phenotype and clinical responses to gemtuzumab ozogamicin. Blood 2001; 98: 988–994.
Naito K, Takeshita A, Shigeno K, Nakamura S, Fujisawa S, Shinjo K et al. Calicheamicin-conjugated humanized anti-CD33 monoclonal antibody (gemtuzumab zogamicin, CMA-676) shows cytocidal effect on CD33-positive leukemia cell lines, but is inactive on P-glycoprotein-expressing sublines. Leukemia 2000; 14: 1436–1443.
Matsui H, Takeshita A, Naito K, Shinjo K, Shigeno K, Maekawa M et al. Reduced effect of gemtuzumab ozogamicin (CMA-676) on P-glycoprotein and/or CD34-positive leukemia cells and its restoration by multidrug resistance modifiers. Leukemia 2002; 16: 813–819.
Paietta E, Andersen J, Racevskis J, Gallagher R, Bennett J, Yunis J et al. Significantly lower P-glycoprotein expression in acute promyelocytic leukemia than in other types of acute myeloid leukemia: immunological, molecular and functional analyses. Leukemia 1994; 8: 968–973.
Estey EH, Giles FJ, Beran M, O'Brien S, Pierce SA, Faderl SH et al. Experience with gemtuzumab ozogamycin (‘mylotarg’) and all-trans retinoic acid in untreated acute promyelocytic leukemia. Blood 2002; 99: 4222–4224.
Lo-Coco F, Cimino G, Breccia M, Noguera NI, Diverio D, Finolezzi E et al. Gemtuzumab ozogamicin (Mylotarg) as a single agent for molecularly relapsed acute promyelocytic leukemia. Blood 2004; 104: 1995–1999.
Gallaher RE . Retinoic acid resistance in acute promyelocytic leukemia. Leukemia 2002; 16: 1940–1958.
Rybner C, Hillion J, Sahraoui T, Lanotte M, Botti J . All-trans retinoic acid down-regulates prion protein expression independently of granulocyte maturation. Leukemia 2002; 16: 940–948.
Kitamura K, Minami Y, Yamamoto K, Akao Y, Kiyoi H, Saito H et al. Involvement of CD95-independent caspase 8 activation in arsenic trioxide-induced apoptosis. Leukemia 2000; 14: 1743–1750.
Takeshita A, Shinjo K, Naito K, Ohnishi K, Sugimoto Y, Yamakawa Y et al. Role of P-glycoprotein in all-trans retinoic acid (ATRA) resistance in acute promyelocytic leukaemia cells: analysis of intracellular concentration of ATRA. Br J Haematol 2000; 108: 90–92.
Takeshita A, Shinjo K, Naito K, Matsui H, Shigeno K, Nakamura S et al. P-glycoprotein (P-gp) and multidrug resistance-associated protein 1 (MRP1) are induced by arsenic trioxide (As(2)O(3)), but are not the main mechanism of As(2)O(3)-resistance in acute promyelocytic leukemia cells. Leukemia 2003; 17: 648–650.
Ohno R, Asou N, Ohnishi K . Treatment of acute promyelocytic leukemia: strategy toward further increase of cure rate. Leukemia 2003; 17: 1454–1463.
Yang C-H, Kuo M-L, Chen J-C, Chen Y-C . Arsenic trioxide sensitivity is associated with low level of glutathione in cancer cells. Br J Cancer 1999; 81: 796–799.
Walter RB, Raden BW, Hong TC, Flowers DA, Bernstein ID, Linenberger ML . Multidrug resistance protein attenuates gemtuzumab ozogamicin-induced cytotoxicity in acute myeloid leukemia cells. Blood 2003; 102: 1466–1473.
Naito K, Takeshita A, Matsui H, Horii T, Maekawa M, Kitamura K et al. Multidrug resistance (MDR)-related protein 1 (MRP1) and lung resistance protein (LRP) are not the main drug resistance mechanisms of gemtuzumab ozogamicin (CMA-676) in AML. Hematol J 2002; 3: 1048a.
We express our sincere gratitude to Ms Satoko Kanomi (Wyeth Pharmaceuticals Inc.) for continuous support, and to Ms Yoshimi Suzuki, Ms Noriko Anma and Dr Kiyoshi Shibata (Equipment Centre at Hamamatsu University School of Medicine) for technical assistance. This study was supported by Japanese grants-in-aid from the Ministry of Health and Welfare (No. 9-2) and the Ministry of Education and Science (No. 14570972).
About this article
Cite this article
Takeshita, A., Shinjo, K., Naito, K. et al. Efficacy of gemtuzumab ozogamicin on ATRA- and arsenic-resistant acute promyelocytic leukemia (APL) cells. Leukemia 19, 1306–1311 (2005) doi:10.1038/sj.leu.2403807
- acute promyelocytic leukemia
- gemtuzumab ozogamicin (Mylotarg)
- all-trans retinoic acid (ATRA)
- arsenic trioxide and drug resistance
Chemical metabolomics for investigating the protective effectiveness of Acanthopanax senticosus Harms leaf against acute promyelocytic leukemia
RSC Advances (2018)
Prolonged treatment with arsenic trioxide (ATO) and all-trans-retinoic acid (ATRA) for relapsed acute promyelocytic leukemia previously treated with ATRA and chemotherapy
Annals of Hematology (2018)
Dasatinib synergizes with ATRA to trigger granulocytic differentiation in ATRA resistant acute promyelocytic leukemia cell lines via Lyn inhibition-mediated activation of RAF-1/MEK/ERK
Food and Chemical Toxicology (2018)
Long-term outcome of acute promyelocytic leukemia treated with all-trans-retinoic acid, arsenic trioxide, and gemtuzumab