Multidrug resistance (MDR) genes in haematological malignancies

Development of in Silico Models for Predicting P-Glycoprotein Inhibitors Based on a Two-Step Approach for Feature Selection and Its Application to Chinese Herbal Medicine Screening

P-glycoprotein (P-gp) is regarded as an important factor in determining the ADMET (absorption, distribution, metabolism, elimination, and toxicity) characteristics of drugs and drug candidates. Successful prediction of P-gp inhibitors can thus lead to an improved understanding of the underlying mechanisms of both changes in the pharmacokinetics of drugs and drug-drug interactions. Therefore, there has been considerable interest in the development of in silico modeling of P-gp inhibitors in recent years. Considering that a large number of molecular descriptors are used to characterize diverse structural moleculars, efficient feature selection methods are required to extract the most informative predictors. In this work, we constructed an extensive available data set of 2428 molecules that includes 1518 P-gp inhibitors and 910 P-gp noninhibitors from multiple resources. Importantly, a two-step feature selection approach based on a genetic algorithm and a greedy forward-searching algorithm was employed to select the minimum set of the most informative descriptors that contribute to the prediction of P-gp inhibitors. To determine the best machine learning algorithm, 18 classifiers coupled with the feature selection method were compared. The top three best-performing models (flexible discriminant analysis, support vector machine, and random forest) and their ensemble model using respectively only 3, 9, 7, and 14 descriptors achieve an overall accuracy of 83.2%-86.7% for the training set containing 1040 compounds, an overall accuracy of 82.3%-85.5% for the test set containing 1039 compounds, and a prediction accuracy of 77.4%-79.9% for the external validation set containing 349 compounds. The models were further extensively validated by DrugBank database (1890 compounds). The proposed models are competitive with and in some cases better than other published models in terms of prediction accuracy and minimum number of descriptors. Applicability domain then was addressed by developing an ensemble classification model to obtain more reliable predictions. Finally, we employed these models as a virtual screening tool for identifying potential P-gp inhibitors in Traditional Chinese Medicine Systems Pharmacology (TCMSP) database containing a total of 13 051 unique compounds from 498 herbs, resulting in 875 potential P-gp inhibitors and 15 inhibitor-rich herbs. These predictions were partly supported by a literature search and are valuable not only to develop novel P-gp inhibitors from TCM in the early stages of drug development, but also to optimize the use of herbal remedies.

An investigation into the cytotoxicity and mode of action of some novel N-alkyl-substituted isatins

A range of substituted N-alkylisatins were synthesized and their cytotoxicity evaluated against several cancer cell lines in vitro. SAR studies indicated that the introduction of an aromatic ring with a one or three carbon atom linker at N1 enhanced the activity from that of the allyl, 2'-methoxyethyl, and 3'-methylbutyl N-substituted isatins. Furthermore, electron-withdrawing groups substituted at the meta or para position of the ring were favored over the ortho orientation. Of the 24 compounds screened, nine displayed sub-micromolar IC50 values and in general demonstrated greater selectivity toward leukemia and lymphoma cell lines over any of the carcinoma cell lines tested. 5,7-Dibromo-N-(p-methylbenzyl)isatin (6) was the most active compound, inhibiting the metabolic activity of both U937 and Jurkat cancer cell lines at 0.49 muM. Various N-alkylisatins were also found to dramatically alter lymphocyte morphology, destabilize microtubules, inhibit tubulin polymerization, induce G2/M cell cycle arrest, and activate the effector caspase-3 and -7.

MDR1 expression in poor-risk acute myeloid leukemia with partial or complete monosomy 7

Expression of the multidrug resistance (MDR1) phenotype, encoded by the MDR1 gene, is an adverse prognostic factor for CR and survival in acute myeloid leukemia (AML). Other prognostic factors, such as specific cytogenetic abnormalities, have been identified in AML. We have investigated the expression of the MDR1 gene in untreated AML patients with monosomy 7 (n = 12), and partial deletions (n = 7) of the long arm of chromosome 7 (respectively -7/7q-), because of the extremely bad prognosis associated with these cytogenetic abnormalities and because of the fact that the MDR1 gene is located on chromosome 7q21.1. The findings were compared with the level of MDR1 expression in a group of 42 other AML patients, matched for age with favourable, neutral or complex cytogenetic abberations. MDR1 mRNA expression, as measured by the RNase protection assay was significantly higher in the -7/7q- group vs other AML patients (median 1.3 vs 0.1 arbitrary units, P = 0.02). Protein expression of MDR1 in the -7/7q- group, as determined with the monoclonal antibody MRK16, was found to be similar to the levels found in the control group. With a functional rhodamine retention assay using the modulator PSC833, increased MDR1 activity was observed in the -7/7q- group as compared to the control group of patients (P = 0.05). Considering the higher MDR1 mRNA expression and equal or slightly elevated level of protein expression of MDR1, we studied the presence of MDR1 genes in this group of -7/7q- patients. Fluorescence in situ hybridization (FISH) studies, using a specific MDR1 probe revealed no loss of an MDR1 allele in any of the deleted q- arms of the seven patients with 7q-, whereas all monosomy 7 patients lacked one MDR1 gene homologue. To determine whether there was selective loss of the MDR1 gene in the -7/7q- patients, the genetic polymorphism of the MDR1 gene was used. Both allelic variants (G and T) were represented in the -7/7q- and in the control group, showing a predominance for GT at position 2677 of the MDR1 gene in the control group. In the 12 monosomy 7 patients loss of the MDR1 allele was random. Methylation studies of the CpG island of the MDR1 gene revealed no hypermethylation in any of the -7/7q- patients. We conclude that MDR1 expression in -7/7q- AML patients is upregulated at transcriptional, but not at translational level, suggesting that mechanisms other than MDR1 are responsible for the poor prognosis in these patients.

P-glycoprotein, metallothionein and NM23 protein expressions in breast carcinoma

Cellular drug resistance and increased metastatic potential are the major obstacles in the successful treatment of cancer with chemotherapy. The aim of this study was to investigate whether the immunohistochemical expression of two proteins implicated in drug resistance (P-glycoprotein and metallothionein) and the product of the suppressor gene nm23 could be related to prognosis in breast cancer. Seventy-two patients with palpable or occult breast carcinoma, not treated with chemotherapy or endocrine therapy, were examined. Immunohistochemical methods were used to determine the expression of P-glycoprotein (PG), metallothionein (MT), nm23, as well as the estrogen receptor (ER), the p53 status, and the Ki67 index. The results were correlated with clinical and morphological features. Cytoplasmic and membrane-specific immunostainings of PG were seen exclusively in tumor cells and identified in 14 of 72 cases (19.4%). Only a statistically significant association with metastases, (p = 0.06) and recurrences (p = 0.1) was observed. MT-positive reaction was identified in the cytoplasm of the tumor cells in 47 (65.3%) cases. Statistical significance was associated with metastases (p = 0.07), but not with death or recurrences. Specific immunostaining of nm23 protein was seen only in the cytoplasm of tumor cells. A positive reaction was observed in 55 of 72 (89.3%) cases. Although a significant association between nm23 protein expression and other morphologic and immunohistochemical variables did not exist, we observed a higher morbidity in patients with the MT-positive/nm23-negative tumor phenotype. Univariate analysis for survival selected the following variables: histologic grade (p = 0.001), ER (p = 0.002), mitotic index (p = 0.005), Ki 67 index (p = 0.068), MT (p = 0.046) and PG (p = 0.085). The Cox model provided the following independent variables: histologic grade (p = 0.021) and metallothionein (p = 0.03). These data confirm the prognosis observed in patients with PG or metallothionein expression as well as the independence of these two variables. It also suggests that nm23 is not necessarily involved in the development of an invasive phenotype.

Preliminary immunocytochemical studies of MDR-1 and MDR-3 Pgp expression in B-cell leukaemias

P-glycoproteins (Pgps) belong to the family of ATP binding cassette (ABC) transporter proteins. In humans two Pgp genes have been identified; mdr-1 and mdr-3. Classical Multiple Drug Resistance (MDR) is associated with over expression of the mdr-1 gene product, P-170. No role for mdr-3 in MDR has yet been proven. However there is evidence that mdr-3 overexpression may be associated with drug resistance in certain B-cell lymphocytic leukaemias. In an immunocytochemical study we have looked at a selection of B-cell leukaemias for mdr-1 and mdr-3 encoded Pgp expression using monoclonal antibodies specific for the mdr-1 and mdr-3 encoded gene products. In B-CLL patients a differential pattern of MDR-3 positive staining was observed; suggesting that MDR-3 positivity may be associated with a more malignant phenotype in B-CLL. This pattern was not observed with MDR-1 positivity. We also observed MDR-3 positivity in an AML stage M5a patient which is the first report of MDR-3 Pgp expression being detected in AML; suggesting that MDR-3 Pgp expression may be limited to particular subtypes of this disease. Results from B-NHL cases were inconclusive with varying expression of MDR-1 and MDR-3 Pgps observed. Work is currently underway to further explain the significance of these findings.

A new monoclonal antibody that specifically recognises the MDR-3-encoded gene product

The MDR-3-encoded P-glycoprotein (Pgp) is highly expressed in liver and is thought to function as a hepatic transporter of phospholipids into bile. However its role, if any, in other tissues remains undefined. Although transfection experiments have indicated that it may be unable to confer drug resistance, there is evidence that it may be involved in drug resistance in certain B-cell leukaemias. To date, most work on clinical samples has been performed at the mRNA level; limited work has been performed using polyclonal antibodies raised to MDR-3 and mdr-2 (the murine equivalent of MDR-3). We have generated a new monoclonal antibody, termed 6/1G, which specifically recognises the human MDR-3 gene-encoded product. Antibody 6/1G was produced by in vitro immunisation of spleen cells from BALB/c mice with a synthetic 12-amino acid peptide. Cells from MDR-3 transgenic mice showed consistent membranous staining with antibody 6/1G. Immunoblotting with 6/1G identified a band at 170 kDa on lysates of MDR-3 transgenic cells. Preliminary results with a range of B-cell leukaemias suggest that MDR-3 Pgp positivity may be a marker for a more malignant phenotype in B-CLL. Antibody 6/1G may be useful in defining a role for MDR-3 in malignancy and drug resistance, as well as in certain liver diseases such as progressive familial intracholeostasis.

Prognostic value of P-glycoprotein and leukocyte differentiation antigens in chronic myeloid leukemia

P-glycoprotein (Pgp) mediated multidrug resistance is often the cause of therapy failure in some tumors. Pgp expression was shown to have prognostic value in several hematological malignancies, especially in acute myeloblastic leukemia (AML) and acute lymphoblastic leukemia (ALL). In chronic myeloid leukemia (CML) Pgp is expressed by peripheral blood (PB) cells more often in the terminal disease stages (20-50% of patients have Pgp+ phenotype). Sequential studies show that Pgp+ cells often disappear from the PB during the course of therapy. Nevertheless Pgp expression has some prognostic value in blast crisis (BC) predicting shorter BC, while CD13 has the same predictive value in BC. 10% of patients formed a distinct group with large numbers of Pgp+CD34+ blasts in the PB and also had shorter BC. Cases with inactive Pgp were found in chronic and accelerated phases of CML but not in BC.

Molecular mechanisms of multidrug resistance in cancer chemotherapy

The occurrence of multidrug resistance (MDR) is one of the main obstacles in the successful chemotherapeutic treatment of cancer. MDR cell lines are resistant to the so-called naturally occurring anti-cancer drugs, such as anthracyclines, Vinca alkaloids and epipodophyllotoxins, but are not cross-resistant to alkylating agents, antimetabolites and cisplatin. So far, three separate forms of MDR have been characterized in more detail: classical MDR, non-Pgp MDR and atypical MDR. Although all three MDR phenotypes have much in common with respect to cross-resistance patterns, the underlying mechanisms certainly differ. Atypical MDR is associated with quantitative and qualitative alterations in topoisomerase II alpha, a nuclear enzyme that actively participates in the lethal action of cytotoxic drugs. Atypical MDR cells do not overexpress P-glycoprotein, and are unaltered in their ability to accumulate drugs. In this review we will focus on classical and non-Pgp MDR. The molecular mechanism of classical and non-Pgp MDR is transcriptional activation of membrane-bound transport proteins. These transport proteins belong to the ATP-binding cassette (ABC) superfamily of transport systems. The classical MDR phenotype is characterized by a reduced ability to accumulate drugs, due to activity of an energy-dependent uni-directional, membrane-bound, drug-efflux pump with broad substrate specificity. The classical MDR drug pump is composed of a transmembrane glycoprotein (P-glyco-protein-Pgp) with a molecular weight of 170 kD, and is, in man, encoded by the so-called multidrug resistance (MDR1) gene. Typically, non-Pgp MDR has no P-gly-coprotein expression, yet has about the same cross-resistance pattern as classical MDR. This non-Pgp MDR phenotype is caused by overexpression of the multidrug resistance-associated protein (MRP) gene, which encodes a 190 kD membrane-bound glycoprotein (MRP). MRP probably works by direct extrusion of cytotoxic drugs from the cell and/or by mediating sequestration of the drugs into intracellular compartments, both leading to a reduction in effective intracellular drug concentrations. For the classical MDR phenotype, evidence is accumulating that it plays a role indeed, in clinical drug resistance, especially in some hematological malignancies (acute myeloid leukemia, multiple myeloma and non-Hodgkin's lymphoma) and solid tumors (soft tissue sarcomas and neuroblastoma). The association of MRP with clinical drug resistance has not been elaborated, yet, and studies on MRP expression in human cancer have just begun. We found that overexpression of MRP, as determined by RNase protection assay as well as by immunohistochemistry, occurs in several human cancers, among which are cancer of the lung, esophagus, breast and ovary, and leukemias. Further studies are indicated to establish whether elevated MRP expression at diagnosis is an unfavorable prognostic factor for clinical outcome of chemotherapy.

Multidrug resistance-associated protein (MRP) in haematological malignancies

The presence of multidrug resistant cells, either acquired or de novo, severely limits treatment outcome in haematological malignancies. Although expression of the Mr 170,000 P-glycoprotein drug pump is likely to play a role in multidrug resistance (MDR) in haematological malignancies, it is now evident that other MDR mechanisms may be operational as well in leukaemias, lymphomas, and multiple myeloma. We determined the expression of a newly recognised drug resistance gene, the Multidrug Resistance-associated Protein (MRP) gene, in peripheral blood cells from healthy volunteers and from patients with haematological malignancies. Expression of MRP mRNA and its Mr 190,000 glycoprotein were estimated by RNase protection assay and immunocytochemistry, respectively. MRP appeared to be ubiquitously expressed at low levels in all nonmalignant haemopoietic cell types. However, some leukaemias showed elevated levels of MRP, probably due to transcriptional activation or increased mRNA stability. High to very high MRP expression levels were frequently found in chronic lymphocytic leukaemia and prolymphocytic leukaemia. Acute myelocytic leukemia often exhibited low but occasionally high MRP expression levels, while in the other acute and chronic leukaemias, lymphomas, and multiple myeloma, predominantly low, basal levels of MRP were found. We conclude that hyperexpression of MRP is observed in leukaemias, and that further studies are needed to assess the clinical relevance of MRP.

High expression of the multidrug resistance-associated protein (MRP) in chronic and prolymphocytic leukaemia

The expression of the multidrug resistance-associated protein (MRP), a new glycoprotein involved in drug resistance, was investigated in tumour samples from 80 patients with chronic B-cell malignancies by a quantitative RNase protection assay. In B-cell chronic lymphocytic leukaemia (B-CLL) (n = 32), either treated (n = 18) or untreated (n = 14), a high percentage of patients (20/32: 63%) had relatively high expression levels of the MRP gene (25U or more). In addition, hyperexpression of the MRP gene was demonstrated in 4/10 (40%) untreated patients with B-cell prolymphocytic leukaemia (B-PLL). In contrast, low MRP mRNA expression levels were detected in hairy cell leukaemia (n = 7), non-Hodgkin's lymphoma (n = 13) and multiple myeloma (n = 18). Statistical analysis of MRP expression in untreated CLL (mean +/- SD 29.2 +/- 18.5 U) versus treated CLL (mean +/- SD 26.7 +/- 13.7 U) did not show significant differences in MRP expression between the two groups. Southern blot analysis did not reveal amplification of the MRP gene in the leukaemia samples with elevated MRP mRNA levels. We conclude that B-PLL and B-CLL frequently display high MRP expression and that this hyperexpression is probably due to transcriptional activation and/or increased mRNA stability.