Search

The study of the cellular, biochemical, and molecular biology and pharmacology of MDR has provided one of the most active and exciting areas within cancer research and one that holds great promise for translation into clinical benefit. While convincing evidence for the functional role of P-gp in mediating clinical drug resistance in humans remains elusive, studies of the clinical expression of P-gp and trials of chemosensitizers with cancer chemotherapy suggest "resistance modification" strategies may be effective in some tumors with intrinsic or acquired drug resistance. However, even if P-gp-associated MDR proves to be a relevant and reversible cause of clinical drug resistance, numerous problems remain to be solved before effective clinical chemosensitization may be achieved. Such factors as absorption, distribution, and metabolism; the effect of chemosensitizers on chemotherapeutic drug clearance; toxicity to normal tissues expressing P-gp; and the most efficacious modulator regimens all remain to be defined in vivo. Clearly, the identification of more specific, potent, and less clinically toxic chemosensitizers for clinical use remains critical to the possible success of this approach. Nonetheless, the finding that a number of pharmacological agents can antagonize a well-characterized form of experimental drug resistance provides promise for potential clinical applications. Further study of chemosensitizers in humans and the rational design of novel chemosensitizers with improved activity should define the importance of MDR in clinically resistant cancer.

The phosphorylation of P-glycoprotein has been appreciated for many years, yet little is known about the factors that initiate this post-translational modification. To determine whether the activation of P-glycoprotein phosphorylation could occur in response to cellular stress and to investigate the possible signal pathways involved, we studied the effect of heat shock on the phosphorylation of P-glycoprotein in sensitive and resistant MCF-7 human breast cancer cells. Treatment of multidrug resistant MCF-7/AdrR cells with heat shock increased the phosphorylation of P-glycoprotein. The response was not seen in the sensitive MCF-7 line, which does not express this drug transporter. Phosphorylation of P-glycoprotein induced by heat shock was not dependent on synthesis of new proteins, since phosphorylation was not inhibited by cycloheximide and the content of P-glycoprotein, as measured by immunoblotting, did not change after heat shock. The activation of P-glycoprotein phosphorylation by heat shock may be initiated through activation of phospholipase C, since heat shock stimulated the activity of this enzyme, as evidenced by increased formation of inositol trisphosphate and diacylglycerol and by phosphorylation of phospholipase C-gamma. U-73122, an inhibitor of phospholipase C and staurosporine, an inhibitor of protein kinase C, both decreased the heat-shock-induced phosphorylation of P-glycoprotein. These results suggest that heat shock induces phosphorylation of P-glycoprotein through the activation of the phospholipase C/protein kinase C pathway.

The mitogenic activity of several growth factors is mediated by calcium-dependent signal transduction. Calmodulin (CaM) binding proteins such as CaM-dependent protein kinases are important components of this pathway and may be altered in diseases characterized by abnormal cell growth. CaM kinase II is believed to regulate the phosphorylation of microtubular-associated proteins and control the initiation of DNA synthesis. Furthermore, drugs that inhibit CaM-mediated signal transduction also inhibit cellular proliferation and are cytotoxic to numerous malignant cell lines, including those established from malignant gliomas. Yet, little is known about CaM-dependent protein kinases in these tumors. Therefore, we have investigated the activity and distribution of CaM-dependent protein kinase II in normal and malignant glial tissues, a kinase believed to play a critical role in cell cycle regulation. C6 and 9L cells contained kinase activities that were activated by Ca2+/CaM and inhibited by trifluoperazine. Tissue extracts from these cell lines and from rat brain white matter phosphorylated exogenous synapsin I in a pattern consistent with the presence of CaM kinase II activity as determined by phosphopeptide mapping. CaM kinase II activity was confirmed using a specific peptide substrate and inhibitor. An unexpected finding was that glioma lines, but not rat brain white matter, also contained a CaM-dependent protein kinase detected by the phosphorylation of a M(r) 100,000 protein, subsequently identified as elongation factor 2, the only known substrate for CaM kinase III.(ABSTRACT TRUNCATED AT 250 WORDS)

Calmodulin (CaM) is believed to play an important role in the regulation of cellular proliferation. The mechanism of regulation, although unknown, may involve CaM-binding proteins, particularly CaM-dependent protein kinases. Previously, we have shown that CaM-dependent protein kinase III phosphorylates elongation factor 2 (EF-2) in proliferating, C6 glioma cells but not in normal white matter, a tissue rich in nonproliferating glia. To determine whether CaM-dependent phosphorylation of EF-2 is linked, in general, to cellular division, we studied the phosphorylation of EF-2 in proliferating and growth-arrested C6 cells and in proliferating, primary cultures of normal glia. Phosphorylation of EF-2 was not detectable in C6 cells arrested in their growth by serum deprivation. When serum-deprived cells were stimulated to proliferate by the re-addition of serum, the amount of phosphorylated EF-2 correlated with levels of [3H]thymidine incorporation into DNA. Primary cultures of dividing, normal glia, obtained from neonatal rats, also demonstrated phosphorylation of EF-2. Therefore, the CaM-dependent phosphorylation of EF-2 appears to be associated with cellular proliferation in normal and malignant glia in the rat.

Antipsychotic drugs that bind to and inhibit the action of calmodulin also inhibit cellular proliferation. In addition these drugs are cytotoxic to most malignant cells and can augment the antiproliferative and cytotoxic effects of bleomycin. They are attractive candidates for use against tumors of the central nervous system since they readily pass the blood-brain barrier and accumulate in the brain. To identify more active derivatives, we studied the effect of a series of phenothiazines and a group of related compounds alone or in combination with bleomycin against rat glioblastoma cell lines. C6 cells were grown for 24 hours prior to a 48 hour exposure to anti-psychotic drug alone or to an IC20 concentration of antipsychotic drug with bleomycin. Cells were stained with methylene blue and enumerated spectrophotometrically. Eight phenothiazines were found to augment the effect of bleomycin by > or = 3-fold. These included 1-chlorpromazine (3.8x), chlorpromazine (3.2x), 3-chlorpromazine (3.0x), 4-chlorpromazine (3.4x), thiomethylpromazine (3.3x), didesmethylchlorpromazine (11x), fluphenazine (5.5x) and trifluoperazine (3.2x). Structurally similar compounds also having activity included trans-flupenthixol (6.0x), 2-chloroimipramine (6.0x), desipramine (22x), and penfluridol (24x). There was a direct correlation between the antiproliferative effect of anticalmodulin compounds and the ability of these drugs to inhibit the activation of calmodulin-sensitive phosphodiesterase. However, there was no correlation between the inhibition of calmodulin and the augmentation of the antiproliferative activity of bleomycin. Penfluridol, one of the most active compounds, was chosen for further study. It increased the activity of bleomycin against L1210 leukemic cells by 90-fold and MCF-7 human breast cancer cells by 4-fold. The effect of penfluridol in combination with bleomycin was due to increased cytotoxicity as measured by clonogenic assay.

We conducted a Phase II study to determine the efficacy of cisplatin, doxorubicin, mitomycin C, and 5-fluorouracil in patients with untreated non-small cell lung cancer. Patients were accrued through the Connecticut Oncology Association (COA), an organization composed of community and university oncologists. Thirteen COA oncologists enrolled 30 patients over 12 months and 26 were eligible for the final analysis. Patients received cisplatin 75 mg/m2, doxorubicin 30 mg/m2, mitomycin C 6.5 mg/m2, and 5-fluorouracil 750 mg/m2 on day 1. The treatment was repeated every 4 weeks, with mitomycin C given during the first 3 cycles and then every other cycle. There were 5 (19%) partial responses, lasting 1.3-7.3 months. The median time to progression was 10.3 months (0.3-12.5 months). Median survival was 7.5 months (0.3-34 months). The major toxicities were related to myelosuppression and there was one septic death. This study demonstrates the limited efficacy of an aggressive regimen using "active" agents in patients with advanced non-small cell lung cancer.

Cross-resistance to chemotherapeutic drugs is a significant problem in the treatment of patients with cancer. The discovery that this phenomenon is associated with the overexpression of a membrane glycoprotein, P-glycoprotein, which acts as a drug efflux pump, has provided a new target for drug development. To develop a model for identifying new compounds which can block the function of P-glycoprotein, we infected P388 mouse leukemic cells with a retrovirus containing a cloned human MDR1 complementary DNA. The new cell line, P388/VMDRC.04, incorporated and overexpressed the human gene as evidenced by Southern blots, increased mRNA and protein synthesis, and recognition by the MRK16 monoclonal antibody. P388/VMDRC.04 was cross-resistant to colchicine, vincristine, and doxorubicin, and the degree of resistance correlated with a reduction in cellular drug accumulation. Unlike many cell lines selected for resistance by growth in increasing concentrations of drug for prolonged periods of time, these cells did not show alternative mechanisms of resistance such as increased synthesis of glutathione or alterations in topoisomerase II. In addition, the sensitivity of P388/VMDRC.04 cells was completely restored by cyclosporin A and trans-flupenthixol. P388/VMDRC.04 cells were subcloned and 10 clones were picked for in vivo evaluation. One subclone grew similarly to parental cells in female BALB/c x DBA/2 F1 mice and showed no responsiveness to therapeutic doses of vincristine or etoposide. The combination of vincristine with cyclosporin A significantly increased the survival of mice inoculated with P388/VMDRC.04 cells. The availability of a cell line that displays the MDR phenotype, overexpresses human P-glycoprotein, but does not contain alterations in at least two well-defined alternative mechanisms of resistance, and that can be grown in simple animal models should facilitate the development of new agents active against this form of chemotherapeutic drug resistance.

Multidrug resistance (MDR) is the phenomenon in which cells become resistant to several classes of structurally and functionally diverse drugs after exposure to a single cytotoxic agent. One form of MDR is associated with the overexpression of a large plasma membrane phosphoglycoprotein, P-glycoprotein, which acts as an energy-requiring drug transport pump. Protein kinase C may participate in MDR through posttranslational modification of P-glycoprotein. The purpose of this study is to critically evaluate P-glycoprotein as a substrate for protein kinase C and to determine whether phosphorylation leads to changes in drug transport. Protein kinase C from rat brain phosphorylated immunoprecipitated P-glycoprotein in a manner dependent on the activation of the exogenous kinase. Phorbol 12-myristate 13-acetate (PMA) increased the phosphorylation of P-glycoprotein 6-fold and selectively decreased the accumulation of vinblastine in resistant MCF-7/AdrR cells. PMA selectively decreased the cellular association of vinblastine with MDR cells after brief periods of incubation, but only after critical concentrations of drug were achieved. The actions of PMA did not require new synthesis of P-glycoprotein. PMA had similar effects in MCF-7/BC-19, a cell line transfected with a cDNA for P-glycoprotein. Staurosporine inhibited the effects of PMA on the phosphorylation of P-glycoprotein and on the accumulation of vinblastine. These data demonstrated that immunoprecipitated P-glycoprotein can be a substrate for protein kinase C, and that phosphorylation of the transporter is associated with significant changes in drug transport.

Certain calmodulin (CaM)-dependent protein kinases phosphorylate substrates have been implicated in regulating cellular proliferation. In this study, CaM-dependent phosphorylation has been examined in normal and tumor tissue from rat brain to determine whether differences exist. Using in vitro phosphorylation reactions, we compared endogenous substrates for Ca2+/CaM-dependent protein kinases in rat brain white matter (RBWM), a tissue rich in normal glia, to those of C6 rat glioma cells. A major phosphoprotein having a M(r) of 100,000 was observed in proliferating C6 cells that was not present in RBWM or in nonproliferating cells. Phosphorylation was stimulated by Ca2+ and CaM and inhibited by trifluoperazine. An antibody to elongation factor 2 (EF-2) immunoprecipitated the M(r) 100,000 protein from C6 cells. EF-2 was present in RBWM but was not phosphorylated. Homogenates of RBWM did not phosphorylate exogenous EF-2, which suggested the absence of CaM kinase III activity in normal glial tissue. Furthermore, the addition of purified, exogenous CaM kinase III to homogenates of RBWM resulted in EF-2 phosphorylation. These data demonstrate that a basal level of EF-2 phosphorylation exists in proliferating glioma cells that is markedly diminished or absent in normal glial tissue and is due to the activity of CaM kinase III.

We conducted a study evaluating the efficacy of simultaneous administration of 5-fluorouracil, mitomycin C, and X-irradiation in unresectable Stage III non-small cell lung cancer. Radiation was delivered in a split course to a total of 55 Gy. In the first course, radiation was administered daily in 250 cGy fractions for 12 treatments. Following a two week rest, an additional 10 treatments of 200 cGy fractions were delivered. Chemotherapy consisted of 10 mg/M2 of mitomycin C on day 1, and 1000 mg/M2 per day of 5-fluorouracil by continuous infusion on days 1 through 5 and 29 through 33. Twenty-one enrolled patients were evaluable. The overall response rate was 52% (95% confidence intervals 31-73%), with a median duration of response of 42 weeks. There were 5 complete responders (24%) and four of these patients remain disease-free with follow-up between 24-54 months. The median survival time for all patients was 59 weeks. For the complete responders the median survival time has not yet been reached, but is in excess of 42 months. Toxicity consisted of moderate stomatitis and myelosuppression. The sustained remissions obtained by 4 of 21 patients (18%) treated with concurrent mitomycin C, 5-fluorouracil and x-irradiation are encouraging and deserve further study.

The ability of malignant cells to develop resistance to chemotherapeutic drugs is a major obstacle to the successful treatment of clinical tumors. The phenomenon multidrug resistance (MDR) in cancer cells results in cross-resistance to a broad range of structurally diverse antineoplastic agents, due to outward efflux of cytotoxic substrates by the mdr1 gene product, P-glycoprotein (P-gp). Numerous pharmacologic agents have been identified which inhibit the efflux pump and modulate MDR. The biochemical, cellular and clinical pharmacology of agents used to circumvent MDR is analyzed in terms of their mechanism of action and potential clinical utility. MDR antagonists, termed chemosensitizers, may be grouped into several classes, and include calcium channel blockers, calmodulin antagonists, anthracycline and Vinca alkaloid analogs, cyclosporines, dipyridamole, and other hydrophobic, cationic compounds. Structural features important for chemosensitizer activity have been identified, and a model for the interaction of these drugs with P-gp is proposed. Other possible cellular targets for the reversal of MDR are also discussed, such as protein kinase C. Strategies for the clinical modulation of MDR and trials combining chemosensitizers with chemotherapeutic drugs in humans are reviewed. Several novel approaches for the modulation of MDR are examined.

Drugs that interfere with the action of P-glycoprotein (P-gp), the membrane efflux pump responsible for multidrug resistance (MDR), should be valuable in the treatment of patients with drug-resistant cancer. We have used one class of drug, the phenothiazines, to study the structural features required for optimum interference with the function of P-gp. The structure-activity relationships revealed three important components including the hydrophobicity of the tricyclic ring, the length of the alkyl bridge and the charge on the terminal amino group. Trans-flupenthixol is a lead compound that conforms to these structural requirements and demonstrates significant activity as a sensitizer of MDR cell lines to drugs affected by the MDR phenotype. Based on these data, we have proposed a model for the binding of modulators to P-gp and have speculated on the structure of the drug-binding domain. We have developed pre-clinical models of MDR that may help predict clinical activity of chemo-modulators. L1210/VMDRC.06 is a murine lymphocytic leukemia line transformed by a retroviral expression vector containing a full-length cDNA for the human mdr1 gene. K562/VBL1-3 are clones of human myeloid blast cells that were transformed with the same vector. Resistance in these lines is not complicated by changes in the cellular content of glutathione or alterations in topoisomerase II. The transformed L1210 line grows in mice as a slowly proliferating non-metastatic peritoneal implant. Both MDR lines are restored to sensitivity by cyclosporin A or trans-flupenthixol, and the K562 clones are induced to differentiate by hemin. These lines should provide simple, sensitive screens for new drugs for use against cancers expressing P-gp. We have proposed a model to explain how the pumping activity of P-gp is activated in response to toxic drugs. In this schema, basal activity of P-gp is modulated through phosphorylation/dephosphorylation reactions mediated by protein kinase C (PKC) and calcium sensitive phosphatases. In response to the activation of phospholipase C by toxic drugs and the local production of 1,2-diacylglycerol, PKC is translocated to the cell membrane where it phosphorylates P-gp. Following the extrusion of drug from the cell membrane, phospholipase C activity returns to baseline, diacylglycerol is metabolized, PKC returns to the cytosol and serine/threonine phosphatases dephosphorylate P-gp returning it to the basal state.

Lung cancer is the major cause of cancer mortality. Locally advanced (Stage III) disease constitutes 30 to 40% of the entire group of non-small cell lung cancer (NSCLC). Surgical resection offers the best opportunity for cure, but resection of disease is possible in only a minority of patients with Stage IIIa disease. Even among patients who have "successful" surgery systemic relapse is common, and the 5-yr survival after complete resection is only 30%. Preoperative (neoadjuvant) chemotherapy is under investigation in an attempt to improve the bleak outcome of patients with Stage IIIa NSCLC. Preliminary trials have shown that this approach is feasible: neoadjuvant treatment can be administered with moderate toxicity and in most cases without compromising the possibility for surgical resection. In some instances, neoadjuvant treatment has produced pathologic complete responses, and in others it has decreased tumor bulk so that inoperable patients became surgical candidates. Whether this latter phenomenon has an impact on survival is unknown. Therefore, the role of neoadjuvant treatment for locally advanced lung cancer will not be known until properly designed randomized trials are conducted.

The phorbol ester receptor (PER), which is believed to be identical to protein kinase C (PKC), has been implicated in the control of the sensitivities of tumor cells to certain forms of cancer chemotherapy. We evaluated the effects of regulating the expression of PER/PKC on the sensitivity of the R1B6 subclone of HL-60 human promyelocytic leukemic cells to anthracyclines. This cell line is resistant to the effects of phorbol esters on cellular differentiation by virtue of a down-regulated PER. R1B6 cells were maintained in phorbol 12, 13 dibutyric acid (PDBu). Twenty four to thirty six hours after removal of PDBu from the medium there was a 3-fold increase in the number of receptors compared to baseline values measured by [3H]-PDBu binding and a marked increase in the activity of PKC as measured by calcium-phospholipid dependent incorporation of [32P] into histone. Despite these changes in the number of PER and in the activity of PKC there was no difference in the cellular accumulation of [3H]-daunomycin or in the sensitivity of cells to the toxic effects of doxorubicin. Furthermore, there was no difference between the R1B6 cells and the parental HL-60 line in their intrinsic sensitivities to doxorubicin. These studies demonstrate that alterations in the expression of PER and the activity of PKC alone are not sufficient to influence anthracycline accumulation or toxicity in R1B6 human leukemic cells in contrast to recently reported results using other cell lines.

Resistance of tumor cells to chemotherapeutic drugs may be due to several mechanisms within a single cell line. Resistance to doxorubicin in the human multidrug resistant breast cancer cell line, MCF-7 AdrR, has been attributed to increased glutathione (GSH) S-transferase and GSH peroxidase activity, as well as to increased expression of the mdr1 gene product, P-glycoprotein. We studied the potentiation of doxorubicin activity in these cells by buthionine sulfoximine (BSO), a specific inhibitor of gamma-glutamylcysteine synthetase, and by verapamil and trans-flupenthixol, agents which interact with P-glycoprotein. Treatment with BSO enhanced the effect of doxorubicin by 1.5-fold, while verapamil or transflupenthixol caused a greater reversal of drug resistance. The combination of BSO with trans-flupenthixol produced no further potentiation of doxorubicin activity. However, the combination of BSO with verapamil and doxorubicin caused up to a 10-fold increment in antiproliferative effect. To explore the mechanism by which BSO interacted with this drug combination, we determined whether or not BSO might potentiate the effects of verapamil. These studies demonstrated that the effects of BSO were predominantly due to an increase in verapamil toxicity rather than to doxorubicin toxicity. In addition, when mice received concentrations of BSO in their drinking water sufficient to deplete GSH and were treated with verapamil, the calcium channel blocker was lethal to 9 of 12 mice receiving BSO compared to 1 of 10 control animals receiving verapamil alone. These studies demonstrate that BSO does not markedly increase the pharmacological effect of doxorubicin against MCF-7 AdrR cells and suggest that alterations in GSH and related enzymes are not a major factor in drug resistance in this cell line. Furthermore, BSO can increase the toxicity of verapamil, a finding which may have important implications for clinical trials.

Several cell lines resistant to alkylating agents possess increased activity of glutathione-S-transferase (GST) drug detoxifying enzymes. Inhibition of certain enzymes of the glutathione redox system may affect cellular sensitivity to alkylators. We report that the (-.)enantiomer of gossypol is a potent and selective inhibitor of GST alpha and GST pi isozymes, and that in combination with buthionine sulfoximine (BSO), causes the enhanced modulation of alkylator resistance in two drug resistant cell lines with increased GST activity. The use of (-)gossypol alone had no effect on the 2-5-fold resistance of MCF-7 Adr and Walker resistant cells to chlorambucil, melphalan and BCNU. Cellular depletion of glutathione with BSO resulted in a 2-4-fold modulation of cell sensitivity to these alkylators. However, the combination of (-)gossypol with BSO resulted in a markedly greater modulation of alkylator sensitivity than with either inhibitor alone. Therefore, the complementary inhibition of glutathione and GST by BSO and (-)gossypol, respectively, produced a synergistic modulation of alkylator cytotoxicity in these drug resistant cell lines. The favorable clinical pharmacokinetics of (-)gossypol suggest its further evaluation for use in combination with BSO and alkylating agents in clinical trials.

Glioblastoma multiforme is a fatal malignancy of the central nervous system, demanding new methods of treatment. The combination of a calmodulin antagonist with bleomycin has shown synergistic activity in several preclinical models and has been evaluated in a Phase I clinical trial. Since phenothiazines reach high concentrations in the central nervous system, and bleomycin has been reported to have antitumoral activity as well, we studied this combination in a Phase II clinical trial. In addition, we purified calmodulin from normal brain and malignant gliomas to determine its biochemical and pharmacological characteristics. Seventeen patients were entered onto this study and all were evaluable. There were no partial or complete responses. There was one case of fatal pulmonary toxicity in a patient showing an objective tumor response. Otherwise, the treatment was well tolerated. Calmodulin purified from the normal brain and gliomas of patients undergoing resection was identical to each other and to calmodulin prepared from rat cerebrum and glioma. These characteristics included elution from a TSK phenyl high pressure liquid chromatography column, migration on 16% sodium dodecyl sulfate gels, amino acid composition, and inhibition by drugs. Therefore, the failure of this combination therapy was not due to a difference in human glioma calmodulin as compared to previously reported studies with calmodulin from murine sources.

We have devised a rapid procedure for the assay of protein kinase C. Reactions (100 microliters) were set up in the wells of a 96-well assay plate and initiated one column at a time by the addition of [gamma-32P]ATP with an eight-channel pipettor. After incubation for 5 min at 30 degrees C, the reactions were terminated by the addition of 100 microliters of 25% trichloroacetic acid. The reaction mixtures were then filtered using a semiautomatic cell harvester, and transferred to scintillation vials using a filter punch apparatus. Direct comparison of this method to traditional techniques revealed a three- to fivefold increase in efficiency with equal sensitivity. This method is suitable for screening large numbers of inhibitors and activators of protein kinase C and appears to be applicable to other enzymes such as calmodulin kinases.

We have previously shown that phenothiazines sensitize multidrug resistant (MDR) cells to chemotherapeutic drugs in a manner related to specific structural features, and have identified structurally related thioxanthenes with increased anti-MDR activity. We have now studied the structure-activity relationships of 16 thioxanthenes in the human breast cancer line MCF-7 AdrR. trans-Thioxanthene stereoisomers were 2- to 7-fold more potent than cis-thioxanthenes for antagonizing MDR. The most potent thioxanthenes possessed a halogenated tricyclic ring connected by a 3-carbon alkyl bridge to a piperazinyl or piperadinyl side group. The chemosensitizing effects of the lead compound, trans-flupenthixol, its stereoisomer cis-flupenthixol, its phenothiazine homologue fluphenazine, and the calcium channel blocker verapamil, were further studied in a series of sensitive and MDR cell lines. trans-Flupenthixol caused a greater reversal of cellular resistance to doxorubicin, vinblastine, vincristine, and colchicine in MCF-7 AdrR, KB-V1, and P388/DOX MDR cells than the other chemosensitizers. In particular, trans-flupenthixol was 2- to 3-fold more potent for reversing MDR than equimolar concentrations of verapamil. Furthermore, trans-flupenthixol fully reversed resistance to doxorubicin, vincristine, and colchicine in MDR MCF-7 and NIH 3T3 cells transfected with the mdr1 gene. None of these agents altered MDR in a non-P-glycoprotein expressing MCF-7 cell line selected with mitoxantrone, nor in any of the parental cell lines. The stereoselective antagonism of the flupenthixol isomers on several putative cellular targets was studied to explore the mechanism of their chemosensitizing activity. cis- and trans-flupenthixol were equally active inhibitors of protein kinase C and calmodulin. Both cis- and trans-flupenthixol were also potent inhibitors of [3H]azidopine binding to P-glycoprotein. The apparent lack of clinical toxicity of trans-flupenthixol makes it an attractive drug for possible use in the modulation of tumor resistance in vivo if appropriate tissue concentrations can be achieved.

Dequalinium belongs to a group of cationic lipophilic drugs believed to be selectively cytotoxic to malignant cells of epithelial origin by virtue of their accumulation in mitochondria. In addition, they are potent inhibitors of calmodulin and, therefore, might sensitize multidrug-resistant cells to chemotherapeutic agents. We compared the responsiveness of multidrug-resistant cells to the effect of dequalinium with that to trifluoperazine, a potent phenothiazine inhibitor of calmodulin. In addition, we studied the effect of these drugs on the responsiveness of multidrug-resistant cell lines to doxorubicin. The effect of drugs on P388 murine leukemic cells was determined by cell counting, [3H]thymidine incorporation into DNA, or soft agar cloning. Drug accumulation was measured by fluorescence spectrophotometry. We found that multidrug-resistant lines were less sensitive than parental cell lines to the intrinsic growth inhibitory effects of dequalinium (IC50, 4.4 versus 0.3 microM in multidrug-resistant and sensitive P388 cells, respectively), whereas they were equally sensitive as the parental line to the effects of trifluoperazine. Following a 3-h exposure of P388/doxorubicin-resistant cells to 0-100 microM doxorubicin with or without either 10 microM dequalinium or 10 microM trifluoperazine, the latter increased the sensitivity to doxorubicin whereas the former had little effect (IC50 values were doxorubicin, 30 microM; doxorubicin plus dequalinium, 25 microM; doxorubicin plus trifluoperazine, 4 microM). Calmodulin prepared from resistant cells were equally sensitive to inhibition by dequalinium and trifluoperazine. P388/doxorubicin-resistant cells accumulated 4.5-fold less dequalinium than P388 cells whereas trifluoperazine was accumulated equally in both. The addition of 4 microM trifluoperazine to resistant cells exposed to 0-100 microM dequalinium completely reversed the alteration in accumulation and resistance to the dequalinium. These studies demonstrate that certain multidrug-resistant lines are cross-resistant to dequalinium and that sensitivity can be completely restored by nontoxic concentrations of trifluoperazine. The resistance appears to be due to changes in drug accumulation and not to be related to an altered sensitivity of calmodulin.

We have characterized the antiproliferative effects of the phenothiazines, a group of antipsychotic drugs possessing a wide range of pharmacological actions. The phenothiazines inhibited both the proliferation and clonogenicity of L1210 leukemic lymphocytes. This effect was dependent on both time of exposure and concentration of drug. Clonogenicity of cells in the logarithmic phase of growth was inhibited by greater than 99% at a concentration of drug that had no effect on cells in the plateau phase of growth. Human and murine cell lines, grown either in suspension or in monolayers, were equally susceptible. Calmodulin (CaM), purified from L1210 cells by preparative polyacrylamide gel electrophoresis, had sensitivity to inhibition by phenothiazines similar to that reported for CaM prepared from brain. The order of potency was trifluoperazine greater than or equal to fluphenazine greater than chlorpromazine greater than chlorpromazine-sulfoxide. As a class, these drugs were less potent antagonists of CaM than was the bee venom polypeptide, melittin. The antiproliferative effects of phenothiazines were similar to the anticalmodulin effects. Thus, the same order of potencies was seen for both effects; the shapes of the dose-response curves were similarly steep and the effects of excess calcium on the inhibition of both were identical. These studies add pharmacological support for CaM being a potential intracellular target for the antiproliferative effect of the phenothiazines.

Certain cationic-lipophilic compounds are known to selectively accumulate in tumor mitochondria and inhibit energy production. Since these substances bear a structural resemblance to known inhibitors of calmodulin, we studied whether rhodamine-123 or a bis-4-aminoquinaldinium could antagonize the action of calmodulin. Rhodamine-123 (IC50 = 58 microM) and dequalinium (IC50 = 1 microM) inhibited the activity of a calmodulin-stimulated cyclic nucleotide phosphodiesterase. Propylinium, a compound similar to dequalinium except for having a 3 rather than 10 carbon alkyl bridge connecting two non-substituted quinoline rings, had no inhibitory effect. Kinetic analysis showed that dequalinium competitively inhibited calmodulin's activation of phosphodiesterase. We also studied the antiproliferative effects of the compounds on the C6 astrocytoma cell line. Rhodamine-123 and dequalinium inhibited the proliferation of this cell line while propylinium had no effect. These studies demonstrate that rhodamine-123 and dequalinium are calmodulin-antagonists and inhibit cellular proliferation.

The effect of bleomycin and calmodulin antagonists on human cells was studied using a clonogenic assay. A 1-h exposure to nontoxic concentrations of the calmodulin antagonists melittin (0.5 microM), pimozide (5 microM), and chlorpromazine (20 microM) increased the lethality of bleomycin to human ovarian carcinoma cells (SK-OV). No increase was seen with chlorpromazine sulfoxide, which lacks calmodulin antagonistic activity. Maximum enhancement of bleomycin lethality by calmodulin antagonists was seen when the antagonist was present simultaneously with bleomycin rather than before or after bleomycin. The cytotoxicity of bleomycin to A-253 head and neck squamous carcinoma cells, which were 10-fold more sensitive to bleomycin alone compared to SK-OV cells, was not markedly altered by the presence of 20 microM chlorpromazine. Chlorpromazine, melittin, or pimozide also increased the toxicity of bleomycin to human granulocyte/macrophage and erythroid stem cell colonies. These results demonstrate that calmodulin antagonists can significantly increase the lethality of bleomycin to some but not all human tumor cells and that nonmalignant hematological human cells may also be affected by this combination.

The calcium-messenger system is involved in the regulation of cellular proliferation. Alterations in any of the components of this system could lead to states of abnormal cellular proliferation. We have focused our attention on drugs that antagonize the actions of calmodulin, a central calcium regulatory protein. Following purification of calmodulin from normal and malignant cells by preparative polyacrylamide gel electrophoresis, we determined the sensitivity of the molecule to several classes of calmodulin antagonists and found a direct correlation between antagonism of calmodulin and inhibition of cellular viability. Fluphenazine mustard was shown to be an irreversible calmodulin antagonist and an extremely potent inhibitor of leukemic cell growth. Dequalinium and rhodamine-123, drugs selectively accumulated in malignant cells, were found to be potent calmodulin antagonists, and inhibited the growth of C6 astrocytoma cells. Further studies led to the observation that the combination of calmodulin antagonists with bleomycin produced synergistic cell-kill that was related to enhanced DNA damage. The combination of trifluoperazine and bleomycin was safely administered to 19 heavily pre-treated cancer patients, of whom four had excellent clinical responses. These studies suggest that calmodulin might be an important target for the development of new antineoplastic drugs.

Cyclosporins have been shown to sensitize multidrug-resistant cells to chemotherapeutic agents but, generally, have minimal effect on sensitive lines. We studied the effect of cyclosporin A (CsA) and two nonimmunosuppressive analogues, 11-methyl-leucine- (11-Me-Leu-CsA) and 6-methyl-alanine-cyclosporin A (6-Me-Ala-CsA), on the action of doxorubicin (DOX) and 4'-epidoxorubicin against parent (AuXB1) and multidrug-resistant (CHRC5) Chinese hamster ovary cells. CsA and its two analogues reduced the IC50 values for DOX in sensitive AuXB1 cells from 0.1 to 0.01-0.02 microM. Cyclosporins reduced the IC50 of DOX in resistant CHRC5 cells from 9 to 0.1 (CsA), 0.7 (6-Me-Ala-CsA), and 1.2 microM (11-Me-Leu-CsA). Similar results were seen when cyclosporins were combined with 4'-epidoxorubicin. The cyclosporins alone had no effect at these concentrations (1-2.0 micrograms/ml). Dose-response curves suggested that CsA was a more potent modifying agent than 11-Me-Leu-CsA towards resistant CHRC5 cells. The ability of the cyclosporins to enhance anthracycline growth inhibition in parent AuXB1 cells may be related to an increase in drug uptake and an increase in anthracycline-induced DNA damage. CsA increased DOX accumulation as well as DOX-associated DNA single-strand breaks in AuXB1 cells over those seen in cells exposed to DOX alone. The degree of increase in DNA breaks paralleled the degree of growth inhibition seen in cells exposed to the same concentrations of drugs. In contrast, CsA had no effect on DOX accumulation or DNA single-strand breaks in CHRC5 cells. These findings imply that, in the resistant cell line, the enhanced anthracycline growth inhibition in the presence of CsA is independent of DOX accumulation and single-strand DNA breaks. These studies demonstrate that CsA and two nonimmunosuppressive analogues can sensitize both sensitive and resistant Chinese hamster ovary cells to anthracyclines. Furthermore, the mechanisms underlying this effect may differ between sensitive and multidrug-resistant cells.

We evaluated the effect of several classes of calmodulin inhibitors on the activity of calmodulin prepared from C6 astrocytoma cells and studied the activity of these drugs as inhibitors of the growth of C6 cells in tissue culture. There was a good correlation between the activity of the drugs as inhibitors of calmodulin and their activity as inhibitors of cell growth. The most potent compounds were calmidazolium and melittin as compared to the phenothiazines, trifluoperazine, chlorpromazine, chlorpromazine-sulfoxide or the diphenylbutylpiperidine, pimozide. The mechanism by which the inhibition of calmodulin leads to the death of cells could not be attributed entirely to inhibition of the calmodulin-sensitive cyclic nucleotide phosphodiesterase. Calmodulin is a heat stable, calcium-binding protein involved in numerous biological processes. Recent evidence indicates that calcium and calmodulin may be important for cellular proliferation. For example, this protein changes in concentration during the cell cycle; is involved in the disassembly of the mitotic apparatus; is increased in concentration in rapidly growing hepatomas and in transformed fibroblasts. Weiss and co-workers demonstrated that phenothiazines and structurally similar drugs are capable of binding to and inhibiting the activity of calmodulin. It has been recently observed that certain drugs that inhibit the activity of calmodulin also inhibit the growth of malignant cells in vitro and in vivo. In these studies, however, there was no direct correlation of the effect of the drugs on the calmodulin from the cell type under investigation with cytotoxicity. To learn more about the relationship between a drug's ability to inhibit calmodulin and its antiproliferative activity, we correlated the effect of drugs on the activity of calmodulin prepared from the C6 astrocytoma cell line with their effect on cellular proliferation. Since many inhibitors of calmodulin readily cross the blood-brain barrier and since no acceptable treatment for malignancies of the central nervous system exist, we chose this cell line as a model for elucidating the potential antineoplastic effects of calmodulin inhibitors.

Phenothiazines and structurally related calmodulin antagonists acted synergistically with bleomycin in killing malignant cells in culture. For exploration of the potential clinical importance of this observation, the effects of administering the combination of chlorpromazine and bleomycin in vitro and in vivo against the transplantable B16 melanoma were studied. The toxicity of the combination to bone marrow and lungs was also determined. Combined chlorpromazine and bleomycin therapy reduced the concentration of chlorpromazine required to inhibit cellular growth of B16 melanoma cells by greater than half (25 vs. 60 microM). Against the transplanted tumor, after 3 weeks of treatment the tumor sizes (in cubic millimeters) were 540 +/- 100 (vehicle), 520 +/- 120 (chlorpromazine), 170 +/- 10 (bleomycin), and 80 +/- 20 (bleomycin + chlorpromazine, P less than .01 vs. bleomycin alone). The median time required to reach a tumor size of 100 mm3 was 15 days for vehicle and chlorpromazine, 17 days for bleomycin, and 26 days for chlorpromazine and bleomycin together. The doubling time of the tumor was also increased 3.5-fold above control with the two-drug combination. At doses of bleomycin and chlorpromazine that increased antitumor activity, no pulmonary fibrosis, measured histologically and biochemically, was detected. At higher doses of bleomycin, the addition of chlorpromazine diminished lung fibrosis. The combination of drugs was not more toxic to hematopoietic precursors than chlorpromazine alone. These studies suggested that the addition of a phenothiazine to bleomycin can augment the therapeutic efficacy of bleomycin in vivo without substantially increasing its toxicity.

Since 1979 a protocol of total body electron beam therapy (3,600 rads; 6 MeV), followed by six monthly cycles of chemotherapy (doxorubicin, 30 mg/M2 given intravenously once monthly, and cyclophosphamide, 100 mg/M2 given orally each day for 14 days), has been used to treat fifty patients with mycosis fungoides (primarily Stages I and II). A group of twenty-four patients, treated by identical high-dose electron beam therapy alone, served as control subjects. Actuarial analysis by the Kaplan-Meier method and statistical analysis by the generalized Wilcoxon test of Gehan demonstrated a significant difference (p = 0.008) in the probability of Stages I and II patients' remaining in complete clinical remission when combination therapy was compared with high-dose electron beam therapy alone. No statistically significant difference was demonstrated in patients in Stages III and IV mycosis fungoides. Although 60% of patients were in "complete clinical remission," the longest follow-up being 75 months, all continued to show karyotypic abnormalities of circulating lymphocytes, and 70% had intermittently and abnormally elevated blood levels of Sézary cells.

One mechanism by which drugs alter the function of enzymes is through chronic inhibition. To determine whether commonly used cancer chemotherapeutic agents could alter protein kinase C (PKC) and thereby modify the calcium-messenger system, we studied the effect of anthracyclines and vinca alkaloids on the activity of PKC. Doxorubicin, daunomycin, vincristine and vinblastine inhibited the activity of PKC by 50% at concentrations of 150, 120, 350 and 140 microM respectively. Furthermore, we demonstrated the potential for this interaction to occur in intact cells, since doxorubicin blocked the binding of the phorbol ester, PDBu, to its receptor, PKC. The mode of inhibition of PKC was due, at least in part, to interference with the activation of the enzyme by phosphatidylserine. The activity of PKC was increased 15 fold in a highly resistant human breast cancer line, but this increase in enzymic activity was not seen in all lines tested. These studies demonstrate that anthracyclines and vinca alkaloids inhibit PKC, and suggest that chronic antagonism could lead to changes in its activity and function.

Cyclosporin A (CsA) has been shown to increase the sensitivity of multidrug resistant (MDR) cells to chemotherapeutic agents. Although the concentration of drug required to produce this effect is clinically achievable, the use of this drug would be hampered by significant immunosuppression. We report a comparison of the effects of 11-methyl-leucine cyclosporin (11-met-leu CsA), a non-immunosuppressive homolog to the parent drug, on MDR cell lines. Both cyclosporins sensitized resistant cell lines to doxorubicin, including P388 murine leukemia and GM 3639 human T-cell leukemia. The action of the cyclosporins was more pronounced with resistant cells than with sensitive ones. 11-Met-leu CsA was less potent than, but equally effective as, the parent drug. Both agents increased the intracellular accumulation and retention of doxorubicin in MDR cells. The sensitization caused by the cyclosporins was independent of their effects on cyclophilin, calmodulin, and protein kinase C. Furthermore, there were no differences in the binding of labelled CsA to MDR cells compared to the binding to sensitive cells, suggesting that P-glycoprotein was also not the molecular site of action. These studies demonstrate that a non-immunosuppressive cyclosporin can modulate multidrug resistance and suggest its further evaluation for use in clinical trials.

The specific activity of cyclic AMP phosphodiesterase and cyclic GMP phosphodiesterase of leukemic lymphocytes was 5-10-fold greater than that of purified normal lymphocytes or of homogenates of spleen, thymus or lymph nodes of normal mice. This rise was demonstrable over a wide range of substrate concentrations. Both normal and leukemic lymphocytes contained a heat-stable, calcium-dependent activator of phosphodiesterase. However, the increased activity of phosphodiesterase in leukemic lymphocytes was not due to this protein activator since (a) phosphodiesterase activity from these cells was not stimulated by this activator and (b) phosphodiesterase activity of leukemic lymphocytes was not inhibited by the calcium chelater, ethylene-glycol-bis,(beta-aminoethylether)-N,N'-tetraacetic acid, suggesting that the enzyme was not already maximally activated. A comparison of several other properties of phosphodiesterase from normal and leukemic lymphocytes showed that the enzymes have similar pH optima, similar stabilities to freezing and thawing and similar sensitivities to inhibition by the phosphodiesterase inhibitors, chlorpromazine, papaverine and isobutylmethylxanthine. However, the subcellular distribution of the phosphodiesterases was different, and the phosphodiesterase of leukemic lymphocytes was significantly more resistant to heat than that of normal lymphocytes. Although no differences were found between the phosphodiesterases of normal and leukemic lymphocytes in their sensitivities to drugs, there were marked differences in drug sensitivity between the phosphodiesterase of lymphocytes and that of other tissue. For example, concentrations of chlorpromazine which inhibited phosphodiesterase of cerebrum by 70% had no effect on phosphodiesterase activity of lymphocytes. On the othere hand, the papaverine-induced inhibition of phosphodiesterase was similar in lymphocytes and cerebrum. Since an optimal concentration of cyclic nucleotides is essential to maintain normal cell growth, these results suggest that the abnormal growth characteristics of leukemic lymphocytes may be explained by their high activity of phosphodiesterase. Furthermore, the qualitative and quantitiative differences between the phosphodiesterases of leukemic lymphocytes and other tissues raise the possibility of selectively inhibiting the phosphodiesterase of the leukemic lymphocytes, thereby reducing their rate of growth, without affecting other tissues.

Both lidocaine (LIDO) and the calmodulin (CaM) antagonists, pimozide (PIM) and trifluoperazine (TFP), enhanced bleomycin (BLM) induced cytotoxicity and DNA damage. The toxicity with the combination of BLM and CaM antagonists represented true pharmacological synergism and was observed with the addition of the CaM antagonist either during or after BLM exposure. Additionally, the DNA damage of BLM and the BLM-like drugs, talisomycin S10b (TAL) or peplomycin (PEPLO), was also enhanced by CaM antagonists. LIDO, which similarly increased the lethal effects and DNA damage of BLM in L1210 cells, was also effective only during or after BLM exposure. The data presented here indicate that the modulation of toxicity seen with these drug combinations is reflected by changes in DNA integrity. Furthermore, these data suggest that the inhibition of DNA repair processes may be at least partially responsible for the enhanced toxicity and DNA damage when CaM antagonists or LIDO are added to BLM.