Enhancement of cisplatin cytotoxicity by terbium in cisplatin-resistant MDA/CH human breast cancer cells

Purpose: The development of cisplatin resistance is a major problem in the treatment of cancer patients with cisplatin chemotherapy. The membrane binding of terbium (Tb³⁺) has been shown to increase the cellular accumulation of cisplatin in breast cancer cells. Therefore, the ability of Tb³⁺ to modulate the cytotoxicity of cisplatin was investigated in cisplatin-sensitive (MDA) and cisplatin-resistant (MDA/CH) MDA-MB-231 human breast cancer cells. Methods: The cytotoxic parameters of cisplatin were determined using live cell microfluorometry and median effect analysis. Results: MDA/CH cells (IC₅₀ = 142 ± 9 μM) were found to be approximately 3.3-fold more resistant to cisplatin than MDA cells (IC₅₀ = 43.5 ± 3.0 μM). In both cell lines, the IC₅₀ value for cisplatin was reduced two-fold in the presence of 80 μM Tb³⁺, thus indicating that the cytotoxicity of cisplatin is increased by Tb³⁺. The cytotoxic activity of cisplatin alone was observed to be 5.7 and 1.6 times more potent than that of Tb³⁺ alone in MDA and MDA/CH cells, respectively. Combination index analyses revealed that the interaction between cisplatin and Tb³⁺ was only synergistic at very low indices of cell death in MDA cells. However, in MDA/CH cells, the two drugs were synergistic up to intermediate levels of cell death. Conclusions: Our results suggest that the enhancement of cisplatin cytotoxicity by Tb³⁺ is more effective in cisplatin-resistant MDA/CH cells than in cisplatin-sensitive MDA cells. Therefore, terbium is potentially useful in cisplatin combination therapy for breast cancer patients, especially for those patients who have developed resistance to the drug. Received: 16 December 1998 / Accepted: 19 January 1999     

N,N-diethyl-2-[4-(phenylmethyl)phenoxy] ethanamine (DPPE), a chemopotentiating and cytoprotective agent in clinical trials: interaction with histamine at cytochrome P450 3A4 and other isozymes that metabolize antineoplastic drugs

Purpose: N,N-diethyl-2-[4-(phenylmethyl)phe- noxy]ethanamine · HCl (DPPE), an intracellular histamine (HA) antagonist with chemopotentiating and cytoprotective properties, is currently in phase 2 and 3 clinical trials in breast and prostate cancer. DPPE modulates growth at in vitro concentrations that antagonize HA binding to cytochromes P450 in rat liver microsomes. HA inhibits P450 metabolism of some drugs. Recent in vitro studies in human colon cancer cells have linked DPPE enhancement of paclitaxel, doxorubicin and vinblastine cytotoxicity to inhibition of the P-glycoprotein (P-gp) pump. Many substrates of P-gp are also substrates of CYP3A4, a P450 isozyme that metabolizes a variety of antineoplastic agents and is highly expressed in some malignant tissues. Therefore, we assessed whether (a) DPPE and HA interact at CYP3A4 and other P450 human isozymes, and (b) DPPE inhibits the catalytic activity of CYP3A4. Methods: Using spectral analysis, we measured DPPE and HA binding to insect microsomes that express human P450 isozymes 1A1, 2B6, 2D6 or 3A4. Employing thin-layer chromatography, we assessed the metabolism of DPPE by each isozyme and DPPE inhibition of testosterone metabolism by CYP3A4 and by rat liver microsomes. Results: (1) DPPE evoked “type I” (substrate site binding) absorbance-difference spectra with CYP2D6 (K_s=4.1 ± 0.4 μM), CYP3A4 (K_s= 31 ±15 μM) and CYP1A1 (K_s=40 ± 9 μM), but not with CYP2B6. (2) In correspondence with the binding studies, DPPE was metabolized by CYP2D6, CYP3A4 and CYP1A1; no metabolism occurred with CYP2B6. (3) HA evoked “type II” (heme iron binding) absorbance-difference spectra with all four isozymes, with K_s values in the range 80–600 μM. DPPE inhibited HA (600 μM) binding to CYP2D6 (IC₅₀=4 μM, 95% CI=1.8–8.9 μM) and CYP1A1 (IC₅₀=135 μM: 95% CI=100–177 μM), but stimulated HA (500 and 1000 μM) binding to CYP3A4 (EC₅₀=155 μM, 95% CI=104–231 μM). DPPE did not affect HA binding to CYP2B6. (4) DPPE inhibited the metabolism of testosterone by CYP3A4. The concentration/effect curve was biphasic: DPPE inhibited metabolism by 30% at the first site (IC₅₀=3 μM, 95% CI=0.5–25.5 μM), and an additional 70% inhibition occurred at the second site (IC₅₀=350 μM, 95% CI=215–570 μM). A similar result was observed with rat liver microsomes. Conclusion: DPPE is a substrate for CYP3A4, CYP2D6 and CYP1A1, but not CYP2B6. DPPE inhibits testosterone metabolism by interacting at two sites on CYP3A4, the first correlating with its K_s value to bind the substrate site and the second, with its EC₅₀ value to enhance HA binding to the heme iron. We postulate that (1) the inhibitory effect of DPPE on CYP3A4 activity is mediated directly at the substrate site and indirectly by its enhancement of the binding of HA to the heme moiety; (2) in tumor cells that express high constitutive levels of CYP3A4, potentiation of chemotherapy cytotoxicity by DPPE results, in part, from inhibition of CYP3A4-mediated metabolism and P-gp-mediated efflux of antineoplastic drugs; (3) in normal cells that express low constitutive levels of the isozyme, cytoprotection by DPPE results, in part, from induction of CYP3A4 and P-gp, resulting in an increase both in metabolism and efflux of antineoplastic drugs. Received: 26 April 1999 / Accepted: 3 September 1999     

Depletion of p185erbB2, Raf-1 and mutant p53 proteins by geldanamycin derivatives correlates with antiproliferative activity

Purpose: Recently, it has been shown that geldanamycin (GA), a benzoquinone ansamycin, is able to deplete mutant p53, p185^erbB2 and Raf-1 proteins in cancer cells. However, the relationship between these activities of GA and its antiproliferative activity is not clear. Here we investigated the effects of 28 GA derivatives in SKBr3, a human breast cancer cell line. Methods: We performed Western blot analysis of Raf-1, p185^erbB2 and mutant p53 proteins following drug treatment and correlated these findings with the cytotoxicity of the various GA derivatives. Results: We found that downregulation of Raf-1, p185^erbB2 and mutant p53 proteins was correlated. Thus, a drug that was active against one oncoprotein was equally active against the two others. Inactive derivatives were identified by their inability to downregulate these oncoproteins, even at a high dose (2 μM). These inactive drugs also had no or minimal antiproliferative activity (IC₅₀ > 3 μM). All other analogs (at a concentration of 2 μM) downregulated p53, p185^erbB2, and Raf-1, and also displayed cytotoxicity (IC₅₀ in the range 6–600␣nM). This category of drugs was further divided into more- and less-active agents by testing at lower doses (40 nM). The drugs that remained active against their molecular targets had an IC₅₀ for antiproliferative activity of less than 40 nM. Maximal effects on mutant p53, p185^erbB2 and Raf-1 were observed at doses that were 4–5 times greater than the cytotoxic IC₅₀. Conclusions: These findings suggest that GA and its derivatives are cytostatic/cytotoxic at concentrations that also downregulate Raf-1, p185^erbB2 and mutant p53, and raise the possibility that depletion of these proteins and the antiproliferative activities of GA have a common mechanism. Received: 6 June 1996 / Accepted: 28 September 1996     

Increased intracellular Ca2+ signaling caused by the antitumor agent helenalin and its analogues

The antitumor sesquiterpene lactone helenalin, which is found in species of the plant genusHelenium, cause a marked potentiation of the increases in intracellular free Ca²⁺ concentration ([Ca²⁺]_i) produced by mitogens such as vasopressin, bradykinin, and platelet-derived growth factor in Swiss mouse 3T3 fibroblasts. Removing external Ca²⁺ partly attenuated the increased [Ca²⁺]_i response. caused by helenalin. The increased [Ca²⁺]_i responses occurred at concentrations of helenalin that inhibited cell proliferation. At higher concentrations, helenalin inhibited the [Ca²⁺]_i responses. No change in resting [Ca²⁺]_i was caused by helenalin even at high concentrations. Other helenalin analogues also increased the [Ca²⁺]_i response. Helenalin did not inhibit protein kinase C (PKC) and PKC appeared to play a minor role in the effects of helenalin on [Ca²⁺]_i responses in intact cells. Studies with saponin-permeabilized HT-29 human colon carcinosarcoma cells indicated that helenalin caused an increased accumulation of Ca²⁺ into nonmitochondrial stores and that the potentiating effect of helenalin on mitogen-stimulated [Ca²⁺]_i responses was due in part to an increase in the inositol-(1,4,5)-trisphosphate-mediated release of Ca²⁺ from these stores.     

Tamoxifen-Induced Increases in Cytoplasmic Free Ca2+ Levels in Human Breast Cancer Cells

Tamoxifen has been shown to increase cytoplasmic free Ca²⁺ levels [Ca²⁺]_i in renal tubular cells and bladder cancer cells, and to alter Ca²⁺ signaling in MCF-7 breast cancer cells. The present study examined the effect of tamoxifen on [Ca²⁺]_i in ZR-75-1 human breast cancer cells using fura-2 as an indicator. Tamoxifen increased [Ca²⁺]_i at a concentration above 2M with an EC₅₀ of 5M. Removing extracellular Ca²⁺ reduced the response by 48±2%. In Ca²⁺-free medium, after tamoxifen-induced [Ca²⁺]_i increased had returned to baseline, adding 3mM Ca²⁺ increased [Ca²⁺]_i in a concentration-dependent manner. Further, pretreatment with 10M tamoxifen abolished the [Ca²⁺]_i increase induced by 1M thapsigargin (an endoplasmic reticulum Ca²⁺ pump inhibitor); and conversely, pretreatment with thapsigargin prevented tamoxifen from releasing more Ca²⁺. Tamoxifen (10M)-induced Ca²⁺ release was not changed by inhibiting phospholipase C activity with 2M U73122. Trypan blue exclusion assay revealed that tamoxifen (1–10M) did not alter viability after 1min of incubation, but killed 10% of cells after 3–10min of incubation. Together, this study shows that tamoxifen (>2M) induced a significant, immediate increase in [Ca²⁺]_i in ZR-75-1 breast cancer cells. Tamoxifen acted by releasing Ca²⁺ from the endoplasmic reticulum Ca²⁺ stores in a manner independent of phospholipase C activity, and by inducing Ca²⁺ entry from extracellular medium. Tamoxifen may be of mild cytotoxicity after acute exposure.     

Sensitivity to camptothecin of human breast carcinoma and normal endothelial cells

Purpose: To assess parameters that might determine resistance to the topoisomerase I inhibitor, camptothecin (CPT), the sensitivities of three established human breast cancer cell lines (ER⁻) and of normal bovine endothelial cells to CPT in the free form and incorporated into liposomes (LCPT), were contrasted with topoisomerase I (topo I) content and activity, and with cell cycle response to CPT treatment. Methods: Drug sensitivities were determined using the tetrazolium dye assay and by ³H-thymidine incorporation. Topo I levels were determined by Western blot analysis, and catalytic activity was determined with a plasmid relaxation assay, using nuclear protein from each cell line. CPT stabilization of cleavable complexes in nuclear extracts was determined using a labeled oligonucleotide with a specific topo I cleavage site. Cell cycle response to CPT was determined by flow cytometric analysis of propidium iodide-stained nuclei. Results: CPT was extremely potent against MDA-MB-157 cells with an IC₅₀ value of 7 nM compared with IC₅₀ values of 150 nM for GI 101A and 250 nM for MDA-MB-231 cells. In contrast, CPT inhibited the incorporation of ³H-thymidine at very low doses in GI 101A and MDA-MB-231 cells with IC₅₀ values of 9 nM and 5 nM, respectively; while MDA-MB-157 cells did not stop incorporating ³H-thymidine until very high doses (500 nM ) of CPT were used. When incorporated into multilamellar liposomes (LCPT), CPT retained its potency, with IC₅₀ values similar to that of the free drug. No correlation was found between CPT-induced cytotoxicity and any of the topo I parameters determined. Cell cycle analysis, however, showed an accumulation of cells in G₂/M phase after 24 h treatment with low doses (5 nM) of CPT in only GI 101A and MDA-MB-231 cells with no arrest in normal endothelial or MDA-MB-157 cells. At higher doses (50 nM), however, a dramatic accumulation of cells in the S phase was observed in MDA-MB-157, MDA-MB-231 and GI 101A cells. In contrast, a G₂/M phase block was seen with the normal bovine endothelial cells using the higher doses of CPT. Conclusions: The results suggest that cell cycle regulation plays an important role in determining the effect of CPT on malignant and normal cells. The possible mechanisms explaining the sensitivities of the two cellular compartments to the action of CPT are discussed. Received: 21 October 1996 / Accepted: 5 March 1997     

Vascular activity of two silicon compounds, ALIS 409 and ALIS 421, novel multidrug-resistance reverting agents in cancer cells

Purpose The aim of this study was to investigate the effects of two novel multidrug-resistance reverting agents, ALIS 409 [1,3-dimethyl-1,3-p-fluorophenyl-1,3(3-morfolinopropyl)-1,3-disiloxan dihydrochloride] and ALIS 421 {1,3-dimethyl-1,3-(4-fluorophenyl)-1,3[3(4-buthyl)-(1-piperazinyl)-propyl]-1,3-disiloxan tetrahydrochloride}, on vascular functions in vitro. Experimental design A comparison of their mechanical and electrophysiological actions in rat aorta rings and single rat tail artery myocytes, respectively, was performed. Results In endothelium-denuded rat aorta rings, ALIS 409 and ALIS 421 antagonized 60 mM K⁺-induced contraction in a concentration-dependent manner with IC₅₀ values of 52.2 and 15.5 μM, respectively. ALIS 409 and ALIS 421 inhibited L-type Ca²⁺ current recorded in artery myocytes in a concentration-dependent manner with IC₅₀ values of 6.4 and 5.6 μM, respectively. In rat aorta, ALIS 409 and ALIS 421 antagonized the sustained tonic contraction induced by phenylephrine with IC₅₀ values of 58.0 and 13.7 μM (endothelium-denuded rings) and of 73.9 and 31.9 μM (endothelium-intact rings), respectively. In endothelium-denuded rings, ryanodine reduced significantly the response to phenylephrine in the absence of extracellular Ca²⁺ whereas nifedipine, ALIS 409 or ALIS 421 did not affect it. Phenylephrine-stimulated influx of extracellular Ca²⁺ was markedly reduced when tissues were pretreated with ALIS 409, ALIS 421 or nifedipine, and stimulated when they were pretreated with ryanodine. Application of ALIS 409 (up to 100 μM) to intact rat aorta rings failed to induce mechanical responses. Conclusions Our results provide functional evidence that the myorelaxing effect elicited either by ALIS 409 or by ALIS 421 involved mainly the direct blockade of extracellular Ca²⁺ influx. This effect, however, took place at concentrations much higher than those effective as modifiers of multidrug resistance in cancer cells.     

On the role of agonist-evoked Ca2+ mobilization in sustaining the ongoing phosphoinositide hydrolysis

Stimulation of SK-N-BE(2) cells with 1 mM carbachol (Cch) elicited phosphoinositide (PPI) hydrolysis and a rapid elevation of cytosolic Ca²⁺ concentration ([Ca²⁺]_i) from 115 nM to about 500 nM, followed by a plateau around 200 nM. In myo [³H]inositol-labelled cells, Cch-evoked accumulation of [³H]inositol phosphates (IPs) was not affected when [Ca²⁺]_i was clamped at resting by cell loading with 10 M BAPTA/AM; under these conditions, maximal 1,4,5-inositol trisphosphate accumulation was not reduced either. When [Ca²⁺]_i was clamped around 700 nM by cell treatment with 600 nM ionomycin, Cch-evoked [³H]IPs accumulation was enhanced by less than 20%, but it was impaired by a 30% and a 55% after [Ca²⁺]_i reduction to about 70 nM and 35—50 nM, by cell loading with 20 M or 40 M BAPTA/AM, respectively. These results show that, in SK-N-BE(2) cells, Cch-activated PPI-specific phospholipase C is sensitive to [Ca²⁺]_i but it already operates under suboptimal conditions at resting [Ca²⁺]_i.     

Metabolism and ribonucleotide reductase inhibition of (E )-2′-deoxy-2′-(fluoromethylene)cytidine, MDL 101,731, in human cervical carcinoma HeLa S3 cells

(E )-2′-Deoxy-2′-(fluoromethylene)cytidine, MDL 101,731, has shown potent antitumor activity against various human xenograft models. Purpose: The purpose of this study was to elucidate the mechanism of the antitumor activity of MDL 101,731 against human carcinoma cells through investigating metabolism and the target enzyme of MDL 101,731. Methods: In respect of the intracellular metabolism of MDL 101,731, the effect on enzymes in the pyrimidine salvage pathway and the intracellular metabolites of MDL 101,731 were investigated. In respect of the target enzyme, the effect on intracellular deoxyribonucleoside triphosphate (dNTP) pools and the inhibition of the enzyme activity were investigated. Results: MDL 101,731 which shows antiproliferative activity against human cervical carcinoma HeLa S₃ cells at nanomolar concentrations (IC₅₀, 30–50 nM ), was hardly metabolized to (E)-2′-deoxy-2′-(fluoromethylene)uridine (FMdU) which had no antiproliferative activity below 100 μM because of resistance to human cytidine deaminase. MDL 101,731 showed low activity against murine lymphocytic leukemia P388^R cells (Ara-C-resistant cells) which contained lower deoxycytidine kinase activity than parental P388 cells. In addition, the antiproliferative activity of MDL 101,731 against HeLa S₃ cells was reversed by deoxycytidine. Studies of the intracellular metabolism of ³H-MDL 101,731 demonstrated that it was rapidly metabolized to the diphosphate and the triphosphate forms without the other metabolites in HeLa S₃ cells. A 3-h treatment with 0.1–10 μM MDL 101,731 decreased intracellular dNTP pools. The recovery of dNTP pools decreased by treatment with 2 μM MDL 101,731 was much slower than the recovery following treatment with 10 mM hydroxyurea, a reversible ribonucleotide reductase inhibitor. At a dose of 250 mg/kg, MDL 101,731 continuously inhibited ribonucleotide reductase activity up to 72 h in a HeLa S₃ xenograft model. Conclusions: This study suggests that the prolonged ribonucleotide reductase inhibition by rapidly activated metabolites of MDL 101,731 in part contributes to the potent antitumor activity of this drug against various xenografts. Received: 18 March 1997 / Accepted: 14 July 1997     

Activity of dolastatin 10 against small-cell lung cancer in vitro and in vivo: induction of apoptosis and bcl-2 modification

Purpose: Dolastatin 10 is a natural cytotoxic peptide which acts through the inhibition of microtubule assembly. Studies have suggested that such agents can induce apoptosis in association with bcl-2 phosphorylation. Since bcl-2 overexpression is common in small-cell lung cancer (SCLC), we evaluated the activity of dolastatin 10 in SCLC cell lines and xenografts. Methods: In vitro growth inhibition was evaluated with a standardized MTT assay and apoptosis with fluorescent microscopy and a TUNEL assay. Immunoblot analysis and phosphatase digestion were used to determine bcl-2 modification. In vivo activity was evaluated in subcutaneous and metastatic SCLC xenograft models in SCID mice. Results: Dolastatin 10 had growth inhibitory activity against four SCLC cell lines (NCI-H69, -H82, -H446, -H510) with IC₅₀ values ranging from 0.032 to 0.184 nM. All four cell lines exhibited evidence of apoptosis after 48 h of exposure to 1.3 nM dolastatin 10. Immunoblot analysis revealed that 1.3 nM dolastatin 10 altered the electrophoretic mobility of bcl-2 in NCI-H69 and -H510 cells within 16 h of treatment. Incubation of protein extract from dolastatin 10-treated NCI-H69 and -H510 cells with calcineurin resulted in the disappearance of the altered mobility species, suggesting dolastatin 10-induced bcl-2 phosphorylation. In in vivo studies, 450 μg/kg of dolastatin 10 IV × 2 given after intravenous injection of NCI-H446 cells completely inhibited tumor formation. In established subcutaneous NCI-H446 xenografts, 450 μg/kg of dolastatin 10 IV induced apoptosis in the majority of tumor cells within 96 h, resulting in a log₁₀ cell kill of 5.2 and an increase in median survival from 42 to 91 days. Conclusions: These findings suggest that dolastatin 10 has potent activity against SCLC and that the modulation of apoptotic pathways deserves further evaluation as an anticancer strategy. Received: 14 July 1998 / Accepted: 3 September 1998     

Cisplatin efflux, binding and intracellular pH in the HTB56 human lung adenocarcinoma cell line and the E-8/0.7 cisplatin-resistant variant

Purpose: Many cell lines resistant to cisplatin (DDP) have reduced DDP accumulation. We postulated that reduced accumulation of DDP in resistant cells might be due to decreased intracellular DDP binding, leading to increased passive efflux. Methods: The total cellular ([T-DDP]), intracellular ultrafiltrable ([F-DDP]) and precipitable cellular bound ([B-DDP]) DDP concentrations were all compared in the HTB56 human lung adenocarcinoma cell line and its E-8/0.7 variant that has acquired DDP resistance. Cells were exposed to 509 μM DDP for 20 min. Ultrafiltration with a 500 molecular weight cut-off separated cellular free from bound cisplatin. Fragmentation by sonication and microcentrifugal spinning precipitated cellular bound cisplatin. Flow cytometry was used to measure the intracellular pH (pH_i) of the HTB56 cell line, the E-8/0.7 cell line, as well as of the OV2008 cell line and its C13 resistant variant. The DNA-bound DDP and protein-bound DDP ([P-DDP]) were also compared when equal [T-DDP] was achieved for both sensitive and resistant cells by exposing them for 1 h to two pairs of DDP concentrations, i.e. 509 vs 911 μM DDP, and 111 vs 666 μM DDP, respectively. Platinum was assayed by flameless atomic absorption spectrophotometry. Results: At time 0 (end of cisplatin exposure), [T-DDP] and [B-DDP] were significantly higher in the sensitive HTB56 parent cell line (P < 0.02 and P < 0.001, respectively), whereas [F-DDP] did not differ significantly (P = 0.62). Two distinct phases of T-DDP efflux were observed. In the first 10 s after DDP exposure, the rate constant for resistant cells (K_R1) was 0.17 s⁻¹, whereas that for sensitive cells (K_S1) was 0.14 s⁻¹. From 10 s to 50 s, however, K_R2 and K_S2 became 0.005 s⁻¹ and 0.004 s, respectively. [T-DDP] remained lower in resistant cells than in sensitive cells at 10, 30 and 50 s (all P < 0.0001). For 1 h drug exposure to 509 vs 911 μM cisplatin concentrations designed to give comparable [T-DDP] in the sensitive and resistant cell lines, only [DNA-bound DDP] was found to be significantly higher in sensitive cells (P = 0.002), whereas both [F-DDP] and [P-DDP] did not differ significantly (P = 0.18, P = 0.75, respectively). On the other hand, there were no significant differences found in [F-DDP], [P-DDP] and [DNA-bound DDP] between the two cell lines when 111 vs 666 μM DDP was used. Flow cytometry data indicated that the pH_i was significantly higher in the E-8/0.7 (P < 0.0186) and C13 (P < 0.0169) resistant variants than in the sensitive parent cell lines. Conclusions: DDP binds more slowly in resistant than in sensitive lung cancer cells, despite comparable amounts of free drug. Early efflux is higher in the resistant variant. Differences between the lines with respect to DNA binding may be DDP concentration-dependent. We speculate that the reduced early binding and increased early efflux in the resistant line may be related to the higher pH in this line. A higher pH is supposed to favor production of neutral hydroxyl metabolites rather than charged aquated metabolites, and these neutral metabolites would be expected to react less readily with intracellular molecules and to efflux more readily across cell membranes. Since we have previously documented a threefold increase in glucose utilization and lactate production in the DDP-resistant variants of the human HTB56 lung cancer cell lines, and this increased lactate production would have been expected to reduce the intracellular pH instead of raising it, it is possible that our alkaline-resistant cells have a higher Na⁺/H⁺ exchanger activity which would protect them from intracellular acidification. Received: 4 September 1998 / Accepted: 3 December 1998     

Role of lipophilicity in determining cellular uptake and antitumour activity of gold phosphine complexes

Purpose: The lipophilic cation [Au(I)(dppe)₂]⁺ [where dppe is 1,2-bis(diphenylphosphino)ethane] has previously demonstrated potent in vitro antitumour activity. We wished to determine the physicochemical basis for the cellular uptake of this drug, as well as of analogues including the 1:2 adducts of Au(I) with 1,2-bis(di-n-pyridylphosphino)ethane (dnpype; n=2, 3 and 4), and to compare in vitro and in vivo antitumour activity. Methods and results: Logarithmic IC₅₀ values for the CH-1 cell line bore a parabolic dependence on drug lipophilicity, as measured either by high-performance liquid chromatography or by n-octanol-water partition. Cellular uptake of drug, as measured by inductively coupled plasma mass spectrometry, varied by over three orders of magnitude over the series. Logarithmic uptake had a parabolic dependence on drug lipophilicity but a linear relationship to logarithmic IC₅₀ values. Free drug concentrations were determined under the culture conditions and logarithmic free drug IC₅₀ values and uptake rates were linearly related to lipophilicity. Uptake of drug in vivo in tissue from murine colon 38 tumours was approximately proportional to the dose administered. Host toxicity varied according to lipophilicity with the most selective compound having an intermediate value. This compound was also the most active of those tested in vivo, giving a growth delay of 9 days following daily intraperitoneal dosing (10 days) at 4 μmol kg⁻¹ day⁻¹. It was also significantly more active than another lipophilic cation, MKT-077. Conclusions: Alteration of lipophilicity of aromatic cationic antitumour drugs greatly affects cellular uptake and binding to plasma proteins. Changes in lipophilicity also affect host toxicity, and optimal lipophilicity may be a critical factor in the design of analogues with high antitumour activity. Received: 11 April 2000 / Accepted: 16 June 2000     

Effects of protein binding on the in vitro activity of antitumour acridine derivatives and related anticancer drugs

Purpose: We set out to measure drug binding to serum proteins. These have been shown to reduce the free plasma concentrations of a number of anticancer drugs, particularly of those of complex organic structure, in both experimental studies and clinical trials. Methods: We have used cultures of murine Lewis lung carcinoma cells as sensors of available drug to measure the effects of two drug-binding plasma proteins, α-acid glycoprotein (AAG) and human serum albumin (HSA), as well as of bovine serum albumin (BSA) on drug activity. Results: The concentrations required for 50% growth inhibition (IC₅₀ values) of a number of anticancer drugs were found to be linear functions of the added proteins. Assuming that cells respond to free drug, the data provide estimates of the product K · n, where K is the binding constant of the protein and n is the number of drug binding sites per protein molecule. Amsacrine, the amsacrine analogue asulacrine, camptothecin, DACA (N-[2-(dimethylamino)ethyl]acridine-4-carboxamide), doxorubicin, etoposide, mitoxantrone, paclitaxel and vincristine were tested. The K · n values for AAG were 30, 2400, 8.7, 340, 29, 290 × 10³ M ⁻¹ and 120 × 10³ M ⁻¹, respectively, and the K · n values for HSA were 16, 580, 530, 10, 6.2, 4.3 × 10³ M ⁻¹ and 0.0 × 10³ M ⁻¹, respectively. The combined data allowed the estimation of free fractions of drug in plasma, assuming that AAG and HSA contributed most to protein binding. The data were in general comparable with that reported using equilibrium dialysis and ultrafiltration. Data for drug binding to BSA were different from those for HSA, in some cases by a large factor with values for HSA generally higher. The applicability of the method to analogue development was illustrated by examining the binding to AAG of a series of DACA analogues, and binding was found to be primarily related to lipophilicity. Conclusion: IC₅₀ determinations provide a rapid means of estimating drug binding to plasma proteins and have utility in the assessment of new anticancer drugs. Received: 20 July 1999 / Accepted: 5 November 1999     

Effects of SCH 59228, an orally bioavailable farnesyl protein transferase inhibitor, on the growth of oncogene-transformed fibroblasts and a human colon carcinoma xenograft in nude mice

The products of the Ha-, Ki-, and N-ras proto-oncogenes comprise a family of 21 kDa guanine nucleotide-binding proteins which play a crucial role in growth factor signal transduction and in the control of cellular proliferation and differentiation. Activating mutations in the ras oncogenes occur in a wide variety of human tumors. Ras proteins undergo a series of posttranslational processing events. The first modification is addition of the 15-carbon isoprene, farnesyl, to a Cys residue near the carboxy-terminus of Ras. Prenylation allows the Ras oncoprotein to localize to the plasma membrane where it can initiate downstream signalling events leading to cellular transformation. Inhibitors of the enzyme which catalyzes this step, farnesyl protein transferase (FPT), are a potential class of novel anticancer drugs which interfere with Ras function. SCH 59228 is a tricyclic FPT inhibitor which inhibits the farnesylation of purified Ha-Ras with an IC₅₀ of 95 nM and blocks the processing of Ha-Ras in Cos cells with an IC₅₀ of 0.6  M. SCH 59228 has favorable pharmacokinetic properties upon oral dosing in nude mice. The in vivo efficacy of SCH 59228 was evaluated using a panel of tumor models grown in nude mice. These included several rodent fibroblast lines expressing mutationally-activated (val¹²) forms of the Ha-Ras oncogene. In some cases, these proteins contain their native C-terminal sequence (CVLS) which directs farnesylation. In one model, the C-terminal sequence was altered to CVLL, making the expressed protein a substrate for a distinct prenyl transferase, geranylgeranyl protein transferase-1. When dosed orally at 10 and 50 mg/kg (four times a day, 7 days a week) SCH 59228 significantly inhibited tumor growth of cells expressing farnesylated Ha-Ras in a dose-dependent manner; over 90% growth inhibition was observed at the 50 mg/kg dose. Tumor growth of cells expressing the geranylgeranylated form of Ha-Ras was less potently inhibited. Growth of tumors derived from a rodent fibroblast line expressing activated Ki-Ras containing its native C-terminal sequence (CVIM), which preferentially directs farnesylation, was also inhibited by SCH 59228. Inhibition in the Ki-Ras model was less than that observed in the Ha-Ras model. In contrast, tumors derived from cells transformed with the mos oncogene were not significantly inhibited even at the highest dose level. SCH 59228 also significantly and dose-dependently inhibited the growth of human colon adenocarcinoma DLD-1 xenografts (which express activated Ki-ras). These results indicate that SCH 59228 possesses in vivo antitumor activity upon oral dosing in tumor models expressing activated ras oncogenes. This is the first report of oral antitumor activity with an FPT inhibitor. These results are discussed in light of recent observations on alternative prenylation of some Ras isoforms. Received: 14 January 1998 / Accepted: 13 May 1998     

Growth-inhibitory effects of the synthetic retinoid CD437 against ovarian carcinoma models in vitro and in vivo

The activity of CD437{6-[3-(1-adamantyl)-4 hydroxyphenyl]-2-naphthalene carboxylic acid}, a relatively selective activator of RAR-γ, was evaluated against four human ovarian-carcinoma cell lines : PE01, PE04 (a Pt-resistant in vivo-derived counterpart of PE01), PE01^CDDP (a Pt-resistant in vitro-derived model of PE01) and PE014. Growth inhibition was observed after 3 and 6 days of exposure to sub-micromolar concentrations as assessed by a reduction in cell number. IC₅₀ values against PE01, PE04, PE01^CDDP and PE014 were 0.09, 0.21, 0.12 and 0.28 μM (day 3) and 0.1, 0.14, 0.07 and 0.17 μM (day 6), respectively. Cisplatin-resistant cell lines were as responsive as cisplatin-sensitive lines, indicating potential activity in resistant disease. CD437 was also evaluated against the PE04 xenograft grown in nude mice using daily doses of 20 (days 0–4) and 10 mg/kg (days 0–4 and 7–11) given either by i.p. delivery or oral administration. Significant growth inhibition (P < 0.05) was obtained for both doses and by both routes. These data provide further support for the view that retinoids have value for the treatment of ovarian cancer. Received: 23 February 1998 / Accepted: 26 March 1998     

Comparative studies on the action of 7-N-[2-[[2-(γ-l-glutamylamino)ethyl]dithio]ethyl]mitomycin C and of mitomycin C on cultured HL-60 cells and isolated phage and plasmid DNA

Summary The mechanism of action of a new mitomycin C (MMC) derivative, KT6149, was studied in human leukemia HL-60 cells and in isolated phage and plasmid DNA, and its effects were compared with those of MMC. Cell growth was markedly inhibited by KT6149, with an IC₅₀ of 2 × 10⁻⁹ M, that for MMC being 2 × 10⁻⁸ M. DNA synthesis of HL-60 cells as determined by incorporation of [³H]-thymidine was also inhibited by KT6149, with an IC₅₀ of 2 × 10⁻⁷ M as compared with 2 × 10⁻⁶ M for MMC. RNA and protein synthesis were less markedly inhibited at low concentrations. Alkaline sucrose density-gradient centrifugation revealed a significant decrease in sedimentation velocity for cellular DNA of the cells after 1 h treatment with KT6149 at concentrations higher than 10⁻⁷ M. In contrast, no such change was observed for DNA of cells treated with MMC, even at a concentration of 10⁻⁵ M. In a cell-free system, analysis by agarose gel electrophoresis patterns showed that the drug induced a decrease in the amount of covalently closed circular DNA of phage PM2 and an increase in that of open circular DNA in the presence of dithiothreitol (DTT), whereas MMC did not cause any change in DNA subfraction amounts. Furthermore, the electrophoretic mobility of linearized pBR322 DNA in alkaline agarose gel was significantly decreased by KT6149 in the presence of DTT and FeSO₄, no such change being observed in the case of MMC. The results clearly indicate that the inhibitory effects of KT6149 on the growth and DNA synthesis of HL-60 cells are more potent than those of MMC and that KT6149-induced DNA damage is due to single-strand scission and to cross-linking of DNA, suggesting a mode of activation different from that of MMC.     

Methyl-β-cyclodextrin in HL-60 parental and multidrug-resistant cancer cell lines: effect on the cytotoxic activity and intracellular accumulation of doxorubicin

The purpose of this work was to determine the role of methyl-β-cyclodextrin (MEBCD) in combination with doxorubicin (DOX) on the cellular proliferation of a sensitive parental and a multidrug-resistant human cancer cell line (HL-60 S and HL-60 R) and to study the effect of MEBCD on DOX intracellular accumulation. The cytotoxicity of DOX at five concentrations (50–50,000 nM ) was evaluated with or without the coadministration of four fixed noncytotoxic concentrations of MEBCD (100, 200, 500, and 1,000 μM ). Intracellular DOX concentrations were determined by a high-performance liquid chromatography (HPLC) method with fluorescence detection. MEBCD applied at 500 and 1000 μM in combination with doxorubicin (DOX) significantly potentiated the activity of DOX used alone on both sensitive and multidrug-resistant cell lines; 50% growth-inhibitory (IC₅₀) ratios (IC₅₀ MEBCD-DOX/IC₅₀ DOX ) were about 3:4 and 1.6:4 for HL-60 S and HL-60 R, respectively. Moreover, intracellular DOX accumulation, determined by HPLC during 6 h of drug exposure, was about 2–4 times higher for cells treated with MEBCD in combination with DOX than in those treated with DOX alone. Similar results were obtained using other paired MCF 7 sensitive and resistant cell lines. Correlation between these results and an MEBCD-cell membrane interaction was discussed. These initial data provide a basis for the potential therapeutic application of MEBCD in cancer therapy. Received: 13 July 1996 / Accepted: 25 March 1997     

Preclinical pharmacology and antitumour activity of the novel sequence-selective DNA minor-groove cross-linking agent DSB-120

 We examined the in vitro cytotoxicity, antitumour activity and preclinical pharmacokinetics of the novel sequence-selective, bifunctional alkylating agent DSB-120, a synthetic pyrrolo[1, 4][2, 1-c]benzodiazepine dimer. DSB-120 was shown to be a potent cytotoxic agent in vitro against a panel of human colon carcinomas [50% growth-inhibitory concentration (IC₅₀) 42±7.9 nM, mean±SE, n=7] and two rodent tumours (L1210 and ADJ/PC6). Antitumour activity was assessed in the bifunctional alkylating-agent-sensitive murine plasmacytoma ADJ/PC6 using a variety of administration protocols. The maximal antitumour eff- ects were observed following a single i.v. dose but the therapeutic index was only 2.6. DSB-120 was less effective when given i.p. either singly or by a daily+5 schedule. After a single i.v. dose at the maximum tolerated dose (MTD, 5 mg kg⁻¹) the plasma elimination was biphasic, with a short distribution phase (t _1/2α 4 min) being followed by a longer elimination phase (t _1/3β 38 min). Peak plasma concentrations were 25 μg ml⁻¹, the clearance was 1.3 ml g⁻¹ h⁻¹ and the AUC_0-∞ was 230 μg ml⁻¹ min. Concentrations of DSB-120 in ADJ/PC6 tumours were very low, showing a peak of 0.4 μg g⁻¹at 5 min. The steady-state tumour/plasma ratio was about 5% and the AUC was only 2.5% of that occurring in the plasma. DSB-120 appeared to be unstable in vivo, with only 1% of an administered dose being recovered unchanged in 24-h urine samples. Plasma protein binding was extensive at 96.6%. In conclusion, the poor antitumour activity of DSB-120 may be a consequence of low tumour selectivity and drug uptake as a result of high protein binding and/or extensive drug metabolism in vivo. Received: 5 November 1995/Accepted: 30 January 1996     

Putrescine-stimulated Intracellular Ca2+ Release for Invasiveness of Rat Ascites Hepatoma Cells

Our previous study showed that treatment of highly invasive rat ascites hepatoma (LC-AH) cells with α-difluoromethylornithine (DFMO), an inhibitor of ornithine decarboxylase, decreased both their intracellular level of putrescine and their in vitro invasion of a monolayer of calf pulmonary arterial endothelial (CPAE) cells, and that both these decreases were completely reversed by exogenous putrescine, but not spermidine or spermine. Here we show that all adhering control (DFMO-untreated) cells migrated beneath CPAE monolayer with morphological change from round to cauliflower-shaped cells (migratory cells). DFMO treatment increased the number of cells that remained round without migration (nonmigratory cells). Exogenous putrescine, but not spermidine or spermine, induced transformation of all nonmigratory cells to migratory cells with a concomitant increase in their intracellular Ca²⁺ level, [Ca²⁺]_i. The putrescine-induced increase in their [Ca²⁺]_i preceded their transformation and these effects of putrescine were not affected by antagonists of the voltage-gated Ca²⁺ channel, but were completely suppressed by ryanodine, which also suppressed the invasiveness of the control cells. The DFMO-induced decreases in both [Ca²⁺]_i and the invasiveness of the cells were restored by thapsigargin, which elevated [Ca²⁺]_i by inhibiting endoplasmic Ca²⁺-ATPase, indicating that thapsigargin mimics the effects of putrescine. These results support the idea that putrescine is a cofactor for Ca²⁺ release through the Ca²⁺ channel in the endoplasmic reticulum that is inhibited by ryanodine, this release being initiated by cell adhesion and being a prerequisite for tumor cell invasion.     

Phase I and pharmacokinetic study of d-limonene in patients with advanced cancer

Purpose: d -Limonene is a natural monoterpene with pronounced chemotherapeutic activity and minimal toxicity in preclinical studies. A phase I clinical trial to assess toxicity, the maximum tolerated dose (MTD) and pharmacokinetics in patients with advanced cancer was followed by a limited phase II evaluation in breast cancer. Methods: A group of 32 patients with refractory solid tumors completed 99 courses of d-limonene 0.5 to 12 g/m² per day administered orally in 21-day cycles. Pharmacokinetics were analyzed by liquid chromatography-mass spectrometry. Ten additional breast cancer patients received 15 cycles of d-limonene at 8 g/m² per day. Intratumoral monoterpene levels were measured in two patients. Results: The MTD was 8 g/m² per day; nausea, vomiting and diarrhea were dose limiting. One partial response in a breast cancer patient on 8 g/m² per day was maintained for 11 months; three patients with colorectal carcinoma had prolonged stable disease. There were no responses in the phase II study. Peak plasma concentration (C_max) for d-limonene ranged from 10.8 ± 6.7 to 20.5 ± 11.2 μM. Predominant circulating metabolites were perillic acid (C_max 20.7 ± 13.2 to 71 ± 29.3 μM ), dihydroperillic acid (C_max 16.6 ± 7.9 to 28.1 ± 3.1 μM ), limonene-1,2-diol (C_max 10.1 ± 8 to 20.7 ± 8.6 μM ), uroterpenol (C_max 14.3 ± 1.5 to 45.1 ± 1.8 μM ), and an isomer of perillic acid. Both isomers of perillic acid, and cis and trans isomers of dihydroperillic acid were in urine hydrolysates. Intratumoral levels of d-limonene and uroterpenol exceeded the corresponding plasma levels. Other metabolites were trace constituents in tissue. Conclusions: d-Limonene is well tolerated in cancer patients at doses which may have clinical activity. The favorable toxicity profile supports further clinical evaluation. Received: 25 June 1997 / Accepted: 6 November 1997