414
Bioconjugate Chem. 2001, 12, 414−420
Synthesis and Preclinical Characterization of a Paclitaxel Prodrug with Improved Antitumor Activity and Water Solubility Andreas Niethammer, Gerhard Gaedicke, Holger N. Lode, and Wolfgang Wrasidlo* Otto-Heubner Zentrum fu¨r Kinder- und Jugendmedizin der Charite´, Department of Pediatrics . Received October 23, 2000; Revised Manuscript Received February 12, 2001
The development of novel chemotherapy strategies based on prodrugs remains a major challenge for effective treatment of malignancies. We tested the hypothesis that this can be achieved by a prodrug of paclitaxel where one biologically active center, represented by the C7 hydroxyl group, was blocked by a dihydroxypropyl side chain which can be hydrolytically cleaved by a pH-dependent, slow-release mechanism. The prodrug was sythesized by condensation of solketal chloroformate with the C7 hydroxyl group of paclitaxel followed by a ring-opening reaction to the dihydroxyl derivative. The cytotoxicity of the prodrug was similar to paclitaxel, when tested in vitro against a variety of human tumor cell lines. In vitro cell cycle analysis indicated that concentrations within the micromolar range of both drug and prodrug are required to induce sufficient G2M arrest. The hydrophilic paclitaxel prodrug proved to be more than 50-fold more water soluble than the parental drug and effectively converted to paclitaxel by pH dependent hydrolysis. Importantly, the prodrug could be used at a 3-fold higher maximum tolerated dose (MTD) and revealed a markedly improved antitumor activity in mice compared to paclitaxel. Taken together, our results demonstrate, that a hydrolytically activated paclitaxel prodrug exhibits greater water solubility and superior antitumor activity than the parental drug.
INTRODUCTION
Paclitaxel is an anticancer agent with established value for clinical application (1) including breast, lung, and ovarian cancer (2-8). However, limited response rates, significant side effects, and low solubility of this drug remain as major obstacles for its optimally effective use in cancer therapy. The chemical structure of paclitaxel, a natural product obtained from the bark of the pacific yew tree (Taxus brevifolia), was first described by Wani at al. in 1971 (9). Like taxotere and unlike other anti-microtubule drugs, such as vinca alkaloids, colchicine, or podophyllotoxin, paclitaxel promotes the irreversible polymerization of tubulin (10-15). The stable and thus dysfunctional microtubules disrupt cell division by cell cycle arrest in the premitotic G2 and the mitotic phases, the so-called G2M-arrest, which is the primary cytotoxic mechanism of this drug (16). A second cytotoxic mechanism of paclitaxel is to assist the induction of tumor necrosis factor R, an event unrelated to the polymerization of microtubules (17). The scope of of paclitaxel’s therapeutic use, especially in combination with other drugs, has been extensively reviewed (36, 37). Importantly, systemic use of paclitaxel in large doses is limited primarily by hematologic toxicity and neurotoxicity. Life-threatening hypersensitivity reactions are reported in 1-3% of patients despite appropriate premedication including dexamethasone, histamine H2 antagonists, and diphenhydramine (18, 19, 35). A second limitation of paclitaxel chemotherapy is neutropenia, starting usually 10 days * To whom correspondence should be adressed. Wolfgang Wrasidlo, Ph.D., Charite´-Humboldt University, Department of Pediatrics, Research Laboratory, Augustenburgerplatz 1, 13353 Berlin, Germany, Phone: +49 30 450 69578; Fax: +49 30 450 66916; e-mail:
[email protected].
after induction of treatment with neutrophil counts below 500/µL in the majority of cases. Severe neutropenia can usually be avoided with doses of less than 200 mg/m2 (1, 19). Neutropenia seems to be more severe with longer infusion times (4, 20). The third common side effect of paclitaxel is dose dependent neurotoxicity, characterized by peripheral symmetric distal loss of sensation, pain, and paresthesia. More severe adverse effects occur at dosages above 250 mg/m2 such as motor and autonomic dysfunction as well as myopathy (26, 27). These symptoms are often judged to be the most discomforting side effects experienced by the patient. Thus, neurotoxicity and neutropenia are considered as principle dose-limiting factors of paclitaxel (19, 21, 22). In view of these limitations of paclitaxel chemotherapy, novel prodrug design maintaining antitumor activity combined with fewer side effects is imperative to allow the administeration of larger doses of paclitaxel or to diminish side effects. Indeed, the chemical modification of the paclitaxel molecule to achieve better tolerance and efficacy has been the goal of numerous projects (23, 33). One such approach is to induce molecular modifications to release paclitaxel by enzymatic or other physicochemical mechanisms. Thus a strategy was developed based on antibody-directed enzymes for prodrug activation, commonly referred to as ADEPT (23). However, inspite of promising initial results, this strategy is severely limited by antigenic heterogenity and immunogenicity of the antibody-enzyme constructs used for tumor targeting. An alternative strategy for tumor specific prodrug activation is provided by acidic conditions in the tumor microenvironment, a common characteristic of many solid tumors (24-29). Here, we describe a novel chemotherapeutic strategy using a hydrolytically activated prodrug with dramatically improved water solubility, allowing the
10.1021/bc000122g CCC: $20.00 © 2001 American Chemical Society Published on Web 04/27/2001
Hydrolytic Activated Prodrug Therapy
Bioconjugate Chem., Vol. 12, No. 3, 2001 415
delivery of higher doses of active drug to the tumor with less toxicity. We demonstrate for the first time that a modified paclitaxel prodrug with slow and pH-dependent release of parental drug exhibits clearly superior antitumor activity in preclinical models. This method of conjugating a natural compound using a multifunctional linker can also be used to couple targeting molecules such as monoclonal antibodies to generate other types of bioconjugates. MATERIAL AND METHODS
The paclitaxel prodrug was synthesized in two steps: 1. Utlization of the 2′ and 7 hydroxyl groups via a condensation with 2,2-dimethyl-1,3-dioxolane-4 methanol chloroformate (solketal chloroformate) to form the bifunctional carbonate intermediate. 2. This intermediate was than hydrolyzed to remove the 2′ carbonate and open the acetal ring to form the final dihydroxypropyl carbonoxy paclitaxel. The following procedure describes the detailed preparation. Synthesis of the Paclitaxel Prodrug. 2′,7-[Bis(2′′,3′′-isopropylideneglycerol carbonoxy)]paclitaxel was synthesized in dry glassware, under a nitrogen atmosphere: paclitaxel (5.0 mg, 5.86 × 10-3 mmol) was dissolved in CH2CL2 (250 µL) and cooled to minus 70 °C. Solketal chloroformate (22.8 mg, 0.117 mmol) was added to the solution, followed by addition of dry pyridine (7.1 mL, 87.8 × 10-3 mmol). The reaction was completed after 4.5 h at room temperature. The reaction mixture was diluted with CH2CL2 (1.0 mL) and brine (1 × 2.0 mL). The organic layer was dried over MgSO4, filtered, and then reduced to 100 mL volume under a flow of N2. The resulting solution was purified by preparative HPLC to yield 5.2 mg of 2′7-[bis(2′′,3′′-isopropylideneglycerol carbonoxy)]paclitaxel (76.5%). Purity by HPLC was greater than 98%. Proof of structure by 1H NMR (500 MHz, CDCL3): δ 8.13 (d, 2H, o-C(O) pH), 7.75 (d, 2H, o-NHC(O) pH), 7.62 (t, 1H, p-OC(O) pH), 7.53-7.49 (band, 3H), 7.43-7.35 (band, 7H), 6.91 (t, 1H, NH), 6.35 (s, 1H, C(10)-H), 6.27 (t, 1H, C(13)- H), 5.99 (t, 1H, C(3′)-H), 5.96 (d, 1H, C(2)H), 5.5l (dd, 1H, C(7)-H), 5.44 (dd, 1H, C(2′)-H), 4.394.25 (band, 3H), 4.24-4.10 (band, 7H) 3,96 (d, 1H), 3.85 (dd, 1H), 3.75 (m, 1H), 2.61 (p, 1H, C(6)-H), 2.46 (d, 3H, C(4)-OAc), 2.41 (m, 1H), 2.23 (m, 1H), 2.14 (d, 3H, C(10)OAc), 2.01 (s, 3H, C(12)-CH3), 1.97 (m, 1H), 1.81 (s, 3H, C(8)-CH3), 1.41 (d, 3H), 1.38 (s, 3H), 1.35 (d, 6H), 1.21 (s, 3H, C(15)-CH3), 1.16 (s, 3H, C(15)-CH3). 7-(2′′,3′′-Dihydroxypropyl carbonoxy)paclitaxel: To a solution of 2′,7-[bis(2′′,3′′-isopropylideneglycerol carbonoxy)]paclitaxel (5.0 mg, 5.86 × 10-3 mmol) in tetrahydrofuran (THF) (100 µL) was added 1 N HCL (100 µL) at room temperature. The reaction was stirred at 35 °C for 3 h to give complete conversion to 2′,7-[bis-2′′,3′′dihydroxypropyl carbonoxy)]paclitaxel. At this point the reaction was stopped and the product isolated by preparative HPLC following an aqueous workup. After allowing the reaction to reach ambient temperature, it was diluted with THF (1.0 mL) and then washed with brine (3 × 2.0 mL). The organic layer was dried over MgSO4 and filtered and the volume reduced to about 100 mL volume under a flow of N2(g) followed by the addition of 0.1 M KH2PO4 (100 µL, pH ) 6.0) at room temperature. The reaction was stirred at ambient temperature for 24 h, diluted with THF (1.0 mL), and washed with brine (3 × 2.0 mL). The organic layer was dried over MgSO4, filtered, and reduced in volume to approximately 100 mL. The resulting solution was purified by preparative HPLC
Figure 1. Schematic depicting structure and hydrolytic activation of paclitaxel prodrug. The hydroxyl group in position 7 of paclitaxel was masked by a hydrophilic side chain (*). The resulting 7-(2′′,3′′-dihydroxypropyl carbonoxy)paclitaxel is activated, pH-dependently, by hydrolytic cleavage of the carbonate moiety, resulting in active paclitaxel, dihydroxypropanol, and CO2.
to yield 4.7 mg of 7-(2′′,3′′-dihydroxypropyl carbonoxy)paclitaxel (82.5%). Puritiy by HPLC was greater than 99%. Proof of structure by 1H NMR (500 MHz, CDCL3): δ 8.11 (d, 2H, o-C(O) pH), 7.76 (d, 2H, o-NHC(O) pH), 7.62 (t, 1H, p-OC(O) pH), 7.52-7.48 (band, 5H), 7.43-7.34 (band, 5H), 7.03 (d, 1H, NH), 6.26 (s, 1H, C(10)-H), 6.19 (t, 1H, C(13)-H), 5.80 (dd, 1H, C(3′)-H), 5.67 (d, 1H, C(2)H), 5.42 (dd, 1H, C(7)-H), 4.95 (d, 1H, C(5)-H), 4.80 (d, 1H, C(2′)-H), 4.58-4.39 (dd, 1H, C2′′-H), 3.72-3.59 (band, 5H), 2.67 (m, 1H, C(6)-H), 2.39 (s, 3H, C(4)-OAc), 2.362.29 (band, 3H), 2.22 (d, 4H), 1.98-1.93 (band, 2H), 1.83 (s, 3H, C(4)-OAc), 1.82 (s, 3H, C(11), CH3), 1.23 (s, 3H, C(15)-CH3), 1.16 (s, 3H,C(15)-CH3). 13 C NMR (500 MHz, CDCl ): 201.28, 201.18, 172.85, 3 171.58, 171.44, 170.70, 167.48, 166.95, 154.46, 154.34, 141.41, 141.34, 138.18, 134.00, 133.84, 132.73, 132.20, 130.37, 129.27, 129.21, 128.93, 128.54, 127.26, 84.02, 81.15, 78.62, 77.48, 77.23, 76.97, 76.57, 76.24, 76.19, 76.12, 74.40, 73.31, 72.39, 70.02, 69.06, 62.96, 62.77, 56.19, 56.16, 55.21, 47.25, 43.47, 35.76, 33.42, 26.80, 22.72, 21.18, 21.12, 21.07, 14.82, 10.92; FABLRMS (NBA/ CsI) m/e 1104, M + Cs+ calculated for C51H57NO18 1104.30; UVmax 230 nm ( 27891), min 274 nm ( 1489). The structure of the prodrug and one mechanism of activation is depicted in Figure 1. Solubility and Partition Coefficients of Paclitaxel and Paclitaxel Prodrug in Octanol/Water. Solubility was determined by sonicating drug and prodrug for 15 min in water at room temperature, followed by centrifugation at 10000g for 2 min. The supernatants were then analyzed by HPLC. The partition coefficients (octanol/water) were determined by vortexing the mixtures for 10 min. The concentrations in each layer were then determined by HPLC. Coefficients were calculated according to [c]octanol/[c]water. Cytotoxicity. Antiproliferative activities of the prodrug and paclitaxel were determined using the XTT cytotoxicity assay. This assay has been described in detail elsewhere (30). Briefly, 1 × 104 cells/well were added to
416 Bioconjugate Chem., Vol. 12, No. 3, 2001
96 well microtiter plates and incubated for 24 h before addition of drug at concentrations ranging from 1 × 10-4 to 1 × 10-12 mol/L by direct addition from stock solutions prepared in DMSO. After 72 h, XTT (2,3-bis(2-methoxy4-nitro-5-sulfophenyl)-5-[(phenylamino)carbonyl]-2H-tetrazolium hydroxide) was added to a final concentration of 280 µg/mL. The plates were then incubated for 4 h, and the inhibition of cell proliferation was quantified by determining the absorbance at 450 nm with a microplate reader (Molecular Devices, San Diego, CA). The drug concentration which inhibited growth by 50% (IC50) was calculated from duplicate runs. Stem cell toxicity was determined by a standard Salmon assay, which has been described in detail before (31, 32). Briefly, CD34 positive stem cells from cord blood were cultured in soft agar plates, followed by an incubation of various concentrations of prodrug and paclitaxel for a period of two weeks (37 °C, 5% CO2). Stem cell toxicity was evaluated by colony counting. Cell Cycle Analysis by Flow Cytometry. The human neuroblastoma cell line SK-N-SH was used as a model to determine the percentage of cells in G2/M phase as a function of prodrug concentration and time. Tumor cells were fixed and permeabilized in the presence of 70% ethanol. Cells were centrifuged (1000g, 5 min) and resuspended in phosphate-buffered saline (PBS, pH 7.4). To reduce background staining, RNAse A (0.3 mg/mL) was added, followed by specific DNA staining using propidium iodide (50 µg/mL). The DNA histograms were determined using a Beckton-Dickenson FACS scan. Animals and Experimental Tumor Xenograft Nude Mouse Models. Male athymic Ncr-nu (nu/nu) mice were obtained from Harlan Sprague Dawley (Indianapolis, IN) at 6-8 weeks of age. Animals were housed in the pathogen-free mouse colony at our institution in groups of four mice each. Mice were fed ad libitum on standard mouse laboratory chow. Animal experiments were performed according to the NIH Guide for The Care and Use of Laboratory animals. Evaluation of Maximum Tolerated Dose (MTD). For this study, MTD was defined as the treatment dose at which all animals survived without significant weight loss, i.e., less than 15% as compared to the control. The MTD for protaxoid and paclitaxel was determined in Balb/c mice, which were injected iv with increasing concentrations of either of these drugs. Doses used included 10, 20, 40, 60 mg/kg in case of the protaxoid and 8, 10, 12, 14, 16, and 18 mg/kg in case of paclitaxel. A total of 80 animals were used, and each group consisted of eight mice. Body weight was monitored and the MTD defined as a >15% loss in body weight compared to controls treated with PBS. Treatment of Human Tumor Xenografts. Three groups of eight mice each were selected randomly for each treatment regimen (PBS, paclitaxel, and prodrug) with each tumor cell line. MT-39, MDA-MB-468, OVCAR-3, and PC-3 tumor cell lines were harvested, and 2 × 106 cells were injected subcutaneously in 200 µL Hanks medium in the right front flank. Mice were examined daily, and the tumors were measured in two dimensions with a microcaliper. Established tumors with mean areas of about 20-50 mm2 were treated with a vehicle control containing DMSO, chremophore, ethanol, and PBS (8%, 6%, 6%, 80%), paclitaxel prodrug (40 mg/kg), and paclitaxel (16 mg/kg), four times on alternative days. Primary tumor size was determined by microcaliper measurements three times weekly and calculated according to width × length × 2/3.
Niethammer et al.
Statistics. Differences between experimental findings were analyzed by the students t-test. Differences were considered statistically significant with p values of
