Bioconjugate Chem. 2001, 12, 711−718
711
N-(2-Hydroxypropyl)methacrylamide Copolymer-6-(3-Aminopropyl)-ellipticine Conjugates. Synthesis, in Vitro, and Preliminary in Vivo Evaluation F. Searle,† S. Gac-Breton,† R. Keane,† S. Dimitrijevic,† S. Brocchini,† E. A. Sausville,‡ and R. Duncan*,† Centre for Polymer Therapeutics, School of Pharmacy, University of London, 29/39 Brunswick Square, London WC1N 1AX, U.K., and Developmental Therapeutics Program, National Cancer Institute, EPN 843, Bethesda, Maryland 20892. Received December 13, 2000; Revised Manuscript Received July 11, 2001
Ellipticine derivatives have potential as anticancer drugs. Their clinical use has been limited, however, by poor solubility and host toxicity. As N-(2-hydroxypropyl)methacrylamide (HPMA) copolymeranticancer conjugates are showing promise in early clinical trials, a series of novel HPMA copolymer conjugates have been prepared containing the 6-(3-aminopropyl)-ellipticine derivative (APE, NSC176328). Drug was linked to the polymer via GFLG or GG peptide side chains. To optimize biological behavior, HPMA copolymer-GFLG-APE conjugates with different drug loading (total APE: 2.3-7% w/w; free APE: 10-fold). HPMA copolymer-GG-APE did not liberate drug in the presence of isolated lysosomal enzymes (tritosomes), but HPMA copolymer-GFLG-APE released APE to a maximum of 60% after 5 h. The rate of drug release was influenced by drug loading; lower loading led to greater release. Whereas free APE (35 µg/mL) caused significant hemolysis (50% after 1 h), HPMA copolymer-APE conjugates were not hemolytic up to 300 µg/mL (APE-equiv). As would be expected from its cellular pharmacokinetics, HPMA copolymer-GFLG-APE was >75 times less cytotoxic than free drug (IC50 ∼ 0.4 µg/mL) against B16F10 melanoma in vitro. However, in vivo when tested in mice bearing s.c. B16F10 melanoma, HPMA copolymer-GFLG-APE (1-10 mg/kg single dose, APE-equiv) given i.p. was somewhat more active (highest T/C value of 143%) than free APE (1 mg/kg) (T/C )127%). HPMA copolymer-APE conjugates warrant further evaluation as potential anticancer agents.
INTRODUCTION
Ellipticine, 5,11-dimethyl-6H-pyrido(4,3-b)carbazole (I, Figure 1) [reviewed in (1, 2)], is a potent cytotoxic agent with an IC50 in the range of 0.3-0.7 µM toward a variety of malignant cell lines in vitro (3). Its mechanisms of action include DNA intercalation (4), topoisomerase inhibition (5), and alkylation (1). Early clinical development was limited by poor drug solubility and in vivo host toxicities including hemolytic activity, decreased heart rate (1), and hepatotoxicity (6). A 9-hydroxy-2-methylellipticinium acetate derivative (II, Figure 1) did however show interesting clinical activity against thyroid and renal tumors, and bone metastases associated with breast cancer [reviewed in (1)]. These observations led us to propose the ellipticines as interesting candidates for evaluation in the form of polymer-drug conjugates. Biocompatible, water-soluble polymers provide a useful platform for tumor-selective drug delivery. Conjugation of doxorubicin (7), paclitaxel (8), and camptothecin (9) to HPMA copolymers improves their water solubility and reduces drug toxicity. Additionally, after intravenous (i.v.) administration, such conjugates selectively accumulate in solid tumor tissue by the “enhanced perme* Correspondence should be addressed to this author at the Centre for Polymer Therapeutics, Welsh School of Pharmacy, Cardiff University, Redwood Building, Cardiff CF10 3XF, EWales, U.K. TEL: +44 (0) 292 087 4180, FAX: + 44 (0) 292 087 4536, E-MAIL:
[email protected]. † University of London. ‡ National Cancer Institute.
Figure 1. Chemical structure of ellipticine (I), 9-hydroxy-2methylellipticinium acetate (II), and 6-(3-aminopropyl)-ellipticine (III).
ability and retention” (EPR) effect (10-12). EPR-mediated targeting arises due to preferential extravasation of circulating conjugate as it passes through hyperpermeable, angiogenic, tumor vessels (12), and the absence of an effective tumor lymphatic drainage encourages intratumoral retention. Those factors found previously to be important for the preclinical optimal design of polymer-drug conjugate have been extensively reviewed (13-16). Several conjugates have now progressed successfully into Phase I/II clinical evaluation. The first, HPMA copolymer-GFLG-doxorubicin (PK1, FCE 28068), displayed greatly reduced toxicity compared to free doxorubicin (7), and also showed antitumor activity in
10.1021/bc0001544 CCC: $20.00 © 2001 American Chemical Society Published on Web 08/30/2001
712 Bioconjugate Chem., Vol. 12, No. 5, 2001
chemotherapy-refractory patients, even at relatively low doses (80, 120, 180 mg/m2 doxorubicin-equiv) (17). The observed PK1 maximum tolerated dose (MTD) was 320 mg/m2 doxorubicin-equiv, which is 4-5 times higher than the usual clinical dose of free doxorubicin (60-80 mg/ m2) (17). In this study, a series of HPMA copolymer-6-(3-aminopropyl)-ellipticine dihydrochloride (APE; NSC176328) (III, Figure 1) conjugates were synthesized with the aim of improving the water-solubility of ellipticine and also improving its therapeutic index. Conjugates were prepared by aminolysis of activated HPMA copolymer precursor (18) to give compounds with either a GFLG linker, since this peptide is degraded by the lysosomal thioldependent proteases (19, 20), or a potentially nonbiodegradable GG linker. The APE content of the conjugate was varied between 2.3 and 7% w/w to allow investigation of the effect of drug loading on conjugate properties. Hemolytic activity, cytotoxicity, and the rate of APE release in the presence of isolated lysosomal enzymes (tritosomes) were all assessed in vitro. Preliminary experiments were also conducted to evaluate the antitumor activity of HPMA copolymer-APE in a subcutaneous (s.c.) murine tumor model in vivo. EXPERIMENTAL PROCEDURES
Materials. HPMA copolymers, carrying GG (5 mol %; Mw ∼30 000 and Mw/Mn ) 1.3-1.5) or GFLG (either 5 or 10 mol %; Mw ∼30 000 and Mw/Mn ) 1.3-1.5) peptide side chains, were supplied by Polymer Laboratories, Shrewsbury, U.K. The C-terminal amino acid in both peptide side chains was esterified with 4-nitrophenol, the content of displaceable activated ester being calculated from the extinction coefficient of bound 4-nitrophenol at 274 nm in dimethyl sulfoxide (DMSO) ( ) 9500 L‚mol-1‚cm-1). APE (NSC176328) was a gift from the National Cancer Institute, Washington, D.C. Dry solvents were supplied by Aldrich, U.K., and dispensed under argon. Chemicals were obtained from Aldrich, U.K., Fischer Chemicals, U.K., or BDH Ltd., U.K., unless otherwise stated. Synthesis of HPMA Copolymer-GFLG(5 mol %)APE (1a), HPMA Copolymer-GFLG(10 mol %)-APE (2a), and HPMA Copolymer-GG(5 mol %)-APE (3a) (Method 1). HPMA copolymer (1 equiv, calculated as 4-nitrophenyl ester) and APE (2 equiv) were dissolved in minimal volumes of dry DMSO. Triethylamine (2 equiv) was added to the copolymer solution to neutralize the hydrochloride protons of the APE solution which was added dropwise to the reaction mixture. Aminolysis was allowed to proceed for 5 h. To complete the aminolysis of any unreacted esters, 1-aminopropan-2-ol (2 equiv) was added to the reaction mixture (1 h). The DMSO was removed at room temperature under high vacuum and the residue redissolved in distilled water. The reaction mixture (comprising conjugate, 4-nitrophenol, and unreacted 1-aminopropan-2-ol and unreacted APE) was then dialyzed against distilled water (Mw cutoff 5000, solvent resistant grade, cellulose ester; Spectrapor, USA, Pierce and Warriner, U.K.). For small-scale preparations, an additional purification step was undertaken. The reaction mixture was applied to a PD10 (Sephadex G25) column in water and the void volume fraction collected and freeze-dried. For larger scale preparations, this step was omitted. Overall yields based on polymer weight were 60-75%. Synthesis of HPMA Copolymer-GFLG(5 mol %)APE (1b-d) (Method 2). HPMA copolymer-GPLG-p-
Searle et al.
nitrophenol ester (5 mol %) (1 g) was dissolved with stirring in dry dimethylformamide (DMF) (25 mL). APE (84 mg, 0.75 equiv) was dissolved in dry DMSO (17 mL). The solutions were mixed, a small aliquot of each (5-10 µL) being retained for thin-layer chromatography (TLC). A dilute solution of dry triethylamine in dry DMF (1:100 v/v, 9.3 mL, 3:1 equiv to APE) was prepared for addition in aliquots (500 µL) at 5 min intervals to the reaction mixture. After each addition, with swirling, samples (10 µL) were taken into a 1 mL cuvette containing Dulbecco’s phosphate-buffered saline (pH 7.0), and the initial absorbance at 400 nm was noted to follow the release of 4-nitrophenol (Shimadzu UV-1601 UV-visible spectrophotometer). The reaction mixture, initially pale yellow, decolorizes after the first six aliquots of triethylamine; then the yellow color progressively deepens. The absorbance at 400 nm reaches a plateau as triethylamine addition is completed, consistent with the displacement of the calculated amount of 4-nitrophenol (0.75 equiv). The disappearance of unbound APE was followed by thinlayer chromatography (TLC) [on Kieselgel ALU 60 F254 silica gel plates from Merck, using chloroform/methanol/ triethylamine 8:1:1 (v/v) as the mobile phase]. The reaction mixture was left overnight in the dark, and then quenched with dilute 1-aminopropan-2-ol (1:100 v/v in dry DMF, 2.2 mL) for 1 h. The solvents were evaporated under high vacuum at 30 °C, and the resulting gum was dissolved in distilled water (60 mL) and dialyzed against three changes of distilled water over 48 h (Spectrapor Mw cutoff 2000, solvent resistant). The contents of the dialysis tubing were freeze-dried to constant weight (750 mg).The content of APE in the preparation (wt %) was determined by UV at 295 nm using APE as a standard. Determination of Impurities in HPMA Copolymer-APE Conjugates by HPLC. Samples of HPMA copolymer-APE conjugates (1-5 mg) or APE (1-10 µg) were dissolved in 900 µL of water. Doxorubicin (100 µL of a 0.3 g/L stock solution) was added to each sample as an internal standard. The pH of the samples was adjusted to 8.5 with ammonium formate buffer (200 µL, 1 M, pH 8.5), and then a mixture of chloroform/propan2-ol at a ratio of 80:20 (8 mL) was added. Samples were thoroughly mixed by hand and then centrifuged at 2000g for 10 min. The lower organic layer was carefully removed and placed in glass test tubes. The solvent was evaporated under nitrogen and the dry residue dissolved in 200 µL of methanol/water 60:40 (v/v). APE content was determined by HPLC using a µBondapak C18 (150 mm × 3.9 mm) column and methanol/water 60:40 (v/v), pH 2.2, adjusted with o-phosphoric acid as the mobile phase delivered at 1 mL/min using an LKB Bromma 2150 HPLC pump. A UV detector (Spectroflow 783 Kratos analytical) with a fixed-wavelength filter (307 nm) was used to monitor APE, ellipticine, and doxorubicin. Determination of the Solubility of APE and HPMA Copolymer-GFLG(5 mol %)-APE (1c) in PhosphateBuffered Saline (PBS). A stock solution of APE in DMSO was prepared (0.1 mg/mL). To obtain a calibration curve (absorbance at 296 nm), samples were diluted using PBS to give an APE concentration range of 0-5 µg/mL. The maximum solubility of APE was determined at 37 °C or at room temperature by preparing a saturated solution in PBS (5 mL) with adjustment of the pH to 7.4 by addition of sodium hydroxide (0.1 M). Either at room temperature or at 37 °C (the saturated solution was left for 10 min to warm, the pH being rechecked) the solution was filtered through a 0.45 µm filter. Triplicate samples (10 µL) were then diluted in PBS (1 mL; 37 °C or room
HPMA Copolymer−APE Conjugates
temperature, respectively), and the absorbance at 296 nm was measured. In similar experiments, the conjugate 1d (50 mg) was dissolved in PBS (2 mL) at 37 °C with pH adjustment and equilibration for 10 min. A clear solution was obtained which was nevertheless filtered (to allow for equivalent losses) and treated as before. Size Exclusion Chromatography (SEC). SEC analysis was performed using two TSK-gel columns in series (G3000 PW followed by G2000 PW) with a guard column (Progel PWXL) and a differential refractometer (Gilson 153 refractive index detector) and a UV-visible spectrophotometer (UV Savern Analytical SA6504) as detectors connected in series. A Tris buffer (Tris, 0.05 M; NaCl, 0.5 M) was used as the mobile phase, delivered at 1 mL/ min by a Jasco PU-980 pump. PowerChrom AD Instrument software was used for recording the detector output signal and for analysis of the data. Evaluation of HPMA Copolymer-APE Conjugate Degradation by Tritosomes (Rat Liver Lysosomal Enzymes). Preparation of Tritosomes and Determination of Their Activity. Rat liver lysosomal enzymes were prepared according to the method described by Trouet (21), and their protein content was determined using bicinchoninic acid assay (22). The tritosome protein content of the tritosomes used here was 1.7 mg/mL. Their proteolytic activity was determined by measuring the release of p-nitroanilide from N-benzoyl-Phe-Val-Arg-pnitroanilide (23). Activity was 25 nM min-1 (mg of protein)-1. Degradation Studies. HPMA copolymer-APE conjugates (at a concentration of 46 µg/mL APE-equiv) to give comparable substrate (peptide spacer-APE concentration) or APE (1-10 µg) were incubated at 37 °C in citratephosphate buffer containing 0.2% Triton X-100 (400 µL) in the presence of EDTA (100 µL, 10 mM) and GSH (100 µL, 50 mM). To start the degradation study, tritosomes warmed to 37 °C (200 µL) were added, and the tube was thoroughly mixed. Aliquots of reaction mixture were taken at times 0-10 h, immediately frozen in liquid nitrogen, and stored frozen in the dark until processed by HPLC. HPLC Evaluation. The samples were mixed with distilled water (800 µL) and ammonium formate buffer (100 µL, 1 M, pH 8.5) and doxorubicin (100 µL of a 0.3 g/L solution) added as an internal standard. The content of free APE was then determined as described above. Throughout this procedure, samples have to be kept on ice because it was noticed that the APE degrades within the organic phase at room temperature to give ellipticine. Evaluation of Rat Red Blood Cell Lysis. HPMA copolymer-GFLG-APE (1a, 1b), HPMA copolymer-GGAPE (3a), APE, and ellipticine (final concentrations between 5 mg/mL and 1 µg/mL) were incubated at 37 °C with rat erythrocytes (2% w/v) for 1 or 24 h. Intact cells and debris were then removed by centrifugation (1500g, 10 min), and the supernatant was removed and used to assess hemoglobin release spectrophotometrically (at 550 nm) using a microtiter plate reader. Data were expressed as the percentage hemolysis observed compared with total hemolysis induced by Triton X-100 (0.5%). Determination of in Vitro Cytotoxicity. In vitro cytotoxicity of APE and the conjugates was evaluated according to a standard, 3-[4,5-dimethylthiazol-2-yl]-2,5diphenyltetrazolium bromide (MTT), cell viability assay (24) using a B16F10 mouse melanoma cell line. B16F10 cells (104) were seeded in 96-well microtiter plates and then left for 24 h to establish. APE or the HPMA copolymer-GFLG(5 mol %)-APE (1c) conjugate was
Bioconjugate Chem., Vol. 12, No. 5, 2001 713
added in fresh medium (concentration range 0.01-70 µg/ mL). Cells grown in medium alone were used as a control. After 67 h of incubation, MTT (20 µL of a 5 mg/mL solution) was added to each well, and the cells were incubated for a further 5 h. After removal of the medium, the precipitated formazan crystals were dissolved in spectroscopic grade DMSO (100 µL), and, after 30 min, the optical density was measured at 550 nm using a microtiter plate reader. Results are expressed as a percentage of cells grown in the absence of drug. Antitumor Activity of APE and HPMA Copolymer-APE in Vivo. All experiments were conducted according to UKCCCR Guidelines governing experiments with neoplasia. Male C57BL/6J mice were inoculated s.c. with 105 viable B16F10 murine melanoma cells. Tumors were allowed to reach a palpable size (∼50-70 mm2 calculated from the product of the two longest diameters), and treatment (saline, free APE, or conjugate 1c at doses calculated as APE-equiv) was administered by singledose i.p. Mice were weighed daily, and the tumor growth was also monitored daily until tumor area reached the maximum allowable 289 mm2. At this time or if the drugs used showed evidence of toxicity, the animals were humanely killed. Data are expressed as T/C values, i.e., the ratio of the mean time to tumor progression of the treated group to the mean time to tumor progression of the control group expressed as a percentage. Animal groups were statistically compared by a Student’s t-test. RESULTS AND DISCUSSION
Synthesis and Characterization. Successful conjugation of APE to HPMA copolymer intermediates was achieved using a standard aminolysis reaction (18) (Figure 2). This was predictable as the 3-dimensional structure of APE (Figure 3) shows that the primary amino group is free from any steric hindrance, standing well outside the plane formed by the unsaturated rings. Conjugates were synthesized to contain a range of APE loadings (Table 1). To maximize yield of conjugation, an improved conjugation method was used to prepare conjugates 1b-d. In the small-scale pilot preparations (1-5 mg of polymer), 4-nitrophenol release (Figure 4) was in molar accordance with that expected. Maintenance of high concentrations of reagents in dry solvents and dropwise addition of the triethylamine to free the 3′amino group of APE slowly from its hydrochloride salt (such that the copolymer remained in excess during the reaction) were both critical for successful conjugation. Also the use of 0.75 equiv of APE (method 2) instead of 2 equiv (as indicated in method 1) gave greatly improved yields of substitution of the peptide linker. For example, synthesis of conjugate 1a had a yield of 25.7% compared with a minimum of 62.9% yield for synthesis of conjugates 1b-d. During the reaction, the disappearance of free APE and the appearance of conjugate were followed by TLC (Rf APE: 0.4, bound APE remains at the origin; Rf 4-nitrophenol: 0.9), and after dialysis no free APE was detectable by TLC. The higher sensitivity of the HPLC procedure showed that there was some free APE, and also small amounts of free ellipticine, present (Table 1). However, the amounts of these impurities detected represent
