Nguyen et al. 10.1073/pnas.0709443104. |
Fig. 6. (Left) Obatoclax (GX15-070). (Right) BH3 peptidomimetic (1YSW) used in the modeling prediction outlined in SI Materials and Methods.
Fig. 7. Interaction of obatoclax with BCL-2. Orientation of GX15-070 in the BH3 binding groove of BCL-2 (see SI Materials and Methods).
Fig. 8. Cellular distribution of obatoclax influenced by BCL-2. Expression of BCL-2 at a specific location in the cell would be expected to influence the intracellular distribution of obatoclax. To investigate this possibility, we selected the lung carcinoma cell line H1299, which expresses low levels of endogenous BCL-2. Ectopic BCL-2 that had been modified to replace the BCL-2 transmembrane segment with that of cytochrome b5 (BCL-2b5) was expressed, resulting in selective targeting of the protein to the membrane of the endoplasmic reticulum (A and B). Exploiting the autofluorescence of obatoclax (absorbance peak, 490 nm; emission peak, 550 nm), the compound was seen to change from predominantly large random puncta in control cells to a distribution representative of the ER in BCL-2b5-expressing cells, including an ER-like perinuclear concentration and the nuclear envelope (C; double staining with an ER marker could not be achieved due to spill over of obatoclax emission across microscope filters). Obatoclax, therefore, appears to preferentially localize to sites in the cells at which BCL-2 concentrates. (A) Targeted expression of BCL-2 at the endoplasmic reticulum. Total extract from H1299 cells stably expressing control vector (-) or BCL-2b5 vector (+) were immunoblotted with anti-BCL-2 antibody or anti-actin antibody. (B) Colocalization of BCL-2b5 with the ER resident protein BAP31. H1299/BCL-2b5 cells were fixed onto a coverslip and probed with hamster anti-BCL-2 antibody and rabbit anti-BAP31 antibody followed by Alexa-488 (BCL-2) or Alexa-594 (BAP 31) conjugated secondary antibodies. (C) Autofluorescent obatoclax localizes to the endoplasmic reticulum in cells expressing BCL-2b5. H1299/BCL-2b5 were seeded onto a coverslip and incubated with 0.1 mM obatoclax for 4 h. Fluorescence was imaged with a widefield fluorescent microscope.
Fig. 9. Release of cytochrome c from mitochondria. Mitochondria were isolated from KB cells and incubated for 30 min at 37°C with DMSO or DMSO containing the indicated concentrations of BID BH3 peptide or obatoclax. Mitochondria were reisolated, and the mitochondrial pellet and postmitochondrial supernatant fractions were analyzed by SDS/PAGE and immunoblot for cytochrome c. Actin and a control (Ctr) protein served as loading controls.
Fig. 10. Obatoclax functionally targets the BCL pathway. (A) Obatoclax overcomes BCL-mediated resistance to BAX and BAK in yeast cells. The W301 yeast strain was transfected with the indicated vectors (control vector designated V), treated with either vehicle (DMSO) or drug, and cell growth monitored as described in SI Materials and Methods. (B) As in A, except that the effects of camptothecin and doxorubicin were recorded as IC50, and compared with the cytotoxic IC50's observed in the human cancer cell lines A549, C33A, PC3, and SW480.
Fig. 11. Anti-tumor activity of obatoclax. (A) SCID mice bearing human C33A cerivical carcinoma tumors were subjected to tail vein injection once a day for 5 days with vehicle or vehicle containing the indicated amount of obatoclax (tartrate salt) per kg of animal body weight, and the mean tumor volume (± SD) was determined for each cohort. (B) Mice bearing tumors derived from the indicated human cell lines were treated with obatoclax (as in A) or cisplatin (five times, once every 3 days by the intraperitoneal route) at the indicated doses. Animal death, animal weight change, and T/C (mean change in tumor size of treatment group ¸ mean change in tumor size of control group) was determined 14 days after the first injection of drug.
SI Materials and Methods
Compounds.
For in vitro or cell-based studies, obatoclax (tartrate or mesylate salts) was prepared as a stock solution in dimethyl sulfoxide (DMSO) and freshly-prepared stock was flash diluted to yield the indicated concentration of compound, while maintaining the final concentration of DMSO constant. ABT-737 was synthesized exactly as described (ref. 1; US Patent US 2004/0192681 A1), and yielded compound with physical properties in concordance with those published (HPLC Purity = 98.36%, retention time = 12.54 min; structure confirmed by proton and carbon NMR spectra for the base and proton NMR spectra for the triflic acid salt); bortezomib was purchased from Millenium Pharmaceuticals; paclitaxel was purchased from Sigma. The compounds were dissolved in DMSO before use. For animal studies, obatoclax (tartrate salt) was formulated at the indicated concentration in 9.6% polyethylene glycol 300, 0.4% polysorbate 20, and 5% dextrose, except for the 4T1 tumor model where it was formulated at a concentration of 0.6 mg/ml in 9.48% polyethylene glycol, 0.38% polysorbate 20, 1.2 mg/ml mannitol, and 5% dextrose. Cisplatin (cis-diamminedichloroplatinum) was purchased as a ready-to-use formulation in 10 ml vials from Faulding (Cat. No: N021858 and DIN = 02126613).Cells and Cell Culture.
Sk-Mel5, PC3, C33A, SW480, 4T1, H1299, KB, H460, and B16-F1 cells were maintained in culture in Roswell Park Memorial Institute (RPMI) medium 1640 supplemented with 10% FBS, penicillin (100 units/ml), streptomycin (100 mg/ml), and 2 mM l-glutamine. Baby mouse kidney cells transformed with adenovirus E1A and dominant-negative mutant p53 were derived from WT or the Bax,Bak double knockout strain were kindly provided by Dr. Eileen White and maintained as described (2), or they were derived from the Bim-/- mouse strain (3), by the same procedure.Molecular Modeling. Prediction of BCL-2 binding mode of obatoclax.
The NMR structure of a BCL-2 / inhibitor complex (1YSW in ref. 4; see SI Fig. 6) was used as the starting point for the modeling study. Structure manipulation and visualization were done using Sybyl 7.0 (Tripos, Inc., St. Louis, MO). The inhibitor was removed and the program Reduce (5) was used to add hydrogen atoms to the protein structure and optimize the orientation of polar hydrogens. The side chain ionization states were based on the pKa of the free amino acid. The protein structure was energy-refined using the AMBER force field with a distance-dependent dielectric constant (4rij), an 8-Å nonbonded cutoff to an RMS gradient of 0.01 kcal/(mol.Å) with restraints applied to all non-hydrogen atoms. The docking of obatoclax was carried out using the rigid docking program FRED (OpenEye Software, New Mexico) in conjunction with a solvated interaction energy (SIE) scoring function parameterized against an independent database of protein-ligand complexes (6). To generate the various ligand conformations used by FRED, Omega v1.81 (OpenEye Software, New Mexico) was used. The docked conformations were then subjected to energy minimization before final scoring. As a control, the method was first tested for its ability to reproduce the known binding mode of a BH3 mimetic (SI Fig. 6) in the BCL-2 binding site, 1YSW (4). Six thousand conformations were generated, docked, and then scored. The conformation matching that of the NMR structure was ranked first overall by this process.Bcl-2 Binding affinity calculation.
A predicted binding affinity for obatoclax binding to BCL-2 was calculated using the SIE scoring function (6). As a control in determining the reliability of the calculation, predicted binding affinities (Ki) were calculated for a set of 12 small molecules with experimentally measured binding affinities to BCL-2 (see SI Table 1).Immunofluorescence.
H1299 cells expressing BCL-2b5 (7) were plated on coverslips, fixed in 4% parafomaldehyde, and probed with hamster anti-BCL-2 antibody (Biomol) and rabbit anti-BAP31 antibody. The bound IgG's were decorated with Alexa-conjugated antibodies (Molecular Probe) and visualized by confocal microscopy. For obatoclax visualization, parental H1299 or H1299/BCL-2b5 cells were plated on coverslips and incubated with 0.1 mM obatoclax for 4 h, and fluorescence was detected by widefield fluorescent microscopy.Cell Plate Assay for Cell Viability.
Cells were plated in logarithmic growth phase at 2,000-4,000 cells per well in 96-well clear bottom plates (Falcon) and cultured for 14 to 16 h before the start of drug treatment. Serial dilutions of obatoclax or companion drug were made in DMSO, diluted 1:50 in RPMI, and then added to tissue culture media at a final concentration of 0.2% DMSO. Cells were typically treated with a dose range of compound from 50 nM to 10 mM for 72 h. Cell viability was then determined using the ViaLight kit (Cambrex Bio Sciences), according to the manufacturer's instructions. To obtain percentage viability, samples are expressed as a percentage of the signal obtained from DMSO-treated cells. Dose-response points were then plotted on a log scale, and IC50 values were determined using a best-fit sigmoidal dose-response curve with variable slope using GraphPad Prism Version 3.00 for Windows (GraphPad Software, Inc., San Diego, CA). The top of the curve was set to 100%.Expression of BCL-2 family members in yeast. MCL-1, BCL-w, BCL-2, and BCL-XL were subcloned into pEG202, a 2-mm plasmid, which contains the ADH promoter and a leu2 selectable marker. BAX and BAK were subcloned into pJG4-5, a 2-mm plasmid, which contains the Gal promoter and the trp selectable marker. Following transformation of the W301 yeast strain with the indicated plasmids, cultures in synthetic complete media containing raffinose (2%) and galactose (2%) were seeded (0.01 OD units) in 96-well plates in a total volume of 150 ml, and cell density was monitored after 48 h shaking at 30°C. Obatoclax or vehicle was included to a final concentration of 50 mM and/or 0.67% DMSO in the growth media.
Tumor Cell Transplantation and Tumor Measurements.
Cells were transplanted s.c. into the flank of female BALB/c or CB17 SCID/SCID mice (6 to 8 weeks of age; Charles River, Inc.) as a suspension in PBS (1.0 ´ 106 cells per ml, 1.5 ´ 106 cells per ml, 2.0 ´ 106 cells per ml, or 5.0 ´ 106 cells per ml for SW480, C33A, PC3, and 4T1 cells respectively. After 7 (SW480), 14 (C33A), or 8 (PC3 and 4T1) days, treatment with drug was initiated, and body weight and tumor size were measured three times per week. The mean relative tumor size and volume (cohort of eight animals per treatment) were calculated as follows: length (mm) ´ [width (mm)]2/2. Formulated obatoclax (tartrate salt) was administered intravenously (tail vein) once a day ´5 and cisplatin once every 3 days ´ 5 by the i.p. route. Obatoclax was formulated at the indicated concentration in 9.6% polyethylene glycol 300, 0.4% polysorbate 20, and 5% dextrose, except for the 4T1 tumor model where it was formulated at a concentration of 0.6 mg/ml in 9.48% polyethylene glycol, 0.38% polysorbate 20, 1.2 mg/ml mannitol, and 5% dextrose. Cisplatin (cis-diamminedichloroplatinum) was purchased as a ready-to-use formulation in 10 ml vials from Faulding (Cat. No: N021858 and DIN = 02126613).1. Bruncko M, Oost TK, Belli BA, Ding H, Joseph MK, Kunzer A, Martineau D, McClellan WJ, Mitten M, Ng SC, et al. (2007) J Med Chem 50:641-662.
2. Degenhardt K, Chen G, Lindsten T, White E (2002) Cancer Cell 2:193-203.
3. Bouillet P, Metcalf D, Huang DC, Tarlinton DM, Kay TW, Kontgen F, Adams JM, Strasser A (1999) Science 286:1735-1738.
4. Oltersdorf T, Elmore SW, Shoemaker AR, Armstrong RC, Augeri DJ, Belli BA, Bruncko M, Deckwerth TL, Dinges J., Hajduk PJ et al. (2005) Nature 435:677-681.
5. Word JM, Lovell SC, Richardson JS, Richardson DC (1999) J Mol Biol 285:1735-1747.
6. Naim M, Bhat S, Rankin KN, Denni S, Chowdhury SF, Siddiqi I, Drabik P, Sulea T, Bayly CI, Jakalian A, Purisima EO (2007) J Chem Inf Model 47:122-133.
7. Breckenridge DG, Stojanovic M, Marcellus RC, Shore GC (2003) J Cell Biol 160:1115-1127.