Controlled Template-Assisted Assembly of Plasmid DNA into

Donald L. McKenzie, Elizabeth Smiley, Kai Y. Kwok, and Kevin G. Rice. Bioconjugate ... Xiao-Xiang Zhang , Thomas J. McIntosh , Mark W. Grinstaff. Bioc...
0 downloads 0 Views 256KB Size
Download PDF
104

Bioconjugate Chem. 2000, 11, 104−112

Controlled Template-Assisted Assembly of Plasmid DNA into Nanometric Particles with High DNA Concentration Ming Ouyang,† Jean-Serge Remy,‡ and Francis C. Szoka Jr.*,† Department of Biopharmaceutical Sciences and Pharmaceutical Chemistry, School of Pharmacy, University of California, San Francisco, California 94143-0446, and Laboratoire de Chimie Genetique Associe CNRS, Universite Louis Pasteur, Faculte de Pharmacie de Strasbourg, F-67401 Illkirch, France

A series of novel cationic detergents that contain cleavable hydrophilic isothiuronium headgroups was synthesized, and their utility in controlled assembly of plasmid DNA into small stable particles with high DNA concentration investigated. The detergents have alkyl chains of C8-C12 and contain hydrophilic isothiuronium headgroups that give relatively high critical micelle concentration (CMC) to the detergents (>10 mM). The isothiuronium group masks a sulfhydryl group on the detergent and can be cleaved in a controlled manner under basic conditions to generate a reactive thiol group. The thiol group can undergo a further reaction after the detergents have accumulated on a DNA template to form a disulfide-linked lipid containing two alkyl chains. The pH-dependent kinetics of cleavage of the isothiuronium group, the CMC of the surfactants, the formation of the complexes, and the transfection efficiency of the DNA complexes have been investigated. Using the C12 detergent, a ∼6 kilo-basepair plasmid DNA was compacted into a small particle with an average diameter of around 40 nm with a ∼ -13 mV ζ-potential at high DNA concentration (up to 0.3 mg/mL). The compounds were well tolerated in cell culture and showed no cytotoxicity under their CMCs. Under appropriate conditions, the small particle retained transfection activity.

INTRODUCTION

Scheme 1. Designed Molecule

How to assemble a small diameter DNA complex is a long-standing problem for nonviral gene delivery systems (1, 2). Previous methods either gave large and polydisperse particle complexes (3-5), typically from 200 nm to many microns, or unstable small particle complexes at low DNA concentration (6-8). In either case, the complexes formed are not optimal for in vivo use. Recently, Blessing and co-workers (9, 10) and Trubetskoy and colleagues (11) used template-directed methods to prepare small particles at the low particle concentration (0.02 mg/mL). In an attempt to increase the DNA concentration at which small particles form, we designed and synthesized a series of novel cationic detergents that are suitable for the template-directed assembly strategy. The detergents contain isothiuronium groups as hydrophilic components (Scheme 1). As a part of the detergent, the isothiuronium group promotes a high CMC. This helps to avoid aggregation of the compact DNA at high DNA concentration because there are no cationic micelles in the solution even at a high detergent concentration. The isothiuronium group also masks the cross-linking group (sulfhydryl group) and eliminates the possibility for premature detergent dimerization leading to larger particles or aggregates of particles. After the detergent associates with DNA, the isothiuronium group can be cleaved by changing the pH or temperature, and thiol groups are generated that can dimerize into lipids with low CMC. This promotes the stability of the particles (Scheme 2). Herein we describe the detailed characterization of the chemical and physical properties of the * Address correspondence to this author. Fax: (415) 476-0688. Phone: (415) 476-3895. E-mail: [email protected]. † University of California. ‡ Universite Louis Pasteur.

Scheme 2. Controlled Template-Assisted Assembly

detergent as well as the DNA complexes that are formed by the procedure. EXPERIMENTAL SECTION

All purchased chemicals were of ACS grade or better and were used without further purification. Melting points are uncorrected. Ultraviolet-visible spectra were carried out on a Hewlett-Packard 8452A diode array

10.1021/bc990101q CCC: $19.00 © 2000 American Chemical Society Published on Web 12/28/1999

Assembly of Plasmid DNA into Nanometric Particles

spectrophotometer. Fluorescence measurements were performed on a Perkin-Elmer luminescence spectrometer LS 50B. Elemental microanalytical data were obtained at Microanalytical Laboratory, College of Chemistry, University of California, Berkeley, CA. NMR spectra were recorded on a GE 300 MHz instrument, and chemical shifts are expressed as parts per million using as internal standard, tetramethylsilane (δ ) 0.0 ppm) for 1H NMR and CDCl3 (δ ) 77.0 ppm) for 13C NMR. Mass spectra were obtained by using a Kratos Tandem MS/ MS mass spectrometer. Bromoethyloctyldimethylammonium Bromide (1a). N,N-Dimethyloctylamine (3.15 g, 20 mmol) was added dropwise into 1,2-dibromoethane (18.79 g, 100 mmol) at 100 °C over 1 h period. The mixture was stirred at 100 °C for another 3 h. After it was cooled to room temperature, acetone was used for recrystallization and gave the product as white crystals (5.66 g, 16.4 mmol, 82%). Rf ) 40 (CHCl3/MeOH/H2O ) 65/25/2). Mp 137138 °C. Anal. calcd for C12H27NBr2 (FW 345.17): C, 41.76; H, 7.88; N, 4.06. Found: C, 41.54; H, 7.68; N, 3.95. 1H NMR (300 MHz, CDCl3): δ 0.89 (3H, t, J ) 6.5 Hz); δ 1.02-1.51 (10H, m); δ 1.69-1.83 (2H, m); δ 3.52 (6H, s); δ 3.68 (2H, t, J ) 8.4 Hz); δ 3.93 (2H, t, J ) 7.4 Hz); δ 4.14 (2H, t, J ) 7.4 Hz). 13C NMR (75 MHz, CDCl3): δ 13.87, 21.94, 22.23, 22.61, 25.80, 28.75, 28.90, 28.99, 31.45, 51.17, 63.55, 64.81. +LSIMS: m/z 264 (M - Br-, 100%). Bromoethylnonyldimethylammonium Bromide (1b). This compound was prepared similarly to bromoethyloctyldimethylammonium bromide (1a) with N,Ndimethylnonylamine to give a white crystals that was recrystallized from acetone (5.46 g, 15.2 mmol, 76%). Rf ) 42 (CHCl3/MeOH/H2O ) 65/25/2). Mp 145-146 °C. Anal. calcd for C13H29NBr2 (FW 359.20): C, 43.47; H, 8.14; N, 3.90. Found: C, 43.34; H, 8.04; N, 3.90. 1H NMR (300 MHz, CDCl3): δ 0.89 (3H, t, J ) 6.5 Hz); δ 1.051.50 (12H, m); δ 1.66-1.81 (2H, m); δ 3.49 (6H, s); δ 3.66 (2H, t, J ) 8.4 Hz); δ 3.92 (2H, t, J ) 7.4 Hz); δ 4.15 (2H, t, J ) 7.4 Hz). 13C NMR (75 MHz, CDCl3): δ 13.77, 21.97, 22.27, 22.57, 25.82, 28.79, 28.92, 28.98, 31.41, 51.14, 63.49, 64.72. +LSIMS: m/z 278 (M - Br-, 100%). Bromoethyldecyldimethylammonium Bromide (1c). This compound was prepared similarly to bromoethyloctyldimethylammonium bromide (1a) with N,Ndimethyldecylamine to give a white crystals that was recrystallized from acetone (6.53 g, 17.5 mmol, 88%). Rf ) 44 (CHCl3/MeOH/H2O ) 65/25/2). Mp 154-155 °C. Anal. calcd for C14H31NBr2 (FW 373.23): C, 45.05; H, 8.37; N, 3.75. Found: C, 44.57; H, 8.50; N, 3.64. 1H NMR (300 MHz, CDCl3): δ 0.87 (3H, t, J ) 6.5 Hz); δ 1.021.47 (14H, m); δ 1.67-1.83 (2H, m); δ 3.49 (6H, s); δ 3.67 (2H, t, J ) 8.4 Hz); δ 3.92 (2H, t, J ) 7.4 Hz); δ 4.20 (2H, t, J ) 7.4 Hz). 13C NMR (75 MHz, CDCl3): δ 13.96, 22.01, 22.31, 22.62, 25.88,29.13, 29.36, 29.40, 31.56, 51.23, 63.58, 65.00. +LSIMS: m/z 292 (M - Br-, 100%). Bromoethylundecyldimethylammonium Bromide (1d). This compound was prepared similarly to bromoethyloctyldimethylammonium bromide (1a) with N,Ndimethylundecylamine to give a white crystals that was recrystallized from acetone (6.04 g, 15.6 mmol, 78%). Rf ) 48 (CHCl3/MeOH/H2O ) 65/25/2). Mp 160-162 °C. Anal. calcd for C15H33NBr2 (FW 387.25): C, 46.53; H, 8.59; N, 3.62. Found: C, 46.63; H, 8.30; N, 3.43. 1H NMR (300 MHz, CDCl3): δ 0.88 (3H, t, J ) 6.5 Hz); δ 1.041.50 (16H, m); δ 1.66-1.82 (2H, m); δ 3.48 (6H, s); δ 3.66 (2H, t, J ) 8.4 Hz); δ 3.95 (2H, t, J ) 7.4 Hz); δ 4.17 (2H, t, J ) 7.4 Hz). 13C NMR (75 MHz, CDCl3): δ 14.00, 22.21,

Bioconjugate Chem., Vol. 11, No. 1, 2000 105

22.55, 22.79, 26.07, 29.23, 29.30, 29.41, 29.55, 31.78, 51.41, 63.67, 64.93. +LSIMS: m/z 306 (M - Br-, 100%). Bromoethyldodecyldimethylammonium Bromide (1e). This compound was prepared similarly to bromoethyloctyldimethylammonium bromide (1a) with N,Ndimethyldodecylamine to give a white crystals that was recrystallized from acetone (6.42 g, 16.0 mmol, 80%). Rf ) 52 (CHCl3/MeOH/H2O ) 65/25/2). Mp 167-168 °C. Anal. calcd for C16H35NBr2 (FW 401.28): C, 47.89; H, 8.79; N, 3.49. Found: C, 47.38; H, 8.85; N, 3.40. 1H NMR (300 MHz, CDCl3): δ 0.89 (3H, t, J ) 6.5 Hz); δ 1.041.50 (18H, m); δ 1.66-1.81 (2H, m); δ 3.50 (6H, s); δ 3.70 (2H, t, J ) 8.4 Hz); δ 3.92 (2H, t, J ) 7.4 Hz); δ 4.19 (2H, t, J ) 7.4 Hz). 13C NMR (75 MHz, CDCl3): δ 14.02, 22.01, 22.57, 22.79, 26.05, 29.21, 29.37, 29.47, 31.78, 51.48, 63.70, 65.03. +LSIMS: m/z 320(M - Br-, 100%). Isothiuroniumethyloctyldimethylammonium Bromide (2a). The mixture of bromoethyloctyldimethylammonium bromide (1a) (1.03 g, 2.98 mmol) and thiourea (1.13 g, 14.8 mmol) prepared at room temperature in 5 mL of DMSO was stirred at 100 °C for 30 min. It was cooled to room temperature, acetone was used to precipitate the product, then the product was dissolved in 1 mL of CHCl3/MeOH (1/1), 2 mL of acetone was added for recrystallization, and the product formed as white crystals (0.65 g, 1.55 mmol, 52%). Rf ) 39 (CHCl3/MeOH/H2O/ HOAc ) 65/25/5/5). Mp 178-179 °C. Anal. calcd for C13H31N3Br2S (FW 421.29): C, 37.06; H, 7.42; N, 9.97; S, 7.61. Found: C, 36.97; H, 7.39; N, 10.12; S, 7.55. 1H NMR (300 MHz, d6-DMSO): δ 0.89 (3H, t, J ) 6.5 Hz); δ 1.06-1.52 (10H, m); δ 1.64-1.75 (2H, m); δ 3.11 (6H, s); δ 3.69 (2H, t, J ) 8.4 Hz); δ 3.94 (2H, t, J ) 7.4 Hz); δ 4.17 (2H, t, J ) 7.4 Hz); 9.39 (s, 4H, disappeared after D2O was added). 13C NMR [75 MHz, CDCl3/CD3OD (4/ 1)]: δ 14.03, 22.62, 22.77, 24.93, 26.15, 29.19, 29.30, 29.45, 29.60, 31.86, 50.82, 62.10, 65.70, 169.40. +LSIMS: m/z 260 (M - HBr - Br-, 100%). Isothiuroniumethylnonyldimethylammonium Bromide (2b). This compound was prepared similarly to Isothiuroniumethyloctyldimethylammonium bromide (2a) with bromoethylnonyldimethylammonium bromide (1b) and gave the product as white crystals (0.71 g, 1.64 mmol, 55%). Rf ) 39 (CHCl3/MeOH/H2O/HOAc ) 65/25/5/5). Mp 181-182 °C. Anal. calcd for C14H33N3Br2S (FW 435.32): C, 38.63; H, 7.64; N, 9.65; S, 7.37. Found: C, 38.57; H, 7.60; N, 9.45; S, 7.32. 1H NMR (300 MHz, d6-DMSO): δ 0.86 (3H, t, J ) 6.5 Hz); δ 1.03-1.45 (12H, m); δ 1.631.73 (2H, m); δ 3.08 (6H, s); δ 3.68 (2H, t, J ) 8.4 Hz); δ 3.93 (2H, t, J ) 7.4 Hz); δ 4.14 (2H, t, J ) 7.4 Hz); 9.37 (s, 4H, disappeared after D2O was added). 13C NMR [75 MHz, CDCl3/CD3OD (4/1)]: δ 14.06, 22.65, 22.74, 24.95, 26.13, 29.19, 29.31, 29.47, 29.58, 31.84, 50.79, 62.11, 65.73, 169.45. +LSIMS: m/z 274(M - HBr - Br-, 100%). Isothiuroniumethyldecyldimethylammonium Bromide (2c). This compound was prepared similarly to Isothiuroniumethyloctyldimethylammonium bromide (2a) with bromoethyldecyldimethylammonium bromide (1c) and gave the product as white crystals (0.60 g, 1.34 mmol, 45%). Rf ) 39 (CHCl3/MeOH/H2O/HOAc ) 65/25/5/5). Mp 182-183 °C. Anal. calcd for C15H35N3Br2S (FW 449.35): C, 40.10; H, 7.85; N, 9.35; S, 7.14. Found: C, 40.21; H, 8.15; N, 9.40; S, 7.09. 1H NMR (300 MHz, d6-DMSO): δ 0.86 (3H, t, J ) 6.5 Hz); δ 1.05-1.45 (14H, m); δ 1.651.73 (2H, m); δ 3.09 (6H, s); δ 3.65 (2H, t, J ) 8.4 Hz); δ 3.96 (2H, t, J ) 7.4 Hz); δ 4.25 (2H, t, J ) 7.4 Hz); 9.50 (s, 4H, disappeared after D2O was added). 13C NMR [75 MHz, CDCl3/CD3OD (4/1)]: δ 14.05, 22.58, 22.78, 24.88, 26.19, 29.15, 29.33, 29.50, 29.63, 31.91, 50.77, 61.98, 65.55, 168.95. +LSIMS: m/z 288 (M - HBr - Br-, 100%).

106 Bioconjugate Chem., Vol. 11, No. 1, 2000

Isothiuroniumethylundecyldimethylammonium Bromide (2d). This compound was prepared similarly to isothiuroniumethyloctyldimethylammonium bromide (2a) with bromoethylundecyldimethylammonium bromide (1d) and gave the product as white crystals (0.55 g, 1.19 mmol, 40%). Rf ) 40 (CHCl3/MeOH/H2O/HOAc ) 65/25/5/5). Mp 184-185 °C. Anal. calcd for C16H37N3Br2S (FW 463.37): C, 41.47; H, 8.05; N, 9.07; S, 6.92. Found: C, 41.58; H, 8.35; N, 9.00; S, 6.77. 1H NMR (300 MHz, d6-DMSO): δ 0.88 (3H, t, J ) 6.5 Hz); δ 1.07-1.47 (16H, m); δ 1.60-1.71 (2H, m); δ 3.07 (6H, s); δ 3.68 (2H, t, J ) 8.4 Hz); δ 3.95 (2H, t, J ) 7.4 Hz); δ 4.18 (2H, t, J ) 7.4 Hz); 9.45 (s, 4H, disappeared after D2O was added). 13 C NMR [75 MHz, CDCl3/CD3OD (4/1)]: δ 13.99, 22.56, 22.72, 24.89, 26.10, 29.17, 29.27, 29.41, 29.56, 31.82, 50.78, 62.05, 65.66, 168.60. +LSIMS: m/z 302 (M - HBr - Br-, 100%). Isothiuroniumethyldodecyldimethylammonium Bromide (2e). This compound was prepared similarly to Isothiuroniumethyloctyldimethylammonium bromide (2a) with bromoethyldodecyldimethylammonium bromide (1e) and gave the product as white crystals (0.65 g, 1.37 mmol, 46%). Rf ) 40 (CHCl3/MeOH/H2O/HOAc ) 65/25/ 5/5). Mp 188-189 °C. Anal. calcd for C17H39N3Br2S (FW ) 477.40): C, 42.77; H, 8.23; N, 8.80; S, 6.72. Found: C, 42.86; H, 8.56; N, 9.02; S, 6.54. 1H NMR (300 MHz, d6DMSO): δ 0.87 (3H, t, J ) 6.5 Hz); δ 1.02-1.46 (18H, m); δ 1.64-1.72 (2H, m); δ 3.10 (6H, s); δ 3.66 (2H, t, J ) 8.4 Hz); δ 3.92 (2H, t, J ) 7.4 Hz); δ 4.13 (2H, t, J ) 7.4 Hz); 9.35 (s, 4H, disappeared after D2O was added). 13C NMR [75 MHz, CDCl /CD OD (4/1)]: δ 13.97, 22.59, 3 3 22.73, 24.89, 26.12, 29.15, 29.27, 29.42, 29.56, 31.83, 50.78, 62.08, 65.67, 169.36. LSIMS: m/z 214 [M - HBr - Br- - (CH2)2SC(NH)NH2, 100%], 316 (M - HBr - Br-, 100%). Kinetics of Hydrolysis of the Isothiuronium Group. The following buffers were used to determine the hydrolysis rate of the isothiuronium group at various pH at 25 ( 0.1 °C: 20 mM sodium acetate-acetic acid, pH 5.0 buffer; 15 mM sodium phosphate, pH 7.0 buffer; 15 mM Tris‚HCl, pH 8.5 buffer; 15 mM Tris‚HCl, pH 9.0 buffer; 15 mM Na2CO3-NaHCO3, pH 10.0 buffer. The 15 mM Tris‚HCl, pH 8.5 buffer was used for testing the hydrolysis rate at various temperature. Each buffer was saturated by argon and contained 10 mM EDTA to protect sulfhydryl group from oxidation catalyzed by heavy metals (12). Free thiol concentration was quantitated spectrophotometrically according to Ellman (13) and Tarnowski (12). The reaction was maintained under argon during the entire procedure to avoid detergent oxidation. For each hydrolysis reaction sample, about 3 mM isothiuronium detergent solution was used, and for each time point, a 125 µL aliquot was removed and mixed with 5 mL of 15 mM Tris‚HCl, 10 mM EDTA, pH 7.5 buffer, and 100 µL of 10 mM fresh 5,5′-dithiobis(2nitrobenzoic acid) solution. For the experiment of hydrolysis rate on DNA template, pLC 0888.143 plasmid DNA containing the luciferase reporter gene was added to a final concentration of 0.25 mg/mL before addition of the detergent. The plasmid was obtained from Valentis, Inc. (The woodlands, TX) and had less than 20 units of endotoxin per mg of DNA (14). All of the points were tested in duplicate. Critical micelle concentrations (CMC) were determinated by a fluorescence enhancement method by using N-phenyl-1-naphthylamine (15). The CMCs of isothiuronium detergents were tested at pH 6.8, in 20 mM sodium phosphate buffer. Mercaptan detergent solutions for CMC determination were obtained by the

Ouyang et al.

following method: 0.2 M argon saturated sodium hydroxide solution was used to dissolve the isothiuronium detergents and to prepare the pH 8.5 stock solutions of the mercaptan detergents. These stock solutions were stored at room temperature for 10 min, and then, a pH 8.5, 20 mM Tris‚HCl buffer containing 10 mM DTT was used to dilute the solutions for the CMC determination. Disulfide lipid solutions for CMC determination were obtained in the following way: the mercaptan detergent stock solutions were oxidized by treatment with 10% excess H2O2 at room temperature for 2 h, and then, a pH 8.5, 20 mM Tris‚HCl buffer containing 10 mM H2O2 was used to dilute the solutions for CMC determination. All of the points were tested in triplicate. DNA-Surfactant Complex Formation. The DNA complexes were formed by mixing a defined concentration of the DNA solution and an appropriate amount of detergent in an oxygen saturated pH 8.5, 20 mM Tris‚ HCl buffer at room temperature. DNA concentration was confirmed by using absorbance at 260 nm: 1 unit ) 50 µg/mL DNA. The charge ratio (() of detergent to DNA nucleotide phosphate is computed and indicated for each experiment. The mixture was placed at room temperature for 24 h, and then the solution was characterized for the formation of particles and for the ability to transfect cell. Fluorescence Method (15) to Test the Accumulation of Detergent in the Complexes. The fluorescence enhancement of N-phenyl-1-naphthylamine (NPN) was used to measure the formation of a hydrophobic environment in the complex. The condition was as follows. Excitation: wavelength 350 nm, slit width 10 mm. Emission: wavelength 410 nm; slit width 10 mm. Integration time: 3 s. A total of 20 µL of 20 µM N-phenyl1-naphthylamine was added to each of the following samples: (1) positive control [400 µL 167 mM isothiuroniumethyldodecyldimethylammonium bromide (2) in 15 mM Tris‚HCl, pH 8.5 buffer], (2) detergent control [400 µL of 0.76 mM isothiuroniumethyldodecyldimethylammonium bromide (2) in 15 mM Tris‚HCl, pH 8.5 buffer], (3) DNA control [400 µL 500 µg/mL (1.52 mM nucleotide phosphate) pLC 0888.143 plasmid DNA in 15 mM Tris‚ HCl, pH 8.5 buffer], (4) DNA complex sample (400 µL of DNA complex solution containing 250 µg/mL (0.76 mM nucleotide phosphate) pLC 0888.143 plasmid DNA and 0.38 mM isothiuroniumethyldodecyldimethylammonium bromide (2) in 15 mM Tris‚HCl, pH 8.5 buffer], (5) blank sample [400 µL of 15 mM Tris‚HCl, pH 8.5 buffer]. The relative fluorescence of NPN in buffer (25 units) was subtracted from the fluorescence of NPN in the presence of the other compounds. All of the points were tested in triplicate. Light Scattering Measurements of Particle Size and ζ-Potential. Particle size was determined on a Zetasizer 1000 (Malvern Instruments Ltd., Spring Lane South, Malvern, Worcs. WR14 1AT, France) at 25 °C with a 90° angle. ζ-Potential was measured on a Zetasizer 4 (Malvern Instruments) at 25 °C in a buffer of 20 mM Tris‚ HCl, pH 8.5. The values for the diameter are based upon the intensity average and determined using the programs that are furnished with the zetasizer. Electron Microscopy. The solution of pLC0888.143 DNA complex compacted by isothiuroniumethyldodecyldimethylammonium bromide (2) in 15 mM Tris‚HCl, pH 8.5 buffer, at 250 µg/mL DNA with 1:1 charge ratio was used in transmission electron microscopy experiment. Negative stain method (9) and freeze-fracture method (16) were used to observe the structure of the complex: (1) Negative Stain Method. Carbon films were prepared

Assembly of Plasmid DNA into Nanometric Particles

by sublimation on freshly cleaved mica and recovered by flotation on Cu/Rh grids (300 mesh, Touzard & Matignon, Courtaboeuf, France). After drying overnight, grids were kept on blotting paper in a Petri dish. Immediately before sample addition, grids were glow-discharged (110 mV, 25 s, 25-30 µA). A drop (5 µL) of sample solution was left on the grid for 1 min. The complex sample was negatively stained with 35 µL of aqueous uranyl acetate (1-2% w/v) for 20 s, and excess liquid was removed with blotting paper. Observations were performed at 80 kV with a Philips EM 410 transmission electron microscope. (2) Freeze-Fracture Method. Freeze-fracture electron micrographs were prepared by Dr. Brigitte SternbergPapahadjopoulos as previously described (16). Thin-Layer Chromatography Identification of the Lipid in the Complexes. A modified Bligh-Dyer method (17, 18) was used to extract the lipids from the DNA complexes. For each extracted sample, two plates were developed in the same solvent pool at the same time to minimize the error. A solvent of CHCl3/MeOH/H2O ) 65/35/4 was used to develop the silica gel TLC. After the chromatography was completed, one TLC plate was sprayed by Ellman’s reagent [10 mM fresh 5,5′-dithiobis(2-nitrobenzoic acid) solution in 20 mM Tris, 0.04 M EDTA, pH 8.5 buffer] to develop the color of the mercaptan spot. Iodine was used to develop the other plate to locate the other material. To determine the Rf of the mercaptan detergent, a mercaptoethyldodecyldimethylammonium bromide sample was obtained by the following method: the isothiuroniumethyldodecyldimethylammonium bromide (2) was dissolved in an appropriate amount of 0.2 M argon saturated sodium hydroxide solution to obtain a final pH 8.5 solution. This solution was stored at room temperature for 10 min, and then the modified Bligh-Dyer method was used to extract the detergent. A sample from the chloroform phase was applied to TLC to obtain the Rf of the mercaptan detergent. To obtain the Rf of the disulfide lipid, the mercaptan detergent solution was oxidized by treatment with 10% excess H2O2 at room temperature for 2 h, and then the Bligh-Dyer method was used to extract the lipid. The lipid in the chloroform phase was applied to the TLC. Cytotoxicity of isothiuroniumethyldodecyldimethylammonium bromide (2) was assessed using Sulforhodamine-B (SRB) assay (19). CV-1 cells were plated in 96 well plates at 3 × 10 3 cells/well in 100 µL of growth medium consisting of DME-H21 with 10% fetal calf serum and antibiotics (100 units/mL penicillin and 100 µg/mL streptomycin). The cells were incubated for 24 h at 37 °C in a 95% air, 5% CO2 atmosphere. The growth medium was then removed, and 225 µL of growth medium was added. Isothiuroniumethyldodecyldimethylammonium bromide (2) was dissolved in 15 mM Trisacetic acid, pH 8.5 buffer, in a concentration serial: 5.0, 2.5, 1.2, 0.6, 0.3, and 0 mM (buffer only). The detergent solutions were added in 25 µL, and the cells were incubated in this medium for 5 h or 24 h. Normally cultured cells provided a blank. All the points were tested in triplicate. Transfection of Monkey Kidney Fibroblast Cells. CV-1 cells were plated in a 24 well plate at 3 × 104 cells/ well in 1 mL of growth medium consisting of DME-H21 with 10% fetal calf serum and antibiotics (100 units/mL penicillin and 100 µg/mL streptomycin). The cells were incubated for 19-24 h at 37 °C in a 95% air, 5% CO2 atmosphere. The growth medium was then removed, the cells were washed with serum free media, and serumfree medium was added (1 mL). The transfection system

Bioconjugate Chem., Vol. 11, No. 1, 2000 107 Scheme 3. Synthesis of the Isothiuronium Detergenta

a n ) 8, 9, 10, 11, 12. (a) Large excess of dibromoethane, 100 °C, ambient, 4 h, 76-88%. (b) Large excess of thiourea, 100 °C, ambient, 30 min, 40-55%.

was then added in 50 µL in 15 mM Tris‚HCl, pH 8.5 buffer, and the cells were incubated in this medium for 3 h. Subsequently, the medium was removed and replaced with complete growth medium, and the cells incubated for a further 21 h. The medium was removed. The cells were lysed and assayed for luciferase activity and total cell protein (Bio-Rad BCA protein assay) performed according to the manufactures’ instruction. The positive controls contained of 1 µg of luciferase DNA and 4 µg of functional dendrimer (4). The surfactant compacted DNA (1 µg) was also mixed with 4 µg of functional dendrimer and applied to a group of wells to associate the negatively charged detergent complex with the cell surface. Normally cultured cells provided a blank. All the points were tested in triplicate. RESULTS

Synthesis. Our approach to the synthesis of isothiuronium detergents employed tertiary amines with alkyl chain of C8-C12 as the starting lipophilic moiety (Scheme 3). The reaction was straightforward, although in the first step to obtain the bromoethyl derivatives, a large excess of dibromoethane was used as the reaction solvent. The tertiary amine was added dropwise to avoid both of the bromo groups on a dibromoethane molecule to react with the tertiary amine. For the synthesis of the isothiuronium detergents, DMSO was used as the reaction solvent because it is a good solvent for a SN2 displacement of bromo group by thiourea (20). A previous paper (21) has used ethanol as the solvent in this kind of reaction; however, we obtained a very poor yield using ethanol. All the products are stable at room temperature when stored as dry powders in a low humidity environment. Kinetics of Hydrolysis of Isothiuronium Group. Base-catalyzed hydrolysis was employed to generate the reactive sulfhydryl required to obtain a cationic two alkyl chain amphiphile with a low CMC. We used isothiuroniumethyldodecyldimethylammonium bromide (2) to determine the kinetics of hydrolysis of the isothiuronium group from the detergent. The preliminary experiments revealed that the hydrolysis was a first-order reaction.

108 Bioconjugate Chem., Vol. 11, No. 1, 2000

Ouyang et al.

Table 1. Hydrolysis Rate of Isothiuronium Group of Isothiuroniumethyldodecyldimethylammonium Bromide (2) at Various pH Value at 25 ( 0.1 °Ca pH (( 0.1)

k (× 10-4 s -1)

t1/2 (min)

5.0 7.0 8.5 9.0 10.0

no reaction observed 0.078 ( 0.006 4.3 ( 0.2 7.4 ( 0.5 170 ( 10

no reaction observed 1500 (110 27 ( 1 16 ( 1 0.68 ( 0.1

a

temp (°C ( 0.1 °C)

k (× 10-4 s-1)

t1/2 (min)

0 15 25 38 60

0.6 ( 0.02 1.4 ( 0.1 4.3 ( 0.2 28 ( 3 130 ( 10

190 ( 6 83 ( 5 27 ( 1 4.1 ( 0.4 0.89 ( 0.06

The rate of the cleavage of the isothiuronium group can be controlled as a function of the pH of the solution. Table 1 summarizes the reaction rate constant (k) and halflife (t1/2) of the hydrolysis at various pH at room temperature. At pH 5.0, the reaction was very slow and no obvious hydrolysis reaction was observed in our experiments over 1 day. However, as the pH was increased from 7.0 to 10.0, the half-life of the hydrolysis was reduced from more than a day to less than 1 min, indicating the hydrolysis of the isothiuronium group is highly pH dependent. According to the data in Table 1, we fitted a formula for the half-life of the reaction as a function of pH value at 25 °C:

(1)

where A ) 25 ( 1, B ) -2.6 ( 0.1, and correlation coefficient r > 0.99. Thus, regulating the pH is an effective way to control the rate of appearance of the mercaptan. The influence of the reaction temperature on the reaction rate constant (k) and half-life (t1/2) of hydrolysis at pH 8.5 is tabulated in Table 2. On the basis of the data in Table 2, we computed the activation energy of the reaction using the Arrhenius equation:

ln k ) ln A - Ea/(RT)

a

t1/2 (min)

4.3 ( 0.2 1.3 ( 0.1

27 ( 1 89 ( 5

Mean ( SD of three measurements.

n

isothiuronium detergent (mM)

mercaptan detergent (mM)

disulfide lipid (mM)

8 9 10 11 12

320 ((30) 190 ((13) 69 ((4) 25 ((2) 10 ((1)

20 ((2) 12 ((1) 8.6 ((0.3) 3.8 ((0.2) 0.21 ((0.03)

7.2 ((0.7) 4.2 ((0.3) 1.3 ((0.2) 0.55 ((0.03) 0.047((0.009)

a

Mean ( SD of three measurements.

ln(t1/2) ) A + B(pH value)

without DNA with DNA

k (× 10 -4 s-1)

Table 4. Critical Micelle Concentration of Surfactants at 25 ( 0.1 °Ca

Mean ( SD of three measurements.

Table 2. Hydrolysis Rate of Isothiuronium Group of Isothiuroniumethyldodecyldimethylammonium Bromide (2) at Various Temperatures at pH 8.5 ( 0.1a

a

Table 3. Hydrolysis Rate of Isothiuronium Group of Isothiuroniumethyldodecyldimethylammonium Bromide (2) on DNA Template at pH 8.5 ( 0.1, 25 ( 0.1 °Ca

(2)

the preexponential factor ln A ) 22 ( 3 and the activation energy Ea ) (7.2 ( 0.7) × 10 4 J mol-1, and correlation coefficient r > 0.99. Thus, the hydrolysis rate was highly temperature dependent. Regulating the temperature is another practical way to control the cleavage rate of isothiuronium group from the detergent. This should permit the controlled formation of the mercaptan and, in this way, the controlled assembly of the DNAsurfactant complex. We examined the effect on the hydrolysis of the isothiuronium group in the presence of DNA in the reaction mixture. The reaction rate constant and the halflife of the hydrolysis at pH 8.5 at 25 °C was reduced between 3 and 4-fold when DNA was present in the solution (Table 3). CMC of the Surfactants. The concept behind the controlled template-directed assembly of plasmid DNA complexes is that the self-assembly of the surfactants can be controlled to permit a stable micelle structure to form only on the DNA template. The control for the assembly is obtained by altering the CMC of the surfactant. Ideally,

Mean ( SD of three measurements.

each step of the process in forming a stable small particle should be associated with a decrease of the CMC of the detergent. To learn if the compounds fulfilled this criterion, we measured the CMCs of the three forms of the surfactant (Table 4). The CMC of the surfactant is significantly decreased when the isothiuronium group is cleaved from the detergent and can be decreased further by oxidizing the mercaptan detergent into the disulfide lipid. By the appropriate selection of the alkyl chain length, we could identify a structure which had a high enough CMC for the isothiuronium detergent to be added to high concentration of DNA without causing aggregation, yet a low enough CMC to allow the corresponding disulfide lipid to form a stable micelle on the DNA template. Complex Formation and Properties. The accumulation of detergent on the DNA template could be demonstrated by following the increase in the fluorescence quantum yield of N-phenyl-1-naphthylamine (NPN) when the fluorophore partitions into a hydrophobic environment (15). The fluorescence quantum yield of NPN increases about 10-fold, and the wavelengths of maximum fluorescence emission is blue-shifted to about 410 nm when NPN partitions into the apolar core of micelle structures from the aqueous phase. The detergent control and the DNA control samples had almost no fluorescence increase over the blank. However, the mixture of the detergent and DNA at the same concentration as the individual components showed a large fluorescence increase as well as a blue-shift of the absorbance spectrum. The fluorescence intensity of the NPN in the complex was about one-half of the fluorescence of NPN in 167 mM detergent solution (where the detergent is in a micelle), although the total concentration of the detergent and the DNA were only 200 nm were observed. This demonstrated that using the isothiuronium detergent for controlled template-assisted assembly is required to form small complex particles, especially at high DNA concentration. Although the conditions required to form the small diameter complexes occur over a narrow range, the precise and predictable control of the process by adjusting the temperature and pH makes the isothiuronium detergents a resolution method to form small diameter complexes. Because the isothiuronium group is positively charged, it provides an additionally electrostatic force to bind DNA. After the isothiuronium group is hydrolyzed, the positive charge on the DNA is decreased, thus negatively charged complexes were expected, and a -13 mV ζ-potential was found. The DNA complex migrated slower than naked DNA on agarose gel electrophoresis but none of the nanometric complexes were fully retarded. This is different from the previous paper (9, 11). We think the reason is that, in all the previous methods, there were an excess of positive charged polymers or detergents in the complex solutions. Cytotoxicity of the compounds is an important aspect of a gene delivery system. In the template-directed method introduced by Trubetskoy and co-workers (11), thermal free radical polymerization initiators were used. The free radical initiators are usually highly toxic. These free radical initiators had to be removed using dialysis before applying the complex to cells. Moreover, the covalent bonds formed during the polymerization were not biodegradable. So the degradation of the polymers formed on the DNA template depended upon the disulfide bonds in the cross-linkers. This limited the type of the cross-linker compounds that was suitable for preparing complexes with this method. The disulfide bond is biodegradable and can be reduced in the cell cytoplasma if it reaches this compartment. Thus, the DNA can be released from the complex for expression. An advantage of using isothiuronium linkers for template-directed DNA compaction is that the degradation product of hydrolysis of the isothiuronium group that forms the reactive sulfhydryl group is urea. Urea does not influence the structure of the complex and is not toxic, so it does not have to be separated from the complex prior to use. Indeed, the isothiuroniumethyldodecyldimethylammonium bromide (2) was toxic to the CV-1 cell line when the concentration was higher than its CMC. After a 24 h incubation with the cells, the thiol detergent was oxidized and was cytotoxic at 0.06 mM, which is near the disulfide lipid CMC (0.047 mM). Thus, these experiments suggest that the acute toxicity of the synthesized compounds is due to detergent effects and are otherwise well tolerated by CV-1 cells. The complexes exhibited detectable but low transfec-

112 Bioconjugate Chem., Vol. 11, No. 1, 2000

tion when added to cells in culture. The activity was consistently greater than naked plasmid DNA when they were treated with proper transmembrane introducing agent. They provided almost the same activity as the original plasmid DNA when a subeffective concentration of the fractured dendrimer were added to cells with them (4) (Figure 4). This finding is consistent with conclusions from results using the original template-directed complex formation methods (9, 11). Controlled assembly of plasmid DNA into nanometric complexes with template-assisted method has a high potential in assembly nonviral gene delivery systems. With this method, the barriers of aggregation and DNA concentration can be overcome. The negative surface charge can protect the complex from rapid elimination in vivo as compared to cationic lipoplexes (5, 14) and also opens the way for further modification, such as liposome encapsulation of the particle or the addition of a targeting ligand which may give efficient gene delivery in vivo. CONCLUSION

A systematical method has been established to control the procedure of DNA compaction to form nanometric DNA complex at high DNA concentration. Using the isothiuronium detergents, a 6 kilo-basepair plasmid DNA can be compacted into a 40 nm particle at a DNA concentration of 0.3 mg/mL. This DNA-surfactant complex is well tolerated to CV-1 cells and maintained high transfection activity when combined with a reagent that increased its association with cells. ACKNOWLEDGMENT

This work was supported by NIH Grant DK 46052 and by State of California Tobacco research Grant 6RT-0109 FS. Mass spectra were obtained by using a Kratos Tandem MS/MS mass spectrometer. It was provided by the UCSF Mass Spectrometry facility (A. L. Burlingame, Director) supported by the Biomedical Research Technology Program of the National Center for Research Resources, NIH NCRR BRTP RR01614 and NSF DIR 8700766. We thank Dr. Brigitte Sternberg-Papahadjopoulos for preparing the freeze-fracture electron micrograph and Lucie Gagne for help in the tissue culture experiments. LITERATURE CITED (1) Felgner, P. L., Heller, M. J., Lehn, P., Behr, J. P., and Szoka, F. C., Eds. (1996) Artificial self-assembling systems for gene delivery, American Chemical Society, Washington DC. (2) Friedmann, T. (1997) Overcoming the obstacles to gene therapy. Sci. Am. 96-101. (3) Nishikawa, M., Takemura, S., Takakura, Y., and Hashida, M. (1998) Targeted delivery of plasmid DNA to hepatocytes in vivo: optimization of the pharmacokinetics of plasmid DNA/galactosylated poly(L-lysine) complexes by controlling their physicochemical properties. J. Pharmacol. Exp. Ther. 287, 408-415. (4) Tang, M. X., Redemann, C. T., and Szoka, F. C. (1996) In vitro gene delivery by degraded polyamidoamine dendrimers. Bioconjugate Chem. 7, 703-714. (5) Kabanov, A. V., Felgner, P. L., and Seymour, L. W. Eds. (1998) Self-assembling complexes for gene delivery - From laboratory to clinical trial, John Wiley & Sons Ltd., Chichester, England. (6) Wagner, E., Cotten, M., Foisner, R., and Birnstiel, M. L. (1991) Transferrin-polycation-DNA complexes: The effect of

Ouyang et al. polycations on the structure of the complex and DAN delivery to cells. Proc. Natl. Acad. Sci. U.S.A. 88, 4255-4259. (7) Tang, M. X., and Szoka, F. C. (1997) The influence of polymer structure on the interactions of cationic polymers with DNA and morphology of the resulting complexes. Gene Ther. 4, 823-832. (8) Kaneda, Y., Iwai, K., and Uchida, T. (1989) Increased expression of DNA co-introduced with nuclear protein in adult rat liver. Science 243, 375-378. (9) Blessing, T., Remy, J. S., and Behr, J. P. (1998) Monomolecular collapse of plasmid DNA into stable virus-like particles. Proc. Natl. Acad. Sci. U.S.A. 95, 1427-1431. (10) Blessing, T., Remy, J. S., and Behr, J. P. (1998) Template oligomerization of DNA-bound cations produces calibrated nanometric particles. J. Am. Chem. Soc. 120, 8519-8520. (11) Trubetskoy, V. S., Budker, V. G., Hanson, L. J., Slattum, P. M., Wolff, J. A., and Hagstrom, J. E. (1998) Self-assembly of DNA-polymer complexes using template polymerization. Nucleic Acids Res. 26, 4178-4185. (12) Tarnowski, G. S., Barclay, R. K., Mountain, I. M., Nakamura, M., Satterwhite, H. G., and Solney, E. M. (1965) Determination of acid-soluble thiols and disulfides in transplanted animal tumors. Arch. Biochem. Biophys. 110, 210216. (13) Ellman, G. L. (1959) Tissue sulfhydryl groups. Arch. Biochem. Biophys. 82, 70-77. (14) Barron, L. G., Meyer, K. B., and Szoka, F. C. (1998) Effects of complement depletion on the pharmacokinetics and gene delivery mediated by cationic lipid-DNA complexes. Human Gene Ther. 9, 315-23. (15) Brito, R. M. M., and Vaz, W. L. C. (1986) Determination of the critical micelle concentration of surfactants using the fluorescent probe N-phenyl-1-naphthylamine. Analy. Biochem. 152, 250-255. (16) Sternberg, B., Hong, K., Zheng, W., and Papahadjopoulos, D. (1998) Ultrastructural characterization of cationic liposome-DNA complexes showing enhanced stability in serum and high transfection activity in vivo. Biochim. Biophys. Acta 1375, 23-35. (17) Felgner, J. H. (1997) Separation and quantitation of cationic liposome components by high performance liquid chromatiography with evaporative light-scattering detection. Pharm. Res. 14, 1269-1271. (18) Bligh, E. G., and Dyer, W. J. (1959) A rapid method of total lipid extraction and purification. Can. J. Biochem. Physiol. 37, 911-917. (19) Skehan, P., Storeng, R., Scudiero, D., Monks, A., McMahon, J., Vistica, D., Warren, J. T., Bokesch, H., Kenney, S., and Boyd, M. R. (1990) New colorimetric cytotoxicity assay for anticancer-drug screening. J. Natl. Cancer Inst. 82, 11071112. (20) Pan, H. L. a. F., T. L. (1968) A facile preparation of sulphydryl compounds in dimethyl sulphoxide. Chem. Ind. 546. (21) Koenig, N. H., Sasin, G. S., and Swern, D. (1958) Organic sulfur derivatives. V. preparation and properties of some longchain mercapto acids and related compounds. J. Org. Chem. 23, 1525-1530. (22) Ruckenstein, E., Huber, G., and Hoffmann, H. (1987) Surfactant aggregation in the presence of polymers. Langmuir 3, 382-387. (23) Melnikov, S. M., Sergeyev, V. G., and Yoshikawa, K. (1995) Transition of Double-Stranded DNA Chains Between Random Coil and Compact Globule States Induced By Cooperative Binding of Cationic Surfactant. J. Am. Chem. Soc. 117, 99519956. (24) Zana, R., and Talmon, Y. (1993) Dependence of aggregate morphology on structure of dimeric surfactants. Nature 362, 228-230. (25) Karaborni, S., Esselink, K., Hilbers, P. A. J., Smit, B., Karthauser, J., van Os, N. M., and Zana, R. (1994) Simulating the self-assembly of gemini (dimeric) surfactants. Science 266, 254-256.

BC990101Q