Article pubs.acs.org/journal/abseba
Parallel Synthesis and Quantitative Structure−Activity Relationship (QSAR) Modeling of Aminoglycoside-Derived Lipopolymers for Transgene Expression Bhavani Miryala,† Zhuo Zhen,‡ Thrimoorthy Potta,† Curt M. Breneman,‡ and Kaushal Rege*,† †
Chemical Engineering, School for Engineering of Matter, Transport, and Energy, Arizona State University, Tempe, Arizona 85287-6106, United States ‡ Department of Chemistry and Chemical Biology, Rensselaer Polytechnic Institute, Troy, New York 12180, United States S Supporting Information *
ABSTRACT: We describe the parallel synthesis of lipopolymers generated by conjugating alkanoyl chlorides to polymers derived from aminoglycoside antibiotic monomers as novel vehicles for transgene delivery and expression in mammalian cells. Parallel screening of lipopolymers led to the identification of six leads that demonstrated higher transgene expression efficacies in several cancer cells, when compared to the parental polymers as well as 25 kDa poly(ethylene imine), a current standard for polymer-mediated transgene expression. Quantitiative structure− activity relationship (QSAR)-based cheminformatics modeling was employed in order to investigate the role of lipopolymer physicochemical properties (molecular descriptors) on transgene expression efficacy. The predictive ability of the QSAR model, investgated using lipopolymers not employed for training the model, demonstrated excellent agreement with experimentally observed transgene expression. Our findings indicate that lipid substitution on aminoglycoside-derived polymers results in high levels of transgene expression compared to unsubstituted polymers. Taken together, these materials show significant promise in nonviral transgene delivery with several applications in biotechnology and medicine. KEYWORDS: gene delivery, aminoglycosides, lipids, cheminformatics, structure−property relationships, combinatorial chemistry
1. INTRODUCTION Delivery of exogenous DNA into mammalian cells has several applications in medicine and biotechnology.1−13 Although viruses are widely investigated for transgene delivery and expression, safety concerns associated with these vectors, particularly in therapeutic applications, necessitate the discovery of effective nonviral systems. Cationic lipids and polymers demonstrate certain advantages over viral vectors, including those associated with lower cost, flexibility in chemical design, and safety.14−20 However, nonviral vectors are often limited by low efficacies and high cytotoxicities.21 Recent efforts have therefore focused on increasing the biocompatibility and enhancing the efficacy of polymeric vehicles using appropriate modifications.22,23 We recently employed combinatorial synthesis and cheminformatics modeling in order to generate and evaluate a library of antibiotics-derived polymers with applications in transgene delivery and expression.24 Several aminoglycoside antibiotics have been approved by the FDA as antibacterial drugs, although concerns about nephrotoxicity and resistance limit their use. Aminoglycosides contain hydrophilic sugar, hydroxyl, and amine groups, which make them highly water-soluble, and therefore attractive in biomedical applications. The presence of multiple hydroxyl and/or amine moieties facilitates straightfor© XXXX American Chemical Society
ward derivatization and polymerization chemistries using aminoglycosides as starting substrates.25 Furthermore, it is likely that polymers derived from aminoglycosides are biodegradable because of the presence of glycosidic linkages in these molecules.26 Given these features, we have employed aminoglycosides as starting materials for generating soluble ligands for biological separations and DNA binding,27 microparticles,28 and more recently, soluble polymers for plasmid DNA delivery and transgene expression.24,29 Several candidates from this aminoglycoside-derived polymer library demonstrated lower cytotoxicities and higher efficacies of transgene expression following delivery of plasmid DNA (pDNA), compared to 25 kDa poly(ethylenimine), or pEI-25, a current standard for polymer-mediated transgene delivery. Quantitative structure−activity relationship (QSAR) models were employed in order to obtain insights into physicochemical factors influencing polymer efficacy by correlating molecular features to transgene expression. Recently, polymers from this library were also employed in order to enhance adenoviral delivery to resistant bladder cancer cells in vitro and in vivo.30 Received: January 27, 2015 Accepted: May 12, 2015
A
DOI: 10.1021/acsbiomaterials.5b00045 ACS Biomater. Sci. Eng. XXXX, XXX, XXX−XXX
Article
ACS Biomaterials Science & Engineering
of 27 lipopolymers in which three aminoglycoside-derived polymers were derivatized with three different alkanoyl chlorides (hexanoyl chloride, myristoyl chloride and stearoyl chloride) were synthesized (Scheme S1). Three different molar ratios, 1:2, 1:5, and 1:10 of the alkanoyl chloride: polymer were employed in the syntheses, leading to the 27-compound library. Lipopolymer syntheses were carried out based on reported methods.41 Briefly, 0.01 mmol of the polymer (48 mg of NG, 45 mg of PG, or 35 mg of AG) were dissolved in 2 mL of DMSO at room temperature by stirring for 30 min. Following this, triethyl amine (0.1 mmol, 10 μL) was added to the solution, which was stirred for an additional 30 min. The mixture was then cooled to 4 °C, and different amounts of alkanoyl chlorides (corresponding to molar ratios of 1:2, 1:5, or 1:10 with respect to the corresponding polymer) were added in a dropwise manner. This mixture was stirred at room temperature for 12 h. The lipopolymer product was collected by precipitation in excess ether. The precipitate was dissolved in water and dialyzed against water for 48 h using a 3500 molecular weight cutoff (MWCO) membrane. The dialyzed material was lyophilized to obtain the lipopolymer product. Abbrevations and nomenclature used for all 27 lipopolymers are provided in Table S1. The structural composition of the lipid-substituted polymers was analyzed using 1H NMR (Varian 400 instrument operating at 400 MHz) in D2O using tetramethyl silane (TMS) as an internal reference. The integrated values of the characteristic resonance shifts corresponding to lipid acid chloride (δ ∼ 0.8 ppm, −CH3) and aminoglycoside polymer (δ ∼ 5.6−6.0 ppm, O−CH−O) were used to obtain the extent of lipid substitution for a given lipopolymer (Table S1). Infrared (IR) spectra of lipopolymers were measured using FT-IR spectroscopy with a Bruker IFS 66 V/S 32 mm instrument using a round cell window (KBr), in order to further confirm the conjugation of lipid onto parental aminoglycoside-derived polymers. 2.2.3. Cell Culture. PC3, PC3-PSMA, and 22Rv1 human prostate cancer cells were cultured in RPMI-1640 media, and MB49 murine bladder cancer cells were cultured in DMEM media supplemented with 10% FBS and 1% penicillin/streptomycin (10000 units/mL) solution in an incubator containing 5% CO2 at 37 °C. At approximately 80% confluence, cells were trypsinized with 0.25% trypsin, seeded at a density of 9000 cells/well in 96-well plates (Corning, Corning, NY, USA), and allowed to attach overnight for subsequent experimentation. 2.2.4. In Vitro Transgene (luciferase) Expression Following Lipopolymer-Mediated Plasmid DNA (pDNA) Delivery. Lipopolymers were screened in parallel for transgene expression efficacy by delivering the pGL4.5 control vector (Promega Corp., Madison, WI), which encodes for the modified firefly luciferase protein; luciferase expression was determined in different cancer cell lines. The pGL4.5 plasmid DNA (pDNA) was prepared using methods described previously.42 Plasmid DNA concentration and purity were determined using a NanoDrop Spectrophotometer (ND-1000; NanoDrop Technologies) by measuring absorbance at 260 and 280 nm. Human prostate cancer cells (PC3, PC3-PSMA and 22Rv1) and murine bladder cancer cells (MB49) were used for transfection studies in vitro. Initial screening of all 27 lipopolymers was carried out in PC3 cells and leads selected from this screen were further evaluated in PC3PSMA, 22Rv1 and MB49 cells. Plasmid DNA (pDNA) (100 ng) was complexed with varying amounts of lipopolymers in 1X PBS buffer for 20 min leading to the formation of lipopolyplexes (i.e., lipopolymer: pDNA complexes). Lipopolymer: pDNA weight ratios were varied from 1:1 to 100:1 for the initial screen in PC3 cells, and from 5:1 to 50:1 for evaluating lead lipopolymers in PC3-PSMA, 22Rv1, and MB49 cells. Cells were seeded at a density of 9000 per well in a 96-well plate for 18−24 h following which, they were incubated with lipopolyplexes for 6h in serum-free media. At this point, media was replaced with fresh 150 μL serum-containing media. After 48 h of further incubation, cells were lysed, and luciferase protein expression was determined as relative luminescence units (RLU) using the Bright GloTM Luciferase assay kit (Promega) using a plate reader (Bio-Tek Synergy 2). Cell lysates were also assayed for total protein content using the BCA Protein Assay kit (Pierce, Rockford, IL, USA). RLU values were
Lead optimization is an approach often employed in drug discovery in order to enhance efficacy of candidates initially identified from parallel/high-throughput synthesis and screening. Our previous QSAR studies indicated that presence of hydrophobic character enhanced the transgene expression efficacy of aminoglycoside-derived polymers.24 Lipid-conjugated cationic polymers combine the benefits of plasmid DNA (pDNA) condensation activity due to presence of positive charge, as well as increased cellular interactions due to presence of lipid moieties on a single carrier system.31−37 Indeed, previous studies have indicated that cationic polymers derivatized with hydrophobic moieties can demonstrate improved transgene expression when compared to parent polymers.38−40 We therefore hypothesized that modification of lead aminoglycoside polymers using aliphatic lipids is a powerful optimization approach for enhancing the transgene expression of parental aminoglycoside-derived polymers. Three polymers, derived from the aminoglycosides neomycin, paromomyicin and apramycin cross-linked with glycerol diglycidylether (GDE), were derivatized with hexanoyl (C6) chloride, myristoyl (C14) chloride and stearoyl (C18) chloride. This combinatorial matrix of lipopolymers was screened in parallel in order to rapidly identify successful candidates that demonstrated high efficacies of transgene expression in different cancer cells. Specifically, the role of lipid chemistry and degree of substitution on transgene expression efficacy were investigated using experimental and computational chemistry methods. QSAR models were developed in order to obtain insights into the physicochemical factors that influence lipopolymer-mediated transgene expression.
2. EXPERIMENTAL SECTION 2.1. Materials. Six aminoglycoside-derived polymers were synthesized using methods described previously. Three alkanoyl chlorideshexanoyl chloride, myristoyl chloride, and stearoyl chloridetriethyl amine, dimethyl sulfoxide (DMSO), and diethyl ether were purchased from Sigma-Aldrich and used without any further purification. Branched poly(ethylene imine), Mw = 25 kDa and Mn = 10 kDa, (here onward referred to as pEI) was also purchased from Sigma-Aldrich. The pGL4.5 control vector, which encodes for modified firefly luciferase protein under the control of an SV promoter, and the Bright-Glo luciferase assay system were purchased from Promega Corporation (Madison, WI). The BCA protein assay kit was purchased from Thermo Scientific Inc. (Rockford, IL). Gel permeation chromatography (GPC) standards were purchased from American Polymer Standards Corporation. 1H NMR spectra were determined with a Varian 400 instrument operating at 400 MHz in the Fourier transform mode. FT-IR spectroscopy was carried out with a Bruker IFS 66 V/S 32 mm instrument using a round cell window (KBr). 2.2. Synthesis, in Vitro Transgene Expression, and Cytotoxicity Studies of Lipopolymers Following Plasmid DNA (pDNA) Delivery. 2.2.1. Synthesis of Aminoglycoside-Derived Polymers. Three aminoglycoside-derived polymers, neomycin-GDE, paromomycin-GDE and apramycin-GDE, were synthesized and characterized using methods described previously.24 Briefly, sulfate-free aminoglycosides (neomycin, paromomycin and apramycin) were reacted with glycerol digylcidyl ether (GDE) in a 1:2 molar ratio in a solvent mixture of water and N,N-dimethylformamide (DMF) (1.5:1) for 5 h at 60 °C. The product obtained upon precipitation using acetone was further purified by dialysis using a 3500 molecular weight cutoff (MWCO) membrane in order to remove unreacted aminoglycosides. The dialyzed material was lyophilized to obtain the polymer product. Henceforth, neomycin-GDE, paromomycin-GDE, and apramycinGDE polymers are referred to as NG, PG, and AG, respectively. 2.2.2. Synthesis and Characterization of Lipid-Conjugated Aminoglycoside-Derived Polymers (Lipopolymers). A small library B
DOI: 10.1021/acsbiomaterials.5b00045 ACS Biomater. Sci. Eng. XXXX, XXX, XXX−XXX
Article
ACS Biomaterials Science & Engineering normalized by the protein content to yield ‘RLU/mg protein’ values, which were employed for comparing different lipopolymers. Untransfected cells and cells transfected with uncomplexed (“free”) pDNA were used as controls. Luciferase expression efficacies of polymers from the library were compared to that obtained using 25 kDa branched pEI. In all cases, the pEI solution was prepared fresh right before all transfection experiments because of possible variability in efficacy and toxicity induced by storage. 2.2.5. Evaluation of Lipopolymer Cytotoxicity. The MTT assay is a metabolic assay for proliferation, and was employed as an indicator for viability. PC3 cell viability was determined following treatment with optimal weight ratios 5:1 to 50:1 for all 27 lipopolymers; cell seeding and lipopolyplex formation was similar to that described in section 2.2.3. Untreated control wells were treated only with
Corporation, MA) and a refractive index detector (Waters 2410), was employed for determining molecular weights of individual lipopolymers. An aqueous solvent containing 0.1% trifloroacetic acid and 40% acetonitrile was used as the mobile phase. The mobile phase in the column was operated at a flow rate of 0.5 mL/min, and a column temperature of 35 °C was employed in the GPC experiments. Poly(2-vinylpyridine) (American Polymer Standards Corporation, OH) Molecular Weights (MWs: 3300, 7600, 12 800, 35 000, and 70 000 Da) were used as standards for calibration. Chromatograms were analyzed using Waters Millennium 32 GPC software. 2.4. Polymer-Mediated EGFP Transgene Expression Observation and Quantification. Lead lipopolymers were employed to deliver EGFP (enhanced green fluorescent protein)-encoding pDNA (pEF-GFP)45 to PC3 cells, using methods described for pGL4.5 plasmid above. GFP expression was visualized using a fluorescence microscope (Zeiss), and quantified using flow cytometry (BD FACS Caliber flow cytometer using the FL1 channel). For flow cytometry studies, the supernatant media was first removed from the plated cells, and each well was then washed with 150 μL, 1X PBS, trypsinized with 150 μL, 0.25% trypsin-EDTA. The entire liquid volume was collected into 1.5 mL microcentrifuge tubes and centrifuged at 8000 rpm for 10 min using Beckman Coulter Microcentrifuge-18. The supernatant was removed and the cell pellet was resuspended in 50 μL, 1X PBS. The percentage of cells exhibiting fluorescence, as well as mean fluorescence of cells in the population, was determined using flow cytometry (Becton Dickinson FACScalibur Flow Cytometer using the FL1 channel); untreated cells were used as controls. 2.5. Lipopolymer-Mediated Transgene Expression in the Presence of Serum. Neomycin-based lipopolymers at 10:1, and apramycin- and paromomycin-based lipopolymers at 50:1 lipopolymer:pDNA weight ratios were investigated for their transgene (luciferase) expression efficacies in the presence of serum. Optimal ratios were used for the lipopolymers in each case. As before, PEI was used as a standard for comparison. Other experimental procedures and determination of luciferase expression are similar to that previously described in section 2.2.3. 2.6. Statistical Analyses of Experimental Data. All cell-based experiments were carried out in triplicate each time, on three different days and the results are expressed as mean ± one standard deviation. A two-way ANOVA analysis was carried out in order to determine statistical significance of difference between group means; p-values