Structural Study of DNA Condensation Induced by Novel

Andrew Lewis , Yiqing Tang , Steve Brocchini , Ji-won Choi and Antony Godwin ... Chengming Li, Niklaas J. Buurma, Ihtshamul Haq, Colin Turner, and Ste...
0 downloads 0 Views 578KB Size
Langmuir 2005, 21, 3591-3598

3591

Structural Study of DNA Condensation Induced by Novel Phosphorylcholine-Based Copolymers for Gene Delivery and Relevance to DNA Protection Y. T. A. Chim,† J. K. W. Lam,‡ Y. Ma,§ S. P. Armes,§ A. L. Lewis,| C. J. Roberts,*,† S. Stolnik,‡ S. J. B. Tendler,† and M. C. Davies† Laboratory of Biophysics and Surface Analysis, School of Pharmacy, The University of Nottingham, NG7 2RD, U.K., Advanced Drug Delivery Group, School of Pharmacy, The University of Nottingham, Nottingham, NG7 2RD, U.K., Department of Chemistry, Sheffield University, Sheffield, S3 7HF, U.K., and Biocompatibles Ltd., Farnham Business Park, Farnham, Surrey, GU9 8QL, U.K. Received October 12, 2004. In Final Form: January 25, 2005 Poly[2-(dimethylamino)ethyl methacrylate-b-2-methacryloyloxyethyl phosphorylcholine] (DMA-MPC) is currently under investigation as a new vector candidate for gene therapy. The DMA block has been previously demonstrated to condense DNA effectively. The MPC block contains a phosphorylcholine (PC) headgroup, which can be found naturally in the outside of the cell membrane. This PC-based polymer is extremely hydrophilic and acts as a biocompatible steric stabilizer. In this study, we assess in detail the morphologies of DNA complexes obtained using the diblock copolymer series DMAxMPC30 (where the mean degree of polymerization of the MPC block was fixed at 30 and the DMA block length was systematically varied) using transmission electron microscopy (TEM) and liquid atomic force microscopy (AFM). Both techniques indicate more compact complex morphologies (more efficient condensation) as the length of the cationic DMA block increases. However, the detailed morphologies of the DMAxMPC30-DNA complexes observed by TEM in vacuo and by AFM in aqueous medium are different. This phenomena is believed to be related to the highly hydrophilic nature of the MPC block. TEM studies revealed that the morphology of the complexes changes from loosely condensed structures to highly condensed rods, toroids, and ovalshaped particles as the DMA moiety increases. In contrast, morphological changes from plectonemic loops to flowerlike and rectangular blocklike structures, with an increase in highly condensed central regions, are observed by in situ AFM studies. The relative population of each structure is clearly dependent on the polymer molecular composition. Enzymatic degradation assays revealed that only the DMA homopolymer provided effective DNA protection against DNase I degradation, while other highly condensed copolymer complexes, as judged from TEM and gel electrophoresis, only partially protected the DNA. However, AFM images indicated that the same highly condensed complexes have less condensed regions, which we believe to be the initiation sites for enzymatic attack. This indicates that the open structures observed by AFM of the DNA complexation by the DMAxMPC30 copolymer series are closer to in vivo morphology when compared to TEM.

1. Introduction DNA typically exists in its extended (rather than condensed) conformations in aqueous solution, behaving as an anionic polyelectrolyte.1 This makes DNA susceptible to enzymatic degradation and also renders it unable to cross biological barriers unaided. For successful gene delivery, therapeutic DNA must be effectively protected from enzymatic degradation and transported through a number of biological barriers to a specific cell target. Inside the target cells, the DNA must escape the enclosing endosome, travel through the cytosol, and cross nuclear membrane to reach the nucleus. Evolution has enabled viruses to overcome this formidable delivery problem. Thus, it is not surprising that viral vectors are currently the largest studied group of gene delivery systems. However, there are significant safety issues for the host, and both the immunogenicity and the limited DNA * Author to whom correspondence should be addressed. E-mail: [email protected]. † Laboratory of Biophysics and Surface Analysis, The University of Nottingham. ‡ Advanced Drug Delivery Group, The University of Nottingham. § Sheffield University. | Biocompatibles Ltd. (1) Park, S. Y.; Harries, D.; Gelbart, W. M. Biophys. J. 1998, 75, 714-720.

packaging capacity of viruses present serious obstacles to their application as optimal gene carriers for human gene therapy.2 This has recently resulted in the extensive studies of a wide range of synthetic vectors, notably cationic block copolymers.3-5 Cationic polymers have been widely used for DNA condensation since they form strong charge-compensated complexes with the anionic phosphate groups on the DNA backbone. The resulting polymer-DNA complexes have colloidal dimensions, which are appropriate for cellular uptake, and the DNA is protected from enzymatic degradation. The ability of a cationic polymer to condense DNA varies according to its chemical structure, polymer architecture, charge density, and molecular weight. Clearly, the degree of DNA condensation is important in determining the effectiveness of the complex to deliver the therapeutic gene into the target cell nucleus.5 Thus, it is important to understand the relationship between (2) Lehn, P.; Fabrega, S.; Oudrhiri, N.; Navarro, J. Adv. Drug Delivery Rev. 1998, 30, 5-11. (3) Bielinska, A. U.; Kukowska-Latallo, J. F.; Baker, J. R. Biochim. Biophys. Acta 1997, 1353, 180-190. (4) Martin, A. L.; Davies, M. C.; Rackstraw, B. J.; Roberts, C. J.; Stolnik, S.; Tendler, S. J. B.; Williams, P. M. Fed. Eur. Biochem. Soc. Lett. 2000, 480, 106-112. (5) Merdan, T.; Kopecek, J.; Kissel, T. Adv. Drug Delivery 2002, 54, 715-758.

10.1021/la047480i CCC: $30.25 © 2005 American Chemical Society Published on Web 03/04/2005

3592

Langmuir, Vol. 21, No. 8, 2005

the polymer properties and the precise nature of the DNA complexes that are formed. Poly[2-(dimethylamino)ethyl methacrylate-b-2-methacryloyloxyethyl phosphorylcholine] (DMA-MPC) diblock copolymer6,7 is currently being evaluated as a potential new synthetic vector for gene delivery. Cationic DMA homopolymers and diblock copolymers have been demonstrated to condense DNA efficiently.8,9 The MPC block contains zwitterionic phosphorylcholine (PC) headgroups, which are commonly found in the outer leaflet of cellular membranes. The biomimetic PC-based polymers have previously shown excellent biocompatibility and are used as coating material for various medical implants such as coronary stent.10,11 A recent study revealed that the hydrophilic MPC effectively provides steric stabilization to the polymer-DNA complex and hence ensures good colloidal stability.12 A series of DMA-MPC diblock copolymers with different compositions were synthesized using atom transfer radical polymerization (ATRP), which allows block lengths to be controlled by varying the monomer/initiator molar ratio.13 In our previous work, it was found that approximately 30 repeat units is the minimum MPC block length to prevent aggregation and provide steric stabilization to the complexes.12 Hence, in the present work, the mean degree of polymerization of the MPC block was fixed at 30 while the DMA block length was varied so as to produce a series of DMAxMPC30 copolymers. A monomer/nucleotide molar ratio of 2:1 was used in this study, as previous studies showed that at this ratio a reasonable degree of DNA condensation was achieved showing a range of DNA condensates12 while the presence of an excess of free polymer is minimized. Relative condensation efficiencies were assessed by comparing the morphologies of the DNA complexes using transmission electron microscopy (TEM) and in situ atomic force microscopy (AFM). Additionally, gel retardation assays were performed to assess the copolymers ability to condense DNA and efficacy to protect from DNase I enzymatic degradation. TEM has been traditionally employed to examine the structure of DNA condensates.14 The recent development of in situ AFM allows biological samples to be imaged directly in an aqueous environment.15,16 AFM is considered particularly appropriate for studying the DMAxMPC30 copolymer series since the MPC block is highly hydrophilic and the dehydration that necessarily occurs under TEM conditions could perturb the structures of the DMAxMPC30-DNA complexes. Additionally, AFM requires minimal sample preparation and no selective staining techniques. Improved understanding of the complex morphology should provide valuable information regarding the potential of these complexes for gene delivery applications. (6) Ma, Y. H.; Tang, Y. Q.; Billingham, N. C.; Armes, S. P.; Lewis, A. L.; Lloyd, A. W.; Salvage, J. P. Macromolecules 2003, 36, 3475-3484. (7) Lewis, A. L. Colloids Surf., B 2000, 18, 261-275. (8) Wetering, P.; Cherng, J. Y.; Talsma, H.; Crommelin, D. J. A.; Hennink, W. E. J. Controlled Release 1998, 53, 145-153. (9) Rungsardthong, U.; Deshpande, M.; Bailey, L.; Vamvakaki, M.; Armes, S. P.; Garnett, M. C.; Stolnik, S. J. Controlled Release 2001, 73, 359-380. (10) Lewis, A. L.; Cumming, Z. L.; Goreish, H. H.; Kirkwood, L. C.; Tolhurst, L. A.; Stratford, P. W. Biomaterials 2001, 22, 99-111. (11) Lewis, A. L.; Tolhurst, L. A.; Stratford, P. W. J Long-Term Eff. Med. Implants 2002, 12, 231-250. (12) Lam, J. K. W.; Ma, Y.; Armes, S. P.; Baldwin, T.; Stolnik, S. J. J. Controlled Release 2004, 100, 293-312. (13) Lobb, E. J.; Ma, I.; Billingham, N. C.; Armes, S. P. J. Am. Chem. Soc. 2001, 123, 7913-7914. (14) Bloomfield, V. A. Curr. Opin. Struct. Biol. 1996, 6, 334-341. (15) Hansma, H. G.; Golan, R.; Hsieh, W.; Lollo, C. P.; Mullen-Ley, P.; Kwoh, D. Nucleic Acids Res. 1998, 26, 2481-2487. (16) Pope, L. H.; Davies, M. C.; Laughton, C. A.; Roberts, C. J.; Tendler, S. J. B.; Williams, P. M. Anal. Chim. Acta 1999, 400, 27-32.

Chim et al.

Figure 1. Chemical structure of poly(2-dimethylamino)ethyl methacrylate-b-2- methyacryloyloxylethyl phopshorylcholine copolymer (DMAxMPCy). A series of copolymers with different x and y ) 30 were created.

2. Materials and Methods Reagents and Buffers. Ethidium bromide solution, PBS tablets, EDTA, glacial acetic acid, Tris acetate, agarose, bromphenol blue, sucrose and uranyl acetate, DNase I, poly(Laspartic acid), and magnesium chloride were purchased from Sigma (Dorset, UK). Tris acetate EDTA (TAE) buffer consists of 40 mM Tris acetate and 20 mM glacial acetate acid. Phosphate buffered saline (PBS) (0.01 M) was prepared by dissolving the PBS tablets in 200 mL ELGA water. DNA loading buffer contained 0.25% w/v bromophenol blue in a 40% sucrose solution. DNA and Polymer. gWiz luc plasmid (6732 bps) containing the luciferase reporter gene was purchased from Aldevron (Fargo, ND). The plasmid was supplied at a concentration of 5.6 mg/mL, which was diluted to 1 mg/mL with autoclaved, sterile distilled water before use. The chemical structure of DMA-MPC diblock copolymer is shown in Figure 1. Each of the DMAxMPC30 diblock copolymers was synthesized using the ATRP protocol described in the literature.6,13 Gel Retardation Assay. Aliquots of aqueous copolymer solution were added to the DNA solution (1 µg/µL). TAE buffer was added to produce a final volume of 8 µL. All complexes were prepared at a monomer/nucleotide molar ratio of 2:1. Free DNA and free polymer were used as controls. The samples were briefly mixed by gentle vortexing and pulse-spun using a benchtop microcentrifuge prior to incubation at room temperature for 30 min. DNA loading buffer (1.5 µL) was added to the samples, which were then mixed and respun again. Samples were loaded into 0.8% agarose gel containing ethidium bromide (1 mg/mL). Electrophoresis was carried out at 70 V in TAE buffer (pH 7.4) for 60 min. The DNA band was visualized under a UV transilluminator. Copolymers were stained by immersing the gel in staining solution (10:50:40 glacial acetic acid, methanol, double distilled water, 0.1% w/v Coomassie blue) for 1 h and washing with destaining solution (10:10:80, glacial acetic acid, methanol, double distilled water) overnight. Enzyme Degradation Study by Gel Electrophoresis. For dissociation of complexes, polymer-DNA complexes (2 µg of DNA) were prepared at a monomer/nucleotide molar ratio 2:1 in TAE × 1 buffer (pH 7.4). The complexes were incubated at room temperature for 30 min. Dissociation of complexes was achieved by addition of excess of p(Asp) (12.5 µg p(Asp)/µg DNA). The mixture was incubated for a further 10 min followed by addition of 3 µL DNA loading buffer. Complexes were analyzed by gel electrophoresis (0.8% agarose gel containing 1µg/mL ethidium bromide) in TAE × 1 buffer (pH 7.4). The electrophoresis was run as mentioned before. For enzyme degradation study, polymer-DNA complexes were prepared as mentioned. MgCl2 was added (to final concentration of 10 mM) in order to activate the enzymes. After 30 min, the complexes were incubated with DNase I (1 U/µg DNA) for 10 min at 37 °C. The activity of DNase I was then stopped by addition of 5 µL EDTA (0.5 M) that chelated with Mg2+. Complexes were dissociated by addition of p(Asp) as described above followed by addition of DNA loading buffer. The samples were then analyzed by gel electrophoresis as mentioned. Free DNA without any treatment, free DNA incubated with DNase I, and DNA ladder λ-Hind III served as controls. TEM Study. DNA solution (2.5 µg) was added to a 10-fold diluted PBS solution followed by addition of the copolymer solution. All the complexes were prepared at a 2:1 monomer/ nucleotide molar ratio, and the total volume of each sample was 250 µL. The samples were incubated at room temperature for 30 min. Drops of each solution were placed onto a copper grid coated with pioloform resin for 30 s. Excess buffer was drawn away

Structural Study of DNA Condensation

Langmuir, Vol. 21, No. 8, 2005 3593

Figure 3. TEM and AFM images of DMA homopolymer-DNA complexes at a monomer/nucleotide molar ratio of 2:1 prepared in 10% v/v PBS. (a) TEM image of DMA homopolymer (Mn ) 12 700)-DNA complexes and (b) TEM image of DMA homopolymer (Mn ) 4100)-DNA complexes, (c) AFM image of DMA homopolymer (Mn ) 4100)-DNA complexes. Image size is 550 nm × 550 nm. Scale bars are 200 nm. Figure 2. Agarose gel retardation assay. Lane 1 is the DNA control. Lanes 2-7 corresponds to polymer-DNA complexes at monomer/nucleotide molar ratios of 2:1. Lanes 8-13 corresponds to polymer control. Lanes 2 and 8 ) DMA homopolymer (Mn ) 4100), lanes 3 and 9 ) DMA homopolymer (Mn ) 12 700), lanes 4 and 10 ) DMA10MPC30, lanes 5 and 11 ) DMA20MPC30, lanes 6 and 12 ) DMA40MPC30, lanes 7 and 13 ) DMA60MPC30. with filter paper. The sample was then stained with uranyl acetate (4% w/v solution of uranyl acetate in 50% ethanol) for 2 min. The grid was then washed in 50% ethanol once and distilled water twice prior to air-drying. The samples were analyzed under TEM (JEOL JEM-1010 TEM). Micrographs were taken at different magnifications. Image analysis was carried out using an analysis Pro 3.1 Soft Imaging System. AFM Study. Copolymer-DNA complexes at a 2:1 monomer/ nucleotide molar ratio in 30 µL of 10% PBS buffer were incubated at room temperature for varying time intervals prior to immobilization onto freshly cleaved muscovite mica (Agar Scientific, Essex, UK) of approximately 1 cm2 and being imaged directly in liquid. An AFM liquid cell and oxidation-sharpened NP-S tips on a V-shaped, silicon nitride cantilever, with a spring constant of around 0.1 N/m (Nanoprobe, Veeco Instruments) and resonant frequency between 8 and 10 kHz were used. All AFM imaging was carried out using a Tapping Mode (TM) AFM on a Veeco Nanoscope (IIIa) MultiMode system (Veeco Instruments, Santa Barbara, CA). Topographical images were taken at 512 × 512 pixel resolution, plane-flattened, and analyzed either by the computer program accompanying the Nanoscope IIIa Multimode AFM or by an offline processing package, SPIP (version 2.2.2) (Image Metrology, Lyngby, Denmark).

3. Results The ability of the copolymers to form DNA complexes was assessed by gel retardation assay at a 2:1 monomer/ nucleotide molar ratio (Figure 2). No fluorescence was detected for the complexes produced using the two DMA homopolymers (both high and low molecular weight) or the DMA60MPC30 copolymer. Fluorescence was observed inside the loading well for complexes prepared using the DMA10MPC30, DMA20MPC30, and DMA40MPC30 copolymers. In addition, fluorescence was detected in the negative side of the gel for complexes based on the DMA10MPC30 and DMA20MPC30 copolymers. Free copolymer was present in all of the copolymer-DNA complexes studied. TEM and AFM images of highly aggregated DNA complexes obtained using the DMA homopolymer are shown in Figure 3. DNA complexes formed with a highmolecular-weight (Mn ) 12 700) and low-molecular-weight (Mn ) 4100) DMA homopolymer, stained with uranyl acetate, were visualized as clumps of dark aggregates by TEM (Figure 3a and b, respectively). DNA complexes formed with a low-molecular-weight (Mn ) 4100) DMA homopolymer observed by AFM (Figure 3c) revealed a large DNA aggregate with smaller DNA aggregates around.

Figure 4. DMA10MPC30-DNA complexes at monomer/nucleotide molar ratios of 2:1 under TEM (a and b) and AFM (c, d, and e) (z ) 10 nm), where images a and c are at 1.5 µm × 1.5 µm and b, d, and e are at 500 nm × 500 nm. Scale bars are 200 nm.

A range of morphologies was obtained for the DNA complexes prepared using the DMAxMPC30 copolymers. For the DMA10MPC30 copolymer, ‘spaghetti-like’ structures were observed under TEM (Figure 4a and b), whereas plectonomic loops were clearly visible in the AFM images (Figure 4c-e). The average width and contour of these ‘spaghetti-like’ structures under TEM were approximately 7 ( 1 and 540 ( 340 nm (n ) 20), respectively. For the DMA20MPC30 copolymer, TEM revealed the presence of long rod and loose ring structures, with a significant fraction of well-defined rods and toroids (Figure 5a and b). In general, both ringlike and linear structures had similar widths of 10.8 ( 2.2 nm (n ) 20). The average length of the linear structures was 141.9 ( 117.0 nm (n ) 20), and the average outer diameter of the toroids was 54.8 ( 12.9 nm (n ) 20). AFM studies of the same complexes indicate flowerlike structures and partial toroids and rods with protruding loose petals consistent with DNA in terms of dimensions (Figure 5c-e). TEM studies revealed both rodlike and toroidal morphologies for complexes prepared using the DMA40MPC30 copolymer, as well as intermediate structures (Figure 6a and b). These toroids and rods had very similar mean widths of around 16.6 ( 2.8 nm (n ) 20). The mean rod lengths were approximately 70.8 ( 19.0 nm (n ) 20), and the outer toroid diameter was approximately 51.8 ( 12.2 nm (n ) 20). AFM studies of the same complexes indicated a mixture of flowerlike structures, partial toroids, and rectangular blocklike structures (tightly packed flowers), often with protruding petals of DNA (Figure 6c-e). Relatively thick rodlike or rectangular blocklike complexes (mean length ) 52 ( 12 nm, mean width ) 32 ( 8 nm; n ) 100) were formed by the DMA60MPC30 copolymer, as judged by TEM (Figure 7a and b). However,

3594

Langmuir, Vol. 21, No. 8, 2005

Figure 5. DMA20MPC30-DNA complexes at monomer/nucleotide molar ratios of 2:1 under TEM (a and b) and AFM (c, d, and e) (z ) 10 nm), where images a and c are at 1.5 µm × 1.5 µm and b, d, and e are at 500 nm × 500 nm. Scale bars are 200 nm.

Figure 6. DMA40MPC30-DNA complexes at monomer/nucleotide molar ratios of 2:1 under TEM (a and b) and AFM (c, d, and e) (z ) 10 nm), where images a and c are at 1.5 µm × 1.5 µm and b, d, and e are at 500 nm × 500 nm. Scale bars are 200 nm.

Figure 7. DMA60MPC30-DNA complexes at monomer/nucleotide molar ratios of 2:1 under TEM (a and b) and AFM (c, d, and e) (z ) 10 nm), where images a and c are at 1.5 µm × 1.5 µm while b, d, and e are at 500 nm × 500 nm. Scale bars are 200 nm.

in situ AFM imaging indicated predominately flowerlike structures and less anisotropic rectangular blocks (mean length ) 163 ( 57 nm, mean width ) 129 ( 51 nm; n ) 41). Figure 8 shows high-resolution images of individual semi-toroidal/rodlike structures observed by TEM for complexes prepared using the DMA20MPC30 (Figure 8ac) and DMA40MPC30 copolymers (Figure 8f-h). It also

Chim et al.

Figure 8. Examples of individual DNA condensates allowing the comparison of two copolymer systems, namely DMA20MPC30 and DMA40MPC30. TEM images of semi-toroid/rods structures formed with DMA20MPC30 (a, b, and c) and DMA40MPC30 (f, g, and h) in the presence of luciferase plasmid at a monomer/ nucleotide molar ratio of 2:1 in 10% v/v PBS. AFM images (z ) 5 nm) of partial toroid and rod structures formed with DMA20MPC30 (d and e) and partial toroid and flowerlike structures formed with DMA40MPC30 (i and j) in the presence of luciferase plasmid at a monomer/nucleotide molar ratio of 2:1 in 10% v/v PBS. All image sizes are 250 nm × 250 nm. Scale bars are 100 nm.

shows a typical flowerlike structure (Figure 8d) and partial rod structure (Figure 8e) of DNA complexes obtained with the DMA20MPC30 copolymer and the partial toroid (Figure 8i) and flowerlike structures (Figure 8j) generally produced by DNA complexes involving the DMA40MPC30 copolymer under AFM. Table 1 summarizes the dimensions of both the rods/ linear and ringlike/toroidal structures obtained from the analysis of the TEM images. As the DMA block length is increased, the mean width of these structures increases, whereas the mean length of the rods/linear structures decreased progressively. However, the outer toroidal diameters formed by the DMA20MPC30 and DMA40MPC30 copolymers were similar. Table 2 shows the estimated relative proportions of toroids and rods. For DNA complexes formed by the DMA20MPC30 copolymer, about half of the population were rods and a significant fraction of the complexes appeared to be intermediate between rods and toroids. A relatively high proportion of rodlike complexes were obtained with the DMA40MPC30 copolymer, with less ‘intermediate’ structures being evident compared to the DMA20MPC30 copolymer. Figure 9 depicts a distribution chart constructed for 370 complexes prepared using various DMAxMPC30 copolymers and observed using in situ AFM at low (1 µg/ mL), intermediate (10 µg/mL) and relatively high (50 µg/ mL) DNA concentrations. A 2:1 copolymer/DNA molar ratio was used in each case. Complexes formed using the DMA20MPC30 copolymer exhibited principally flowerlike structures, with partial toroids also observed at intermediate and high DNA concentrations and some evidence for rodlike structures observed at high DNA concentration. Although flowerlike structures were predominately observed for DMA40MPC30-DNA complexes at low and intermediate DNA concentrations, rectangular blocklike structures were prevalent at high DNA concentration. Partial toroids and rods were also observed at intermediate and high DNA concentrations. Flowerlike and rectangular blocklike structures were observed at low DNA concentrations, and only rectangular blocklike structures were observed at intermediate and high DNA concentrations for the DMA60MPC30-DNA complexes. Figure 10 shows the polymer-DNA complexes were all dissociated by p(Asp) to release DNA. The topologies of all released DNA were similar to the DNA control. When the complexes were subjected to DNase I incubation, only the DMA homopolymer offered some protection. However, the presence of low-molecular-weight DNA fragments

Structural Study of DNA Condensation

Langmuir, Vol. 21, No. 8, 2005 3595

Figure 9. A percentage distribution chart of the different complex structures observed with different polymers at different DNA concentrations using AFM. The monomer ratio of polymer to DNA was 2:1 in all cases. Total number of complexes, n ) 370. For DMA20MPC30, n ) 64; DMA40MPC30, n ) 172 and DMA60MPC30, n ) 134. Table 1. Measurements of Thickness, Contour Lengths of Toroids and Rods, and the Outer Diameter of DNA Condensates with DMA10MPC30, DMA20MPC30, and DMA40MPC30 under TEMa

a

copolymer system

thickness/width (nm) ( sd

contour length of rods/linear structure (nm) ( sd

contour length of toroid (nm) ( sd

outer diameter of toroid (nm) ( sd

DMA10MPC30 DMA20MPC30 DMA40MPC30

7.1 ( 1.3 10.8 ( 2.2 16.6 ( 2.8

537.4 ( 342.3 141.9 ( 117.0 70.8 ( 19.0

n/a 155.2 ( 37.1 136.7 ( 29.5

n/a 54.8 ( 12.9 51.8 ( 12.2

(n ) 20).

Table 2. Comparison of the Relative Populations of Toroids and Rods for DNA Condensates with DMA20MPC30 and DMA40MPC30 under TEMa copolymer system

% toroids

% rods

% semi toroids/rods

DMA20MPC30 DMA40MPC30

22.7 16.7

49.3 74.0

28.0 9.3

a

(n ) 300).

suggests that partial breakdown of DNA still occurs. For the DMA-MPC copolymers, the original DNA topology could no longer be seen. Only low-molecular-weight DNA fragments were observed with DMA60MPC30 and DMA40MPC30 systems. In the case of DMA20MPC30, DMA10MPC30, and naked DNA, fluorescence could not be detected at all, suggesting complete degradation of DNA. 4. Discussion Physicochemical evaluation of DMA-MPC diblock copolymers confirmed their ability to form DNA complexes, with the MPC block conferring improved colloid stability via steric stabilization.12 However, the extent of DNA condensation appeared to be adversely affected by the presence of the zwitterionic MPC block, which seems to hinder the condensation process.12 The MPC block is a highly hydrated structure, whereby each monomer unit

Figure 10. Enzyme degradation assay by gel electrophoresis. Lanes 1-6 correspond to polymer-DNA complexes at monomer/ nucleotide molar ratios of 2:1 after dissociation using poly(Laspartic acid). Lane 7 corresponds to DNA control. Lanes 9-15 correspond to lane 1-7 but with DNase I. Lanes 1 and 9 ) DMA homopolymer (Mn ) 4100), lanes 2 and 10 ) DMA homopolymer (Mn ) 12 700), lanes 3 and 11 ) DMA60MPC30, lanes 4 and 12 ) DMA40MPC30, lanes 5 and 13 ) DMA20MPC30, lanes 6 and 14 ) DMA10MPC30, and lanes 7 and 15 ) control DNA. Lane 8 corresponds to the DNA ladder.

associates with approximately 12 water molecules.17,18 The associated water molecules surrounding the MPC chains, in addition to the space that MPC occupies, may be creating (17) Ishihara, K.; Nomura, H.; Mihara, T.; Kurita, K.; Iwasaki, Y.; Nakabayashi, N. J. Biomed. Mater. Res. 1998, 39, 323-330. (18) Konno, T.; Kurita, K.; Iwasaki, Y.; Nakabayashi, N.; Ishihara, K. Biomaterials 2001, 22, 1883-1889.

3596

Langmuir, Vol. 21, No. 8, 2005

a steric barrier, which hinders DNA complexation. PEO is another hydrophilic polymer typically used to conjugate with cationic polymers to provide steric stabilization of DNA conplexes. Compared to MPC, PEO blocks appear to be less influential in DNA condensation of DMA-PEO diblock copolymers.9 It may be possible that as the PEO chain is extremely flexible, has smaller monomer units than MPC, and has only two water molecules associated with each residue that it occupies less physical volume than MPC and hence creates less of a steric barrier. To assist our understanding on the MPC effect on the complexation process, morphologies of copolymer-DNA complexes were studied by dry TEM and solution AFM. DNA complexes formed by the DMA homopolymer tend to aggregate in aqueous solution, as judged by TEM and AFM analysis and confirmed by dynamic light scattering studies.12 Introduction of the MPC block provides a steric barrier that prevents aggregation and enables formation of smaller, more colloidally stable complexes. However, depending on their block composition, these DMAxMPC30 copolymers produce complexes exhibiting a wide range of morphologies. In general, both imaging techniques indicate that more efficient DNA condensation occurs as the DMA block length is increased. TEM studies revealed striking morphological differences for DNA complexes prepared with different block copolymers. The shortest DMA block length (i.e., DMA10MPC30) within the DMAxMPC30 copolymer series forms ‘spaghetti-like’ DNA complexes, suggesting incomplete DNA condensation. As the DMA block length is increased, the linear structures become progressively thicker and shorter. This indicates successive folding and twisting of DNA, hence the more efficient DNA condensation. Plectonemic structures comprising long rods and loose rings with a proportion of well-defined rods and toroids were observed for DMA20MPC30 complexes. For DMA40MPC30 complexes, mainly ‘classical’ rods and toroids were observed. The average outer diameter of the toroids observed for the DMA20MPC30 and DMA40MPC30 copolymers were 54.8 ( 12.9 and 51.8 ( 12.2 nm, respectively, which is consistent with the toroidal dimensions of 5055 nm reported by other groups.14,19,20 It is generally agreed that toroids are the most common and stable form of condensed DNA, and it is the most commonly reported morphological structure. Given the range of morphologies of the DNA complexes formed by the DMA20MPC30 and DMA40MPC30 block copolymers, it is worth emphasizing that the frequent co-existence of rods and toroids is well documented,14,15 but the relationship between rods and toroids is not fully understood. It has been claimed that toroids and rods are formed by different condensation pathways,21 but subsequent evidence suggests that rods and toroids are actually interchangeable.4 It has been postulated that DNA rods may bend around22 or open up to form toroids20,23 or that toroids collapse to give rods.24 In our study, both DMA20MPC30 and DMA40MPC30 are seen to form rod and toroid structures. However, the relative dimensions of the rods and toroids for each polymer are different, suggesting that the polymer (19) Hud, N. V.; Downing, K. H.; Balhorn, R. Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 3581-3585. (20) Golan, R.; Pietrasanta, L. I.; Hsieh, W.; Hansma, H. G. Biochemistry 1999, 38, 14069-14076. (21) Bloomfield, V. A. Biopolymers 1998, 53, 329-341. (22) Dunlap, D. D.; Maggi, A.; Soria, M. R.; Monaco, L. Nucleic Acids Res. 1997, 25, 3095-3101. (23) Arscott, P. G.; Bloomfield, V. A. Ultramicroscopy 1990, 33, 127131. (24) Erbacher, P.; Zou, S.; Bettinger, T.; Steffan, A. M.; Remy, J. S. Pharm. Res. 1998, 15, 1332-1339.

Chim et al.

structure influences the process of condensation formation and structures formed. For example, the rods and toroids formed by the DMA20MPC30 copolymer have similar widths (Table 1) and contour lengths (141.9 ( 117 and 155.2 ( 37.1 nm, respectively), while for the DMA40MPC30 copolymer, the contour length of the rods is approximately half that of the toroids (70.8 ( 19 and 136.7 ( 29.5 nm, respectively) while both structures have very similar thickness. We cannot from these data with certainty assess whether rods and toroids are indeed interchangeable structures, progress from each other, or indeed are separate final points of the complexation process. Interestingly, some apparently transitional structures partly resembling toroids and rods were also observed for DNA complexes prepared using DMA20MPC30 and DMA40MPC30 block copolymers by both imaging techniques (Figure 8). The DMA20MPC30 data are clearly consistent with the possibility of a 1:1 relationship between the rods and toroids, whereas the DMA40MPC30 data does not suggest such a clear interdependence. These observations are consistent with our previous physicochemical study12 demonstrating a higher level of DNA condensation with DMA40MPC30 rather than DMA20MPC30. Therefore, for the DMA20MPC30, more flexible condensates are possible and, hence, a tendency to structural interchangeability. Whereas for the higher levels of DNA condensation seen with DMA40MPC30,12 it might be expected that rods will be further twisted and folded into the shorter structures observed. Although toroids have been reported to be the most common morphology for DNA complexes,14 rods clearly predominate for both the DMA20MPC30 and DMA40MPC30 copolymers (Table 2). Other vectors such as chitosan also produce a high proportion of rods.25 Theoretical studies indicate that chain stiffness of macromolecules plays an important role in coil-globule transitions. This can be applied to the structural behavior of DNA condensates.26,27 A toroidal morphology is more likely for a stiff chain, whereas rods are favored when the chain is flexible. Our results therefore suggest that the chains of DNA condensates are relatively flexible due to DNA complexation with the DMA-MPC copolymers, allowing coexistence of toroids and rods to occur. As the mean DMA block length was further increased (DMA60MPC30), the copolymerDNA complexes appeared to be highly condensed since compact, oval-shaped particles or thick, short rods were observed instead of toroids. These progressive changes indicate that a higher degree of DNA condensation is achieved, which is in agreement with gel electrophoresis study. The dramatic morphological changes within this series of copolymer-DNA complexes can be correlated to the relative proportion of MPC residues in the copolymer. Although the MPC block length was kept constant at 30 monomeric units, the proportion of MPC within the system varied. From the gel electrophoresis study, it was clear that both low- and high-molecular-weight DMA homopolymers (with mean degrees of polymerization of 27 and 81, respectively) were able to condense DNA efficiently at a 2:1 monomer/nucleotide molar ratio. Similarly, both TEM and AFM images (Figure 3) confirm condensation and aggregation of DNA-DMA homopolymer complexes. Thus, we suggest that differences in complex morphologies are not solely due to variation of the DMA block length (25) Danielsen, S.; Vårum, K. M.; Stokke, B. T. Biomacromolecules 2004, 5, 928-936. (26) Stevens, M. J. Biophys. J. 2001, 80, 130-139. (27) Noguchi, H.; Yoshikawa, K. J. Chem. Phys. 1998, 109, 50705077.

Structural Study of DNA Condensation

Langmuir, Vol. 21, No. 8, 2005 3597

but are also due to the relative proportions of MPC present in the copolymer systems. It has been demonstrated previously that MPC acted as a steric barrier to prevent the aggregation of copolymer-DNA complexes.12 According to the present studies, the MPC-containing complexes remained discrete and aggregation was not observed. However, the presence of the bulky, noncondensing, and highly hydrophilic MPC moiety appears to interfere with the condensation process. This was particularly apparent in the AFM images of the DMA20MPC30-DNA and DMA40MPC30-DNA complexes. Similarly, as the DMA block length increased, the relative proportion of MPC is reduced, leading to more efficient condensation and the formation of more compact condensates. This is evident from the studies of the DNA condensates formed by the DMA60MPC30 copolymer and is consistent with the corresponding gel retardation assays. In liquid AFM studies, stable DNA complexes formed using two extreme copolymer compositions, DMA10MPC30 and DMA60MPC30, had plectonomic loops and rectangular blocklike structures, respectively (see Figure 9). Whereby, a mixture of intermediates, including flowerlike structures, partial toroids, and rods with petal-like extensions were observed for DNA complexes prepared with DMA20MPC30 and DMA40MPC30 copolymers. A number of repeated experiments with the two latter copolymers revealed that their DNA complexes are comprised of mixtures of various structures with differing degrees of DNA condensation, the relative population of each structure being clearly dependent on the polymer molecular composition. The effect of varying the DNA concentration was also assessed while maintaining a 2:1 monomer/ nucleotide molar ratio. As shown in Figure 9, there is a gradual increase in the proportion of more highly condensed complexes (partial rods, toroids, and rectangular blocks) as the DNA concentration was increased; a similar phenomenon has previously been described by Liu et al.28 The DNA condensates observed in the presence of copolymers with longer DMA block lengths (DMA40MPC30 and DMA60MPC30) contained areas of high density (nucleation centers), which are believed to indicate highly condensed and/or possibly aggregated DNA. Although the general structures of these complexes did not change during AFM observation, an increase in the degree of nucleation and aggregation of preformed complexes was observed over incubation periods of 30-120 min (data not shown). The novel rectangular blocklike complexes observed for the DMA60MPC30 copolymers may be an intermediate state between the flowerlike structures and the rods, rather than a precursor structure. There is currently no literature evidence to suggest that the flowerlike structures and rodlike structures arise from the same, or indeed different, condensation pathways. The rectangular blocks may be a variant of the rodlike structures induced by the steric stabilization imparted by the MPC blocks. It is not clear at this stage whether these rectangular blocklike structures are useful condensates for intracellular DNA delivery. When comparing the dehydrated (TEM) and hydrated (AFM) images, the respective structures obtained from the extreme copolymer compositions in the series, DMA10MPC30 and DMA60MPC30, were similar. The DNA is barely condensed in the DMA10MPC30-DNA complexes and is tightly condensed in the DMA60MPC30-DNA complexes; the structures were perhaps less affected by the differing sample preparation protocols and environments applied

by these two imaging techniques. However, there are certainly significant structural differences between the TEM and AFM data for the intermediate polymers. This may be simply because dried samples were studied under TEM and hydrated/liquid samples under AFM. The MPC chains are highly hydrophilic and bind approximately 1012 molecules of water per MPC unit;17,18 hence, dehydration of the complexes during drying would be expected to have dramatic impact on the morphology of MPC chain and consequently the complex morphology. Furthermore, heavy metal staining is necessary to enable visualization of the complexes under the TEM. In this study, a 50% ethanolic solution of uranyl acetate was used as a stain for the DNA and the image produced is actually the shadow of the stain. However, it has been shown that the presence of ethanol can affect the extent of DNA condensation.29 Thus, the TEM images need to be interpreted with caution. In contrast, neither drying nor staining is necessary for AFM studies, and one can assume that this less invasive sample preparation significantly decreases the probability of artifacts. However, the complexes still require adsorption onto a substrate (and hence confined to a twodimensional plane), and those complexes (or parts thereof) that do not adhere cannot be imaged. Thus, repeated AFM experiments under the same conditions may reveal different structures adsorbed onto the mica surface. This was particularly true for condensates formed in the presence of the DMA20MPC30 and DMA40MPC30 copolymers, for which a range of variably condensed DNA complexes apparently co-exist in solution. It is not yet clear whether the complexes observed by AFM are truly representative of the entire population present in the bulk solution. The co-existence of various copolymer-DNA complex morphologies with each DMAxMPC30 copolymer makes it difficult to determine the order in which these structures are formed. However, we suggest on the basis of the TEM and AFM data that there are four major categories of DNA condensation of different stable/ metastable structures observed for these DMAxMPC30-DNA complexes. These are (the order here does not indicate an order in terms of structural transitions) (i) the formation of loops, (ii) the twisting and initiation of DNA condensation at different points/regions of the DNA backbone, (iii) formation of various structures such as flowers, rods and toroids, and (iv) the compaction of condensed regions leading to structures with an increase in high-density center regions. It is essential for DNA condensation within these complexes to be strong enough to allow protection against biological enzymes but yet be sufficiently reversible to release the DNA into the target cell nucleus for transcription of the therapeutic gene. The enzyme degradation assay shows that all DNA complexes can be dissociated in the presence of a competitor polyelectrolyte, poly(Laspartic acid), to release the DNA. However, a comparison of structural and electrophoretic data in this study with the complexes resistance to the enzymatic degradation shows that only the DMA homopolymers provided effective DNA protection against enzymatic degradation, although some low-molecular-weight DNA fragments were also observed in the gel for DMA40MPC30 and DMA60MPC30 complexes. No fluorescence could be detected for DMA10MPC30 and DMA20MPC30 complexes and naked DNA, suggesting complete DNA degradation. For DMA10MPC30 and DMA20MPC30, the gel electrophoresis study, as well as the morphology studies using TEM and AFM, all indicate low levels of DNA condensation, and the absence

(28) Liu, D.; Wang, C.; Lin, Z.; Li, J.; Xu, B.; Wei, Z.; Wang, Z.; Bai, C. Surf. Interface Anal. 2001, 32, 15-19.

(29) Fang, Y.; Spisz, T. S.; Hoh, J. H. Nucleic Acids Res. 1999, 27, 1943-1949.

3598

Langmuir, Vol. 21, No. 8, 2005

Chim et al.

of highly condensed regions within these complexes renders the DNA exposed to the surrounding environment and susceptible to enzymatic degradation. For DMA40MPC30 and DMA60MPC30 complexes, however, gel electrophoretic and TEM data both indicate high levels of DNA condensation, as no fluorescence probe inclusion (gel electophoresis study) or loosely condensed regions (TEM images) could be observed. One would, hence, expect that these complexes would provide an effective protection of DNA against enzymatic degradation. AFM images, however, show that these complexes contain high-density regions and loose DNA around their perimeter in the form of petals. We thus propose that the exposed ‘petals’ regions of DNA seen under the AFM are the initiation sites of enzymatic attack by DNase I.30 This is probably affecting the morphology of the complexes and further exposing DNA to the enzyme, while only the highly condensed area resisted the attack. As a result, partly degraded DNA fragments are seen in the gels. Nevertheless, the protection offered by these two copolymers were insufficient, as the structures of DNA were not preserved, possibly leading to the loss of the genes’ therapeutic information.

MPC30 copolymer series is highly dependent on the block composition of the synthetic vector. In general, copolymers with longer DMA blocks produced more condensed structures, whereas systems with relatively longer MPC components appeared to hinder electrostatic complexation between the DNA and the copolymer. For a given DMAMPC copolymer, the morphology of the dehydrated DNA complex observed by TEM was significantly different to the in situ AFM images obtained in the aqueous environment. These differences are most likely attributable to the highly hydrophilic nature of the MPC block. Dehydration of this stabilizing block may have a significant impact on the resulting morphology, thus explaining the discrepancy between liquid AFM and dried TEM images. Although the DMA block does bind DNA to some degree, the highly hydrophilic nature of the MPC block appears to sterically hinder the DNA condensation process, thus explaining why no classical tight condensates such as toroids and rods were visualized under liquid AFM. Furthermore, the loosely condensed petal-like DNA structures observed under liquid AFM maybe the origin of enzyme degradation susceptibility.

Conclusions This study demonstrates that the structure and morphology of plasmid DNA complexes formed by the DMAx-

Acknowledgment. Y.T.A.C. would like to thank the BBSRC for funding. J.L. would like to thank Biocompatibles, UK and The University of Nottingham, UK for funding.

(30) Abdelhady, H. G.; Allen, S.; Davies, M. C.; Roberts, C. J.; Tendler, S. J. B.; Williams, P. M. Nucleic Acids Res. 2003, 31, 4001-4006.

LA047480I