Columnar Mesophases of the Complexes of DNA with Low

Biomacromolecules 2009, 10, 773–783

773

Columnar Mesophases of the Complexes of DNA with Low-Generation Poly(amido amine) Dendrimers Chun-Jen Su,†,‡ Hsin-Lung Chen,*,† Ming-Chen Wei,† Shu-Fen Peng,† Hsing-Wen Sung,† and Viktor A. Ivanov§ Department of Chemical Engineering, National Tsing Hua University, Hsin-Chu 30013, Taiwan, and Physics Department, Moscow State University, Moscow 117234, Russia Received October 29, 2008; Revised Manuscript Received January 5, 2009

The electrostatic complexes of polyanionic DNA with dendrimers have been considered as a class of nonviral vector for gene therapy. The gene transfection efficiency has been believed to be influenced by the structure of the complex. In this study, we have systematically characterized the supramolecular structures of the complexes of DNA duplexes with poly(amido amine) (PAMAM) dendrimers with generation two (G2) and three (G3) in pure water using small-angle X-ray scattering. The structures were examined as a function of the charge density of the dendrimer expressed by its degree of protonation (dp) and the molar ratio of the amine groups of dendrimer to the phosphate groups of DNA (N/P). The DNA chains in all complexes under study were found to selforganize into two-dimensional hexagonal or square lattice. In general, hexagonal phase was the favorable structure for G2 complexes, while the DNA in G3 complexes tended to organize into a square lattice. Interesting transitions between the columnar mesophases with respect to the changes of N/P ratio and dp have been identified. The geometric features of the dendrimer molecules accommodated within the interstitial tunnels of the DNA lattices have also been revealed. The B conformation of DNA was effectively retained in the complexes in spite of the influence of the electrostatic interaction with the dendrimers.

Introduction The electrostatic complexes of polyanionic DNA with various cationic agents, including lipids, macrocations, polyelectrolytes, and amphiphilic block copolymers, have received much attention in recent years due to the effort in developing nonviral vectors for gene therapy.1 The complexation is driven mainly by the electrostatic attraction between DNA and the cationic species coupled with the entropic gain from counterion release and it usually results in significant aggregation of DNA chains, leading to the formation of submicrometer-sized particles.2,3 Two levels of the structure can hence be defined for the complexes: (1) the “colloidal level” characterized by the topological feature (e.g., the shape and size) and the surface charge of the particles at the length scale of several hundred nm or above; (2) the “supramolecular level” characterized by the organization of DNA chains and the cationic agent within the particles at the characteristic length scale of several nm. It is believed that the gene transfection efficiency is influenced by the structure of the complex;4,5 consequently, resolving the structures at both levels and the strategy for tuning them by various parameters such as charge ratio, ionic strength, pH, and temperature have been regarded as an important fundamental task for the realization of effective nonviral gene vector. The present study concerns the supramolecular structure of the complexes of DNA with cationic dendrimers in pure water. Dendrimer is a class of hyperbranched macromolecule composing of layers of monomer units irradiating from a central core. 6,7 Each complete grafting cycle is called a “generation” * To whom correspondence should be addressed. E-mail: [email protected]. edu.tw. † National Tsing Hua University. ‡ Present address: National Synchrotron Radiation Research Center, HsinChu, Taiwan. § Moscow State University.

(denoted by Gn with n being the generation number). The dendrimers possessing amine groups at the surface and/or the interior can be protonated to controlled level under acidic aqueous environment. The macrocations thus formed have been considered as the carriers for macromolecular drug and gene delivery. In recent years, the high-generation dendrimers have been studied for pH-controlled drug delivery systems8 and as gene delivery agents.9-14 Lower-generation dendrimers have been investigated as a delivery system for plasmid DNA15,16 due to low cell cytotoxicity. Due to the similarity in geometric feature between dendrimers and histone protein,17-19 it was widely postulated that DNA chain should coil around the cationic dendrimer upon complexation, leading to the “beads-on-string” type of supramolecular structure, as that found for DNA/histone complex constituting the nucleosome particle.20 Recent simulation21,22 works have revealed that the ability of a DNA chain to wrap around a spherical macrocation and the resultant wrapping mode depended on the balance among the electrostatic attraction, energy cost of bending the DNA chain and the intrachain repulsion of DNA. Tight wrapping around the macrocation enhances the charge matching but increases the bending free energy; therefore, the beads-on-string configuration may not be accessible for the complexes with dendrimers of relatively low generations, because the weaker electrostatic attraction (due to smaller charge density of the dendrimer) may be outweighed by the high bending energy cost for wrapping around the dendrimer molecules with much smaller size than the persistence length of DNA (ca. 50 nm in the absence of salt). Alternatively, the complexation merely induces aggregation of the relatively straightened DNA chains bridged by the dendrimer. The DNA chains within the aggregates may attain certain degree of orientational and positional order to reduce their excluded volume interaction and to regulate the charge matching with

10.1021/bm801246e CCC: $40.75  2009 American Chemical Society Published on Web 03/18/2009

774

Biomacromolecules, Vol. 10, No. 4, 2009

the dendrimer. The complexation hence yields the “columnar mesophase” characterized by the spatial order of DNA. The formation of this type of mesophase has indeed been disclosed recently, where the DNA chains in the complexes with poly(amido amine) (PAMAM)23 and polypropylene(imine) (PPI)24 dendrimers of intermediate generations (G4 and G5) were found to organize into columnar mesophases with in-plane square or hexagonal symmetries (referred as the “square columnar phase” and the “hexagonal columnar phase”, respectively). In this study, we systematically investigate the supramolecular structures of the complexes of DNA with PAMAM dendrimers of low generations (G2 and G3) using small-angle X-ray scattering (SAXS). The structure will be examined as a function of the charge density of the dendrimer (expressed by its nominal degree of protonation (dp)) and the molar ratio of the amine groups of dendrimer to the phosphate groups of DNA (N/P), where the effect of the former has not been studied in detail before. It will be shown that all complexes under study selforganize into columnar mesophases, including square columnar phase and hexagonal columnar phase, depending on the generation number, N/P and dp. Detailed morphological maps will be constructed for identifying the structure associated with a given combination of N/P and dp. Attempt will also be made to elucidate the organization of the dendrimer molecules situating in the interstitial tunnels of DNA in the columnar mesophases.

Experimental Section Materials. Linear DNA type XIV from herring testes sodium salt (Na content 6.2%) was purchased from Sigma and used without further purification. Its molecular weight determined by gel electrophoresis was found to have a polydispersity value between 400 and 1000 base pairs (bp) with a center of distribution at about 700 bp.25 Ethylenediamine (EDA) core poly(amidoamine) dendrimers (generations 2 and 3) were obtained from Dendritic Nanotechnologies Inc. (DNT) as methanol solutions. After thorough drying, the solids were weighed and then redissolved in distilled water to produce stock solutions of 0.1% (w/w).The solutions were stored at 4 °C until use. Complex Preparations. To complex with the polyanionic DNA, the amine groups in PAMAM dendrimer were first protonated by adding prescribed amount of 0.1 N HCl solution. Therefore, the degrees of protonation (dp) of PAMAM G2 and G3 dendrimer were controlled by the amount of HCl added. This dendrimer solution was mixed with the aqueous solution containing prescribed amount of DNA to obtain the complex. The concentration of DNA aqueous solution was 2 mg/ mL. The complexaton was usually manifested by visually observable precipitation. Small Angle X-ray Scattering (SAXS) Measurements. The selfassembled structures of the complexes in the fully hydrated state were probed SAXS at room temperature. The aqueous suspensions of the complexes were directly introduced into the sample cell comprising of two ultralene windows. SAXS measurements were performed using a Bruker NanoSTAR SAXS instrument, which consisted of a Kristalloflex K760 1.5 kW X-ray generator (operated at 40 kV and 35 mA), crosscoupled Go¨bel mirrors for Cu KR radiation (λ ) 1.54 Å), resulting in a parallel beam of about 0.05 mm2 in cross section at the sample position and a Siemens multiwire type area detector with 1024 × 1024 resolution mode. All data were corrected by the empty beam scattering and the sensitivity of each pixel of the area detector. The area scattering pattern has been circularly averaged to increase the efficiency of data collection. The intensity profile was output as the plot of the scattering intensity (I) versus the scattering vector, q ) 4π/λ sin (θ/2) (θ ) scattering angle). The exposure time of the X-ray measurements was 3 h, and the samples remained stable after the measurements because of the relatively weak intensity of the X-ray source.

Su et al.

Results and Discussion 1. DNA-PAMAM G2 Complex. PAMAM G2 dendrimer carries 30 amine groups and its diameter is around 2.5 nm. The composition of the complex denoted by N/P ratio expresses the molar ratio of the amine groups (irrespective whether they are protonated) of the dendrimer to the anionic phosphate groups of DNA prescribed by the feed ratio of the two components in the complex preparation. The actual composition of the complex, denoted by (N/P)a, may deviate from this prescribed composition. The charge density of the dendrimer was prescribed by its nominal degree of protonation (dp), which represents the number fraction of the protonated amine groups of the dendrimer molecule calculated from the amount of HCl added. The primary amine groups at the outer surface of the dendrimer tended to be protonated first because their basicity (pKa = 9.0) was higher than that of the interior tertiary amines (pKa = 5.8);26 in principle, dp ) 0.5 corresponded to full protonation of the surface amine groups. SAXS was employed here to probe the supramolecular structure of the complexes in pure water. The SAXS intensity of the complex can be factorized into three partial structure factors, namely,27 2 I(q) ) ∆FDNA SDD(q) + 2∆FDNA∆FdenSDd(q) + ∆F2denSdd(q) (1)

where SDD(q), Sdd(q), and SDd(q) are the partial structure factors associated with DNA-DNA, dendrimer-dendrimer, and DNA-dendrimer correlations, respectively; ∆FDNA and ∆Fden are the scattering length density (SLD) contrast of DNA and dendrimer relative to the water medium, respectively. The X-ray SLDs of DNA, dendrimer, and water are 15 × 1010, 11.3 × 1010, and 9.3 × 1010 cm-2, respectively, ∆FDNA is, hence, 2.85 times ∆Fden. ∆Fden was further reduced by the significant penetration of water into G2 dendrimer with opened structure.28 As a result, the SAXS profile of the complex was dominated by SDD(q), and the scattering curve yields mainly the information about the spatial organization of DNA chains within the complex particles. a. Effect of N/P Ratio. Figure 1a shows the room temperature SAXS profiles of DNA-G2 (dp ) 0.5) complexes with different N/P ratios. At N/P g 1.5 a primary scattering peak was observed at about 2.0 nm-1 along with a small higher-order peak situating at about 3.5 nm-1. The ratio of the positions of these two peaks was about 1:31/2, which indicated that DNA in the complex packed into a hexagonal lattice. This hexagonal columnar phase was denoted as the “H phase”. The solid curve superposing on the SAXS profile of the complex with N/P ) 20 represents the fit using the paracrystalline model of hexagonally packed cylinders considering the polydispersities of the cylinder radius and length, the domain (grain) size of the lattice, and the lattice distortion by thermal fluctuations.29,30 The primary scattering peak underwent a large shift to higher q when N/P was decreased from 1.5 to 1.0, signaling a significant drop of the interhelical distance (i.e., the nearest distance between the DNA chains, dDNA) in the complex. The peak at N/P ) 1.0 was also broader and the higher-order peak was not discernible. In this case, the paracrystalline model still yielded satisfactory fit to the observed scattering profile (cf. the solid curve superposing on the profile); however, the internal order of the lattice became poorer as evidenced by the larger value of g factor () 0.094) compared with that (g ) 0.083) associated with the complex with N/P ) 20, where g factor

Columnar Mesophases of the Complexes of DNA


775

Figure 1. (a) Room temperature SAXS profiles of the DNA-PAMAM G2 dendrimer complexes with different N/P ratios. The dp of the dendrimer was 0.5. The solid curves superposing on the scattering profiles of the complexes with N/P ) 1.0 and 20 represent the fits using the paracrystalline model of hexagonally packed cylinders considering the polydispersities of the cylinder radius and length, the domain (grain) size of the lattice, and the lattice distortion by thermal fluctuations. (b) Variation of dDNA with N/P.

Figure 2. Schematic illustrations of the organization of dendrimer molecules in the 3-fold interstitial tunnels of the hexagonal lattice formed by DNA: (a) the top view; (b) the side view. Dcc represents the center-to-center distance between dendrimer molecules along the DNA long axis (i.e., z axis).

represents the ratio of the mean displacement of lattice points to dDNA.29,30 For the convenience of discussion, we denote this hexagonal phase without showing clear higher-order diffraction peak as the “distorted hexagonal columnar phase” (DH). Figure 1b shows dDNA as a function of N/P ratio. The interhelical distance was obtained by multiplying the Bragg’s spacing (2π/qm, with qm being the position of the primary scattering peak) by 2/3. It can be seen that dDNA increased

progressively with increasing N/P, but it started to level off at N/P ≈ 1.5. This implied that the actual composition of the complex, that is, (N/P)a, reached a saturated value of (N/P)a ≈ 1.5 when the prescribed N/P ratio exceeded 1.5. In other words, nearly all DNA and dendrimer added for the complex preparation were consumed for complexation at the prescribed N/P ratio of 1.5, and further increase of N/P did not add more dendrimer into the complex. This was verified by the very low concentrations of free DNA and free dendrimer remained in the supernatant for the complex prepared with the prescribed N/P of 1.5 (see Supporting Information). The composition saturation was accompanied with a structure transformation from DH phase to H phase. At low N/P ratio where the supply of dendrimer was limited, all the dendrimer macrocations were consumed for complexation with DNA. The amount of dendrimer in the complex was however not sufficient to “glue” the position- and orientation-fluctuating DNA chains very cohesively, such that the hexagonal lattice underwent a greater distortion by thermal fluctuations. Moreover, the G2 dendrimer molecules contained in the interstitial regions between DNA tended to elongate along the longitudinal direction of DNA to enhance the charge matching with DNA,31 and such an elongational deformation led to a decrease of dDNA in the DH phase (to be discussed in further detail later). With the increase of N/P ratio the amount of dendrimer accommodated within the complex also increased, but it saturated as the prescribed N/P reached 1.5, where further increase of N/P did not add more dendrimer into the complex. In this case, the complexes with (N/P)a ≈ 1.5 were formed. Here the amount of dendrimer accommodated in the complex was sufficient to glue the DNA chains cohesively so as to yield a more ordered hexagonal lattice. The saturated composition corresponded well to the “isoelectric point” measured by the zeta potential, where the surface charge of the complex particles was fond to inverse from negative to positive at N/P ≈ 1.5. Knowing the actual composition, that is, (N/P)a ≈ 1.5, for the complexes showing H phase, it is possible to elucidate the organization and the geometry of the dendrimer molecules situating in the interstitial tunnels of the hexagonal lattice. The number of phosphate groups of DNA allocated for one den-

776


Su et al.

drimer molecule in the complex is given by

nph )

30 amine groups in a G2 molecule ) 20 (N/P)a

(2)

Therefore, each dendrimer molecule was able to cover the 20 phosphate groups associated with a pitch of the DNA duplex. Under this condition and the fact that the DNA chains arranged in a hexagonal lattice, the accommodation of the dendrimer molecules in the 3-fold tunnels, as depicted in Figure 2, was the most plausible arrangement for the dendrimer. This model postulates that each dendrimer molecule was in contact with three different DNA chains each allocating about 20/3 ≈ 7 phosphate groups for the dendrimer molecule. These 20 phosphate groups were however not all bound directly with the dendrimer considering the mismatch between the surface curvature of DNA and that of the dendrimer (even though the dendrimer may elongate along the DNA chain axis) as well as

the mistmatch between the spatial distribution of the amine groups on the dendrimer and that of the spirally arranged phosphate groups around the DNA duplex. Moreover, the average center-to-center distance, Dcc, of the dendrimers along the DNA longitudinal direction (i.e., z axis in Figure 2) should correspond approximately to the pitch length of DNA duplex, that is, 3.4 nm.32 Similar structure has been identified for the complex of Actin, a rodlike anionic polyelectrolyte, with lysozyme, a cationic globular protein with ellipsoidal geometry,33 where the electrostatic attraction played an important role for the structural organization. In that system, lysozymes were found to pack closely along the long axis of Actin, generating a scattering peak characterizing the longitudinal spacing between the lysozymes situating in the 3-fold tunnels of the hexagonal lattice of Actin.33 Such a dendrimer-dendrimer correlation peak was, however, not observable for the DNA-PAMAM G2 complexes, because the significant penetration of water molecules into the dendrimers diminished the electron density

Figure 3. (a) Room-temperature SAXS profiles of the DNA-PAMAM G2 (dp ) 0.3) dendrimer complexes as a function of N/P. (b) Variation of dDNA with N/P.

Figure 4. (a) Room-temperature SAXS profiles of the DNA-PAMAM G2 (dp ) 0) dendrimer complexes as a function of N/P. (b) Variation of dDNA with N/P.



777

Figure 3a shows the SAXS profiles of DNA-G2 (dp ) 0.3) complexes prepared with different N/P ratios. At N/P g 1.6 the presence of the two lattice peaks with the relative position of 1:31/2 showed the formation of H phase. The primary peak shifted to higher q and the 31/2 peak vanished at N/P ) 1.0, signaling a significant distortion of the hexagonal lattice. The corresponding variation of dDNA with N/P displayed in Figure 3b revealed an abrupt increase followed by leveling off at N/P ≈ 1.7. According to our foregoing discussion, this composition was the saturated composition or (N/P)a of the complex showing H phase. Assuming again the 3-fold arrangement of dendrimer in the complex, the corresponding Dcc was 3.0 nm calculated according to the following formula

Figure 5. Schmatic illustration of the organization of dendrimer in H′ phase.

contrast between dendrimer and water medium.28 This was corroborated by the small angle neutron scattering (SANS) result, where the SANS profiles of the G2 and G3 complexes did not show discernible dendrimer-dendrimer correlation peak although the SANS SLD contrast of the dendrimer (assuming complete free of water) was 1.85 times that of DNA (see Supporting Information). Now we proceed to analyze the geometry of dendrimer molecules accommodated in the hexagonal lattice. We assumed that the pervaded volume of a dendrimer molecule was represented by an ellipsoid with the half-lengths of the two axes being a and c (with c being the rotating axis), as illustrated in Figure 2a. Considering the 3-fold arrangement of dendrimer molecules, dDNA in the hexagonal lattice is given by dDNAH ) 31/2(a + RDNA) with RDNA () 1 nm) being the radius of DNA. The value of c can then be calculated via c ) (3Vden)/(4πa2) knowing the value of a and the volume of a PAMAM G2 molecule Vden ≈ 8.18 nm3.34 The values of a and c thus obtained were about 1.15 and 1.48 nm, respectively, which prescribed the aspect ratio of c/a ) 1.29. This supported our foregoing postulate that the dendrimer molecules were elongated into prolates along z axis to enhance the adhesion on the relatively planar surface of DNA chain.

nph )

Dcc × 3.4 nm per pitch in the DNA duplex (20 phosphate groups per pitch) (3)

where nph ) 30/(N/P)a ) 17.7. Figure 4 presents the SAXS profiles and the N/P variation of dDNA for the system with dp ) 0. It is interesting to note that although no HCl was added to protonate G2 dendrimer here, the complexation still took place spontaneously. A recent small angle neutron scattering study by Chen et al. revealed that PAMAM dendrimer only carried an extremely small amount of positive charge in pure water (0.29 charge per molecule).35 Therefore, the hydrogen bonding between the primary amine groups of PAMAM and the phosphate groups of DNA was considered as the main binding force responsible for the complexation.36,37 According to the SAXS profiles, the boundary between H phase and DH phase was located at N/P ≈ 2.0 for this series of complex; however, in contrast to the systems with nonzero dp where the saturated composition defined the boundary between these two structures, the corresponding dDNA was found to increase rather smoothly with increasing N/P ratio until it leveled off at N/P ≈ 6. We hypothesized that this N/P ratio corresponded to the saturated composition; in this case, the amount of dendrimer accommodated in the hexagonal lattice was four times that in the complex prepared with dp ) 0.5. If the dendrimer molecules still arranged in the 3-fold tunnels,

Figure 6. (a) Room temperature SAXS profiles of the DNA-PAMAM G2 dendrimer complexes with N/P ) 6 as a function of dp. (b) Variation of dDNA with dp.

778


Su et al.

Table 1. Geometric Parameters of G2 Dendrimer Molecules Accommodated in the Interstitial Tunnels of the Hexagonal Lattice as a Function of dpa dp

a (nm)

c (nm)

aspect ratio (c/a)

0 0.01 0.1 0.2 0.3 0.4 0.5 0.6

1.48 1.46 1.35 1.33 1.26 1.22 1.13 0.98

0.90 0.92 1.08 1.11 1.23 1.32 1.53 2.05

0.61 0.63 0.8 0.82 0.98 1.08 1.35 2.09

a

N/P ratio of the complexes were fixed at 6.0.

Figure 8. Morphologic map of DNA-PAMAM G2 complexes in terms of N/P and dp. The red dashed line shows the boundary between H′, H, and DH phase.

Figure 7. Variation of the dDNA as a function of dp for the complexes with N/P ) 2, 6, 14, and 40.

then the corresponding Dcc was only 0.9 nm, which was smaller than the diameter of a G2 molecule. To avoid the overlap of the dendrimer, we proposed that a given layer of dendrimer molecules (defined by the triangle formed by connecting molecules 1, 2, and 3 in Figure 5) rotated around z axis by 60° relative to the adjacent layer (defined by the triangle formed by molecules 1′, 2′, and 3′), such that molecule 1′ could be fitted into the valley between molecules 1 and 2 and so forth. This packing mode would prescribe a smaller Dcc than the dendrimer diameter while preventing overlap of the hard-core volume of the dendrimer. A large amount of dendrimer molecules could hence be accommodated in the complex once the weak repulsion between the nearly uncharged dendrimers was outweighed by their attractive interaction with DNA. This hexagonal structure was denoted as the “H′ phase”. b. Effect of dp. We further examined the effect of dp (under a fixed N/P ratio) on the supramolecular structure of the complex. Figure 6a displays the room-temperature SAXS profiles of the complexes with PAMAM G2 dendrimer with different dp, where the N/P ratio was fixed at 6.0. The 31/2 peak was observable up to dp ) 0.53, showing the formation of H phase. The primary scattering peak became broad and the higher-order peak was invisible at dp ) 0.7, indicating a significant distortion of the hexagonal lattice. The corresponding variation of dDNA with dp is displayed in Figure 6b. dDNA was seen to decrease with the increase of dp until dp reached 0.8. Table 1 lists the geometric parameters of the dendrimer molecules calculated from the vales of dDNA for the complexes forming H phase. It should be noted that here we have adopted

a fixed value of Vden (≈ 8.18 nm3) for the calculation. This appears to be a plausible assumption considering that recent SANS studies have revealed that the radii of gyration of PAMAM dendrimers were hardly affected by pH irrespective of the generation number.35,38 It can be seen from Table 1 that the aspect ratio of the G2 dendrimer was smaller than 1.0 at dp 1.4 and N/P < 1.2, respectively, while these two structures coexisted between these two N/P ratios. The variation of dDNA with N/P displayed in Figure 12b revealed that the complex composition saturated at N/P ≈ 1.8. However, in contrast to the G3 complex with dp ) 0.5 and the G2 complexes with nonzero dp, where the saturated composition defined the boundary between the ordered and distorted columnar phase, the present system still largely retained S phase at the NP ratios lying between 1.8 and 1.4, despite that the dDNA in this composition range was apparently smaller than that associated with the saturated composition. This implied that the DNA chains already arranged in a highly ordered square lattice although the amount of G3 dendrimer in the square lattice had not reached saturation. Thus, two types of S phase could be distinguished here, namely, “Sa phase” for the complex with the saturated composition and “Sus phase” formed by the complex with unsaturated composition. Both types of S phases were observable over the dp range of 0.1∼0.6. Figure 13a shows the SAXS profiles of the complex with unprotonated G3 dendrimer (i.e., dp ) 0). In contrast to the foregoing G3 complexes showing Sa, Sus or DS phase, the packing of the DNA chains in the present system was found to transform from square to hexagonal lattice near N/P ) 5.5. Figure 13b plots the corresponding dDNA as a function of N/P. Two abrupt transitions of dDNA associated with the structure transformation from Sus to Sa phase and from Sa phase to H phase were identified. In a previous study of the complexes of DNA with fully protonated PPI G4 dendrimer, Evans et al. observed a transition from S phase to H phase with the increase of N/P.24 They proposed that the lattice structure was governed by the competition between the long-range electrostatic cohesion that favored the square packing and the short-range electrostatic adhesion by counterion release that favored the hexagonal structure. Nevertheless, because the charge density of the unprotonated dendrimer was extremely low, these two effects were not expected to be operative here. Alternatively, we proposed that the transformation from Sa to H phase was driven by the tendency of the system to accommodate more dendrimer



781

Figure 12. (a) Room temperature SAXS profiles of the DNA-PAMAM G3 (dp )0.4) dendrimer complexes with different N/P ratios. (b) Variation of dDNA with N/P.

Figure 13. (a) Room temperature SAXS profiles of the DNA-PAMAM G3 (dp ) 0.0) dendrimer complexes with different N/P ratios. (b) Variation of dDNA with N/P.

molecules due to the favorable hydrogen bonding. The saturated composition of the complex displaying Sa phase was N/P ) 2.7, which prescribed the Dcc of 4.03 nm. This distance was already very close to the diameter of G3 dendrimer () 3.4 nm). To accommodate more dendrimer molecules, the packing of DNA transformed into a hexagonal lattice with the dendrimer molecules placed in the 3-fold tunnels in accordance to the arrangement in H’ phase found for G2 complexes (cf. Figure 5). Because dDNA in the H’ phase started to level off at N/P ) 5.5, this composition was considered as the saturated composition of the H′-forming complexes. In this case, the corresponding Dcc was 1.98 nm. b. Effect of dp. Figure 14a presents the room temperature SAXS profiles of DNA-G3 dendrimer complexes as a function of dp (with N/P fixed at 6). At dp < 0.06, the complexes displayed H′ phase and the structure transformed into S phase at 0.075 < dp < 0.6. S and DS phase coexisted at 0.6 < dp < 0.65, and as dp exceeded 0.65, the observation of only a broad correlation peak indicated that the majority of the DNA chains packed in a distorted square lattice. The SAXS results hence revealed the structure transformation from H to S phase and

finally to DS phase with increasing dp. Figure 14b revealed the corresponding variation of the dDNA with dp. dDNA decreased with increasing dp and showed clear transitions corresponding to the transformations of the mesophases. The geometric parameters of G3 dendrimer in the complexes as a function of dp are listed in Table 2. Similar to G2 system, G3 dendrimer was oblate in shape (aspect ratio < 1) at low dp (dp < 0.068), and with the increase of dp, they were deformed into prolates to enhance the charge matching. The observed variation in the degree of dendrimer deformation was again attributable to the interplay between the electrostatic repulsion between the DNA chains and the electrostatic attraction between dendrimer and DNA. Figure 15 displays the morphological map of DNA-PAMAM G3 complexes in terms of N/P and dp. Four types of columnar phases were distinguished here, namely, H′ phase, saturated square phase (Sa), unsaturated square phase (Sus), and distorted square phase (DS). The boundaries between these structures were drawn by he dashed lines. Our SAXS results have revealed that in the cases where ordered columnar phases were formed, the complexes of DNA

782


Su et al.

Figure 14. (a) Room temperature SAXS profiles of the DNA-PAMAM G3 dendrimer complexes with N/P ) 6 as a function of dp. (b) Variations of dDNA as a function of dp. Table 2. Geometric Parameters of G3 Dendrimer Molecules Accommodated in the Interstitial Tunnels of the Hexagonal or Square Lattice as a Function of dpa structure

dp

a (nm)

c (nm)

aspect ratio (c /a)

H′ H′ H′ H′ H′ S S S S S S S

0 0.01 0.025 0.044 0.068 0.068 0.075 0.1 0.2 0.3 0.4 0.51

1.77 1.74 1.7 1.64 1.63 1.76 1.72 1.69 1.67 1.58 1.5 1.4

1.56 1.63 1.7 1.82 1.86 1.59 1.66 1.71 1.77 1.97 2.17 2.49

0.88 0.94 1 1.11 1.14 0.9 0.97 1.01 1.06 1.25 1.45 1.78

a

N/P ratio of the complexes were fixed at 6.0.

governed by the interaction energies associated with DNA-DNA repulsion, dendrimer-dendrimer repulsion, and the dendrimerDNA attraction interaction. H′ phase, being able to accommodate the largest amount of dendrimer, may maximize the DNA-dendrimer interaction, but the dendrimer molecules within the interstitial tunnels might experience a strong repulsion, such that this structure was only observed at very low dp. S phase was the structure offering the second effective charge matching, while allowing the dendrimer molecules to separate from each other by a certain Dcc (>dendrimer diameter) along the DNA longitudinal direction to alleviate their mutual repulsion. The charge matching was the least effective in H phase, but here the DNA-DNA repulsion may be alleviated due to larger dDNA. In the case of DNA-G2 system, the DNA-DNA repulsion played a significant role because of less binding sites per dendrimer molecules; consequently, DNA chains selforganized into hexagonal packing to alleviate their mutual repulsion. For the DNA-G3 complexes, the higher positive charge density per dendrimer molecule made the DNA-dendrimer attraction the dominant factor; therefore, the DNA chains arranged into square lattice to enhance their charge matching with protonated dendrimer.

Conclusions

Figure 15. Morphologic map of DNA-PAMAM G3 complexes. The blue dash line indicates the boundary of saturated square columnar phase and unsaturated square columnar phase. The red dash line defines the boundary of H′, S, and DS phases.

with protonated G2 dendrimers in general tended to form H phase, while in the complexes with protonated G3 the DNA chains tended to organize into a square lattice We proposed that the type of columnar mesophase formed was mainly

We have systematically investigated the supramolecular structures of the complexes DNA with low-generation PAMAM dendrimers, G2 and G3, in pure water as a function of N/P ratio and dendrimer dp. All complexes were found to self-organize into columnar mesophases with positional and orientational order of DNA. The beads-on-string structure was not observed here due to the high bending energy cost for the DNA chain to wrap around the dendrimers with small size. Six types of mesomorphic structures, namely, H, H′, DH, Sa, Sus, and DS phases, were distinguished based on the SAXS profiles. In general, H phase was the favorable structure for G2 complexes, while G3 complexes tended to form S phase. Distorted columnar phases appeared as the N/P ratio was small or when dp of the dendrimer was large. When dp of the dendrimer approached zero, the complexes formed H′ phase for accommodating a large amount of dendrimer within the hexagonal lattice. The dendrimer molecules were found to undergo different degrees of deformation in the complexes. At low dp, they were basically oblate in


shape to alleviate the DNA-DNA repulsion, while they tended to elongate along the DNA longitudinal direction above moderate dp to enhance the adhesion with the rather planar surface of DNA. Finally, we would like to note that the DNA chains in all the complexes with both G2 and G3 dendrimers adopted B conformation, irrespective of the N/P ratio and dp of the dendrimers. Acknowledgment. This work was supported by a joint grant from the National Science Council, Taiwan and Russian Foundation for Basic Researches (RFBR; NSC 97-2923-E-007001-MY3). Supporting Information Available. Demonstration of nearly complete consumption of DNA and PAMAM dendrimer for complexation at the saturated N/P ratio by UV-vis spectroscopy; a SANS profile of DNA-PAMAM G3 (dp ) 0.06) complex with the prescribed N/P ratio of 6.0. This material is available free of charge via the Internet at http://pubs.acs.org.

References and Notes (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)

(13) (14) (15) (16) (17)

Niidome, T.; Huang, L. Gene Ther. 2002, 9, 1647–1652. Bloomfield, V. A. Biopolymers 1991, 31, 1471–1481. Bloomfield, V. A. Biopolymers 1997, 44, 269–282. Koltover, I.; Salditt, T.; Ra¨dler, J. O.; Safinya, C. R. Science 1998, 281, 78–81. Safinya, C. R. Curr. Opin. Struct. Biol. 2001, 11, 440–448. Tomaila, D. A.; Baker, H.; Dewald, J.; Hall, M.; Kallos, G.; Martin, S.; Roeck, J.; Ryder, J.; Smith, P. Polym. J. 1985, 17, 117–132. Esfand, R.; Tomaila, D. A. Drug DiscoVery Today 2001, 6, 427–436. Pistolis, G.; Malliaris, A.; Tsiourvas, D.; Paleos, C. M. Chem.sEur. J. 1999, 5, 1440–1444. Bielinska, A.; Kukowska-Latallo, J. F.; Johnson, J.; Tomalia, D. A.; Baker, J. R., Jr. Nucl. Acid. Res. 1996, 24, 2176–2182. Tang, M. X.; Redemann, C. T.; Szoka, F. C., Jr. Bioconjugate Chem. 1996, 7, 703–714. Bielinska, A. U.; Chen, C.; Johnson, J.; Baker, J. R., Jr. Bioconjugate Chem. 1999, 10, 843–850. Toth, I.; Sakthivel, T.; Wilderspin, A. F.; Bayele, H.; O’Donnell, M.; Perry, D. J.; Pasi, K. J.; Lee, C. A.; Florence, A. T. STP Pharma Sci. 1999, 9, 93–99. Braun, C. S.; Vetro, J. A.; Tomalia, D. A.; Koe, G. S.; Koe, J. G.; Middaugh, C. R. J. Pharm. Sci. 2005, 94, 423–436. Braun, C. S.; Fisher, M. T.; Tomalia, D. A.; Koe, G. S.; Koe, J. G.; Middaugh, C. R. Biophys. J. 2005, 88, 4146. Zinselmeyer, B. H.; Mackay, S. P.; Schatzlein, A. G.; Uchegbu, I. F. Pharm. Res. 2002, 19, 960–967. Hollins, A. J.; Benboubetra, M.; Omidi, Y.; Zinselmeyer, B. H.; Schatzlein, A. G.; Uchegbu, I. F.; Akhtar, S. Pharm. Res. 2004, 21, 458–466. Prosa, T. J.; Bauer, B. J.; Amis, E. J. Macromolecules 2001, 34, 4897– 4906.


783

(18) Tomalia, D. A.; Hall, M.; Hedstrand, D. M. J. Am. Chem. Soc. 1987, 109, 1601–1603. (19) Bielinska, A. U.; Chen, C.; Johnson, J.; Baker, J. R., Jr. Bioconjugate Chem. 1999, 10, 843–850. (20) Stryer, L. Biochemistry; W.H. Freeman and Company: San Francisco, 1981. (21) Kunze, K. K.; Netz, R. R. Phys. ReV. Lett. 2000, 85, 4389–4392. (22) Kunze, K. K.; Netz, R. R. Phys. ReV. E: Stat., Nonlinear, Soft Matter Phys. 2002, 66, 011918. (23) Liu, Y. C.; Chen, H. L.; Su, C. J.; Liu, H. K.; Liu, W. L.; Jeng, U. Macromolecules 2005, 38, 9434–9440. (24) Evans, H. M.; Ahmad, A.; Ewert, K.; Pfohl, T.; Martin-Herranz, A.; Bruinsma, R. F.; Safinya, C. R. Phys. ReV. Lett. 2003, 91 (7), 0755011. (25) Dias, R.; Mel’nikov, S.; Lindman, B.; Miguel, M. G. Langmuir 2000, 16, 9577–9583. (26) Cakara, D.; Kleimann, J.; Borkovec, M. Macromolecules 2003, 36, 4201–4207. (27) Roe, R. J. Methods of X-ray and Neutron Scattering in Polymer Science; Oxford Unversity Press: Oxford, 2000. (28) Maiti, P. K.; Cag˘in, T.; Lin, S. T.; Goddard, W. A., III. Macromolecules 2005, 38, 979–991. (29) Fo¨rster, S.; Timmann, A.; Schellbah, C.; Meyer, A.; Funari, S. S.; Mulvaney, P.; Knott, R. J. Phys. Chem. B 2005, 109, 1347–1354. (30) For the complex with N/P ) 20, the values of the morphological parameters obtained from the fit using the paracrystalline model for hexaginally-packed cylinders were dDNA ) 3.63 nm, domain size ) 34.3 nm, g ) 0.083, RDNA ) 0.92 nm, and σR ) 0.06. For the system with N/P ) 1.0, the parameters were dDNA ) 3.20 nm, domain size ) 30.2 nm, g ) 0.090, RDNA ) 0.92 nm, and σR ) 0.05. (31) Ottaviani, M. F.; Sacchi, B.; Turro, N. J.; Chen, W.; Jockusch, S.; Tomalia, D. A. Macromolecules 1999, 32, 2275–2282. (32) Strick, T. R.; Allemand, J. F.; Bensimon, D.; Croquette, V. Biophys. J. 1998, 74, 2016–2028. (33) Sanders, L.; Gua´queta, C.; Angelini, T. E.; Lee, J. W.; Slimmer, S. C.; Luijten, E.; Wong, G. C. L. Phys. ReV. Lett. 2005, 95, 108302. (34) Tomalia, D. A.; Naylor, A. M.; Goddard, W. A., III. Angew. Chem., Int. Ed. Engl. 1990, 29, 138–175. (35) Chen, W. R.; Porcar, L.; Liu, Y.; Butler, P. D.; Magid, L. J. Macromolecules 2007, 40, 5887–5898. (36) Guo, Z.; Sadler, P. J.; Zang, E. Chem. Commun. 1997, 32, 27–28. (37) Kondo, S.; Hiraoka, Y.; Kurumatani, N.; Yano, Y. Chem. Commun. 2005, 1720–1722. (38) Porcar, L.; Liu, Y.; Verduzco, R.; Hong, K.; Butler, P. D.; Magid, L. J.; Smith, G. S.; Chen, W. R. J. Phys. Chem. B 2008, 112, 14772. (39) Schneider, B.; Patel, K.; Berman, H. M. Biophys. J. 1998, 75, 2422– 2434. (40) Thomas, T. J.; Thomas, T. Nucleic Acids Res. 1989, 17, 3795–3810. (41) Robinson, H.; Wang, A. H., J. Nucleic Acids Res. 1996, 24, 676–682. (42) Müller, J. J. J. Appl. Crystallogr. 1983, 16, 74–82. (43) For the complex with N/P ) 14, the values of the morphological parameters obtained from the fit using the paracrystalline model for square-packed cylinders were dDNA ) 3.35 nm, domain size ) 30.3 nm, g ) 0.089, RDNA ) 0.95 nm, and σR ) 0.09. For the system with N/P ) 1.0, the parameters were dDNA ) 2.78 nm, domain size ) 20.3 nm, g ) 0.126, RDNA ) 1.05 nm, and σR ) 0.1.

BM801246E

Columnar Mesophases of the Complexes of DNA with Low

Recommend Documents