Temperature- and Composition-Dependent DNA Condensation by Thermosensitive Block Copolymers.

Successful intracellular delivery of genes requires an efficient carrier, as genes by themselves cannot diffuse across cell membranes. Because of the toxicity and immunogenicity of viral vectors, nonviral vectors are gaining tremendous interest in research. In this work, we have investigated the temperature-dependent DNA condensation efficiency of various compositions of a thermosensitive block copolymer viz., poly(N-isopropylacrylamide)-b-poly(2-(diethylamino)ethyl methacrylate) (PNIPA-b-PDMAEMA). Three different copolymer compositions of varying molecular weights were successfully synthesized via the RAFT polymerization technique. Steady-state fluorescence and circular dichroism (CD) spectroscopies, dynamic light scattering (DLS) and zeta potential measurements, agarose gel electrophoresis, and atomic force microscopy techniques were utilized to study the interaction of the copolymers with DNA at temperatures above and below the critical aggregation temperature (CAT). All these experiments revealed that, above the CAT, there was formation of highly stable and tight polymer-DNA complexes (polyplexes). The size of polyplexes was dependent on the temperature up to a certain charge ratio, as determined by the DLS results. The results obtained from temperature-dependent fluorescence spectroscopy, CD, and gel electrophoresis indicated that the DNA molecules were shielded more from aqueous exposure above the CAT because of the formation of relatively more compact complexes. The polyplexes also exhibited changes in the particle morphology below and above the CAT, with particles generated above CAT being more spherical in morphology. These results suggested at the possibility of modulating the complex formation by temperature modification. The present biophysical studies would provide new physical insight into the design of novel gene carriers.

Introduction

Nonviral gene delivery vectors are becoming increasingly attractive as gene carriers[1,2] because of their high flexibility with respect to the size and nature of DNA delivered, low immunogenicity, besides relatively low cost of production, etc.[3−5] Among the nonviral vectors, polycations can easily condense the DNA chains containing negatively charged phosphate groups into compact complexes, thus facilitating the entry of the complex into the cells via endocytosis and subsequent dissociation of complexes for the release of DNA.[6−9] For example, poly(ethylenimine) (PEI), mostly a highly branched structure that is commercially available, is one of the most efficient polycation systems studied widely for gene delivery.[9−13] However, PEI exhibits disruption of the cell membrane and a higher cell cytotoxicity, partly because of a significantly high cation content, and also shows lower efficiency in terms of transcription and transfection as compared to most viral vectors.[14] Linear poly((2-dimethylamino)ethyl methacrylate) (PDMAEMA)[15−18] is another example of a popular polycation for gene therapy because of its relatively lower cytotoxicity than PEI[19] and easy preparation. There are also other polycations, which may be used as gene vehicles, such as 2-(dimethylamino) ethyl methacrylate (DMAEMA)-based star copolymers,[19,20] cationic peptides,[21] dendrimer polycations,[22−24] hyperbranched poly(amino ester)s,[25,26] and polysaccharide-based cationic carriers.[27]

The main objectives for the development of nonviral vectors are enhanced responsivity,[28,29] specificity, transfection efficiency, and reduction of cell cytotoxicity.[30,31] It has been reported that above the normal body temperature, a direct cytotoxic effect on cancer cells has been noticed, whereas at lower temperatures, an immunostimulatory effect as well as radiation- and chemosensibilization occur.[32] Again, an additional advantage of application of heat at the targeting site is to achieve increase of the permeability of the tissue, allowing cell entry of gene carriers. In principle, this effect can also be effectively used for gene-delivery systems, for example, polyplexes,[33,34] by facilitating easy penetration of gene into the cells.

One of the potential routes by which such DNA-binding modulation or switching can be attained is by using responsive or “smart” polymers capable of conformational or phase changes under the effect of varying pH and temperatures.[35−37] Haladjova et al.[38] have recently reported the preparation of mesoglobules using thermosensitive polymers as potential reservoirs, vehicles, and transferring agents for biologically active substances such as DNA etc. Thermosensitive star-shaped copolymers having branched chains were designed by Zhou et al., which can potentially serve as gene carriers with improved gene transfection efficiency.[39] Hinrichs et al.[40] showed that the transfection efficiency as well as cytotoxicity were influenced by changing both the size and zeta potential of thermosensitive copolymer/plasmid complexes by tuning the temperature. Again, Takeda et al.[41] reported that the affinity of temperature-sensitive polymeric carriers incorporating hydrophobic monomers is reduced by lowering the incubation temperature below the lower critical solution temperature (LCST), suggesting that a thermally regulated gene expression can enable DNA release from the polymeric carrier in a more effective manner.

Poly(N-isopropylacrylamide) (PNIPA), one of the most widely studied responsive polymers, is known to undergo a sharp coil–globule transition in water at 32 °C (LCST) resulting from the hydrophilic to hydrophobic phase transition.[42] This property has been utilized for biomedical applications, ranging from pulsatile drug release to control of cell adhesion, mainly because the LCST is close to our body temperature.[43] Incorporation of PNIPA in a block polycation may lead to the formation of a compact and stable complex at a higher temperature because of collapsing of the PNIPA chains. Thus, by simply changing the temperature, it is possible to control the transition of the polyplex from tight or compact (prevents DNA degradation) to loosely bound complexes (suitable for transfection).[44,45] In this regard N-isopropyl acrylamide (NIPA) and DMAEMA- based thermosensitive block polycations have been explored for gene-delivery applications.[39,40]

Although thermosensitive polycations based on NIPA and DMAEMA have been already explored for gene-delivery applications, including cell-cytotoxicity and transfection efficiency study of polyplexes, detailed biophysical investigations on the influence of thermosensitive components in the block polycation on the properties of polyplexes are needed. Moreover, in this study, we have synthesized block copolymers (BCPs) of DMAEMA and NIPA of different compositions and molecular weights via the RAFT polymerization technique. This technique is more efficient than free radical copolymerization to get more control over the molecular weight and dispersity.[40,46,47] The structure of the copolymers used here for studying the interaction behavior is linear, whereas the previous studies were of a different architecture. Here, we report the interaction behaviors of the synthesized thermosensitive BCPs with calf thymus DNA (ctDNA) in detail using various biophysical techniques for the evaluation of the potential of these BCPs as gene carriers.

Results and Discussion

Synthesis of Cationic BCPs

The diblock copolymers of NIPA and DMAEMA were synthesized by the RAFT polymerization process (Scheme ). At first, PNIPA macro-chain transfer agent (CTA) was synthesized by polymerizing NIPA using S-1-dodecyl-S′-(α,α′-dimethyl-α″-acetic acid)trithiocarbonate (DDMAT) as the primary CTA. The number average molecular weight (Mn) of the synthesized PNIPA macro-CTA was determined by proton nuclear magnetic resonance (1H NMR) spectroscopy. The intensity of the peak at 0.88 ppm corresponding to the three protons of the terminal methyl group of the dodecyl unit of DDMAT was compared with the intensity of one proton at 4.1 ppm of NIPA repeat units (Figure S1, Supporting Information). Mn of PNIPA macro-CTA was obtained as 8600 g mol–1 that corresponds to the presence of 76 NIPA units in the macro-CTA. In the next step, DMAEMA was copolymerized in the presence of PNIPA macro-CTA to synthesize the desired BCPs. The 1H NMR spectrum of one of the copolymers is provided in Figure . The 1H NMR spectra of the other two copolymers are given in the Supporting Information (Figures S2 and S3). The composition of the BCPs was determined by comparing the intensities of two methylene protons (a, at ∼4.06 ppm) adjacent to oxygen of the ester group plus two methylene protons (b, at ∼2.6 ppm) adjacent to the amine group in the PDMAEMA block with the intensity of one tertiary proton (d, at 4.06 ppm) of the isopropyl group in the PNIPA block. The gel permeation chromatography (GPC) traces of the three BCPs are provided in the Supporting Information (Figure S4). All polymers showed unimodal peaks having narrow dispersity (Đ) values, (<1.30), confirming the formation of BCPs. The molecular weight and composition of the three synthesized diblock copolymers (NIDM108, NIDM135, and NIDM158) are given in Table .

Scheme 1

Schematic Representation of the Synthetic Pathway of Diblock Copolymers (PNIPA-b-PDMAEMA)

In the cartoon picture, the purple units represent the cationic PDMAEMA block, and dark orange units represent the thermosensitive PNIPA block. At pH = 4.2, the PDMAEMA block becomes cationic due to the protonation of tertiary amine groups.

Figure 1

1H NMR spectra of NIDM108 in CDCl3.

Table 1

Details of Molecular Weight and Composition of the Three Synthesized Diblock Copolymers

polymer abbreviation	polymer composition	no. of DMAEMA units per polymer chain estimated from feed ratio	no. of DMAEMA units per polymer chain determined from 1H NMR	Mn (from1H NMR) (g mol–1)	Mn (from GPC) (g mol–1)	dispersity (Đ) (from GPC)
NIDM108	(PNIPA)76-b-(PDMAEMA)108	125	108	25 600	23 240	1.26
NIDM135	(PNIPA)76-b-(PDMAEMA)135	150	135	29 800	28 650	1.23
NIDM158	(PNIPA)76-b-(PDMAEMA)158	175	158	33 400	32 050	1.18

Thermally Induced Self-Assembly of BCPs

Temperature-induced phase transition of the PNIPA-b-PDMAEMA copolymers in 10 mM potassium phosphate buffer (pH = 4.2) was investigated by measuring the transmittance at 500 nm as a function of the temperature. On increasing the temperature above 32 °C, transparent copolymer solutions turned hazy, which was manifested in the decreasing value of transparency (Figure a). However, the reduction in the transmittance value for these copolymers was less sharp than those observed for PNIPA macro-CTA (PNIPA homopolymer) that shows LCST at ∼32 °C.[42,48] This is due to the presence of the hydrophilic PDMAEMA block in the copolymer. Moreover, unlike the PNIPA homopolymer (where the transmittance values decreased to near zero and the polymer became insoluble), for the BCPs, the transmittance values did not decrease to near zero, indicating the formation of self-assembled nanoaggregates having a comparatively higher solubility and stability above a certain temperature. All BCP solutions were kept at 40 °C for longer than a week, and no precipitation was observed. After the critical aggregation temperature (), the PNIPA block of the BCPs became hydrophobic but the PDMAEMA block remained hydrophilic, which resulted in the formation of self-assembled aggregates. This established the water-soluble nature of the BCP aggregates at this temperature. The critical temperature above which the aggregation was formed (CAT) was found to be ∼34, ∼35.5, and ∼36.5 °C for NIDM108, NIDM135, and NIDM158, respectively. The temperature corresponding to 50% drop in transmittance was considered as the CAT (Figure a). Dynamic light scattering (DLS) measurement was also performed to follow the temperature-induced self-association of these BCPs. With DLS, we can determine the average hydrodynamic diameters (DH) of the BCPs; from the change of DH with the temperature, we can also detect their CAT. The average hydrodynamic diameters (DH) of all polymers were found to be ∼20 nm at 25 °C. The size increased with the increase in the temperature above the transition point because of the formation of self-assembled aggregates. The size of the aggregates was more or less similar for the three copolymers. The average hydrodynamic diameter (DH) of the copolymers at various temperatures is shown in Figure b. The zeta potential values of the copolymer solutions also displayed an abrupt increase at around the CAT (Figure c). The increase in the zeta potential is most likely due to the self-assembly of the copolymer molecules. Above the CAT, the PNIPA molecules form the core and the charged PDMAEMA forms the shell of the nanoparticles that results in a higher surface charge concentration and consequently a higher zeta potential.[35,49]

Figure 2

(a) Plot of transmittance (%) at 500 nm of 0.1 mM PNIPA-b-PDMAEMA copolymers and the PNIPA homopolymer in the phosphate-buffered saline (PBS) buffer (pH = 4.2). Variation of the (b) average hydrodynamic diameter (DH) and (c) zeta potential of the BCPs with temperature in the PBS buffer (pH = 4.2).

Interaction of ctDNA with the BCPs

Interaction between ctDNA and the PNIPA-b-PDMAEMA copolymers was studied in 10 mM potassium phosphate buffer (pH = 4.2) at various solution temperatures. At pH = 4.2, the amino groups of DMAEMA units are protonated, and the polymers become cationic, facilitating the electrostatic interaction between negatively charged DNA molecules. In the present work, the binding between the BCPs and ctDNA was studied using the turbidity measurement, ethidium bromide (EB) dye exclusion assay (steady-state fluorescence), DLS, zeta potential measurements, agarose gel electrophoresis, CD, and atomic force microscopy (AFM) studies.

Colloidal Stability of the Polyplexes

Turbidity measurements were performed to record the stability of the polyplexes at 25 °C (below CAT) and 40 °C (above CAT). The results showed insignificant turbidity values at 25 °C for all polyplexes, which were nearly unchanged for 10 days, suggesting no precipitation (Supporting Information, Figures S5 and S6). Therefore, it is evident that the BCPs had the ability to condense DNA efficiently, forming nanosized polyplexes with an enhanced colloidal stability. At 40 °C, a very little increase of turbidity was observed for the polyplexes. The stability at lower and higher charge ratios was provided predominantly by the net anionic or cationic charges in the polyplexes, as indicated by the zeta potential values of the complex (discussed later). Neutral or nearly neutral polyplexes were stabilized by the lyophilizing effect of the nonionic hydrophilic PNIPA blocks present in the copolymers. The later stabilization effect is quiet comparable with a report published by Ambardekar et al.[50] They reported that polyethylenimine-g-poly(ethylene glycol) (PEI–PEG) graft copolymer produced stable neutralized/electropositive polyplexes with siRNA because of the solubilizing effect of the neutral hydrophilic PEG chain.

Steady-State Fluorescence Spectroscopic Studies

EB dye exclusion assay was used to monitor the interaction between the copolymers and ctDNA. It has been well-known that EB can easily bind to DNA by intercalation into the hydrophobic G–C and A–T base pairs, thereby resulting in a high fluorescence intensity.[24] Intercalated EB in DNA molecules can be effectively dislodged from the DNA double-helix by polycations that include cationic BCPs. The fluorescence intensity of EB in the DNA–EB complex gets reduced on addition of the cationic agents, owing to the change in the surrounding of the EB dye from hydrophobic (DNA helix) to hydrophilic (buffer medium). In this particular study also, the decrease in the fluorescence intensity as a result of displacement of EB molecules from the DNA double-helix by the NIDM copolymers was utilized to monitor the copolymer–ctDNA polyplex formation.

Polyplexes were prepared in the range of Z+/– values of 0–6; fluorescence spectra were taken at four different temperatures; 15 min of equilibration time was provided at each temperature before recording the spectra. The change in the fluorescence intensity of EB with increasing Z+/– ratios is shown in Figure . A reduction in the fluorescence intensity was observed till Z+/– ≈ 1, above which the EB displacement profile reached a plateau. This particular Z+/– value did not change significantly with temperature for a particular copolymer. But, the decrease in the fluorescence intensity was the steepest for NIDM108. Formation of copolymer–ctDNA polyplexes can be attributed to the ionic interaction of the positive charges on the polymer with the negatively charged phosphate groups of ctDNA. As the length of the cationic block was the shortest in NIDM108, for a given Z+/– value, the number of such polymers present in the solution was more. This results in relatively easier access to the negatively charged phosphate groups in DNA for the cationic charges in the block.

Figure 3

(a) Fluorescence spectra of the EB–DNA complex in the presence of NIDM108 corresponding to Z+/– = 1 at 25 and 40 °C. (b–d) present the relative fluorescence intensity of the EB–DNA complex in the presence of various amounts of NIDM108, NIDM135, and NIDM158, respectively, at four different temperatures.

Although the nature of the EB exclusion curves was not affected by temperature variation, the extent of this polymer–DNA interaction was clearly temperature-dependent, with a higher amount of EB displaced at temperatures above the CAT of the polymers. The fluorescence intensity of EB for polyplexes with NIDM108 decreased maximum by ∼85%; the decrease for the other two copolymers NIDM135 and NIDM158 were ∼78 and ∼73%, respectively, at 25 °C. NIDM108 showed the highest displacement of EB, whereas NIDM158 displaced the lowest amount of EB in the series at the same charge ratio (Z+/–). The displacement of EB for all BCPs was higher at 40 and 45 °C (i.e., above CAT) than that at 25 and 28 °C. A comparison of data presented in Figure b–d showed that the effect of temperature on EB displacement was the most prominent for NIDM108 and the least for NIDM158. The same data can be replotted to infer that the relative difference in the EB-displacement capability among the three BCPs gets amplified at a temperature above the CAT (please see Figure S7, Supporting Information). Significant electrostatic interaction continued to happen between the cationic PDMAEMA block and DNA till the charge neutralization occurs. Once this neutralization process is achieved, further hydrophobic interaction between the PNIPA chain and DNA moiety starts taking effect. As the electrostatic interaction between the cationic PDMAEMA block and DNA brings the PNIPA chains to a close proximity to the DNA chains, the local concentration of PNIPA chains around the DNA got increased, resulting in a hydrophobic interaction between PNIPA and hydrophobic DNA bases, especially above the CAT. Therefore, PNIPA might induce additional condensation of DNA in a synergistic manner. Similar observations were reported by us while studying the interaction of DNA with PEGylated cationic BCPs.[51,52] For a given charge ratio, the NIDM108 copolymer would have a higher concentration of temperature-sensitive PNIPA chains, helping in higher compaction of the polyplexes, which forced more EB to be released.

DLS Measurements

The DLS technique provides useful information about the average hydrodynamic diameters and polydispersities of DNA–cationic BCP polyplexes. The sizes of the polyplexes are dependent on the charge ratios. The average size versus charge ratios for the three copolymers at two different temperatures, 25 °C (below CAT) and 40 °C (above CAT), are presented in Figure (the data for other temperatures can be seen in Figure S8, Supporting Information), and the polydispersities of polyplexes along with the intensity-weighted average hydrodynamic diameters are tabulated in Table (intensity-weighted size distributions are given in the Supporting Information, Figure S9). For both the temperatures, a gradual decrease in the average size occurred from ∼700–750 nm (the size of free ctDNA) to ∼200–300 nm for the ctDNA–BCP polyplexes at Z+/– ≈ 1.0 for NIDM108 and at Z+/– ≈ 1.5 for NIDM135 and NIDM158. After this charge ratio, the average size of the polyplexes increased drastically to about 1000–1100 nm at Z+/– = 6 for all three polymers. From the studies of the size distribution (please see Figure S9 in the Supporting Information and Table ) of polyplex particles, it was also noted that an inverse relationship between Z+/– and the mean polydispersity of the corresponding polyplexes was observed for most of the polyplex formations up to the effective charge neutralization, Z+/– ≈ 1.0 for NIDM108 and Z+/– ≈ 1.5 for NIDM135 and NIDM158. Above these Z+/– values, the polydispersity gets increased because of aggregation between the polyplexes, as discussed above. For all Z+/– values studied, the polyplexes exhibited an acceptable (<0.420) mean polydispersity. Moselhy et al.[53] reported that the sizes of the polyplexes formed from the thermosensitive cationic nanogel (based on NIPA and DMAEMA) with salmon sperm DNA were around 450–700 nm, which were bigger than our system at the neutralization point.

Figure 4

Hydrodynamic diameter of DNA–cationic BCP polyplexes at a fixed DNA concentration of 25 μM and varying concentrations of copolymers at (a) 25 and (b) 40 °C.

Table 2

Particle Size (DH) and Polydispersity Index (PDI) of NIDM/ctDNA Polyplexes Measured by DLS; (Mean ± SD, n = 3)

	0.2		0.6		1.0		1.5		2.0
charge ratio, Z+/–	DH (nm)	PDI	DH (nm)	PDI	DH (nm)	PDI	DH (nm)	PDI	DH (nm)	PDI
	NIDM108/DNA
25 °C	470 ± 6	0.282	351 ± 1	0.228	246 ± 7	0.192	269 ± 2	0.229	450 ± 10	0.251
40 °C	448 ± 5	0.250	313 ± 4	0.213	204 ± 2	0.191	293 ± 4	0.261	543 ± 5	0.298
	NIDM135/DNA
25 °C	497 ± 7	0.420	393 ± 11	0.310	316 ± 1	0.302	238 ± 8	0.301	419 ± 7	0.310
40 °C	490 ± 14	0.244	363 ± 10	0.247	283 ± 4	0.236	252 ± 2	0.298	502 ± 4	0.512
	NIDM158/DNA
25 °C	516 ± 8	0.382	405 ± 2	0.323	338 ± 4	0.313	252 ± 2	0.291	449 ± 7	0.350
40 °C	512 ± 15	0.299	388 ± 3	0.314	318 ± 3	0.303	267 ± 6	0.277	502 ± 4	0.567

At lower Z+/– values, polyplexes from NIDM108 had the lowest sizes among the three polyplexes. These results indicate that NIDM108 had the highest efficiency among all the copolymers in condensing the DNA molecule till the size reached a minimum. As discussed earlier, as the length of the cationic block was the shortest in NIDM108, for a given charge ratio, the number of such polymers present in the solution was more. This probably provided relatively easier access to the negatively charged phosphate groups in DNA for the cationic charges in NIDM108, resulting in higher compaction. It was also observed that for the charge ratios near and below the neutralization point, compaction was more above the CAT in comparison to below the CAT (Figure S9, Supporting Information). As discussed previously, this increased compaction was due to the PNIPA-induced additional condensation of DNA chains. However, after the neutralization, the size of the polyplexes above the CAT was larger than that below the CAT. This may be due to increased aggregation between the individual polyplexes through the hydrophobic interaction of PNIPA chains above the CAT. At higher Z+/– values (>1.5), polyplexes from NIDM108 had larger sizes compared to other polyplexes. For a given charge ratio, NIDM108 provided maximum number of PNIPA chains for aggregate formation, causing formation of large-sized aggregates. This result is slightly amplified at a higher temperature.

Zeta Potential Measurements

The zeta potential (ζ) values of different polyplexes in the buffer at four different temperatures 25, 28, 40, and 45 °C are depicted in Figure . The zeta potential of free ctDNA was found to be around −20 mV because of the presence of negatively charged phosphate groups. With the addition of the cationic polymers, the zeta potential of the polyplexes gradually increased until a plateau value of around 15–25 mV was reached at Z+/– = 3–6.[41,54] The increased zeta potential with increase in Z+/– may be possibly due to a gradual covering of the negatively charged DNA by the positively charged polymer. Here, we can see that neutrally charged polyplexes were formed at different charge ratios for different polymers. Charge ratios for the formation of neutral polyplexes corresponds to a Z+/– value of ∼0.7 for NIDM108 and between 1.0 and 1.5 for both NIDM135 and NIDM158 at all temperatures. The formation of neutral polyplexes by NIDM108 at a lower charge ratio explains a more compact polyplex formed because of the reasons explained earlier. The same data can be replotted to infer that the zeta potentials of polyplexes formed by the three BCPs gets amplified at a temperature above the CAT (Figure S10, Supporting Information).

Figure 5

Zeta potential (ζ) of DNA–cationic BCP polyplexes at a fixed DNA concentration of 25 μM and varying concentrations of copolymers (a) NIDM108, (b) NIDM135, and (c) NIDM158, respectively. Zeta potentials were determined at 25, 28, 40, and 45 °C.

Agarose Gel Electrophoresis

The binding capability of the BCPs has been investigated qualitatively via agarose gel electrophoresis.[44,45] In this experiment, we prepared the polyplexes at 25 and 40 °C, and the electrophoresis assay was also conducted at 25 and 45 °C. The concentration of the dye was much lower (1 μg/mL) in these experiments than in the EB exclusion assays, and the dye was added to the loading buffer after polyplex formation rather than before (as in the dye exclusion assay). Images of gels obtained on electrophoresis of the polyplexes are shown in Figure . Here, we can see that the migration of polyplexes was retarded with an increase in the charge ratio from 0.25 to 3, and much more DNA was restricted in the wells at 25 °C. These retardations imply the neutralization of negatively charged DNA by positively charged copolymers as well as the higher molecular weight of the formed polyplexes. A complete reduction of the migration of DNA was observed at a charge ratio less than 1.0 for NIDM108 and just higher than 1.0 for the other two copolymers as a consequence of charge neutralization. The charge ratio values were in agreement with the zeta potential data. Earlier retardation of NIDM108 polyplexes through the gel at lower charge ratios is thus indicative of higher DNA binding capability of NIDM108. Electrophoresis was also carried out at temperatures above the CAT for these copolymers. Between a temperature below or above the CAT of thermosensitive copolymers, a noticeable change in the retardation bands was observed for NIDM108/DNA complexes, but slight shifts in the electrophoretic mobility of the other two polymer–DNA complexes were observed. It was also noticed that, while at 40 °C, the movement of NIDM108/DNA complexes was completely restricted at a charge ratio of 0.8, same could not be said at 25 °C. This implied the formation of more stable NIDM108/DNA complexes at a higher temperature. Similar results were also observed for the EB exclusion assay. Additionally, bands of the polyplexes were found to be of lower intensity above the CAT, implying the formation of DNA polyplexes into which the EB dye could intercalate to a lesser extent because of the collapse of the PNIPA side chain. This indicates the formation of more tight polyplexes above the CAT than those formed below the CAT.

Figure 6

Agarose gel retardation assays for the BCPs. Complexes were prepared in the PBS buffer (pH 4.2) at different charge ratios. The complexes were allowed to get stabilized for 15 min prior to gel loading. (a–c) NIDM108, NDM135, and NIDM158, respectively; polyplexes were formed and run at 25 °C. (d–f) NIDM108, NDM135, and NIDM158, respectively; polyplexes were formed and run at 40 °C. Lane 1, free DNA; lanes 2–7, polyplexes at various charge ratios: lane 2, Z+/– = 0.3; lane 3, Z+/– = 0.8; lane 4, Z+/– = 1.0; lane 5, Z+/– = 1.5; lane 6, Z+/– = 2.0; and lane 7, Z+/– = 3.0.

Circular Dichroism (CD) Spectroscopy

CD spectroscopy has been widely used to monitor the conformational change of DNA upon complexation with thermosensitive gene delivery vectors.[55] None of the copolymers used here showed any significant CD signal within the UV region investigated, which means that the observed signals arise entirely from the DNA molecules. The CD spectrum of free ctDNA is typical of the B conformation showing a positive signal at 275 nm due to stacking of bases and a negative minimum near 245 nm due to the helical structure of the polynucleotide.[56] A significant change in the DNA conformation was observed when DNA was complexed with cationic polymers at different charge ratios. Typically, the CD spectra (Figure a–c) showed a gradual decrease in the molar ellipticity values of the positive signal, concomitant with a red shift as the polymer fraction in the complexes was increased. In most cases, the negative signal also shifted to higher wavelengths, with decreased molar ellipticity values. For the NIDM108 copolymer, CD signals reached a plateau level, but the other two did not exhibit saturation. Again, it is clearly observed from Figure that the CD intensity at 275 nm decreased to a certain degree when the temperature was increased from 25 to 40 °C.

Figure 7

Influence of the temperature-induced phase transition of the BCPs on the CD spectra of polyplexes. (a–c) On the left-hand side, CD spectra of ctDNA in the absence or presence of various amounts of BCPs recorded at 25 and 40 °C. (d–f) On the right-hand side, dependence of θ (275 nm) on the charge ratios determined at 25 and 40 °C.

Generally, DNA remained as the B conformation in the complexes both at 25 and 40 °C. The ellipticities at 275 nm [θ (275 nm)] of the polyplexes are plotted[45] as a function of the charge ratios in Figure d–f. These temperature-induced peak intensity changes can be explained as follows. At temperatures below the CAT, a loose structure of polyplex is formed because of the hydration of PNIPA chains rendering more DNA exposure. But, at 40 °C (above the CAT), the collapsed PNIPA chains tightly cover the surface of the polyplex, protecting DNA from exposure. As a result, the CD signals are decreased at a higher temperature.

Effect of Temperature on the Morphology of Polyplexes Using AFM

AFM images were obtained to gain further insight into the temperature-dependent behavior of the polyplexes. These studies demonstrated the potential of temperature-sensitive polymers in DNA delivery. We find that the degree of DNA condensation can be controlled by temperature. The AFM images suggest (Figure ) that above the CAT, the polyplexes were reasonably homogeneous in size, and that, they possessed a nearly compact spherical and globular morphology due to heavy dehydration of PNIPA chains; but below the CAT, these were slightly elongated in morphology. Bao et al.[57] summarized through transmission electron microscopy images that a more uniform morphology of particles can be achieved by the complexation of thermosensitive chitosan-based terpolymers with DNA above the CAT. Globules are, for the design of a gene carrier, the most desired morphology for gene transfection and cell entry. The condensing ability of the copolymers also decreased from NIDM108 to NIDM158 because of the gradual decrease of the weight content of NIPA units in copolymers; also, a higher NIPA content introduced more reduction in the size. Polyplex aggregation also occurred in the form of larger aggregates at a higher charge ratio.

Figure 8

AFM images of polyplexes deposited on mica at a charge ratio of 1.0. (a–c) Polyplexes are formed from NIDM108, NDM135, and NIDM158, respectively; polyplexes are prepared and deposited at 25 °C. (d–f) Polyplexes are formed from NIDM108, NDM135, and NIDM158, respectively; polyplexes are prepared and deposited at 40 °C.

Effect of pH on the Polyplexes

To find out the effect of pH on the complexation behavior of the BCPs and ctDNA, EB exclusion assay was performed at different pH values at 25 °C by taking NIDM108/ctDNA polyplex at Z+/– = 1.0. At pH 7.4, the relative fluorescence intensity was found to be 0.34, which was significantly higher than the value at pH = 4.2 (0.19). This confirmed a comparatively weak complexation behavior at pH = 7.4 because of a lower charge density in the PDMAEMA block. Further, to find the usefulness of the polyplex formed at pH = 4.2, the polyplex was first prepared at pH = 4.2, following which the pH was step-wise increased to 7.4 by the addition of dilute NaOH, and the fluorescence spectra (EB exclusion assay) and hydrodynamic size were recorded after allowing 15 min of equilibration time at each pH. With the increase in pH, more EB was excluded from DNA, and the size of the polyplex was gradually increased, although the EB exclusion at pH = 7.4 was lower than the value when the complexation was carried out directly at pH = 7.4 (Figure S11, Supporting Information). This means that the total time provided during the pH change process was not sufficient to reach equilibrium value at pH = 7.4. To get some idea about the time required to reach equilibrium, the polyplex was prepared at pH = 4.2, and then, the pH was increased to 7.4 at once, and the EB exclusion and hydrodynamic size were monitored with time. As it can be seen from Figure S12 (Supporting Information), even after 30 min, there was no significant change in the size or EB exclusion, after which the EB release and size increase happened gradually. This indicates that, potentially, the polyplex formed at pH = 4.2 can maintain its integrity for the first 30 min in physiological condition before slowly loosening up.

Summary

In this study, BCPs of DMAEMA and NIPA were evaluated as thermosensitive carrier systems for DNA. The BCPs formed self-assembled nanoparticles above a certain temperature (CAT). The copolymers effectively condensed DNA into neutrally or slightly charged tight particles of size around 200–300 nm for the charge ratio between 1.0 and 1.5. This size range is optimum for efficient transfection. The complexation behavior between DNA and the copolymers was found to be dependent on the surrounding temperature. Although at temperatures below the CAT, comparatively loosely bound polyplexes were formed, above the CAT, relatively tight complexes were formed because of temperature-induced collapse of PNIPA chains of the copolymers in aqueous medium. The results suggest that the synthesis of thermosensitive polymers may enable compaction of DNA into nanosized particles that may potentially be rapidly taken up by the cells. Further, a relatively easy release of DNA from the loosely packed polyplexes may be favorable for transcription. This thorough biophysical investigation provides valuable insight into the DNA–thermosensitive polymer complexation phenomenon, which has direct consequences on the release behavior of DNA from the gene-delivery vectors.

Experimental Section

Materials

DMAEMA, NIPA, 2,2′-azobisisobutyronitril (AIBN), sodium salt of ctDNA, and EB were purchased from Sigma-Aldrich and used as such. DDMAT was made according to a reported procedure with some modification.[58] All experiments were performed in 10 mM potassium phosphate buffer (pH = 4.2) prepared using Milli-Q water. All of the other chemicals used were of analytical reagent grade and used without any purification. Unless otherwise specified, the concentrations of DNA solutions are provided in mol L–1 in terms of the negatively charged phosphate groups in the DNA backbone.

Synthesis and Characterization of BCPs

Synthesis of Diblock Copolymers (PNIPA-b-PDMAEMA) (Scheme )

Poly(N-isopropylacrylamide) macro-CTA was synthesized in the first step, using AIBN as the initiator at 70 °C in the presence of DDMAT as the primary CTA. Polymerization was conducted under a N2 atmosphere for 18 h in 1,4-dioxan medium. The ratio of the monomer (NIPA) to CTA, that is, [M]0/[CTA]0 was 100:1 and the CTA to initiator ratio, that is, [CTA]0/[I]0 was 3:1. The reaction was quenched by cooling the polymer solution in liquid nitrogen, followed by dilution with a small quantity of tetrahydrofuran (THF) and subsequent precipitation into excess ice-cold diethyl ether to isolate the polymer. The process of precipitation was repeated thrice, and finally, the pure solid polymer was isolated by filtration followed by vacuum drying for ∼12 h. Thereafter, this poly(N-isopropylacrylamide) macro-CTA was used for polymerizing DMAEMA in 1,4-dioxan at 70 °C for 24 h under a nitrogen atmosphere to yield three BCPs with the desired number of DMAEMA units. The polymerization reaction was quenched in the usual way using liquid nitrogen, and the polymer was isolated by precipitation. This was followed by dialysis using cellulose membranes (molecular weight cutoff value ∼12 kDa) against distilled water for 3 days with frequent change of water (four times a day). The pure copolymer solution was then lyophilized, freeze-dried, and characterized by 1H NMR spectroscopy and GPC for its final composition.

1H NMR spectra were taken in CDCl3 using a Bruker DPX-400 MHz NMR spectrometer. The spectra were calibrated using the residual solvent signal as the internal standard. The number-average (Mn) molecular weight, weight-average molecular weight (Mw), and dispersity (D̵ = Mw/Mn) of the synthesized BCPs were determined by a GPC (Viscotek TDAmax) system equipped with refractive index, differential pressure viscometry, and dual-angle light scattering (λ = 670 nm, 90° and 7°) detectors and an isocratic pump (Agilent 1200), using THF (high-performance liquid chromatography grade) as the mobile phase with a flow rate of 1 mL/min at 33 °C. Triple detector array assembly is advantageous as chromatographic calibrations are not necessary. The light scattering detector was calibrated using polystyrene standards having a low polydispersity (Mn = 105 164, Mw/Mn = 1.02, and [η] = 0.48 dL g–1 at 33 °C in THF, dn/dc = 0.185 mL g–1) provided by the supplier Viscotek.

Turbidity measurement was done from the transmittance data at 500 nm obtained by using a Cary 5000 UV–vis–NIR spectrophotometer with a digital temperature controller (Varian Scientific Instruments). The concentrations of the aqueous polymer solutions were maintained at 1.0 g L–1. The solution was heated from 18 to 50 °C, and transmissions were recorded. A plot of % transmittance versus temperature was used to estimate the cloud point of the BCP solution. DLS measurement was used to determine the hydrodynamic diameter (DH) of the BCPs in aqueous solutions, using a Malvern Nano ZS instrument equipped with a thermostated sample chamber, using a 4 mW He–Ne laser (λ = 632.8 nm). The detector angle was fixed at 173°. The zeta (ζ) potential of the BCPs was measured in the temperature range of 25–45 °C using a Malvern Nano ZS instrument equipped with a 15 mV solid-state laser operating at a wavelength of 635 nm.

Characterization of Polymer–ctDNA Polyplexes

Preparation of Polymer–ctDNA Polyplexes

A ctDNA stock solution was prepared in 10 mM potassium phosphate buffer (pH = 4.2) at 25 °C. The concentration of the DNA stock solution was measured using a UV–visible spectrophotometer. The concentration of ctDNA (with respect to the negatively charged phosphate groups) was 712 μM, measured from its absorbance data at 260 nm with a molar extinction coefficient (ε) of 6600 M–1 cm–1. The concentration of DNA with respect to the base pairs is exactly half the concentration of the phosphate groups (ε = 13 200 M–1 cm–1). The ratio of the absorbance values of the DNA solution at 260 and 280 nm was 1.82, whereas the absorbance measured at 320 nm was negligible, suggesting the absence of any protein contamination. The EB stock solution was made by dissolving 2.2 mg in 1 mL of phosphate buffer. UV–visible spectrophotometer (ε = 5600 M–1 cm–1 at 480 nm) was used to determine the concentration. The EB solutions were stored in the dark at 4 °C prior to use. The BCP stock solutions (5150 μM) were prepared by dissolving a known weight of a given copolymer in the required volume of 10 mM phosphate buffer solution (pH = 4.2). The polymer–DNA polyplexes were prepared by adding the required amount of polymer solution to a ctDNA solution in 10 mM potassium phosphate buffer (pH = 4.2) to achieve an appropriate polymer-to-DNA charge ratio (Z+/–). This was followed by vortexing and equilibration for 1 h.[59−61]Z+/– was expressed as the ratio of equivalents of cationic DMAEMA units (from 1H NMR) to the number of phosphate groups (negatively charged) in DNA. The final concentration of ctDNA was kept fixed at 25 μM (except for CD spectroscopy, where 100 μM of ctDNA solution was taken) with respect to the phosphate groups. Polyplexes were prepared in the range of Z+/– values of 0–6.

Determination of Colloidal Stability of Polyplexes

Turbidity assays were used to study the colloidal stabilities of the polyplexes formed at Z+/– = 0.2, 0.5, 1.0, 2.0, and 3.0. The polyplexes were prepared at 25 °C, vortexed, and equilibrated for 1 h. Turbidity values were monitored for 10 days by taking the transmittance at 550 nm by using a Cary 5000 UV–vis–NIR spectrophotometer (Varian Scientific Instruments) fitted with a digital temperature controller. Similar measurements were also done at 40 °C.

Steady-State Fluorescence Spectroscopic Studies

The steady-state fluorescence spectra were recorded with help of a Jobin Yvon Fluorolog Spectrofluorometer using an excitation wavelength of 480 nm, and the emission spectra were recorded in the range of 500–700 nm wavelength. The excitation and emission slits were fixed at 5 and 2 nm, respectively. To make the DNA–EB complex, the DNA and EB stock solutions were mixed (1 EB:1 bp) in the phosphate buffer and equilibrated for 10 min. To a ctDNA–EB mixture (1 mL) in a quartz cuvette, the desired amounts of NIDM stock solutions were added. The steady-state fluorescence spectra were recorded at four different temperatures 25, 28, 40, and 45 °C. After addition of a given NIDM solution, the resultant mixture was equilibrated for 10 min at these four different temperatures before recording the spectrum. The addition of NIDM solutions continued so that the charge ratio could be varied from Z+/– = 0 to Z+/– = 6.

The size of ctDNA and DNA–BCP complexes were characterized by means of DLS measurements using a Malvern Nano ZS instrument with a thermostated sample chamber employing a 4 mW He–Ne laser operating at a wavelength of 632.8 nm and an avalanche photodiode detector. All measurements were performed at four different temperatures following the same procedure as discussed in fluorescence spectroscopic studies. The concentration of ctDNA was kept fixed at 25 μM in terms of negatively charged phosphate groups. The addition of copolymer solutions continued so that the charge ratio was varied from Z+/– = 0 to Z+/– = 6. After mixing the polymer with DNA in a cuvette, 10 min was allowed for equilibration to each sample mixture for four different temperatures before recording the spectrum. The average hydrodynamic diameter was provided by the software itself. The software used cumulant algorithm to calculate the distribution and averages.

Zeta (ζ) Potential Measurements

Zeta potentials of ctDNA and DNA–cationic copolymer complexes prepared in phosphate buffer (pH = 4.2) were measured at the same four different temperatures as mentioned before using a Malvern Nano ZS instrument equipped with a 15 mV solid-state laser operating at a wavelength of 635 nm.

The interaction between ctDNA and the cationic polymers was investigated by gel electrophoresis on agarose gel using 0.8% agarose gel containing EB (1 μg/mL). All polyplexes were prepared at different charge ratios (Z+/–) at 25 °C, and the electrophoresis was done at 25 °C. For the run at 40 °C, the agarose gel chamber was maintained at 40 °C, the polyplex solutions were preheated to 40 °C before injecting in the wells. The gel running buffer was 40 mM tris acetate (pH adjusted to 4.2) and 1 mM ethylenediaminetetraacetic acid. The gel was run at 80 V for 1 h, following which the DNA was visualized on a UV transilluminator (254 nm).

CD Spectroscopy

The thermosensitive conformational changes of polymer–DNA complexes compared to free ctDNA were studied by the analysis of their CD spectra. CD spectra were recorded using a JASCO 815 (Japan) CD spectrophotometer equipped with a Peltier temperature controller to monitor the temperature of the sample. All CD spectra were recorded in the range of 210–320 nm with a scan speed of 50 nm/min and a spectral band width of 10 nm. An average value obtained from three scans was taken for all experiments. The background spectrum of the buffer solution (phosphate buffer, pH = 4.2) was subtracted from the spectra of DNA and polymer–DNA complexes. Typically, a solution of 100 μM ctDNA was titrated with the appropriate cationic polymer solution in 10 mM phosphate buffer from Z+/– = 0 to Z+/– = 5 at 25 °C, and then, the temperature was increased to 40 °C as described before.

AFM Characterization

AFM imaging was done using an Agilent 5500 microscope without pretreatment of the sample. To take the images of polymer–DNA complexes, the contact mode in air was conducted on mica. The microfabricated Si-type NCH cantilevers had a nominal spring constant of 0.2 N/m and a nominal resonance frequency of 13 kHz. The scan rate employed was below 2 Hz to obtain good tracking of the surface morphology. The polymer–DNA complexes were prepared at Z+/– = 1 and kept at two different temperatures 25 and 40 °C for 30 min, and then, a drop of the sample solution was allowed to settle on the mica for 5 min. The samples were air-dried overnight at respective temperatures. The images were further autoflattened and analyzed using Agilent PicoView software.