Thermosensitive polymers as carriers for DNA delivery

Journal of Controlled Release 60 (1999) 249–259

Thermosensitive polymers as carriers for DNA delivery W.L.J. Hinrichs 1 , N.M.E. Schuurmans-Nieuwenbroek, P. van de Wetering, W.E. Hennink* Department of Pharmaceutics, Utrecht Institute for Pharmaceutical Sciences, Faculty of Pharmacy, Utrecht University, P.O. Box 80.082, 3508 TB Utrecht, The Netherlands Received 24 September 1998; received in revised form 4 March 1999

Abstract Copolymers of 2-(dimethylamino)ethyl methacrylate (DMAEMA) and N-isopropylacryl amide (NIPAAm) of various monomer ratios and molecular weights were evaluated as carrier systems for DNA delivery. All copolymers, even with a low DMAEMA content of 15 mol%, were able to bind to DNA at 258C. Light-scattering measurements indicate that complexation is accompanied by precipitation of the (co)polymer in the complex caused by a drop of the lower critical solution temperature of the (co)polymer. The (co)polymer / plasmid ratio at which complexes with a size of around 200 nm were formed increased with increasing NIPAAm content of the copolymer and was independent of molecular weight of the (co)polymer. However, complexes containing (co)polymers of low molecular weight or high NIPAAm content prepared at 258C aggregated rapidly when the temperature was raised to 378C, whereas complexes containing (co)polymers of high molecular weight or lower NIPAAm content were relatively stable at 378C. The zeta potential of the complexes was also independent of molecular weight of the (co)polymer and increased with increasing (co)polymer / plasmid ratio until a plateau value was reached. The (co)polymer / plasmid ratio at which this plateau was reached increased with increasing NIPAAm content. The plateau values decreased from around 26 mV to around 13 mV when the NIPAAm content of the copolymer was increased from 0 to 85 mol%. The cytotoxicity of the complexes strongly decreased with increasing NIPAAm content and was independent of molecular weight of the (co)polymer. The transfection efficiency of complexes with poor stability was in general much lower than that of complexes with good stability. The transfection efficiency as a function of the (co)polymer / plasmid ratio showed a bell-shaped curve. The (co)polymer / plasmid ratio at which the transfection efficiency was maximal increased with increasing NIPAAm content, while the maximum transfection efficiency strongly decreased with increasing NIPAAm content of the copolymer. The results of this study show that the formation of stable (co)polymer / plasmid complexes with a size of around 200 nm is a prerequisite for efficient transfection. Furthermore, the transfection efficiency and cytotoxicity strongly decreased with decreasing zeta potential. Therefore, besides the size, the zeta potential can also be used as a characteristic to predict the behavior of this type of (co)polymer / plasmid complexes in transfection. Copolymers of DMAEMA and NIPAAm provided with a homing device may be interesting carrier systems for gene targeting because these copolymers can condense DNA to small particles, and the resulting complexes show a low cytotoxicity and aspecific transfection.  1999 Elsevier Science B.V. All rights reserved. Keywords: Cationic polymers; DNA; Thermosensitive polymers; Transfection *Corresponding author. Tel.: 131-302536964; fax: 131-302517839. E-mail address: [email protected] (W.E. Hennink) 1 Present address: Department of Pharmaceutical Technology and Biopharmacy, Groningen Institute for Drug Studies, Groningen University, Antonius Deusinglaan 1, 9713 AV Groningen, The Netherlands. 0168-3659 / 99 / $ – see front matter  1999 Elsevier Science B.V. All rights reserved. PII: S0168-3659( 99 )00075-9

250

W.L. J. Hinrichs et al. / Journal of Controlled Release 60 (1999) 249 – 259

1. Introduction In gene therapy, DNA is considered as a prodrug for the synthesis of a therapeutic protein in cells [1,2]. To act as a prodrug, DNA has to be introduced into the nucleus of target cells. For an efficient DNA delivery, a suitable carrier is required. Distinction can be made between viral and nonviral carriers [3–5]. Because the application of viral carriers comprises several disadvantages such as undesired immune response, limited size of DNA, and upscaling problems, the development of nonviral carriers has gained increasing attention over the last few years. One class of nonviral carriers consists of watersoluble cationic polymers [6–11]. Within our laboratories, the application of a synthetic cationic polymer, poly(2-(dimethylamino)ethyl methacrylate) (polyDMAEMA), as a gene carrier is currently under investigation [12–16]. In a comparative study, it was shown that the in vitro transfection of OVCAR-3 and COS-7 cells with polyDMAEMA / pCMV–LacZ plasmid complexes is more efficient than with complexes of plasmid with well known carriers such as poly( L-lysine) and DEAE-dextran [15]. In another study, copolymers of DMAEMA and ethoxytriethylene glycol methacrylate or N-vinyl-pyrrolidone were evaluated as transfection agents [14,16]. The results indicate that both size and zeta potential of (co)polymer / plasmid complexes are important parameters determining the transfection efficiency and cytotoxicity. In this paper, copolymers of DMAEMA and Nisopropylacryl amide (NIPAAm) were evaluated as transfection agents. PolyNIPAAm is an uncharged thermosensitive polymer which is highly water soluble at low temperatures but precipitates when the temperature is raised above 31–328C [17]. This transition temperature is referred to as the lower critical solution temperature (LCST). The LCST of copolymers containing NIPAAm depends on the nature of the comonomer, copolymer composition and copolymer architecture [17–22]. The aim of the present study was to investigate whether (co)polymers of DMAEMA and NIPAAm have the potential to be used in nonviral gene delivery. It is expected that changing DMAEMA / NIPAAm ratio of the copolymer will affect the complex size. Since a thermosensitive component is incorporated, the com-

plex size may be temperature dependent. Furthermore, changing DMAEMA / NIPAAm ratio of the copolymer will affect the complex charge. It is expected that changing complex size and / or charge will affect the transfection efficiency and cytotoxicity.

2. Materials and methods

2.1. Materials Ammonium peroxodisulfate (APS), dextran standards, DMAEMA, and 2,29-azobisisobutyronitril (AIBN) were purchased from Fluka (Zwijndrecht, The Netherlands). Amphotericin B, plain Dulbecco’s modified Eagles medium (DMEM), L-glutamine, penicillin, RPMI 1640 medium, streptomycin, and 5-bromo-4-chloro-3-indolyl-b-D-galactopyranoside (X-Gal) were obtained from Gibco (Breda, The Netherlands). Sodium 39-[1-(phenylaminocarbonyl)3,4-tetrazolium]-bis(4-methoxy-6-nitro)benzene sulfonic acid (XTT) was obtained from Sigma (Zwijndrecht). NIPAAm was purchased from Aldrich (Zwijndrecht). Fetal calf serum (FCS) was obtained from Integro (Zaandam, The Netherlands). All other chemicals were of analytical grade. All compounds were used without purification except when mentioned.

2.2. ( Co)polymer synthesis and characterization Copolymers of DMAEMA and NIPAAm of various monomer ratios were synthesized by free radical polymerization. Appropriate amounts of freshly distilled DMAEMA and NIPAAm were dissolved in either 1,4-dioxane, benzene, toluene, or water (adjusted to pH 5 with HCl) to obtain solutions of 10–20% (w / v). The polymerizations in the organic solvents and water were initiated by adding AIBN and APS, respectively. The monomer / initiator molar ratio was in all cases 100 / 1. The reactions were carried out at 608C under a blanket of oxygen-free nitrogen for 16–22 h. Thereafter, the solvents of the polymerizations in 1,4-dioxane and benzene were evaporated after which the crude products were dissolved in acetone. Subsequently, these products and the products of the reactions in toluene were purified by precipitation in a large excess of diethyl-

W.L. J. Hinrichs et al. / Journal of Controlled Release 60 (1999) 249 – 259

ether or petroleum ether and then dried in vacuo. The reaction mixtures of the polymerizations in water were dialyzed against 0.1 M acetic acid for 1 day and thereafter for 6 days against reversed osmosis water. The products were collected by lyophilization. The composition of the copolymers was determined by 1 H NMR (Gemini 300-MHz spectrometer, Varian Associates NMR instruments, Palo Alto, CA, USA). The molecular weight was determined by means of gel permeation chromatography (GPC). GPC measurements were performed using two columns in series (Shodex Ollpak KB-802 and KB-80, Showa Denko, Japan). The columns were calibrated with dextran standards. An aqueous solution of 0.7 M NaNO 3 and 0.1 M tris(hydroxymethyl) aminomethane of pH 7.2 was used as eluents. The LCST of the (co)polymers was determined by static light scattering (SLS). Therefore, the temperature of 0.33-mg / ml solutions of the (co)polymers in Hepes-buffered saline (HBS, aqueous solution of 20 mM Hepes, 150 mM NaCl adjusted to pH 7.4) was raised at a rate of 0.5–1.08C / min from 25 to 808C. Scattering of light with a wavelength of 650 nm was measured at an angle of 908 using a Luminescence Spectrometer (LS-50B, Perkin-Elmer, Nieuwekerk a / d Ijssel, The Netherlands). The LCST was defined as the temperature at which the degree of scattering started to increase.

2.3. Plasmid and cell line pCMV-LacZ containing a bacterial LacZ gene proceeded by a nuclear signal under control of a CMV promoter (size, 7.8 kB) was amplified and purified as described before [12,23]. The human ovarian cancer cell line NIH:OVCAR3 was originally obtained from Dr Hamilton (National Cancer Institute, Bethesda, MD, USA). The cells were routinely cultured in DMEM supplemented with 10% (v / v) FCS, L-glutamine (2 mM), L-glucose (4.5 g / l), penicillin (100 IU / ml), streptomycin (100 mg / ml), and amphotericin B (0.25 mg / ml) in a 5% CO 2 humidified atmosphere.

251

measurements, the stock solutions were diluted with HBS to solutions of appropriated concentrations. Subsequently, 100 ml of a 25-mg / ml solution of plasmid in HBS was added to 400 ml of (co)polymer solution and then vortexed for 5 s. In routine measurements, the Z-average particle sizes of the (co)polymer / plasmid complexes were determined after 45–75 min of complexation at 258C by means of dynamic light scattering (DLS). A Malvern 4700 system equipped with an argon laser (10 mW, 488 nm; NEC, Tokyo, Japan) and the automeasure version 1.34 PCS software (Malvern Ltd., Malvern, UK) was used. For the data analysis, the viscosity and refractive index of water were used. The instrument was calibrated with a latex of known particle size. As a measure of particle size distribution, the system reports a polydispersity index (pd). This index ranges from 0 for a monodisperse sample up to 1 for an entirely polydisperse sample. In some experiments, the complexation time and / or temperature was varied. Furthermore, in some experiments the particle size was measured at 378C. For zeta potential measurements, the (co)polymer / plasmid complexes were prepared in the same way as ascribed above, except that the stock solutions were diluted with 20 mM Hepes, pH 7.4. Furthermore, a three times higher plasmid concentration and two times larger volume were used. Thus, 200 ml of a 75-mg / ml plasmid solution was added to 800 ml of a polymer solution. The electrophoretic mobility was measured in an experimental dipping cell with a Malvern zeta-sizer 2C unit and the automeasure version 1.34 PCS software (Malvern) at a temperature of 258C. The zeta potential was calculated with the Smoluchowski equation: z 512.83me in which z is the zeta potential and me is the electrophoretic mobility. The instrument was calibrated with a dispersion of carboxyl-modified polystyrene latex with a zeta potential of 250 mV (DTS5050, Malvern). For transfection experiments, the (co)polymer / plasmid complexes were prepared in the same way as for particle size determinations except that dilutions were made with RPMI.

2.4. Preparation and characterization of (co) polymer /plasmid complexes

2.5. In vitro transfection and cytotoxicity

Copolymers were dissolved in HBS, pH 7.4 to obtain stock solutions of 1.0 mg / ml. For particle size

The transfection efficiency and cytotoxicity of the (co)polymer / plasmid complexes were determined

W.L. J. Hinrichs et al. / Journal of Controlled Release 60 (1999) 249 – 259

252

using OVCAR-3 cells as described before [12]. Briefly, cells were seeded in a 96-well plate (1310 4 per well (0.38 cm 2 )) and cultured under standard conditions. After 24 h (70–80% confluency), the cells were washed with phosphate-buffered saline and then incubated with 200 ml of a (co)polymer / plasmid complex dispersion for 1 h at 378C in a 5% CO 2 humidified atmosphere. Thereafter, the (co)polymer / plasmid complex dispersion was replaced by complete DMEM medium and the cells were cultured for an additional period of 48 h. Two series of experiments were performed simultaneously, one series for determining the transfection efficiency and the other series for determining the influence on the cell viability. The transfection efficiency was determined by means of a X-Gal assay as described before [12]. With this assay, the number of cells in which the reporter gene (b-galactosidase) is expressed can be established. The cytotoxicity was determined by measuring the number of viable cells using a XTT colorimetric assay as described before [12]. The percentage of viable cells was calculated using a calibration curve

established with wells containing known numbers of living cells (0–1.5310 5 cells per well).

3. Results

3.1. Polymer synthesis and characterization Different solvents were used to synthesize copolymers of DMAEMA and NIPAAm of various compositions and molecular weights. The characteristics are listed in Table 1 Polymerizations in 1,4-dioxane and toluene yielded copolymers of lower molecular weights than polymerizations in benzene. This can be ascribed to the high chain transfer activities of 1,4-dioxane and toluene as compared to benzene [24]. As shown before, (co)polymers of high molecular weights could be synthesized in aqueous solutions [15]. The lower critical solution temperature (LCST) of poly(DMAEMA-co-NIPAAm) was determined by means of SLS at pH 7.4 (Table 1). The LCST of poly(DMAEMA-co-NIPAAm) gradually increased with increasing DMAEMA content and was indepen-

Table 1 Overview of (co)polymers Code

DMAEMA/NIPAAM in feed (mol/mol)a

DMAEMA/NIPAAm in (co)polymer (mol/mol)b

Solvent

Yield (%)

LCST (8C)

LMW DMAEMA/NIPAAm 80/20 LMW DMAEMA/NIPAAm 30/70 LMW DMAEMA/NIPAAm 15/85 polyNIPAAm HMW DMAEMA/NIPAAm 80/20 HMW DMAEMA/NIPAAm 30/70 HMW DMAEMA/NIPAAm 15/85 (UHMW) polyDMAEMA UHMW DMAEMA/NIPAAm 80/20

80/20

75/25

Toluene

108

.80

30/70

24/76

1,4-Dioxane

82

15/85

13/87

1,4-Dioxane

0/100 80/20

– 67/33

30/70

a

Mn d (kDa)

91

9

43.560.5

195

12

85

38.360.6

202

12

1,4-Dioxane Benzene

81 36

31.860.3 .80

nd e 235

nd e 22

30/70

Benzene

89

45.761.5

380

17

15/85

14/86

Benzene

87

40.260.8

289

14

100/0 80/20

– 79/21

Water/HCl Water/HCl

85 78

309 1237

75 138

The monomer to initiator ratio was 100 / 1 (mol / mol) in all cases. Determined by 1 H NMR. c Number average molecular weight as determined by GPC. d Weight average molecular weight as determined by GPC. e Not determined. b

Mw c (kDa)

.80 .80

W.L. J. Hinrichs et al. / Journal of Controlled Release 60 (1999) 249 – 259

253

dent of the molecular weight. At a monomer ratio DMAEMA / NIPAAm of 80 / 20 and higher, no LCST could be detected. Apparently, the LCST of these (co)polymers is higher than 808C or not present. From these results, it can be concluded that under these conditions, DMAEMA acts as a hydrophilic comonomer [17–20]. This can be expected since polyDMAEMA is highly soluble at pH 7.4 and can be ascribed to the fact that polyDMAEMA has a pKa of 7.5 [25,26] which implies that at physiological pH more than halve of the tertiary amine groups of the polymer are protonated. Furthermore, the increase of the LCST with increasing the DMAEMA content of the copolymer indicates that the copolymerization of DMAEMA and NIPAAm yield random copolymers [21,22]. It has indeed been claimed that the polymerization of a related monomer system consisting of NIPAAm, 2-(diethylamino)ethyl methacrylate (DEAEMA) and n-butyl methacrylate (BMA) also proceeds randomly [18].

3.2. Size of (co)polymer /plasmid complexes The interaction between poly (DMAEMA-coNIPAAm) and plasmid was investigated by DLS and zeta potential measurements to determine the size and charge of the complexes, respectively. As found before, free plasmid or free polyDMAEMA dissolved in HBS showed no significant light scattering. However, with mixtures of plasmid and polyDMAEMA, light scattering was substantial indicating complexation [12,15]. Formation of polymer / plasmid complexes can be ascribed to ionic interactions between the positive charges of the polymer and the negatively charged phosphate groups of the plasmid. The size of the complexes depended on the polymer / plasmid ratio (see Fig. 1a). At polymer / plasmid weight ratios smaller than two, relatively large aggregates with a pd larger than 0.3 were formed, whereas at polymer / plasmid weight ratios larger than two, the particles size and pd were respectively 150–200 nm and 0.1–0.2. The formation of large aggregates at a low polyDMAEMA / plasmid ratio can be ascribed to cross-linking of plasmid by the polymer [12,15]. At higher polyDMAEMA / plasmid ratios, enough polyDMAEMA is present to maximally cover the plasmid with polymer by which crosslinking is prevented. As a

Fig. 1. (a, b) Size of poly (DMAEMA-co-NIPAAm) / plasmid complexes at a fixed plasmid concentration of 5 mg / ml and varying concentrations of (co)polymer. Particles sizes were determined at 258C (n52–5, 6S.D.).

consequence, the plasmid structure is condensed to small particles. As expected, solutions containing mixtures of polyNIPAAm and plasmid showed no substantial light scattering at a temperature below the LCST of polyNIPAAm (258C) indicating that no complexes were formed. Therefore, polyNIPAAm was excluded from further studies. With all mixtures of poly(DMAEMA-co-NIPAAm) and plasmid, however, substantial light scattering was observed at 258C indicating complexation (solutions of copolymer without plasmid showed no significant light scattering). The size of these complexes as a function of the copolymer / plasmid ratio shows the same trend as that of the polyDMAEMA / plasmid complexes: at a low copolymer / plasmid ratio, relatively large par-

254

W.L. J. Hinrichs et al. / Journal of Controlled Release 60 (1999) 249 – 259

ticles with a high pd (larger than 0.3) were formed, whereas above a certain critical copolymer / plasmid ratio, particles with a size of around 200 nm and a pd of 0.1–0.2 were detected (Fig. 1a,b). Complexes with UHMW DMAEMA / NIPAAm 80 / 20 and LMW DMAEMA / NIPAA 30 / 70 showed anomalous behavior at low copolymer / plasmid ratios as small particles were measured. However, the pds were very large (.0.4) indicating a very large size distribution. Therefore, the particles can be considered as aggregates. The critical copolymer / plasmid ratio above which particles of 200 nm were formed increased with increasing NIPAAm content of the copolymer. No major differences were observed between LMW, HMW, and UHMW copolymers of the same composition. To study the effects of temperature, the size of complexes prepared at a plasmid concentration of 5 mg / ml and a (co)polymer concentration of 40 mg / ml (80 and 160 mg / ml for LMW and HMW DMAEMA / NIPAAm 15 / 85) was measured at 25 and 378C. Raising the temperature from 25 to 378C had no effect on the complex size for each (co)polymer composition and for varying molecular weights of the (co)polymers (Table 2). Also, the stability of the complexes prepared at a plasmid concentration of 5 mg / ml and a (co)polymer concentration of 40 mg / ml (80 and 160 mg / ml for

Table 2 Effect of temperature on particle size a DMAEMA / NIPAAm (mol / mol)b

LMW

HMW

UHMW

100 / 0 c 80 / 20 c 30 / 70 c 15 / 85 d 15 / 85 e

– 1.0560.25 0.9860.10 0.8360.08 0.9560.04

– 0.9760.12 0.9460.16 1.0860.17 0.9760.06

1.0160.14 1.0160.11 – – –

a Poly(DMAEMA-co-NIPAAm) / plasmid complexes were prepared at a fixed plasmid concentration of 5 mg / ml at room temperature. After 45–75 min the size of the complexes were determined at 258C. Immediately thereafter, the temperature was raised to 378C and the size of the complexes was determined again. Given values are the size of the complexes at 378C relative to their size at 258C (n52–5, 6S.D.). b Feed ratios. c Polymer / plasmid58 (w / w). d Polymer / plasmid516 (w / w). e Polymer / plasmid532 (w / w)).

Table 3 Stability of polymer / plasmid complexes a DMAEMA / NIPAAm (mol / mol)b

LMW

HMW

UHMW

100 / 0 c 80 / 20 c 30 / 70 c 15 / 85 d 15 / 85 e

– Aggregates f Aggregates f Aggregates f Aggregates f

– 1.1060.64 g 1.1060.16 g Aggregates f Aggregates f

1.0660.14 g 1.0360.83 g – – –

a Poly(DMAEMA-co-NIPAAm) / plasmid complexes were prepared at fixed plasmid concentration of 5 mg / ml at room temperature. After 45–75 min the size of the complexes were determined at 378C. Subsequently the samples were stored at 378C. At different time intervals the size of the complexes were determined again at 378C. b Feed ratios. c Polymer / plasmid58 (w / w) d Polymer / plasmid516 (w / w) e Polymer / plasmid532 (w / w). f Within a few hours. g Given values are the size of the complexes after 24 h storage at 378C relative to their size after 45–75 min of complexation (n52–5, 6S.D.).

LMW and HMW DMAEMA / NIPAAm 15 / 85) was evaluated. After complex formation, the complexes were incubated at 378C and their size was measured at different time intervals (Table 3). It was found that complexes with LMW copolymers and HMW DMAEMA / NIPAAm 15 / 85 were highly unstable. Within a few hours large aggregates were formed. In contrast, complexes with all other (co)polymers were stable: no increase in size was observed within 24 h.

3.3. Zeta potential of (co)polymer /plasmid complexes The zeta potential of free plasmid was found to be around 220 mV. As found before, with increasing polyDMAEMA / plasmid ratios the zeta potential gradually increased until a plateau value of around 26 mV was reached at a polymer / plasmid weight ratio of 2–4 [12] (see Fig. 2a). The increased zeta potential with increased polymer / plasmid ratios can be ascribed to a gradual covering of the negatively charged plasmid by the positively charged polymer. Above a certain copolymer / plasmid ratio, the zeta potential remains the same. This indicates that the

W.L. J. Hinrichs et al. / Journal of Controlled Release 60 (1999) 249 – 259

255

Fig. 3. Maximum zeta potential of poly (DMAEMA-coNIPAAm) / plasmid complexes as a function of the NIPAAm contents of the (co)polymer as determined with 1 H NMR (n52–3, 6S.D.).

gradually decreased with increased amounts of NIPAAm incorporated in the copolymer (see Fig. 3).

3.4. Transfection and toxicity of the (co)polymer / plasmid complexes.

Fig. 2. (a, b) Zeta potentials of poly(DMAEMA-co-NIPAAm) / plasmid complexes at a fixed plasmid concentration of 15 mg / ml and varying concentrations of (co)polymer. Zeta potentials were determined at 258C (n52–3, 6S.D.).

extra amount of polymer will be present as free polymer [12]. Like the zeta potential of complexes containing polyDMAEMA, the zeta potentials of complexes containing poly(DMAEMA-co-NIPAAm) increased with increasing copolymer / plasmid ratios until a maximum value is reached after which it remained constant (see Fig. 2a, b). The zeta potential was independent of the molecular weight of the copolymers. Incorporation of NIPAAm in the copolymer had two distinct effects. Firstly, the copolymer / plasmid ratio at which the plateau value was reached increased with increased NIPAAm content of the copolymer. Secondly, the maximum zeta potential

To evaluate copolymers of DMAEMA and NIPAAm of varying monomer ratios as gene carriers, transfection experiments were carried out at a fixed plasmid concentration and varying concentrations of the (co)polymers (see Figs. 4a, b and 5a, b). PolyDMAEMA was used as a control. As found before with polyDMAEMA, the number of transfected cells increased with increasing polymer concentration until a maximum was reached at a polymer / plasmid ratio of 2–4 after which it decreased [12,15]. At maximum, around 10% of the cells were transfected. The low transfection efficiency at polymer / plasmid ratios smaller than 2 can be related to the size of the complexes. Apparently, the complexes are too large to be taken up by the cells. At polymer / plasmid ratios higher than 4 the transfection efficiency decreased with increasing polymer concentrations despite the fact that the physicochemical characteristics remain the same. This decrease can be ascribed to the increased cytotoxicity caused by the presence of increasing amounts of free polymer [12,15]. The behavior of complexes containing HMW and UHMW DMAEMA / NIPAAm 80 / 20 with respect to

256

W.L. J. Hinrichs et al. / Journal of Controlled Release 60 (1999) 249 – 259

Fig. 4. (a, b) Transfection efficiency of poly(DMAEMA-coNIPAAm) / plasmid complexes at a fixed plasmid concentration of 5 mg / ml and varying concentrations of (co)polymer (n53–4, 6S.D.).

transfection was similar to that of polyDMAEMA / plasmid complexes. However, the maximum number of transfected cells was about 30% of that obtained with polyDMAEMA. Furthermore, transfection efficiency for these two copolymers was maximal at a higher polymer / plasmid ratios than for polyDMAEMA. LMW DMAEMA / NIPAAm 80 / 20 was found to be a poor transfection agent. The maximum number of transfected cells was around 10% as compared to polyDMAEMA. The cytotoxicity of the three DMAEMA / NIPAAm 80 / 20 copolymers was about the same and slightly less than polyDMAEMA. The performance of LMW and HMW DMAEMA / NIPAAm 30 / 70 to act as a transfection agent was

Fig. 5. (a, b) Cytotoxicity of poly(DMAEMA-co-NIPAAm) / plasmid complexes at a fixed plasmid concentration of 5 mg / ml and varying concentrations of (co)polymer (n53–4, 6S.D.).

poor as compared to polyDMAEMA: the maximum number of transfected cells was about 10% using both copolymers as compared to polyDMAEMA. Furthermore, the maximum number of transfected cells was reached at a higher copolymer / plasmid ratio in both cases than with (co)polymers of higher DMAEMA content. Like the (co)polymers of higher DMAEMA content, the transfection efficiency increased with increasing copolymer / plasmid ratio until a maximum is reached. However, a decrease of transfection efficiency at further increasing of the copolymer / plasmid ratio was not observed. The cytotoxicity of both copolymers was about the same and substantially less than (co)polymers of higher DMAEMA content. HMW DMAEMA / NIPAAm 15 / 85 showed very limited capacities to act as a

W.L. J. Hinrichs et al. / Journal of Controlled Release 60 (1999) 249 – 259

transfection agent. Furthermore, the copolymer displayed no significant cytotoxicity.

4. Discussion In this study, random copolymers of DMAEMA and NIPAAm were evaluated as carriers for DNA delivery. All copolymers, even with a low DMAEMA content of 15 mol%, were able to bind to DNA. Above a critical (co)polymer / plasmid ratio, all complexes had a size of around 200 nm (Fig. 1a, b). Furthermore, above this critical (co)polymer / plasmid ratio, the zeta potentials of the complexes had constant values (Fig. 2a, b). The critical ratios increased with increasing NIPAAm content of the copolymer. Since at these ratios the plasmid is maximally covered with copolymer, it can be concluded that with increasing NIPAAm content of the copolymer, more copolymer is able to bind to the plasmid. This increase can be explained by the fact that with increasing amounts of NIPAAm incorporated in the copolymer, the charge density of the copolymer decreases. The maximum zeta potential of the (co)polymer / plasmid complexes decreased with increasing NIPAAm content of the copolymer (Fig. 3). This decrease indicates that, although the maximum amount of copolymer which can complex with plasmid increased with increasing amounts of NIPAAm incorporated in the copolymer, the maximum number of DMAEMA groups which can complex with plasmid decreased. All copolymers had a LCST higher than 378C (Table 1). Also polyDMAEMA exhibits LCST behavior which is pH dependent [26–28]. At low pH, polyDMAEMA has a good water solubility. On the other hand, at high pH, polyDMAEMA is insoluble because the tertiary amine groups are deprotonated. Feil found that the LCST of a positively charged copolymer composed of random tercopolymers of NIPAAm, DEAEMA and BMA strongly decreased upon addition of negatively charged dextran sulfate or poly(styrene sulfonate) [18]. This decrease was associated with the shielding of the positive charges of DEAEMA by the polyanion by which the copolymer is rendered hydrophobic. Due to the similarity of the systems, this mechanism will also be

257

applicable to the complexation of poly(DMAEMAco-NIPAAm) with plasmid. In all cases DLS measurements showed substantial light scattering at 258C indicating a decrease of the LCST to below 258C. In other words, complexation of (co)polymers and plasmid can be regarded as precipitation of the polymer caused by a substantial drop of the LCST. Consequently, when the temperature is raised from 25 to 378C, no LCST is passed. Therefore, upon this temperature shift, no change in particle size is observed (Table 2). At 378C, complexes containing the low molecular weight copolymers and / or HMW DMAEMA / NIPAAm 15 / 85 showed a poor stability (Table 3). Copolymer chains consisting of low molecular weight poly(DMAEMA-co-NIPAAm) and / or containing high amounts of NIPAAm contain relatively few positive charges which can interact with plasmid. Because of these few interactions, it is likely that copolymer chains which are complexed with plasmid may be easily exchanged by free copolymer or copolymer complexed with another plasmid molecule. These processes may lead to aggregation. Apparently, poly(DMAEMA-co-NIPAAm) of HMW or UHMW and low NIPAAm content contain sufficient positive charges to prevent these exchange processes by which stable complexes are formed. Transfection experiments showed that HMW DMAEMA / NIPAAm 15 / 85 and all LMW copolymers are poor transfection agents (Fig. 4a, b). This can be ascribed to the poor stability of the complexes since aggregation will prevent cell entrance. The results with the three DMAEMA / NIPAAm 80 / 20 copolymers show that molecular weight does not affect the transfection efficiency as long as the complexes are stable. Van de Wetering et al. [14,16] who studied the application of polyDMAEMA and copolymers of DMAEMA and ethoxytriethylene glycol methacrylate (triEGMA) as transfection agents, found that the transfection efficiency increased with increasing molecular weight of the (co)polymer. However, the stability of the (co)polymer / plasmid complexes was not studied. Furthermore, it was found that in contrast to the present study, the complex size changed with varying the molecular weight of the (co)polymer. In the studies mentioned above as well as in the present study, the (co)polymer / plasmid ratio at

258

W.L. J. Hinrichs et al. / Journal of Controlled Release 60 (1999) 249 – 259

which the transfection efficiency was maximal increased with increasing comonomer (triEGMA or NIPAAm) content of the copolymer. This trend as found for poly(DMAEMA-co-NIPAAm) can be related to the size of the complexes as the (co)polymer / plasmid ratio at which the plasmid condensed to particles of around 200 nm in size also increased with increasing NIPAAm content of the copolymer. Therefore, it can be concluded that the formation of stable (co)polymer / plasmid complexes with a size of around 200 nm is a prerequisite for efficient transfection. At higher (co)polymer / plasmid ratios, the transfection efficiency decreased due to the cytotoxicity caused by increasing amounts of free (co)polymer. Therefore, a maximum transfection efficiency will be achieved when the plasmid is maximally covered with (co)polymer while the amount of free (co)polymer is minimal. As shown with particle size and zeta potential measurements, this maximum amount increases with NIPAAm contents of the copolymer and correlates well with the results of the transfection experiments. The results clearly indicate that besides particle size, the zeta potential plays an important role in transfection. With decreasing zeta potential both the transfection efficiency and cytotoxicity strongly decreased (Figs. 4a, b and 5a, b). A decrease of cytotoxicity with a decrease in DMAEMA content of the copolymer can be expected. Inherently to cationic polymers polyDMAEMA is cytotoxic [29,30]. Because with increasing NIPAAm content of the copolymer, the charge density decreases, the zeta potential and therefore also the cytotoxicity decreases. As can be seen from Fig. 5a, b, the decrease in cytotoxicity is even more than what can be expected based on the decrease of the number of DMAEMA groups. Somehow, NIPAAm seems to mask the cytotoxicity of the DMAEMA groups. Similar results were found with copolymers of DMAEMA and triEGMA or N-vinylpyrrolidone [14,16]. It can only be speculated why the zeta potential plays an important role in the transfection efficiency as not much is known about the complicated process of transfection. It has been proposed that the positively charged (co)polymer / plasmid complexes electrostatically interact with the negatively charged cell membrane after which the complexes are taken up by endocytosis [2,6,9]. With increasing NIPAAm

content, the zeta potential of the complexes decreases which will be associated with a less strong interaction with cells. This may result in less uptake of the complexes and therefore in a decreased transfection efficiency. Because aspecific uptake is reduced, poly(DMAEMA-co-NIPAAm) may be interesting for the design of targeted systems. For example, HMW DMAEMA / NIPAAm 30 / 70 / plasmid complexes showed low cytotoxicity and a poor transfection efficiency. However, as these complexes are stable and have a size of around 200 nm, they have the potential to be taken up by cells. When provided with a homing device, these complexes will show limited aspecific transfection and cytotoxicity, but substantial transfection of cells bearing receptors for the homing device is expected to occur.

5. Conclusions In this study copolymers of DMAEMA and NIPAAm were evaluated as carrier systems for DNA delivery. It is shown that the formation of stable (co)polymer / plasmid complexes with a size of around 200 nm is a prerequisite for efficient transfection. Furthermore, it is shown that with decreasing zeta potential of the (co)polymer / plasmid complexes, both the transfection efficiency and cytotoxicity is decreased. Maximum transfection efficiencies were found when the plasmid is maximally covered with (co)polymer while the amount of free (co)polymer is minimal. Although the transfection efficiency of complexes with copolymers of DMAEMA and NIPAAm is reduced compared to those with polyDMAEMA, these copolymers may be interesting for gene targeting.

Acknowledgements The authors like to thank M.J. van Steenbergen (Department of Pharmaceutics, Utrecht Institute for Pharmaceutical Sciences, Faculty of Pharmacy, Utrecht University) for his technical assistance.

References [1] C.R. Middaugh, R.K. Evans, D.L. Montgomery, D.R. Casi-

W.L. J. Hinrichs et al. / Journal of Controlled Release 60 (1999) 249 – 259

[2] [3]

[4] [5]

[6]

[7]

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

[16]

miro, Analysis of plasmid DNA from a pharmaceutical perspective, J. Pharm. Sci. 87 (1998) 130–145. F.D. Ledley, Pharmaceutical approach to somatic gene therapy, Pharm. Res. 13 (1996) 1595–1614. P. Lehn, S. Fabrega, N. Oudrhiri, J. Navarro, Gene delivery systems: bridging the gap between recombinant viruses and artificial vectors, Adv. Drug. Delivery Rev. 30 (1998) 5–11. D.D. Lasic, N.S. Templeton, Liposomes in gene therapy, Adv. Drug Delivery Rev. 20 (1996) 221–266. J. Smith, Y. Zhang, R. Niven, Towards development of a non-viral gene therapeutic, Adv. Drug Delivery Rev. 26 (1997) 135–150. W. Zauner, M. Ogris, E. Wagner, Polylysine-based transfection systems utilizing receptor-mediated delivery, Adv. Drug Delivery Rev. 30 (1998) 97–113. M. Hashida, S. Takemura, M. Nishikawa, Y. Takakura, Targeted delivery of plasmid DNA complexed with galactosylated poly( L-lysine), J. Control. Release 53 (1998) 301– 310. P. Erbacher, A.C. Roche, M. Monsigny, P. Midoux, The reduction of positive charges of polylysine by partial gluconoylation increases the transfection efficiency of polylysine / DNA complexes, Biochim. Biophys. Acta 1324 (1997) 27–36. J.-S. Remy, B. Abdallah, M.A. Zanta, O. Boussif, J.-P. Behr, B. Demenaix, Gene transfer with lipospermines and polyethylenimines, Adv. Drug Delivery Rev. 30 (1998) 85–95. V. Toncheva, M. Wolfert, P.R. Dash, D. Oupicky, K. Ulbrich, L.W. Seymour, E.H. Schacht, Novel vectors for gene delivery formed by self-assembly of DNA with poly( L-lysine) grafted with hydrophilic polymers, Biochim. Biophys. Acta 1380 (1998) 354–368. O. Boussif, F. Lezoualch, M.A. Zanta, M.D. Mergny, D. Scherman, B. Demeneix, J.-P. Behr, A versatile vector for gene and oligonucleotide transfer into cells in culture and in vivo, Proc. Natl. Acad. Sci. USA 92 (1995) 7297–7301. J.-Y. Cherng, P.V.d. Wetering, H. Talsma, D.J.A. Crommelin, W.E. Hennink, Effect of size and serum protein on transfection efficiency of poly ((2-dimethylamino)ethyl methacrylate)-plasmid nanoparticles, Pharm. Res. 13 (1996) 1038– 1042. P. Van de Wetering, E.E. Moret, N.M. E. Nieuwenbroek, M.J. van Steenbergen, W.E. Hennink, Structure–activity relationship of water-soluble cationic methacrylate / methacrylamide polymers as non-viral gene delivery agents, Gene Ther. (submitted). P. Van de Wetering, N.M. E. Nieuwenbroek, M.J. van Steenbergen, W.E. Hennink, Copolymers of 2-dimethylamino)ethyl methacrylate with ethoxytriethylene glycol methacrylate or N-vinyl-pyrrolidone as gene transfer agents, J. Control. Release (submitted). P. Van de Wetering, J.-Y. Cherng, H. Talsma, W.E. Hennink, Relation between transfection efficiency and cytotoxicity of poly(2-(dimethylamino)ethyl methacrylate) / plasmid complexes, J. Control. Release 49 (1997) 59–69. P. Van de Wetering, J.-Y. Cherng, H. Talsma, W.E. Hennink,

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29]

[30]

259

2-(Dimethylamino)ethyl methacrylate based (co)polymers as gene transfer agents, J. Control. Release 53 (1998) 145–153. J.H. Priest, S.L. Murray, R.J. Nelson, A.S. Hoffman, Lower critical solution temperatures of aqueous copolymers of Nisopropylacrylamide and other N-substituted acrylamides, in: P.S. Russo (Ed.), Reversible Polymeric Gels and Related Systems, Am. Chem. Soc, Washington, DC, 1987, pp. 254– 264. H. Feil, Thermosensitive polymers and hydrogels based on N-isopropylacrylamide, Ph.D. thesis, University of Twente, The Netherlands (1994). Y.H. Bae, T. Okano, R. Hsu, S.W. Kim, Thermo-sensitive polymers as on-off switches for drug release, Makromol. Chem. Rapid Comm. 8 (1987) 481–485. Y.G. Takei, T. Aoki, K. Sanui, N. Ogata, T. Okano, Y. Sakurai, Temperature-responsive bioconjugates 2. Molecular design for temperature-modulated bioseparations, Bioconj. Chem. 4 (1993) 341–346. S.K. Vakkalanka, N. Peppas, Swelling behavior of temperature- and pH-sensitive block terpolymers for drug delivery, Polym. Bull. 36 (1996) 221–225. G. Chen, A.S. Hoffman, Graft polymers that exhibit temperature-induced phase transitions over a wide range of pH, Nature 373 (1995) 49–52. A. Bout, D. Valerio, B.J. Scholte, In vivo transfer and expression of the LacZ gene in the mouse lung, Exp. Lung Res. 19 (1993) 193–202. C.-A. Cole, J. S.M.Schreiner, H. Priest, N. Monji, A.S. Hoffman, N-Isopropylacrylamide and N-acryloxysuccinimide copolymer, in: P.S. Russo (Ed.), Reversible Polymeric Gels and Related Systems, Am. Chem. Soc, Washington, DC, 1987, pp. 245–254. P. Van de Wetering, N.J. Zuidam, M.J. van Steenbergen, O.A. G.J. van de Houwen, W.J. M. Underberg, W.E. Hennink, A mechanistic study of the hydrolytic stability of poly(2-(dimethylamino)ethyl methacrylate), Macromolecules (submitted). S. Wen, W.T.K. Stevenson, Synthetic sensitive polyampholyte hydrogels: a preliminary study, Colloid Polym. Sci. 271 (1993) 38–49. S.H. Yuk, S.H. Cho, S.H. Lee, pH / temperature-responsive polymer composed of poly((N,N-dimethylamino)ethyl methacrylate-co-ethylacrylamide), Macromolecules 30 (1997) 6856–6859. S.H. Cho, M.S. Jhon, S.H. Yuk, H.B. Lee, Temperatureinduced phase transition of poly(N,N-dimethylaminoethyl methacrylate-co-acrylamide), J. Polym. Sci. Part B 35 (1997) 595–598. S. Choksakulnimitr, S. Masuda, H. Tokuda, Y. Takakura, M. Hashida, In vitro cytotoxicity of macromolecules in different cell culture systems, J. Control. Release 34 (1995) 233–241. D.M.L. Morgan, J. Clover, J.D. Pearsson, Effects of polycations on leucine incorporation, lactate dehydrogenase release, and morphology of human umbilical vein endothelial cells, J. Cell Sci. 91 (1988) 231–238.