ARTICLE IN PRESS

Biomaterials 25 (2004) 17711777

Preparation and characterization of cationic PLGA nanospheres as DNA carriers

M.N.V. Ravi Kumar*,1, U. Bakowsky, C.M. Lehr
Department of Biopharmaceutics and Pharmaceutical Technology, Saarland University, Saarbrucken, D 66123, Germany . Received 11 May 2003; accepted 11 August 2003

Abstract Nanoparticles formulated from biodegradable polymers such as poly(lactic acid) (PLA) and poly(lactide-co-glycolide) (PLGA) are being extensively investigated as non-viral gene delivery systems due to their controlled release characteristics and biocompatibility. PLGA nanoparticles for DNA delivery are mainly formulated by an emulsion-solvent evaporation technique using PVA as a stabilizer generating negatively charged particles and heterogeneous size distribution. The objective of the present study was to formulate cationically modied PLGA nanoparticles with dened size and shape that can efciently bind DNA. An Emulsion-diffusion-evaporation technique to make cationic nanospheres composed of biodegradable and biocompatible copolyester PLGA has been developed. PVA-chitosan blend was used to stabilize the PLGA nanospheres. The nanospheres were characterized by atomic force microscopy (AFM), photon-correlation spectroscopy (PCS), and Fourier transform infrared spectroscopy (FTIR). Zeta potential and gel electrophoresis studies were also performed to understand the surface properties of nanospheres and their ability to condense negatively charged DNA. The designed nanospheres have a zeta potential of 10 mV at pH 7.4 and size under 200 nm. From the gel electrophoresis studies we found that the charge on the nanospheres is sufcient to efciently bind the negatively charged DNA electrostatically. These cationic PLGA nanospheres could serve as potential alternatives of the existing negatively charged nanoparticles. r 2003 Elsevier Ltd. All rights reserved.
Keywords: Biodegradable; Chitosan; Gene therapy; Nanoparticles; PLGA

1. Introduction Biodegradable colloidal particles have received considerable attention as a possible means of delivering drugs and genes by several routes of administration. Special interest has been focused on the use of particles prepared from polyesters like PLGA, due to their biocompatibility and to their resorbability through natural pathways [1]. Various methods have been reported for making nanoparticles viz., emulsion-evaporation [2], salting-out technique [3], nanoprecipitation [4], cross-ow ltration [5] or emulsion-diffusion technique [6,7]. Indeed PLGA particles are extensively investigated for drug [810] and gene delivery [11,12], but still improvements in the existing methods are
*Corresponding author. Tel.: +91-172-214683 ext 2057; fax: +91172-214692. E-mail address: mnvrkumar@yahoo.com (M.N.V. Ravi Kumar). 1 Present address: Department of Pharmaceutics, National Institute of Pharmaceutical Education and Research (NIPER), SAS Nagar, Sector 67, Punjab 160062, India. 0142-9612/$ - see front matter r 2003 Elsevier Ltd. All rights reserved. doi:10.1016/j.biomaterials.2003.08.069

needed to overcome the difculties in terms of reproducibility, size, and shape. The size and shape of the colloidal particles are inuenced by the stabilizer and the solvent used. Most investigated stabilizers for PLGA lead to negatively charged particles and the plasmid incorporation is achieved via double emulsion technique during particle preparation. This could pose problems in the stability and biological activity of the plasmid due to the involvement of organic solvents during the preparation process. This can be overcome by using cationically modied particles that can bind and condense negatively charged plasmids by simply adding nanoparticles to plasmid or vice versa. Literature suggests PVA as most popular stabilizer for the production of PLGA nanoparticles leading to negatively charged particles, nevertheless, investigations have been carried out using other stabilizers as well [13]. Vila et al. investigated double emulsion technique for making PLGA-lecithin nanoparticles for protein delivery using PVA-chitosan blend as coating material [14]. The particle size and charge reported were 500729 nm and 21.871.1 mV

ARTICLE IN PRESS
1772 M.N.V. Ravi Kumar et al. / Biomaterials 25 (2004) 17711777

respectively [14]. However, sole cationic PLGA nanospheres with chitosan as a modier for gene delivery can hardly be seen in the literature. Therefore, attempts were made to develop a technique that produces uniform and much smaller nanospheres with cationic surface modication, which can readily bind and condense DNA. Chitosan was selected in these studies because, other than its cationic charge it has been recognized for its mucoadhesivity, biodegradability and ability to enhance the penetration of large molecules across mucosal surfaces [15]. To our knowledge, not many studies were reported describing such a well dened shape and size (o200 nm) of the nanospheres, particularly when PLGA and high molecular weight polymers like chitosan were used.

manufacturers instructions (QIAGEN, Hilden, Germany). The DNA was precipitated in 70% ethanol and reconstituted in water to 1 mg/ml. All other chemicals and reagents used in this study were from AldrichSigma, Germany. 2.2. Preparation of PLGA nanospheres Nanospheres were prepared by a new emulsiondiffusion-evaporation technique as shown in the Fig. 1. The methodology in brief goes as follows: 200 mg of PLGA is dissolved in 10 ml ethyl acetate at room temperature. The organic phase is then added to an aqueous stabilizer mixture containing 100 mg of PVA and 30 mg of chitosan in 10 ml water under stirring. The emulsion is stirred at room temperature for 3 h before homogenizing at 13,500 rpm for 10 min using an UltraTurrax T25 homogenizer (Janke and Kunkel GmbH KG, Staufen, Germany). To this emulsion water is added under stirring resulting in nano-precipitation. Stirring is continued on a water bath maintained at 40 C to remove organic solvent. 2.3. FTIR spectroscopy To assess the modication of the polymer surfaces an FT-IR (ATR) spectrometer (Perkin Elmer system 2000) was used. For the measurements, the particles in solution were spread directly onto the ATR crystal

2. Materials and methods 2.1. Materials Poly(l-lactide-co-glycolide) (PLGA) (70:30 lactide: glycolide) and Poly(vinyl alcohol) were obtained from Polysciences, Inc. and MoWiol, Germany, respectively. Chitosan Hydrochloride (Seacure 210, 83% deacetylated) was obtained from Pronova Biopolymer, Norway. The b-galactosidase expression plasmid pCMVb was purchased from ATCC (Manassas, VA, USA) and transformed into E.coli DH5a: Gigaprep from 2500 ml of over-night culture was performed according to the

Ethyl acetate + PLGA

2 hours Stirring 1000 rpm Passed through 0.2 m filter

3 hours Stirring 1000 rpm

Add organic to aqueous 2 hours Stirring 1000 rpm Passed through 0.2 m filter

Mixing NH2 NH2

NH2 NH2 NH2 NH2

NH2 NH2 NH2

Aqueous PVA-Chitosan
NH2 NH2 NH2 NH2 NH2 NH2 NH2 NH2 NH2 NH2 NH2 NH2 NH2 NH2 NH2 NH2 NH2 NH2 NH2 NH2 NH2 NH2 NH2

NH2 NH2

NH2

NH2 NH2 NH2

NH2

NH2 NH2

NH2

Water

NH2

NH2 NH2

Stirring 2 hours Water bath, 40 oC

NH2 NH2 NH NH2 2 Homogenize NH2 NH2 NH2NH2 NH210 min 13,500 NH2 NH2 NH2 NH2 NH2 NH2 NH2 NH2 NH2 NH2

rpm

Fig. 1. Schematic representation of PLGA nanospheres preparation process.

ARTICLE IN PRESS
M.N.V. Ravi Kumar et al. / Biomaterials 25 (2004) 17711777 1773

(ATMS 45 , 7 cm in longitude). The water was evaporated by a nitrogen stream. The spectrum was collected in a range between 4500 and 850 cm1 with a resolution of 1 cm1 (100 scans per sample). 2.4. Photon correlation spectroscopy Particle size was determined by photon correlation spectroscopy (PCS) on an ALV 5000 (Laser Vertriebsgesellschaft mbH, Langen, Germany) at a scattering angle of 90 (sampling time 200 s). Autocorrelation was performed using the contin method. For PCS measurements, all samples were diluted 50 fold in demineralized water, resulting in comparable viscosities. Hence, no corrections for the effect of the additives were necessary. 2.5. Zeta potential measurements Surface charge of nanoparticles was determined by zeta potential measurement on a Malvern Zetasizer 2000 HS (Malvern, UK) with a ow measurement cell connected to a Mettler DL 25 (Mettler-Toledo, Giessen, Germany) auto-titrator via a circulating system. Within the 250 ml sample container at the titrator, 510 ml of nanoparticle samples were diluted with demineralized water to a nal volume of 200 ml. The pH was adjusted to 3 by using HCl (1 n ) before titration to pH 10 with NaOH (0.1 n ). Measurements of the zeta-potential were carried out at 0.5 pH increments at 25 C. The instrument was calibrated routinely with a 50 mV latex standard. 2.6. Gel electrophoresis and determination of unbound DNA NanoparticleDNA complexes were prepared by mixing the nanoparticles with plasmid at a concentration of 10 mg/ml in 25 mm Hepes (pH 7.4) as well as in deionized water (pH 6.0). The complex formation studies were performed at room temperature and allowed to stand for 15 min to attain complexes. The nanoparticleDNA complexes were electrophored on an agarose gel (1% ethidium bromide included for visualization) for 90 min at 5 V/cm. Images were acquired using a Geldoc 2000 gel documentation system (Bio-Rad, Munich, Germany) equipped with a UV transluminator. Molecular Analyst, version 1.1 software (Bio-Rad) was used for band integration and background correction. 2.7. Atomic force microscopy The size and surface morphology of the PLGA particles was analyzed by atomic force microscopy (AFM) Nanoscope IV Bioscopet (Digital Instruments, Veeco) in tapping mode using a Si3N4 cantilever with a

spring constant of about 34 N/m and a resonance frequency of about 200 kHz. Scanning was performed at a scan speed of 0.5 Hz with a resolution of 512 512 pixels. The tip loading force was minimized to avoid structural changes of the sample.

3. Results and discussions 3.1. Nanospheres formation The routine emulsion-solvent evaporation technique being used for formulating PLGA nanoparticles is believed to produce heterogeneous size distribution [16]. Various formulation factors and characteristics of the nanoparticles have a key role to play in biological applications. The foremost factor that could have an inuence on the transfection and cellular uptake is the size of the nanoparticles. Prabha et al. [16] have studied the size-dependency of nanoparticle-mediated gene transfection with fractionated nanoparticles. Recent reports suggests that a fraction of the stabilizer PVA always remains associated with the nanoparticles despite repeated washings because PVA forms an interconnected network with the polymer at the interface [17]. We came across similar factors while formulating nanoparticles using PVA as a stabilizer and discussed in the following sections. Above all, the stability and biological activity of the plasmid have been major concerns due to the involvement of organic solvents during the preparation process [16,17]. Keeping the above factors in mind we developed a method for formulation of cationically modied PLGA nanoparticles. An emulsion-diffusion-evaporation technique using ethylacetate as organic solvent and PVA-chitosan blend as a stabilizer yielded uniform spherical cationic nanospheres. We have screened several solvents and found that the particle size is at lower end when ethylacetate was employed (data not shown). Stabilizers (PVA and Chitosan) concentration has been optimized for the smallest particle size for this method (data not shown). We believe that the nanospheres formation involves the mechanism as described: Stirring causes the dispersion of the solvent as irregular sized globules in equilibrium with the continuous phase and the stabilizer is then absorbed on the larger interface created; homogenization further results in the smaller globules; the addition of water and the heating step destabilizes the equilibrium and causes to diffuse the organic solvent to the external surface. During the transport of solute, new smaller globules less than 200 nm are produced; the heating step also helps to have a nal colloidal suspension free of organic solvent and more uniform in size. Nanospheres were also prepared by eliminating one or two steps of the complete method and the results obtained are presented as Table 1. Eliminating either of

ARTICLE IN PRESS
1774 M.N.V. Ravi Kumar et al. / Biomaterials 25 (2004) 17711777 Table 1 Eliminating one or two steps of the method and the resultant particle size No. 1 2 3 O/W emulsion stirring 1000 rpm Yes Yes Yes Homogenization 13,500 rpm No Yes Yes Add. water & evaporation No No Yes Particle size by PCS (nm) 884717 40378 18173

Results are presented as mean (n 3)7standard deviation.

the steps resulted in increase in the particle size, which is in agreement to our discussion. 3.2. FTIR characterization The nanospheres were characterized by FTIR. The characteristic peaks obtained from PVA, chitosan and PLGA were compared with the peaks resulted from nanospheres. The characteristic peaks at 1511 and 3015 cm1 due to amino groups from chitosan were also found in the nanospheres prepared from PVAchitosan blend, suggesting the cationic modication. 3.3. PCS measurements The nanospheres when analyzed by dynamic light scattering demonstrated a unimodal size distribution for PVA alone and PVA-Chitosan blend formulated by emulsion-diffusion-evaporation technique (Fig. 2). However, there is no indication of nanosphere formation when chitosan was used alone, which is in agreement with the reported studies that PVA is necessary to stabilize PLGA particles [18]. Prabha et al. [16] in their recent report compared the difference between the PCS vs. TEM measurements in terms of particle size and found huge difference. The PVA is known to form layers of aggregates (B5 layers) around the surface of nanoparticles contributing towards the hydrodynamic diameter of nanoparticles [19,20]. The discrepancy in the size of nanoparticles could be that the dynamic light scattering method gives the hydrodynamic diameter rather than the actual diameter of nanoparticles, therefore a comparison of the particle size with other techniques as well is worth it. The mean hydrodynamic particle diameter was found to be 111.774.2 nm when PVA was alone used as stabilizer, whereas, 181.573 nm when a combination of PVA and chitosan were used as a blend. The increase in size is expected and attributed to the high molecular weight chitosan. We have compared the size of the particles as analyzed by PCS and AFM techniques and presented as Table 2. 3.4. Zeta potential measurements The zeta potential value is an important particle characteristic as it can inuence both particle stability as
Size Distribution (%)

100

80

PVA alone PVA-Chitosan Blend Chitosan alone

60

40

20

0 0.1 10 100 1000

Particle Size

Fig. 2. Particle size distributions of the nanospheres as measured by PCS. Table 2 Nanospheres as measured by PCS and AFM No. 1 2 3 Stabilizer PVA Chitosan PVA-chitosan PCS measurement (nm) 111.774.2 n 3 Not detectable 181.573 n 3 AFM measurment (nm) 100.276.2 n 47 24.972.7 n 117 187714.4 n 112

PCS: Number in parenthesis represents number of replicates; AFM: number in parenthesis represents number of particles measured. Results are presented as mean7standard deviation.

well as particle mucoadhesion. In theory, more pronounced zeta potential values, being positive or negative, tend to stabilize particle suspension. The electrostatic repulsion between particles with the same electric charge prevents the aggregation of the spheres [21]. Mucoadhesion, on the other hand, can be promoted by a positive zeta potential value. The mucus layer itself is at a neutral pH value an anionic polyelectrolyte [22]. Consequently, the presence of the positively charged groups on the particles could lead to electrical charge interactions between the mucus and the particles. In the present studies nanospheres were made with PVA alone and a blend of PVA-Chitosan to attain surface modication. The particles made of PVA (1% w/v) alone were negatively charged (8 mV at pH 7.4). Zeta potential titration provided proof of successful cationic surface modication when a blend of PVAchitosan (1.3% w/v) was used. The nal nanoparticle suspension using PVA-chitosan blend has a pH of 4 and a zeta potential of 36 mV, which suggests that the

ARTICLE IN PRESS
M.N.V. Ravi Kumar et al. / Biomaterials 25 (2004) 17711777 1775

suspension would be stable. The zeta potential at pH 3.0 was 46 mV; however, it decreased with increase in pH and reached to 10 mV at pH 7.4 (Fig. 3). AFM images show uniform cationic modication, which is evident through uniform DNA coating onto the nanospheres due to the electrostatic interaction between phosphate groups of DNA and the NH2 groups of chitosan on the surface (Fig. 5E). This has been conrmed by gel electrophoresis studies in the later sections. 3.5. Gel electrophoresis The binding of the cationic PLGA nanospheres to the polyanionic DNA was studied using analysis of the electrophoretic mobility of the DNA within an agarose gel. Efcient complexation of pCMVbeta by cationic PLGA nanospheres leads to immobilisation. These new PLGA nanospheres were able to immobilise pCMVbeta plasmid (Fig. 4). Negligible amounts of free DNA in the lane of 100 particles:1 DNA and no free DNA there

after is the proof of good complex at the ratio 100:1 and beyond. 3.6. AFM measurements The size and surface morphology was analyzed by AFM. When PVA was alone used in the preparation the particle size is about 100 nm (Fig. 5A) and the reasons for the discrepancy of the size between the two measurements was discussed under Section 3.3. It appears that lot of PVA is adhered to the particle surface (Fig. 5A), which is a similar nding to the reported studies, irrespective of the method used [16]. When chitosan was used, AFM analysis did show the particle size to be very small (24.972.7 nm) (Fig. 5B), which is unlikely with high molecular weight polymers like chitosan. Moreover, the particle shape is not well dened and fused. We could not detect any particle size by PCS. It appears that the nanospheres were uniform and spherical in shape with smooth surfaces when PVAchitosan blend was used in the preparation (Fig. 5C and D). Also the AFM pictures show no free/unbound material when PVA-chitosan blend was used (Fig. 5C and D). DNA is uniformly coated onto the nanospheres (DNA shell of 22.472.1, n 65 nm) (Fig. 5 panel E) due to the electrostatic interaction between phosphates groups of DNA and the NH2 groups of chitosan on the surface as shown in the Fig. 1. The size of the nanospheresDNA complexes is smaller and more uniform when compared to the reported DNA-polymer (in particular when chitosan is used) self-assemblies [2325]. To our knowledge such a high resolution AFM image clearly showing the electrostatic interaction between positively charged PLGA nanospheres and negatively charged DNA has not been shown before. Many reports on PLGA particles are entirely based on PCS studies while discussing size and very few reports
100

50

40

Zeta Potential mV

30

20

10

-10

3.5

4.5

5.5

6.5 pH

7.5

8.5

9.5

10

Fig. 3. Zetapotential titration curve of PLGA nanospheres coated with PVA-chitosan blend.

B 0

1 10 20 25 50100120140160 B M

90 80

Free DNA (%) M= MARKER DNA B = BLANK SPACE

70 60 50 40 30 20 10 0 0 1 10 20 25 50 100 120 140 160

Particle to DNA ratio

(A)

(B)

Fig. 4. PLGA-DNA complexes with increasing amounts of PLGA nanospheres were prepared and analysed for DNA immobilisation ability. The amounts of free DNA were related to un-complexed DNA (100% mobile) run on the same gel. To quantify the DNA-immobilisation ability, the cationic PLGA:DNA ratios (w/w) required for 100% immobilisation are compared in this graph. (Solid bars=Percentage of free DNA; white bars=100% immobilization).

ARTICLE IN PRESS
1776 M.N.V. Ravi Kumar et al. / Biomaterials 25 (2004) 17711777

(A)

(B)

(C)

(D)

(E)

Fig. 5. AFM images of nanospheres (A) nanospheres with PVA alone as stabilizer (B) with chitosan alone (C) PVA-chitosan blend; (D) surface morphology; (E) nanosphere-DNA complex (bar represents 150 nm (A, B, D & E); 500 nm (C)).

have shown visual images of the nanoparticles. From the present studies its clear that one should not base only on PCS analysis for the particle formation or size.

way. Subsequently, investigations on scale-up process will be performed.

Acknowledgements 4. Conclusion From these investigations it is evident that this method forms uniform cationic PLGA nanospheres that can bind DNA readily by electrostatic interaction. These cationic surface modied PLGA nanospheres avoid the usage of the plasmid during the particle preparation process, where it has to stay in contact with organic solvents for quite a while. PVA alone could not give the cationic charge needed and chitosan alone could not stabilize the particles, therefore, a blend of these two is needed. PVA-chitosan blend not only giving the net positive surface charge, but also produced particles with uniform size and spherical shape, as observed by AFM. Investigations were performed using as low as 50 mg and as high as 500 mg of polymer and found the technique is reproducible irrespective of the polymer amount, which is one of the key ndings of the study. Gene transfection and cellular uptake studies in cultured cells are under MNVRK is grateful to Alexander von Humboldt foundation, Germany for providing with a personal fellowship. U. Bakowsky wishes to thank Stiftung Deutscher Naturforscher Leopoldina (BMBF/LPD9901/8-6).

References
[1] Lemoine D, Francois C, Kedzierewicz F, Preat V, Hoffman M, Maincent P. Stability study of nanoparticles of poly(epsiloncaprolactone), poly(d,l -lactide) and poly(d ,l-lactide-co-glycolide). Biomaterials 1996;17:21917. [2] Gurny R, Peppas NA, Harrington DD, Banker GS. Development of biodegradable and injectable latices for controlled release of potent drugs. Drug Dev Ind Pharm 1981;7:125. [3] Allemann E, Gurny R, Doelker E. Preparation of aqueous polymeric nanodispersions by a reversible salting-out process: inuence of process parameters on particle size. Int J Pharm 1992;87:24753.

ARTICLE IN PRESS
M.N.V. Ravi Kumar et al. / Biomaterials 25 (2004) 17711777 [4] Fessi H, Puisieux F, Devissaguet JP, Ammoury N, Benita S. Nanocapsules formation by interfacial polymer deposition following solvent displacement. Int J Pharm 1989;55:R14. [5] Quintanar-Guerrero D, Ganem-Quintanar A, Allemann E, Fessi H, Doelker E. Inuence of the stabilizer coating layer on the purication and freeze-drying of poly(D,L-lactic acid) nanoparticles prepared by an emulsion-diffusion technique. J Microencapsul 1998;15:10719. [6] Choi SW, Kwon HY, Kim WS, Kim JH. Thermodynamic parameters on poly(d ,l-lactide-co-glycolide) particle size in emulsion-diffusion process. Colloids Surf A: Physicochem Eng Aspect 2002;201:2839. [7] Niwa T, Takeuchi H, Hino T, Kunou N, Kawashima Y. Preparations of biodegradable nanospheres of watersoluble and insoluble drugs with dl -lactide/glycolide copolymer by a novel spontaneous emulsication solvent diffusion method, and the drug release behavior. J Control Release 1993;25: 8998. [8] Schachter DM, Kohn J. A synthetic polymer matrix for the delayed or pulsatlie release of water-soluble peptides. J Control Release 2002;78:14353. [9] Lamprecht A, Ubrich N, Hombreiro Perez M, Lehr CM, Hoffman M, Maincent P. Biodegradable monodispersed nanoparticles prepared by pressure homogenization-emulsion. Int J Pharm 1999;184:97105. [10] Lamprecht A, Ubrich N, Yamamoto H, Schafer U, Takeuchi H, Maincent P, Kawashima Y, Lehr CM. Biodegradable nanoparticles for targeted drug delivery in treatment of inammatory bowel disease. J Pharmacol Expt Ther 2001;299:77581. [11] Cohen-Sacks H, Najareh Y, Tchaikovski V, Gao G, Elazer V, Dahan R, Gati I, Kannan M, Waltenberger J, Golomb G. Novel PDGFbR antisense encapsulated in polymeric nanospheres for the treatment of restenosis. Gene Therapy 2002;9:160716. [12] Panyam J, Zhou WZ, Prabha S, Sahoo SK, Labhasetwar V. Rapid endo-lysosomal escape of poly(dl-lactide-co-glycolide) nanoparticles: implications for drug and gene delivery. FASEB J 2002;16:121726. [13] Vandervoort J, Ludwig A. Biocompatible stabilizers in the preparation of PLGA nanoparticles: a factorial design study. Int J Pharm 2002;238:7792. [14] Vila A, Sanchez A, Tobio M, Calvo P, Alonso MJ. Design of biodegradable particles for protein delivery. J Control Release 2002;78:1524. 1777 [15] Illum L. Chitosan and its use as pharmaceutical excipient. Pharm Res 1998;15:132631. [16] Prabha S, Zhou W-Z, Panyam J, Labhasetwar V. Size-dependency of nanoparticle-mediated gene transfection: studies with fractionated nanoparticles. Int J Pharm 2002;244:10515. [17] Sahoo SK, Panyam J, Prabha S, Labhasetwar V. Residual polyvinyl alcohol associated with poly(d ,l -lactide-co-glycolide) nanoparticles affects their physical properties and cellular uptake. J Control Release 2002;82:10514. [18] Murakami H, Kawashima Y, Niwa T, Hino T, Takeuchi H, Kobayashi M. Inuence of the degrees of hydrolyzation and polymerization of poly(vinylalcohol) on the preparation and properties of poly(d ,l-lactide-co-glycolide) nanoparticles. Int J Pharm 1997;149:439. [19] Zambaux MF, Bonneaux F, Gref R, Maincent P, Dellacherie E, Alonso MJ, Labrude P, Vigneron C. Inuence of experimental parameters on the characteristics of poly(lactid acid) nanoparticles prepared by double emulsion method. J Control Release 2000;50:3140. [20] Konan YN, Cerny R, Favet J, Berton M, Gurny R, Allemann E. Preparation and characterization of sterile sub-200 nm mesotetra(4-hydroxylphenyl)porphyrin-loaded nanoparticles for photodynamic therapy, European. J Pharm Biopharm 2003; 55:11524. [21] Feng S, Huang G. Effects of emulsiers on the controlled release of paclitaxel (Taxol) from nanospheres of biodegradable polymers. J Control Release 2001;71:5369. [22] Bayems V, Gurny R. Chemical and physical parameters of tears relevant for the design of ocular drug delivery formulations. Pharm Acta Helv 1997;72:191202. [23] MacLaughlin FC, Mumper RJ, Wang J, Tagliaferri JM, Gill I, Hinchcliffe M, Rolland AP. Chitosan and depolymerized chitosan oligomers as condensing carriers for in vivo plasmid delivery. J Control Release 1998;56:25972. [24] Mao HQ, Roy K, Troung-Le VL, Janes KA, Lin KY, Wang Y, August JT, Leong KW. Chitosan-DNA nanoparticles as gene carriers: synthesis, characterization and transfection efciency. J Control Release 2001;70:399421. [25] Koping-Hoggard M, Guan IT, Edwards K, Nilsson M, Varum KM, Artursson P. Chitosan as a nonviral gene delivery system. Structure-property relationships and characteristics compared with polyethylenimine in vitro and after lung administration in vivo. Gene Therapy 2001;8:110821.