Pegylated nanoparticles based on poly(methyl vinyl ether-co-maleic anhydride): preparation and evaluation of their bioadhesive properties

European Journal of Pharmaceutical Sciences 24 (2005) 411–419

Pegylated nanoparticles based on poly(methyl vinyl ether-co-maleic anhydride): preparation and evaluation of their bioadhesive properties Krassimira Yonchevaa,1 , Elena Lizarragab , Juan M. Irachea,∗ a

Centro Gal´enico, Dep. Farmacia y Tecnolog´ıa Farmac´eutica, Universidad de Navarra, Apartado 177, Irunlarrea, 31080 Pamplona, Spain b Departamento de Quimica Organica y Farmaceutica, CIFA, Universidad de Navarra, 31080 Pamplona, Spain Received 3 September 2004; received in revised form 1 December 2004; accepted 10 December 2004 Available online 2 February 2005

Abstract Pegylated nanoparticles based on poly(methyl vinyl ether-co-maleic anhydride) (PVM/MA) were prepared by simple solvent displacement method, in the absence of catalysts or specific chemical conditions. Pegylation efficiency increased with the increasing of molecular weight and bulk concentration of poly(ethylene glycols) (PEGs) investigated. In fact, the use of PEG with molecular weight less than 1000 Da did not lead to its attachment. 1 H NMR spectroscopy was performed in order to estimate the conformation state of PEG-chains and to predict the nanoparticle structure. Pegylation with PEG 2000 gave surface modified nanoparticles (“brush” conformation), while the chains of PEG 1000 were distributed either in the core or physically adsorbed on the nanoparticle surface. The capacity of nanoparticles to adsorb mucin at pH 7.4 was significantly higher for PEG 1000-NP than for PEG 2000-NP. The “brush” layer seemed to decrease the interaction between PEG 2000-NP and mucin, which facilitated their penetration through the mucus gel. As a consequence, PEG 2000-NP displayed higher capacity to develop adhesive interactions with rat intestinal mucosa in vivo. Independent on the weaker bioadhesive potential of PEG 1000-NP, both types of pegylated nanoparticles demonstrated very high affinity to the intestinal mucosa rather than to the stomach wall, which could be established for drug targeting to the small intestine. © 2004 Elsevier B.V. All rights reserved. Keywords: Poly(ethylene glycol); Pegylated nanoparticles; Bioadhesion; Oral delivery

1. Introduction Colloidal nanoparticles are widely investigated as carriers for oral delivery of drugs over the last years. When nanoparticles are administered by oral route, only a small portion of the given dose appears to reach the gastrointestinal mucosa since they have to avoid eventual interactions with proteins or food components in the lumen and then to overcome the mucus barrier layer. Nanoparticles which interact with the mucus compounds could be entrapped in the mucus gel and subsequently removed by the normal mucus-turnover (Ponchel and Irache, 1998). If nanoparticles are able to diffuse through ∗

Corresponding author. Tel.: +33 948 425600; fax: +33 948 425649. E-mail address: [email protected] (J.M. Irache). 1 Department of Pharmaceutical Technology, Faculty of Pharmacy, 2 Dunav Str., 1000 Sofia, Bulgaria. 0928-0987/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.ejps.2004.12.002

the mucus network, they may adhere to the cell surface and, in some cases, to be taken up and translocated (Florence and Hussain, 2001). Despite of these phenomena, it has been clearly demonstrated the efficacy of particulate carriers to improve the oral bioavailability and activity of a number of drugs. The enhanced bioavailability can be related to the capacity of nanoparticles to interact strongly with the upper regions of the gut, including chitosan (Chen et al., 2004), poly(lactic acid) (PLA) (Florence and Hussain, 2001), gliadin (Arangoa et al., 2001) and poly(methyl vinyl ether-co-maleic anhydride) (Arbos et al., 2002a) nanoparticles. However, the design of nanoparticulate systems able to adhere to the small intestine mucosa and/or to target specific areas within the gut appears to be more difficult. During the last years, different strategies have been developed including the use of surfactants (Jung et al., 2000), lectins (Hussain et al., 1997),

412

K. Yoncheva et al. / European Journal of Pharmaceutical Sciences 24 (2005) 411–419

invasins (Dawson and Halbert, 2000) and vitamin B12 derivatives (Russell-Jones et al., 1999). Therefore, a present challenge is to design new particulate systems that would be able either to interact with the intestinal mucosa and, thus, prolong their residence time within the gut or to target specific regions for vaccination or translocation purposes. In this context, one of the most interesting approaches would be the use of pegylated nanoparticles. These carriers seem to promote specific bioadhesive interactions with mucosal tissues due to the ability of PEG-chains to interdiffuse across mucus network (Huang et al., 2000). Similarly, other studies have demonstrated that PEG-coating layer around PLA-nanoparticles modulated their interaction with biological surfaces such as nasal and intestinal mucosa (Tobio et al., 1998, 2000). Further, improvement of the oral bioavailability of various drugs as well as target drug delivery in the gastrointestinal tract could be achieved. In addition, several investigations have reported improved drug stability due to the protective properties of PEG-coating layer (Iwanaga et al., 1997; Tobio et al., 2000). Thus, pegylated nanoparticles could be considered as suitable carriers providing an efficient transport and oral bioavailability of unstable macromolecular drugs as peptides, proteins and genes. However, the main difficulty of the pegylation strategy is the preparation of nanoparticles with a stable PEGattachment (Peracchia et al., 1997). Stability of PEGattachment to desorption/displacement in vivo is important because the configuration and conformational mobility of PEG-chains could be changed. It was reported that surface density, flexibility and chain length have a great effect on the interactions of particles with the biological media (Bazile et al., 1995; Torchilin, 1998; Mosqueira et al., 2001). The mode of PEG-attachment to nanoparticles is carried out mainly by physical adsorption or by covalent grafting (Stolnik et al., 1995; De Jaeghere et al., 2000). However, the drawback of simple adsorption is an easy displacement of the coating layer in vivo (Neal et al., 1998). Since covalently binding of PEGchains is preferable way, most of the pegylated nanoparticles have been prepared on the base of di- or tri-block-copolymers of PEG with various monomers including lactide, glycolide and caprolactone (Gref et al., 1995). The process of copolymerization however requires the use of various catalysts and specific chemical conditions. In addition, a possible presence of organic solvents (such as methylene chloride, toluene, etc.) remaining in the final nanoparticles may be another problem. The aim of the present work was to develop pegylated nanoparticles and to estimate their bioadhesive potential as oral drug delivery system. Poly(methyl vinyl ether-co-maleic anhydride) (PVM/MA) was considered as a nanoparticle carrier taking into account its low toxicity and excellent biocompatibility. Because of the availability of free functional groups either in the PVM/MA molecules or in the PEG-molecules the association of PEG under mild conditions was expectable. The reaction may occur by nucleophile attack of anhydride groups in the copolymer by hydroxyl or amino groups. In this view, PVM/MA copolymer could be a suitable nanoparticle

carrier allowing ligand attachment without the need of special chemical reagents (i.e. glutaraldehyde or carbodiimide derivatives) (Arbos et al., 2002b). In the present study PEGs with different molecular weights (400, 1000 and 2000 Da) as well as various initial concentrations were examined aiming to optimize pegylation degree.

2. Materials and methods 2.1. Materials Poly(methyl vinyl ether-co-maleic anhydride) (PVM/ MA) (Gantrez AN 119; Mw of 200 000 Da) was a gift from ISP (Barcelona, Spain). Poly(ethylene glycol) with Mw of 400, 1000 and 2000 Da (PEG 400; PEG 1000, PEG 2000) were provided by Fluka (Switzerland) and Micro BCA Protein Assay Reagent Kit was from Pierce (Rockford, USA). Rhodamine B isothiocyanate (RBITC) and pig gastric mucin (Type II: crude) were supplied by Sigma (St. Louis, USA). 2.2. Preparation of pegylated PVM/MA-nanoparticles Pegylated nanoparticles were prepared by modification of solvent displacement method previously established for manufacture of PVM/MA-nanoparticles (Arbos et al., 2002b). In the present study, 100 mg of PVM/MA copolymer and various amounts of PEG were dissolved in acetone (5 ml) and incubated simultaneously under stirring (700 rpm) for 1 h. After their incubation, water/ethanol mixture (1:1 v/v) was added to the organic phase. The solvents were eliminated under reduced pressure (Buchi R-144, Switzerland) and the resulting nanoparticles were purified by twice centrifugation at 17000 rpm for 20 min (Sigma 3K30, Germany). Finally, nanoparticles were lyophilized using sucrose (5% w/v) as cryoprotector (Genesis 12EL, Virtis, USA). Conventional PVM/MA-nanoparticles (NP) were prepared paralelly applying the same protocol but in the absence of PEG. For in vivo studies, control (NP) and pegylated nanoparticles (PEG-NP) were fluorescently labelled with RBITC (Arbos et al., 2002a). For this purpose, the aqueous nanosuspensions were incubated at room temperature for 5 min with 1.25 mg RBITC prior their purification by centrifugation and lyophilization. 2.3. Characterization of the pegylated nanoparticles The shape and morphology of the nanoparticles were assessed by scanning electron microscopy (SEM). For this purpose, freeze-dried microparticles were covered with 9 nm molecular gold (Emitech K550 Equipment, UK) and SEM photographs made with a Zeiss DSM 940 A (USA). Nanoparticle size and zeta-potential were determined by photon correlation spectroscopy and electrophoretic laser Doppler anemometry using a Zetamaster analyzer (Malvern

K. Yoncheva et al. / European Journal of Pharmaceutical Sciences 24 (2005) 411–419

Instruments, UK). Samples were diluted with 0.05 M phosphate buffered saline (pH 7.4) and measured at 25 ◦ C with a scattering angle of 90◦ . Each batch was analysed in triplicate. The interaction between PEG and PVM/MA during preparation was followed by nuclear magnetic resonance spectroscopy (1 H NMR) (Brucer Avance 400, Germany) and thin layer chromatography (TLC). The 1 H NMR spectrum of the organic phase was made in deuterated acetone. The mobile phase used for TLC was methylene chloride:methanol (90:10 v/v) and the spots were visualised after treatment of the plate in iodine environment. The way of PEG-attachment to the nanoparticles was examined by infrared spectroscopy (IR) using Nicolet-FTIR Avatar 360 connected with OMNIC ESP software. The resolution was 2 cm−1 and the spectra were collected at 100 scans. The amount of PEG attached to the resulted nanoparticles was estimated by 1 H NMR spectroscopy. For this purpose, exactly weighted amounts of pegylated nanoparticles (5 mg) were dissolved in deuterated DMSO (0.5 ml) and the spectra were obtained at ns = 6400. 1 H NMR spectra of free PEGs were performed using the same ratio and experimental conditions. The quantity of PEG-attached to nanoparticles was calculated by the ratio between peak areas of the protons of ethylene units (3.51 ppm) detected in the spectra of pegylated particles and in the spectra of free PEG, respectively. The amount of the RBITC loaded in the nanoparticles was determined by colorimetry at wavelength 540 nm (Labsystems iEMS Reader MF, Finland). The quantity of loaded RBITC was estimated as a difference between its initial concentration added and the concentration found after total hydrolysis of certain amount of nanoparticles in 0.1N NaOH medium. For the calculations standard curve of RBITC prepared in 0.1N NaOH medium was used (concentration range of 4–30 ␮g/ml; r > 0.999). In vitro release of RBITC was studied by dispersing of nanoparticles (5 mg) in 0.5 ml simulated gastric or intestinal fluid (USP XXIII) at 37 ◦ C. At predetermined time intervals, the nanosuspensions were centrifuged (17 000 rpm, 20 min) and the amount of RBITC released was measured in the supernatants by colorimetry (λ = 540 nm). 2.4. In vitro interaction between nanoparticles and mucin The interaction was studied by incubation of mucin and nanoparticles (1:4 weight ratio) either in acidic (pH = 1.2, solution of hydrochloric acid) or in neutral (pH = 7.4, PBS) medium. The incubation was carried out under stirring (700 rpm) at temperature of 37 ◦ C (Variomag, Germany). The dispersions were centrifuged at predetermined time (17 000 rpm, 20 min) and 150 ␮l of each supernatant was placed in a test plate. Micro BCA Protein Assay Reagent Kit (150 ␮l) was added to the supernatants and the plate was incubated for 2 h at 37 ◦ C. According to this procedure, the absorbance of mucin was measured by colorimetry at

413

wavelength of 570 nm. The amount of the mucin adsorbed to the nanoparticles was determined as a difference between its initial concentration and the concentration found in the dispersion after incubation and centrifugation. The calculations were made on the base of mucin standard curves which linearity were r > 0.997 (pH = 1.2) and r > 0.994 (pH = 7.4) in the concentration range 100–500 ␮g/ml. 2.5. In vivo bioadhesive study The study was performed in compliance with the regulations of the responsible committee of the University of Navarra in line with the European legislation on animal experiments (86/609/EU). Male Wistar rats (average weight of 220 g; Harlan, Barcelona, Spain) were housed under normal conditions. The animals were divided in three different groups and placed in metabolic cages and fasted overnight to prevent coprophagia but allowing free access to water. Each animal received a single oral dose of 1 ml aqueous suspension containing 10 mg of the nanoparticles loaded with RBITC (approximately 45 mg particles/kg body weight). The animals were sacrificed by cervical dislocation 1 and 3 h post-administration. The abdominal cavity was opened and the stomach, small intestine and cecum were removed, opened lengthwise along the mesentery and rinsed with phosphate saline buffer (pH 7.4). Further, the stomach, small intestine and cecum were cut into segments of 2 cm length and digested in 1 ml 3N NaOH for 24 h (Arbos et al., 2002a). RBITC was extracted from the digested samples by addition of 2 ml methanol, vortexed for 1 min and centrifuged at 4000 rpm for 10 min. Aliquots (1 ml) of the supernatants were assayed for RBITC by spectrofluorimetry at λex 540 nm and λem 580 nm (GENios, Austria) to estimate the fraction of adhered nanoparticles to the mucosa. For calculations, standard curves of RBITC were prepared by addition of RBITC-solutions in 3N NaOH (0.5–10 ␮g/ml) to tissue segments following the same treatment steps (r > 0.996). 2.6. Statistical analysis The data were compared using the non-parametric Kruskal–Wallis test. P values of <0.05 were considered significant. All calculations were performed using a statistical software program (SPSS® 6.1.2, Microsoft).

3. Results and discussion 3.1. Preparation of the pegylated nanoparticles The search of the most suitable method to prepare pegylated nanoparticles in an easy way led us to investigate the effect of PEG incorporation during the formation process of nanoparticles. We hypothesized that the absence of water during the incubation of the copolymer and PEG would decrease the opening of the anhydride groups of PVM/MA

414

K. Yoncheva et al. / European Journal of Pharmaceutical Sciences 24 (2005) 411–419

Table 1 Influence of the initial amount and molecular weight of PEG on the degree of its association to the nanoparticles Bulk concentration (mg)a

Associated PEG 400 (%)

Associated PEG 1000 (%)

Associated PEG 2000 (%)

2.5 5 10 25 50

– – n.d.b n.d. n.d.

0.79 ± 0.10 1.51 ± 0.15 1.75 ± 0.11 1.98 ± 0.21 –

– 2.85 ± 0.10 3.45 ± 0.14 3.02 ± 0.40 4.46 ± 0.25

The amount of PEGs was determined by 1 H NMR spectroscopy and expressed in percents. Data represent mean ± S.D. (n = 3). a Bulk concentration of PEG into the organic phase during preparation. b n.d.: not detectable.

and, thus, would increase the possibilities for the interaction between functional groups of PEG and anhydride residues of the copolymer. Ladaviere et al. (1999a) have reported that, in an aqueous medium, PVM/MA hydrolyses to a diacid form with pKa -values of yielded COOH-groups are 7.5 and 3.5, respectively. Here, 1 H NMR spectrum of the acetone phase containing PVM/MA and PEG did not show a signal for protons of COOH-groups (12.5 ppm, data not shown) which denoted that the anhydride groups were transformed rather to ester derivatives than to COOH-groups. The opening of anhydride groups and their esterification with OH-groups of PEG in the acetone phase was also followed by TLC. The chromatograms of the organic phase in the beginning of incubation showed availability only of product with Rf-value similar to that of free PEG (Rf = 0.45). However, 1 h after incubation in the organic phase appeared another product (Rf = 0.76) giving suggestion for coupling of PEG to the copolymer. In preliminary experiments, direct coating of pre-formed PVM/MA-nanoparticles with PEG in aqueous phase was examined. According to this method, the association was expectable assuming binding of PEG-chains with the carboxyl residues extending from the surface of freshly formed PVM/MA-nanoparticles. Data showed that three times higher pegylation degree was achieved when PEG was incorporated in the organic phase than coating of freshly prepared nanoparticles. The main reason for the failure could be a predominant role of hydrolysis of anhydride moieties in water over the PEG-association (Ladaviere et al., 1999b). Hence, the coating procedure was also possible although the high affinity of PEG to the water phase and the rapid hydrolysis of the polymer groups led to lower association. The influence of both bulk concentration and molecular weight of PEGs on the pegylation efficiency was studied. Table 1 summarizes the association degrees of the different PEGs at various initial concentrations. As it could be

seen, the pegylation with PEG 400 did not lead to its attachment at any of the bulk concentrations investigated. Efficient pegylation was achieved only with both PEGs possessing molecular weights higher than 400 Da. In addition, the pegylation of nanoparticles with PEG 2000 was found to reach approximately two times higher association degree compared to PEG 1000. The results also showed that PEG-attachment slightly increased with increase of its bulk concentration – e.g. raise from 2.90 to 4.46% corresponding to initial PEG 2000-amounts of 5 and 50 mg, respectively. However, with increasing of initial PEG-concentration the nanoparticle yield decreased (data not shown). 3.2. Characterization of the nanoparticles Table 2 summarizes the main physicochemical characteristics of the different formulations prepared at PEG bulk concentration of 25 mg. Under these conditions, nanoparticles displayed a similar size of about 280–300 nm. Their morphological characterization by SEM showed sphericalshaped particles with size similar to that obtained by photon correlation spectroscopy (Fig. 1). This fact suggested that PEG-attachment did not influence the size of the resulting carriers. However, the pegylation of nanoparticles significantly decreased the negative charge of conventional nanoparticles (P < 0.05). This tendency was more pronounced for PEG 2000-NP than for PEG 1000-NP, as it could be seen from the zeta-potential measurements. The way of PEG-attachment was examined by IR- and by 1 H NMR-spectroscopy. IR-spectra of PVM/MA and pegylated nanoparticles presented bands at 1770–1850 cm−1 which were recognised as typical for anhydride groups (Fig. 2). The IR-spectrum of pegylated nanoparticles showed additionally a small peak at 1730–1740 cm−1 (corresponding to a C O group) missing in the spectrum of PVM/MA. This

Table 2 Physicochemical characteristics of the nanoparticles selected for the further investigations Sample

Size (nm)

Zeta-potential (mV)

RBITC loading (␮g/mg)a

PEG-attached (␮g/mg)b

NP PEG 1000-NP PEG 2000-NP

289 ± 11 271 ± 10 299 ± 22

−58.8 ± 4.5 −50.4 ± 2.2 −44.1 ± 4.0

10.33 ± 0.87 11.47 ± 0.50 10.37 ± 0.09

– 19.8 ± 2.05 30.2 ± 4.0

Pegylated nanoparticles were prepared at bulk PEG-concentration of 25 mg. Data express mean ± S.D. (n = 3). a The amount of RBITC (␮g/mg nanoparticles) was determined by colorimetry. b The amount of PEG (␮g/mg nanoparticles) was determined by 1 H NMR spectroscopy.

K. Yoncheva et al. / European Journal of Pharmaceutical Sciences 24 (2005) 411–419

415

Fig. 1. SEM-photograph of PEG 2000-NP obtained at bulk PEG-concentration of 25 mg. Image contributed by Prof. R. Jordana (Universidad de Navarra, Spain).

peak illustrated that coupling reaction with probable formation of ester bonds occurred when nanoparticles were made. On the other hand, 1 H NMR-spectroscopy provided more details about the coupling as well as the conformation state of PEG-chains following the quantitative changes of the protons typical for PEG-groups. The 1 H NMR spectra of the pegylated nanoparticles revealed two essential peaks – the first one was typical for the protons of ethylene units (3.51 ppm) and the second corresponded to the protons of the hydroxyl groups (4.58 ppm) (Fig. 3). The latter peak was assumed to belong to the protons of OH-groups of PEG regarding two facts. First, this peak always appeared in the spectra of PEGs made in DMSO and disappeared when deuterated water was added to the sample. Second, this peak did not exist neither in the spectrum of non-pegylated nanoparticles nor in the spectrum of free PVM/MA-copolymer (not shown). Thus, comparative ratio between the peak area of protons of ethylene groups and protons of OH-groups for both free PEG (as control) and pegylated nanoparticles was formulated, aiming to quantify the participation of PEG-functional groups in coupling reaction with anhydride residues of PVM/MA. The values of the ratio for nanoparticle samples and free PEGs are summarized in Table 3.

The ratio between peak areas of both types of protons was almost two times higher for PEG 2000-NP than for the free PEG 2000 due to the extreme reduction of the peak area of the hydroxyl protons (Table 3 and Fig. 3). These data suggested two times lower availability of hydroxyl protons in the nanoparticle spectrum, which could be a result from the esterification between these OH-groups and carboxyl residues of PVM/MA. The esterification took place predominantly between OH-groups of PEG and COOH-groups with pKa = 3.5 (Fig. 4a), although esterification of the other COOH-groups having pKa = 7.5 also might occur (Fig. 4b). However, the es-

Table 3 Quantitative data obtained by 1 H NMR spectra of the pegylated nanoparticles and free PEGs as controls Sample

Peak A (3.51 ppm)

Peak B (4.58 ppm)

Ratio A/B

PEG 1000 PEG 1000-NP PEG 2000 PEG 2000-NP

262.0 × 108 5.71 × 108 753.5 × 108 20.35 × 108

3.10 × 108 0.08 × 108 7.36 × 108 0.11 × 108

84.5 70.7 102.4 184.5

Abbreviations: Peak A corresponded to protons of ethylene units ( OCH2 CH2 , δ = 3.51 ppm) and peak B to protons of hydroxyl groups of PEGs (δ = 4.58 pm).

416

K. Yoncheva et al. / European Journal of Pharmaceutical Sciences 24 (2005) 411–419

Fig. 2. Infrared spectra of PEG 2000, PEG 2000-NP and PVM/MA. The arrow shows the new peak appeared in the spectrum of pegylated nanoparticles missing in the spectrum of PVM/MA.

ter bonds were always formed only between one OH-group of PEG-chains and one anhydride residue of the copolymer. Simultaneous esterification of both COOH-groups with both OH-groups of PEG-chain was excluded by 1 H NMR

spectroscopy. The presence of visible hydroxyl protons of PEG (4.58 ppm) in the spectrum of pegylated nanoparticles demonstrated that one of the OH-groups of PEG was free (Fig. 4). Consequently, the PEG-chains anchored by one functionalized end to the polymer whereas the other one (terminating with OH-group) extended from the nanoparticle surface. According to these data, a small part of OH-groups did not react which supposed that some PEG-chains would be distributed into the nanoparticle core or physically adsorbed

Fig. 3. 1 H NMR spectra of PEG 2000-NP (top) and free PEG 2000 (bottom) performed in DMSO at ns = 6400. In the small figure: increased image of the triplet at δ = 4.58 ppm in the 1 H NMR spectrum of PEG 2000-NP.

Fig. 4. Schematic presentation of the probable ester bonds between PEGchains (ˆˆˆˆˆ) and anhydride residues of PVM/MA depending on pKa -values: coupling at residues with pKa = 3.5 (a) or with pKa = 7.5 (b).

K. Yoncheva et al. / European Journal of Pharmaceutical Sciences 24 (2005) 411–419

Fig. 5. In vitro interaction between pig gastric mucin and nanoparticles (1:4 weight ratio) investigated in acidic (pH = 1.2) or in basic medium (pH = 7.4) after 1 h incubation; mean ± S.D. (n = 3).

on the surface. However, the larger part was extended outwards in a “brush” conformation. On the other hand, the ratio between peak areas of these protons for PEG 1000 and PEG 1000-NP was almost similar (Table 3). The results assumed that all protons of OH-groups kept their conformation and these OH-groups did not participate in eventual binding process. Hence, in the case of PEG 1000-NP, either interior distribution or physical adsorption on the nanoparticle surface could occur. Furthermore, the data from zeta-potential measurements corresponded with the results obtained by 1 H NMR spectroscopy. Since the chains of PEG 1000 were distributed into nanoparticle core or physically adsorbed on the surface, the surface charge of PEG 1000-NP was less modified (Table 2). The more pronounced reduction of the negative zeta-potential observed in the case of PEG 2000-NP gave another evidence for the surface location of PEG 2000 chains suggested by 1 H NMR spectra. This is in agreement with other studies describing significant reductions in the zeta-potential of pegylated nanoparticles (Tobio et al., 1998; Vila et al., 2004). In fact, the presence of a PEG-coating may shift the shear plane of the diffusion layer to a greater distance from the nanoparticles, thus resulting in a decrease in the absolute value of zeta-potential.

417

These results can be explained with the compact conformation state of PVM/MA-copolymer in acidic medium (Dubin and Strauss, 1967). With increasing the pH-value of the medium, the water insoluble anhydride form conversed to expanded configuration. On the other hand, part of PEG-chains might be displaced due to hydrolysis of the ester bonds formed with COOH-groups with the lower pKa -value (Fig. 4a). The affinity of nanoparticles to adsorb mucin in basic medium followed rank order: PEG 1000-NP > NP > PEG 2000-NP (P < 0.05). The high degree of mucin adsorption to the surface of NP was expectable taking in account the bioadhesive potential of the copolymer (Esposito et al., 1994). The more important point was the different behaviour of both types of pegylated nanoparticles. The difference in the molecular weight of both PEGs could not be a factor since physical entanglement with mucin chains does not occur at such a low molecular weight (Lee et al., 2000). The lower level of mucin association to PEG 2000-NP was more probably due to the steric hindrance provoked by surface extending chains of PEG 2000. This is in consistency with Sanders et al., who have reported that pegylated GL67 lipoplexes did not interact with mucus components whereas non-pegylated formed aggregates (Sanders et al., 2002). Protein repellent properties of PEG-coated surface also might participate in the prevention of mucin association regarding the protein parts of mucin chains (Gref et al., 1994). Similarly, Tobio et al. (2000) have reported that PEG-coated nanoparticles interacted less than non-coated with pepsin and pancreatin. In this view, the absence of surface located chains or their physical adsorption to PEG 1000-NP could facilitate the interaction between mucin and acid groups of PVM/MA copolymer. This can occur by development of hydrogen bonds, Van der Waals, hydrophobic or ionic interactions (Lehr, 1996). However, hydrogen binding between carboxyl residues of PVM/MA-copolymer (first acidity with pKa = 3.5) and sialic mucin residues (pKa of 2.6) would be limited at pH = 7.4 due to their ionization (Willits and Saltzman, 2001). Consequently, the interactions between mucin protein chains and nanoparticles probably occurred by hydrophobic bonds. In this case, the presence of hydrophilic “brush” layer on the surface of PEG 2000-NP additionally hindered the hydrophobic interactions. 3.4. In vivo bioadhesive study

3.3. In vitro interaction between mucin and nanoparticles Nanoparticle formulations summarized in Table 2 were tested for in vitro and in vivo experiments. In vitro interaction between pegylated nanoparticles and mucin were investigated either in acidic (pH 1.2) or in basic (pH 7.4) medium. Under acidic conditions (pH = 1.2), both pegylated and conventional nanoparticles demonstrated similar behaviour characterised by a weak interaction with the pig gastric mucin (Fig. 5). On the contrary, more intensive interaction and some differences between nanoparticles appeared in basic medium.

In vivo bioadhesive studies were performed using pegylated nanoparticles loaded with a fluorescent marker (RBITC). To ensure that the fluorescence determined in the gastrointestinal parts was due to the RBITC-associated nanoparticles, in vitro release of RBITC was examined preliminary. Since the percentage of RBITC released during the first 2 h was low – between 7–15% in simulated gastric fluid and 20–25% in simulated intestinal fluid (data not shown), it could be assumed that the fluorescence measured would be due to the RBITC-associated to the nanoparticles. Further, because the in vitro release profiles of RBITC were similar

418

K. Yoncheva et al. / European Journal of Pharmaceutical Sciences 24 (2005) 411–419

for NP than for pegylated nanoparticles. In addition, the ratio between the adhered fractions in the small intestine and stomach was about four times higher for pegylated nanoparticles than for NP. One possible explanation would be the entrapment of non-modified nanoparticles in the mucus network hindering their penetration to the intestinal surface. Further, the physiological turnover of the mucus layer with the time could enable nanoparticle elimination which explained the lower levels of adhesion observed for these nanoparticles. This is in agreement with our previous results concerning similar elimination rates for both conventional nanoparticles and copolymer solubilised in water (Arbos et al., 2002a). In any case, these results confirm that pegylated nanoparticles can be appropriate carriers to target the mucosa of the small intestine. This can occur because of their ability to avoid “permanent” interactions with soluble proteins. On the other hand, it appears that the pegylated nanoparticles penetrate easily through the mucus layer, allowing their proximity with the cells of the intestinal wall. It seems that the physicochemical characteristics of pegylated nanoparticles may influence their behaviour, such as the time for maximal interaction and the residence time within the mucosa. However, these facts require further investigations.

4. Conclusion Fig. 6. Adhered nanoparticle fractions in the different parts of rat gastrointestinal tract 1 and 3 h after oral administration. The adhered fractions are presented as a percent of the initial dose applied; mean ± S.D. (n = 3).

for the three types of nanoparticles, an eventual difference in their in vivo behaviour would be consequence of their different bioadhesive properties (Fig. 6). Thus, 1 h post-administration, all formulations possessed similar ability to interact with the stomach mucosa (P < 0.05), whereas, in the small intestine where the pH-value of the fluids rises to 7.4, the adhesive capacity of the nanoparticles appeared to be different. As it could be seen in Fig. 6, PEG 2000-NP displayed higher capacity to develop adhesive interactions with the intestinal mucosa than NP and PEG 1000-NP (P < 0.05) and this tendency was observed during the first 3 h of the study. In addition, the ratio between the adhered fractions in the small intestine and stomach was also higher for PEG 2000NP than for the other two formulations (about 4 versus 2). This fact can be explained by the lower interaction of PEG 2000-NP with mucin (Fig. 5), which may facilitate and increase their adhesion to the intestinal wall. Huang et al. (2000) have reported that PEG-chains grafted on the hydrogel surface strongly facilitated its diffusion across the mucus network due to interdiffusion phenomenon. Three hours after administration, the amount of nanoparticles remaining adhered within the gut was decreased for all formulations, however, this decrease was more pronounced

The present study has demonstrated the possibility to prepare pegylated nanoparticles based on the poly(methyl vinyl ether-co-maleic anhydride) under mild conditions. Pegylation efficiency lowered with the decreasing of the molecular weight of PEGs. Pegylation with PEG 2000 gave surface modified nanoparticles with a “brush” conformation of the chains, while the chains of PEG 1000 were distributed inside or physically adsorbed on the nanoparticle surface. Thus, the different surface characteristics of PEG-layer seemed to influence bioadhesive properties of the formulations. The “brush” layer appeared to decrease the interaction between PEG 2000-NP and mucin, which further facilitated their contact with the intestinal mucosa. In any case, both types of pegylated nanoparticles possessed very high affinity to adhere to the small intestine rather than to the stomach mucosa. The latter could be adapted for the need of drug targeting to the distal regions of gastrointestinal tract.

Acknowledgements This work was supported by “Ministerio de Ciencia y Tecnolog´ıa” of Spain (Project SAF2001-0690-C03 and AGL2004-07088-C03-02), Fundaci´on Roviralta and Fundaci´on Universitaria de Navarra (Pamplona, Spain). The authors would also like to thank Prof. Rafael Jordana (Zoology Department, University of Navarra, Spain) for his collaboration in the scanning electron microphotograph.

K. Yoncheva et al. / European Journal of Pharmaceutical Sciences 24 (2005) 411–419

References Arangoa, M.A., Campanero, M.A., Renedo, M.J., Ponchel, G., Irache, J.M., 2001. Gliadin nanoparticles as carriers for the oral administration of lipophilic drugs. Relationship between bioadhesion and pharmacokinetics. Pharm. Res. 18, 1521–1527. Arbos, P., Arangoa, M.A., Campanero, M.A., Irache, J.M., 2002a. Quantification of the bioadhesive properties of protein-coated PVM/MA nanoparticles. Int. J. Pharm. 242, 129–136. Arbos, P., Wirth, M., Arangoa, M., Gabor, F., Irache, J.M., 2002b. Gantrez® as a new polymer for the preparation of ligand-nanoparticle conjugates. J. Contr. Rel. 83, 321–330. Bazile, D., Prud’homme, C., Bassoullet, M.T., Marlard, M., Spenlehauer, G., Veillard, M., 1995. Stealth Me.PEG-PLA nanoparticles avoid uptake by the mononuclear phagocytes system. J. Pharm. Sci. 84, 493–498. Chen, J., Yang, W.L., Li, G., Qian, J., Xue, J.L., Fu, S.K., Lu, D.R., 2004. Transfection of mEpo gene to intestinal epithelium in vivo mediated by oral delivery of chitosan-DNA nanoparticles. World J. Gastroenterol. 10, 112–116. Dawson, G.F., Halbert, G.W., 2000. The in vitro cell association of invasin coated polylactide-co-glycolide nanoparticles. Pharm. Res. 17, 1420–1425. De Jaeghere, F., Allemann, E., Feijen, J., Kissel, T., Doelker, E., Gurny, R., 2000. Cellular uptake of PEO surface-modified nanoparticles: evaluation of nanoparticles made of PLA:PEO diblock and triblock copolymers. J. Drug Target. 8, 143–153. Dubin, P., Strauss, U., 1967. Hydrophobic hypercoiling in copolymers of maleic acid and alkyl vinyl ethers. J. Phys. Chem. 71, 2757–2759. Esposito, P., Colombo, I., Lovrecich, M., 1994. Investigation of surface properties of some polymers by a thermodynamic and mechanical approach: possibility of predicting mucoadhesion and biocompatibility. Biomaterials 15, 177–182. Florence, A.T., Hussain, N., 2001. Transcytosis of nanoparticle and dendrimer delivery systems: evolving vistas. Adv. Drug Del. Rev. 50, S69–S89. Gref, R., Minamitake, Y., Peracchia, M.T., Trubetskoy, V., Torchilin, V., Langer, R., 1994. Biodegradable long-circulating nanospheres. Science 263, 1600–1603. Gref, R., Domb, A., Quellec, P., Blunk, T., Muller, R.H., Verbavatz, J.M., Langer, R., 1995. The controlled intravenous delivery of drugs using PEG-coated sterically stabilized nanospheres. Adv. Drug Deliv. Rev. 16, 215–233. Huang, Y., Leobandung, W., Foss, A., Peppas, N., 2000. Molecular aspects of muco- and bioadhesion: tethered structures and site-specific surfaces. J. Contr. Rel. 65, 63–71. Hussain, N., Jani, P.U., Florence, A.T., 1997. Enhanced oral uptake of tomato lectin-conjugated nanoparticles in the rat. Pharm. Res. 14, 613–618. Iwanaga, K., Ono, S., Narioka, K., Morimoto, K., Kakemi, M., Yamashita, S., Nango, M., Oku, N., 1997. Oral delivery of insulin by using surface coating liposomes. Improvement of stability of insulin in GIT. Int. J. Pharm. 157, 73–80.

419

Jung, T., Kamm, W., Breitenbach, A., Kaiserling, E., Xiao, J.X., Kissel, T., 2000. Biodegradable nanoparticles for oral delivery of peptides: is there a role for polymers to affect mucosal uptake? Eur. J. Pharm. Biopharm. 50, 147–160. Ladaviere, C., Delair, T., Domard, A., Pichot, C., Mandrand, B., 1999a. Covalent immobilization of biological molecules to maleic anhydride and methyl vinyl ether copolymers – a physico-chemical approach. J. Apll. Polym. Sci. 71, 927–936. Ladaviere, C., Delair, T., Domard, A., Pichot, C., Mandrand, B., 1999b. Covalent immobilization of bovine serum albumin onto (maleic anhydride-alt-methyl vinyl ether) copolymers. J. Apll. Polym. Sci. 72, 1565–1572. Lee, J.W., Park, J.H., Robinson, J.R., 2000. Bioadhesive-based dosage forms: the next generation. J. Pharm. Sci. 89, 850–866. Lehr, C.M., 1996. From sticky stuff to sweet receptors – achievements, limits and novel approaches to bioadhesion. Eur. J. Drug Metab. Pharmacokinet. 21, 139–148. Mosqueira, V., Legrand, P., Gulik, A., Bourdon, O., Gref, R., Labarre, D., Barratt, G., 2001. Relationship between complement activation, cellular uptake and surface physicochemical aspects of novel PEGmodified nanocapsules. Biomaterials 22, 2967–2979. Neal, J.C., Stolnik, S., Schacht, E., Kenawy, E.R., Garnett, M.C., Davis, S.S., Illum, L., 1998. In vitro displacement by rat serum of adsorbed radiolabeled poloxamer and poloxamine copolymers from model and biodegradable nanospheres. J. Pharm. Sci. 87, 1242–1248. Peracchia, M.T., Vauthier, C., Passirani, C., Couvreur, P., Labarre, D., 1997. Complement consumption by poly(ethylene glycol) in different conformations chemically coupled to poly(isobutyl 2-cyanoacrilate) nanoparticles. Life Sci. 61, 749–761. Ponchel, G., Irache, J.M., 1998. Specific and non-specific bioadhesive particulates for oral delivery. Adv. Drug Deliv. Rev. 34, 191–219. Russell-Jones, G.J., Arthur, L., Walker, H., 1999. Vitamin B12 -mediated transport of nanoparticles across Caco-2 cells. Int. J. Pharm. 179, 247–255. Sanders, N.N., De Smedt, S.C., Cheng, S.H., Demeester, J., 2002. Pegylated GL67 lipoplexes retain their gene transfection activity after exposure to components of CF mucus. Gene Ther. 9, 363–371. Stolnik, S., Illum, L., Davis, S.S., 1995. Long circulating microparticulate drug carriers. Adv. Drug Deliv. Rev. 16, 195–214. Tobio, M., Gref, R., Sanchez, A., Langer, R., Alonso, M.J., 1998. Stealth PEG-PLA nanoparticles as protein carriers for nasal administration. Pharm. Res. 15, 270–275. Tobio, M., Sanchez, A., Vila, A., Soriano, I., Evora, C., Vila-Jato, J.L., Alonso, M.J., 2000. The role of PEG on the stability in digestive fluids and in vivo fate of PEG-PLA nanoparticles following oral administration. Coll. Surf. B: Biointerf. 18, 315–323. Torchilin, V.P., 1998. Polymer-coated long-circulating microparticulate pharmaceuticals. J. Microencapsulation 15, 1–19. Vila, A., Gill, H., McCallion, O., Alonso, M.J., 2004. Transport of PLAPEG particles across the nasal mucosa: effect of particle size and PEG coating density. J. Contr. Rel. 98, 231–244. Willits, R.K., Saltzman, W.M., 2001. Synthetic polymers alter the structure of cervical mucus. Biomaterials 22, 445–452.