Thiomer nanoparticles: Stabilization via covalent cross-linking

Abstract

The purpose of this study was the development of stable thiomer nanoparticles for mucosal drug delivery. Chitosan-thioglycolic acid (chitosan-TGA) nanoparticles (NP) were formed via ionic gelation with tripolyphosphate (TPP). In order to stabilize the NP inter- and intra-molecular disulfide bonds were formed via controlled oxidation with hydrogen peroxide (H₂O₂). Thereafter, stability was investigated in saline and simulated body fluids at pH 2 and pH 5.5 via optical density measurements. The mucoadhesive properties were evaluated in vitro on freshly excised porcine intestinal mucosa via the rotating cylinder method. Particles had a mean size of 158 ± 8 nm and a zeta potential of ~ + 16 mV. Three different degrees of oxidation were adjusted by the addition of H₂O₂ in final concentrations of 10.60 µmol (chitosan-TGA (ox1)), 21.21 µmol (chitosan-TGA (ox2)), and 31.81 µmol (chitosan-TGA (ox3)) leading to 60%, 75%, and 83% of oxidized thiol groups, respectively. More than 99% of chitosan-TGA (ox3) NP, 70% of chitosan-TGA (ox2) NP, and 50% of chitosan-TGA (ox1) NP were stable over a 60-min period in simulated gastric fluid. In contrast, only 10% of unmodified chitosan and chitosan-TGA NP which were just ionically cross-linked remained stable in the same experiment. The adhesion times of covalently cross-linked chitosan-TGA (ox1), chitosan-TGA (ox2), and chitosan-TGA (ox3) were ~ 41-fold, 31-fold, and 25-fold longer in comparison to unmodified ionically cross-linked chitosan. The method described here might be useful for the preparation of stable nanoparticulate drug delivery systems.

Keywords::

• Thiolated chitosan
• nanoparticles
• stability
• mucoadhesion
• delivery system

Introduction

Nanotechnology is enabling technology that deals with nanometer sized objects to create new materials and devices with a vast range of applications (Salata, 2004). One common field of application is the pharmaceutical industry which occupies, for example with the development of advanced drug delivery systems, new therapies and in vivo imaging. Drug delivery focuses on maximizing bioavailability both at specific target sites in the body and over a prolonged period of time. Among various strategies applied in order to reach this goal are nanoparticulate drug carrier systems (LaVan et al., 2003). Such carrier systems consist of lipids or biopolymers. In particular, biodegradable biopolymers produced by microorganisms such as alginate, hyaluronic acid, and chitosan are commonly used in the pharmaceutical industry because they are generally recognized as safe. In this study chitosan was chosen because of its high mucoadhesive properties based on its cationic character (Luessen et al., 1996), permeation-enhancing properties (Artursson et al., 1994), and well-established techniques to formulate NP out of it (Hu et al., 2008). The most commonly used method to produce chitosan NP is by in situ gelation of the cationic polymer with polyanions (Csaba et al., 2009). Nevertheless, there are also disadvantages of ionically cross-linked chitosan NP such as rapid disintegration under low pH conditions and a rapid release of encapsulated drugs. Consequently, their practical application is strongly limited (Pan et al., 2002; Dhawan et al., 2004; Zhang et al., 2004). To that purpose the NP must stay stable until they arrive at their target. Furthermore, the positive charge of chitosan is neutralized by addition of polyanions in general, which results in a loss of mucoadhesive properties (Luessen et al., 1996). A promising strategy to overcome these shortcomings, as described above, thiolated chitosan was chosen for this study, because of its greater stability by forming disulfide bonds due to the addition of H₂O₂ to generate a stable matrix carrier system. Further benefits of thiolated chitosans are comparatively higher mucoadhesive, efflux pump inhibitory and permeation enhancing properties. Accordingly, thiolated particles should display greater stability (Bernkop-Schnürch et al., 2006a), comparatively higher mucoadhesion (Bernkop-Schnürch & Steininger, 2000; Grabovac et al., 2005), and controlled drug release (Greimel et al., 2007) based on the formation of stabilizing disulfide bonds. However, as disulfide bonds within thiolated chitosan NP are on the one hand responsible for stabilization and remaining free thiol groups are on the other hand responsible for improved mucoadhesive properties with cystein-rich sub-domains of glycoproteins (Luessen et al., 1996), the ratio between oxidized and non-oxidized thiol groups on the polymer has to be well-balanced. In consequence, the aim of this study was to investigate the pH impact on the stability of chitosan and chitosan-TGA NP by measuring the optical density in contrast to previously realized studies (Bernkop-Schnürch et al., 2006a, b). Therefore, quantities of the oxidized thiol groups which are necessary to ensure stability were determined. Moreover, the correlation between non-oxidized thiol groups and mucoadhesive properties was evaluated. In this way, nanoparticulate stability was characterized under oral and intravesical simulated conditions, and in vitro mucoadhesion studies with the rotating cylinder method were carried out in order to evaluate the suitability of chitosan-TGA NP as a mucosal drug delivery system.

Materials and methods

Materials

Chitosan (middle-viscous), thioglycolic acid (TGA), 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDAC), 5.5-dithiobis (2-nitrobenzoic acid), Tris (2-carboxyethyl) phosphine hydrochloride (TCEP− HCl), sodium nitrite, and hydrogen peroxide were purchased from Sigma-Aldrich (Austria). All other chemicals were of analytical grade.

Preparation of low molecular weight chitosan

First, 4 g of chitosan was dissolved in 400 ml of acetic acid (6% (v/v)) to get a clear solution. Subsequently, a solution of 160 mg sodium nitrite in 10 ml demineralized water was added. After 1 h incubation under continuous stirring, chitosan was precipitated by addition of 5 M sodium hydroxide solution until pH 9 was reached. The precipitate was filtered and washed by cold acetone. The received low molecular weight chitosan was resolubilized in 200 ml of 0.1 M acidic acid and exhaustively dialyzed against demineralized water. The product was lyophilized under vacuum 0.2 mbar and at −45°C (Benchtop 2K, VirTis, NY).

Synthesis of chitosan-TGA conjugates

The covalent attachment of TGA to chitosan was achieved by the formation of amide bonds between primary amino groups of the polymer and the carboxylic acid groups of TGA mediated by EDAC, as described by our research group previously (Greindl, 2008).

Determination of thiol/disulfide groups

The degree of free thiol groups immobilized on the polymer was quantified by Ellman’s method (Bernkop-Schnürch et al., 1999). In brief, 0.5 mg of thiolated chitosan was dissolved in 500 µl of 0.5 M phosphate buffer pH 8 and 500 µl of Ellman’s reagent (3 mg of 5.5-dithiobis (2-nitrobenzoic acid) in 10 ml 0.5 M phosphate buffer pH 8) were added. The samples were incubated at 37°C and protected from light for 4 h. After that, 100 µl of the sample was analyzed by a microplate reader (FluoStar Galaxy, BMG, Offenburg, Germany) at a wavelength of 450 nm to determine the content of thiol groups. The content of disulfide bonds was measured after reduction with NaBH₄ and determined with Ellman’s reagent as described above.

Preparation of chitosan-TGA NP

The NP were prepared in aqueous solution via in situ gelation of unmodified chitosan and chitosan-TGA, respectively, with TPP according to a method described previously (Bernkop-Schnürch et al., 2006b). Ionic gelation is based on electrostatic interactions between positively-charged amine groups of chitosan and negatively-charged groups of polyanions such as TPP. In brief, unmodified and thiolated low molecular weight chitosan, respectively, were dissolved in 0.05 M acetic acid/sodium acetate buffer solution (pH 6.2) at a final concentration of 0.2% (w/v). Afterwards a 0.2% (w/v) TPP solution in demineralized water was added to the low molecular weight chitosan solution to create ionically cross-linked NP, which were spontaneously formed under mechanical stirring.

Stabilization of chitosan-TGA NP

Afterwards the particle suspensions were partially oxidized by the addition of 0.5% (v/v) H₂O₂ solution with different amounts and the mixture was incubated for 1 h under continuous stirring at room temperature. Thereafter, TPP and H₂O₂ were removed by exhaustive dialysis in tubings (molecular weight cut-off 12 kDa; dialysis tubing, cellulose membrane; Sigma-Aldrich, Austria) for 1 day 5-times at 10°C in the dark against 5 mM HCl. Subsequently, 3% (w/v) of trehalose were added to the suspension in order to avoid the aggregation of NP and to enhance the suspendability after the lyophilization (Benchtop 2K, VirTis, NY).

Particle characterization

The amount of thiol groups on the particles was quantified via Ellman’s method as described above. Size distribution and zeta potential of particles were determined with a particle sizer (Zeta Potential/Particle Sizer, Nicomp^™ 380 ZLS, PSS, Santa Barbara, CA). Measurements were performed via dynamic light scattering analyses of particle suspensions in demineralized water at room temperature.

Stability studies of unmodified chitosan NP and chitosan-TGA NP

Different degrees of oxidation were adjusted by the addition of H₂O₂ and TCEP-HCl. This resulted in not at all oxidized NP by reduction with TCEP-HCl (Getz et al., 1999), to a certain extent oxidized NP due to the addition of H₂O₂ in the final concentrations of 10.60 µmol (chitosan-TGA (ox1)), 21.21 µmol (chitosan-TGA (ox2)), and 31.81 µmol (chitosan-TGA (ox3)) and entirely oxidized NP with 881.83 µmol of H₂O₂. The stability of the obtained lyophilized NP towards pH and ionic strength changes was tested. On the one hand NP were exposed to different simulated body fluids of 0.1 M buffer solutions at pH 2 (potassium chloride/hydrogen chloride buffer) and 5.5 (acetic acid/sodium acetate buffer) and on the other hand to 0.9% (w/v) sodium chloride solution. Therefore, each lyophilized nanoparticle suspension as described above was resuspended in simulated gastric fluid, simulated bladder fluid, and sodium chloride solution. Subsequently, 100 µl of each suspension was removed and added to a 96-well microplate. Afterwards the optical density at 600 nm was determined with a microplate reader (FluoStar Galaxy, BMG, Offenburg, Germany) over a 60-min period to quantify the stability of NP. Physiological conditions between the measurements were given by continuous shaking of the microplate and a temperature of 37°C in the microplate reader.

Tablets manufacture

Lyophilized covalently cross-linked chitosan-TGA polymers and ionically cross-linked chitosan, which was used as control, were compressed into 30 mg, 5.0 mm diameter flat-faced test discs (hand hydraulic press PW 30, Paul Otto Weber GmbH, Germany). The compaction pressure (force of 11 kN) was kept constant during the preparation of test discs.

In vitro mucoadhesion studies with the rotating cylinder method

In order to evaluate the in vitro mucoadhesion an Erweka DT 700 (Erweka GmbH, Heusenstamm, Germany) dissolution tester was used. Test discs as described above were attached to a freshly excised intestinal porcine mucosa, which has been spanned on a stainless steel cylinder (diameter: 4.4 cm; height: 5.1 cm). The vessel was filled with 900 ml of 0.1 M phosphate buffer pH 6.8 at 37°C ± 0.5°C. Thereafter, the cylinder was placed into the dissolution apparatus at a rotating speed of 125 rpm. The detachment of test discs was determined visually by digital recording with a webcam placed in front of the dissolution tester until the last test discs detached (Bernkop-Schnürch & Steininger, 2000).

Statistical data analysis

Statistical data analyses were performed using the student t-test, with p < 0.05 as the minimal level of significance. Calculations were done using the software Minitab version 15.1.3.

Results

Preparation and characterization of chitosan-TGA

Low molecular mass chitosan with a mean molecular mass of 10 kDa was obtained according to a method described previously (Schmitz et al., 2007). Thioglycolic acid was covalently attached to this low molecular mass chitosan by formation of amide bonds. The obtained polymer conjugate was white, odorless, of fibrous structure, and easily soluble in aqueous solution at pH 6 and lower. The resulting amount of free thiol groups immobilized on 1 g of chitosan was 1527 ± 13 µmol.

Particle preparation

The chitosan-TGA NP were formed spontaneously via ionic gelation by mixing 0.2% (w/v) chitosan-TGA solution with 0.2% (w/v) TPP solution in a ratio of 7.5:1. The NP displayed a mean particle diameter of 158 ± 8 nm with a narrow size distribution, which is illustrated in Figure 1. Generally, the sizes of NP before dialysis were comparatively smaller (data not shown) than after exhaustive dialysis. Lopez-Leon et al. (2005) investigated the destabilization of chitosan NP during storage. The results of this study demonstrated that particle size and standard deviation increased over time. This suggested that ionically cross-linked NP with TPP are a metastable system. Therefore, they must be stored lyophilized and only prepared when required. In order to increase the stability of chitosan-TGA NP the formation of disulfide bonds was achieved by oxidation due to addition of increasing amounts of H₂O₂. The more H₂O₂ was added, the more disulfide bonds were formed, as shown in Table 1.

Table 1.  Amount of thiol groups and disulfide bonds immobilized on the basic thiomer chitosan–TGA and nanoparticles after ionic gelation with TPP and different degrees of oxidation with H₂O₂, respectively. Indicated values are means ± SD (n ≥ 3).

	H2O2 [µmol]	–SH [µmol/g]	–S–S– [µmol/g]	S-SH [µmol/g]
Chitosan-TGA	—	1427	50	1527 ± 13
Chitosan-TGA (+TPP)	—	764	380	1524 ± 10
Chitosan-TGA (ox1)	10.60	309	607	1523 ± 14
Chitosan-TGA (ox2)	21.21	189	668	1525 ± 12
Chitosan-TGA (ox3)	31.82	133	695	1523 ± 9

Figure 1.  Size distribution by intensity of ionically cross-linked NP based on chitosan [◊] and chitosan-TGA [□], as well as covalently cross-linked NP based on thiolated chitosan obtained by oxidation with H₂O₂ (ox1 [x], ox2 [♦], ox3 [○]). Indicated values are means ± SD of last three experiments.

Stability studies

The stability of chitosan-TGA NP towards changes in pH and ionic strength was investigated. Lyophilized NP were exposed to different simulated body fluids at pH 2 and 5.5 and their stability was determined by measuring the optical density. Results are illustrated in Figures 2 and 3. The particles were defined as stable when the measured optical density showed no deviation of more than 5% from initial value. In case of chitosan-TGA NP, it was shown that higher disulfide bond content resulted in higher stability at lower pH levels. Ionically cross-linked NP rapidly disintegrated in aqueous solution at pH 2 and pH 5.5, whereas the same NP remained stable, being covalently cross-linked via disulfide bonds. Due to this strongly improved stability of chitosan-TGA NP via covalent cross-linking, the ionic-cross-linker TPP was not necessary anymore. Finally, the developed NP indicated a zeta potential of – +16 mV, which is posed in Table 2. For ionic strength changes the NP were exposed to 0.9% sodium chloride solution to induce colloidal destabilization. Results of this study suggested that ionically cross-linked particles disintegrated in the presence of low or moderate salt concentrations, whereas cross-linked particles via disulfide bonds were stable, as shown in Figure 4.

Table 2.  Mean particle diameter and zeta potential of chitosan–TGA nanoparticles obtained by ionic gelation with TPP and followed by different oxidation with H₂O₂, respectively. Indicated values are means ± SD (n ≥ 3).

Nanoparticles	Mean particle diameter (nm)	Zeta potential (mV)
Ionically cross-linked
Chitosan	197 ± 5	7 ± 1
Chitosan-TGA	185 ± 3	7 ± 1
Covalently cross-linked
Chitosan TGA (ox1)	160 ± 11	16 ± 1
Chitosan TGA (ox2)	156 ± 6	16 ± 4
Chitosan TGA (ox3)	157 ± 9	16 ± 2

Figure 2.  Stability of different modified NP in simulated gastric fluid of pH 2. Studies were carried out with completely oxidized chitosan-TGA [Δ], chitosan-TGA (ox3) [○], chitosan-TGA (ox2) [♦], chitosan-TGA (ox1) [x], reduced chitosan-TGA [□], and chitosan [◊]. Indicated values are means ± SD (n ≥ 3).

Figure 3.  Stability of different modified NP in simulated bladder fluid of pH 5.5. Studies were carried out with completely oxidized chitosan-TGA [Δ], chitosan-TGA (ox3) [○], chitosan-TGA (ox2) [♦], chitosan-TGA (ox1) [x], reduced chitosan-TGA [□], and chitosan [◊]. Indicated values are means ± SD (n ≥ 3).

Figure 4.  Stability of different modified NP in a moderate sodium chloride concentration of 0.9%. Studies were carried out with chitosan-TGA (ox3) [○], chitosan-TGA (ox2) [♦], chitosan-TGA (ox1) [x], reduced chitosan-TGA [□], and chitosan [◊]. Indicated values are means ± SD (n ≥ 3).

Mucoadhesion studies

This study was performed to demonstrate the influence of the ionic-cross-linker TPP on mucoadhesion. For this purpose the covalently cross-linked polymers were distinguished based on the amount of disulfide bonds as described above and were compressed into flat-faced test discs (30 mg, 5.0 mm diameter). Chitosan, ionically cross-linked with TPP, was used as control. The adhesion time of the cross-linked polymers was more than 41-fold (chitosan-TGA (ox.1)), 31-fold (chitosan-TGA (ox2)), and 25-fold (chitosan-TGA (ox3)) increased, in comparison to the ionically cross-linked chitosan. The control test discs displayed an insufficient adhesion time on mucosa as a consequence of the ionic interactions of the cross-linker with the amino groups of chitosan. Due to the addition of TPP, the positive charge of the amino groups was neutralized. For that reason mucoadhesive properties were strongly reduced, which is illustrated in Figure 5.

Figure 5.  Comparison of the residence times on the intestinal porcine mucosa. Studies were carried out with chitosan-TPP, chitosan-TGA (ox3), chitosan-TGA (ox2), and chitosan-TGA (ox1). Indicated values are means ± SD (n ≥ 3).

Discussion

The preparation of chitosan-TGA NP was carried out by mixing TPP solution with chitosan-TGA solution under stirring. The resulting polycation–multivalent anion complex relies on an ionic gelation technique between positively-charged chitosan and negatively-charged TPP (Mi et al., 1999). The effects on the ionic reaction were shown to be dependent on molecular mass and concentration of chitosan, pH of the solution, and mass ratio of chitosan to TPP (Hu et al., 2008). In order to obtain favorable preparation conditions, low molecular mass chitosan-TGA, a concentration of 0.2% (w/v), a pH of 6.2, and a chitosan–TPP ratio of 7.5:1 were identified according to results of several experiments. This preparation technique of chitosan–TGA NP displayed a mean particle diameter of 158 ± 8 nm with a narrow size distribution, which is illustrated in Figure 1.

Stability studies in artificial body fluids demonstrated that the more intra- and inter-particular disulfide bonds were formed, the greater was the stability of oxidized NP. More than 99% of chitosan-TGA (ox3) NP, 70% of chitosan-TGA (ox2) NP, and 50% of chitosan-TGA (ox1) NP remained stable in simulated gastric fluid, which is illustrated in Figure 2. In contrast, 90% of unmodified chitosan NP and reduced chitosan-TGA NP without disulfide bonds rapidly disintegrated within the first 10 min. Under simulated bladder conditions ~ 99% of chitosan-TGA (ox3) NP, chitosan-TGA (ox2) NP, and 85% of chitosan-TGA (ox1) NP were stable, as shown in Figure 3. In comparison to unmodified chitosan NP and reduced chitosan-TGA NP, only 50% disintegrated in the same period as in gastric fluid due to the higher pH value in the bladder fluid. In order to characterize the stability under moderate salt concentrations, studies in a sodium chloride solution of 0.9% were carried out. Thereby, the disulfide cross-linked chitosan-TGA NP exhibited reliability in contrast to ionic cross-linked NP, which is illustrated in Figure 4. This outcome was in good agreement with previous studies performed with chitosan–TPP NP by adding moderate salt concentrations to induce colloidal destabilizations (Lopez-Leon et al., 2005). To conclude, covalent bonds within NP confer more stability to the nanoparticulate drug delivery system than weak ionic interactions between cationic polymers and anionic cross-linkers. In case of oral drug delivery, the disulfide cross-linked chitosan-TGA NP do not need to be additionally enteric coated as stability in gastric fluids is obviously provided. Regarding to the biodegradation of stabilized particles numerous studies were previously published by our research group. It was shown that unmodified polymers as well as thiomers were degraded under physiological conditions by the addition of lysozyme. Results of these studies demonstrated a faster degradation of unmodified polymers in comparison to thiomers (Kast & Bernkop-Schnürch, 2001; Kafedjiiski et al., 2007; Schmitz et al., 2008). With regard to covalently cross-linked particles by the formation of disulfide bonds, Groll et al. (2009) proved the cleavage of disulfide bonds in the presence of glutathione under physiological conditions. Consequently, particle disintegration is not affected.

In former studies the advantages of cross-linked thiolated NP were already discussed by our research group but not yet investigated in detail (Bernkop-Schnürch et al., 2006a). On the one hand, disulfide bonds are responsible for the stability of thiolated chitosan NP, but on the other hand free thiol groups are essential for improved mucoadhesive and permeation enhancing properties (Bernkop-Schnürch et al., 2004; Roldo et al., 2004). The improved mucoadhesive properties of thiomers are explained by the formation of covalent bonds between thiol groups of the polymer and cysteine-rich sub-domains of glycoproteins in the mucus layer (Bernkop-Schnürch, 2005). In order to achieve that goal, the oxidation has to be well controlled during the preparation process to maintain enough free thiol groups, as shown in Table 1. Controlled oxidations could be reached by target-oriented addition of H₂O₂ in final concentrations of 10.60 µmol (chitosan-TGA (ox1)), 21.21 µmol (chitosan-TGA (ox2)), and 31.81 µmol (chitosan-TGA (ox3)). Thereby, amounts of 309 µmol/g (chitosan-TGA (ox1)), 189 µmol/g (chitosan-TGA (ox2)), and 133 µmol/g (chitosan-TGA (ox3)) sulfhydryl groups were obtained. Results of this study indicate that even an amount of 133 µmol/g of free thiol groups after controlled oxidation is sufficient to provide superior mucoadhesion compared to ionically cross-linked chitosan, which is illustrated in Figure 5. The adhesion time of covalently cross-linked chitosan-TGA (ox1), chitosan-TGA (ox2), and chitosan-TGA (ox3) were ~ 41-fold, 31-fold, and 25-fold increased, while unmodified ionically cross-linked chitosan detached after 42 min. Accordingly, with increasing amounts of sulfhydryl groups the mucoadhesion time ascends.

Otherwise the unique presence of disulfide bonds to induce nanoparticulate stability enabled under acetic conditions the protonation of the amino groups for additional interactions with anionic sialic acid moieties of the mucus. This contributed also to higher mucoadhesion by formation of non-covalent bonds (Bernkop-Schnürch, 2000). The increase of zeta potential due to the removal of TPP exemplifies the availability of protonated amino groups. Results are shown in Table 2. Another benefit of the cationic mucoadhesive polymer is the possibility to provide sustained release for anionic drugs. In the case of ionic polymers, a sustained release can be guaranteed if the therapeutic agent possesses the opposite charge of the polymer (Bernkop-Schnürch & Freudl, 1999).

Altogether the developed nanoparticulate drug carrier system based on thiolated chitosan affords besides the stability of NP an adequate mucoadhesion as a result of controlled oxidation with H₂O₂. In contrast to ionically cross-linked NP, covalently cross-linked NP are not influenced by factors such as ionic strength and pH. With the aid of this drug delivery system it is possible to cover a larger area of applications for mucosal membranes.

Conclusion

Within this study the impact of the degree of cross-linking on stability and mucoadhesive properties of thiolated chitosan NP was investigated. Due to this covalent cross-linking the NP displayed a greater stability and superior mucoadhesion than the corresponding ionically cross-linked particles via TPP. Based on these results the ratio between oxidized and non-oxidized thiol groups on the polymer can be adjusted to the surrounding milieu at the application area in order to obtain adequate dosage forms for mucosal drug delivery. The controlled oxidation technique described here might be a useful tool for the design of nanoparticulate drug delivery systems in order to achieve sufficient stability and mucoadhesion.

Acknowledgment

This work was supported by the BFS project (Bayrische Forschungsstiftung) (ID 3084) and the Nano-Health project (No. 0200) as part of the Austrian Nano-Initiative being financed by the Austrian FFG (Forschungsförderungsgesellschaft mbH) (Project No. 819721).

Declaration of Interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.