Elaboration of targeted nanodelivery systems based on colloidal polyelectrolyte complexes (PEC) of chitosan (CH)-dextran sulphate (DS)

Materials

Chitosan was from Mahtani chitosan PVT, Veraval, India, batch 113, Mw 430,000 g/mol and Degree of Acetylation of 5% extracted from squid pen chitin. The sample was first purified by filtration through Millipore membranes (Millipore, Molsheim, France). Purified high molar mass chitosans were N-acetylated with acetic anhydride in a hydro-alcoholic mixture according to the procedure previously described by Vachoud [26]. After re-acetylation, chitosans were neutralised, rinsed with deionised water, and then freeze-dried. In addition, controlled nitrous deaminations [27] were carried out to produce low molar mass polymers. Chitosans were dissolved at 0.5% (w/v) in a 0.2 M acetic acid/0.1 M sodium acetate buffer. A 0.15 M sodium nitrite solution was added to chitosan solutions to obtain a nitrite/glucosamine unit molar ratio of 0.5. The reaction was run under moderate magnetic stirring for 1 h to obtain Mw lower then 150,000 g/mol. Dextran sulphate with a Mw of 740,000 g/mol (DS 500 k) were provided by Sigma (Saint Quentin Fallavier, France) and used without further purification. The antiretroviral drug model tenofovir (99% purity) was purchased from R&D Systems (Lille, France). The antibodies Immunoglobulin A (IgA) anti-hCEA and anti-α4β7 were provided by B-Cell Design (Limoges, France). Their molecular weights and isoelectric points were ca 170 kg/mol and 6.8, respectively, as provided by the producer. Their concentrations were confirmed by BCA Assay according to the procedure provided by Pierce (Thermo Fischer Scientific, Courtabœuf, France), phosphate buffer solution (PBS) and water from Invitrogen^® and citric acid from Sigma (Saint Quentin Fallavier, France).

Polymer characterization

Degrees of acetylation were determined on purified chitosans by ¹H NMR spectroscopy (Varian, 500 MHz, Palo Alto, USA), according to the method developed by Hirai [28]. The water content was determined by Thermogravimetric analysis (TGA) (SETARAM, Caluire, France). The weight-average molecular weight and the polydispersity index (PI) were measured by size exclusion chromatography (SEC) (3000 and 6000 PW TSK gel columns, 7.8 mm inner diameter and 300 mm length) coupled on line with a differential refractometer (Waters 410) and a multi-angle-laser-light-scattering (Waters, Guyancourt, France) spectrophotometer equipped with a 5 mW He/Ne laser operating at k=632.8 nm. Analyses were performed in micro-batch mode using the K5 flow cell. A degassed 0.2 M acetic acid/0.15 M ammonium acetate buffer (pH 4.5) was used as eluent. The flow rate was maintained at 0.5 mL/min. Refractive index increments (dn/dc) were determined from a master curve previously established under identical conditions, in the same solvent and with an interferometer (NFT ScanRef, Gottingen, Germany). The chitosan used in this investigation had a DA of 48% and an Mw of 130,000 g/mol. The molecular characteristic of DS was also determined by SEC using the same system as described above with a PL aquagel-OH mixed column. A 0.1N NaNO₃ solution adjusted at pH 7.0 was used as solvent and eluent.

Polyelectrolyte complex formation

Chitosan was dispersed at 0.1% (w/w) in Versol^® water (Aguettant, France) containing 50 mM NaCl [17], taking into account the residual water. Dissolution was achieved by adding a stoichiometric amount of acetic acid, with respect to the free amino functions. Then, solutions were adjusted to pH 4.0 with 0.1 M sodium hydroxide or hydrochloric acid. Dextran sulphate solutions, at 0.1% (w/w), were prepared directly in Versol^® water, 50 mM sodium chloride was added to obtain the required ionic strength and pH was adjusted at 4.0. Both solutions were filtered through 0.22 μm Millipore membranes before use.

Colloidal PECs were formed in non-stoichiometric conditions at a molar charge ratio R (n⁺/n^–)=2, at room temperature using chitosan as starting solution. The solution containing DS was added in one shot to the starting solution, under a constant magnetic stirring of 750 rpm. The final volume of the particles dispersion was 37.4 mL (30 mL CH and 7.4 mL DS) with a solid content of 0.1% w/w. In all experiments, the initial amino and sulphate concentrations in the starting solutions of chitosan and dextran sulphate, respectively, were set at 6×10^–3 M corresponding to a weight concentration close to 0.1% for both polymer solutions. Because of non-stoichiometric conditions, the polymer in excess was not completely consumed, thus, a low amount of free polymer still remained in solution. To remove it, particles were separated from the continuous phase by centrifugation at 7000 g for 30 min at 20°C. The supernatant was discarded and the particles were re-dispersed in three different buffers: PBS pH 7.4, 150 mM NaCl or 0.1 M citric acid, pH 5 at different concentrations from 2% to 0.05%.

For the preparation of tenofovir loaded CH-DS nanoparticles, 10 mg of TF were dissolved in 1 mL of 1 M NaOH. The drug solution was dropped into the chitosan solution under magnetic stirring (10–60 μL of the TF solution added to 10 mL of 0.1%, chitosan solution) followed by the addition of the dextran sulphate solution. Nanoparticles were recovered by centrifugation at 10,000 g and 20°C for 60 min. The supernatant was used to determine the drug encapsulation efficiency (EE%).

Nanoparticles characterisation

The solid content (PSC) was defined by the ratio between the weight of dried particles at 60°C for 24 h to the initial weight of the dispersion.

Dynamic light scattering measurements of PECs dispersions were carried out using a Malvern Nanosizer SZ (Malvern, Orsay, France) equipped with a 10 mW He/Ne laser beam operating at λ=633 nm (at 173° scattering angle). All measurements were performed in triplicate at 25°C. The self-correlation function was expanded in a power series (cumulants method). For a monodisperse colloid, the polydispersity index should be below 0.05, but values up to 0.5 can be used for comparison purposes [29].

Zeta potentials were derived from electrophoretic mobility measurements using Smoluchowski’s equation. Particles electrophoretic mobilities (μE) were determined at 25°C with the Malvern Nanosizer SZ. μE was expressed as the average of 10 measurements with a relative error of 5% and were performed by suspending PECs dispersions in 10^–3 M NaCl solutions.

Thermal analyses DSC runs of PECs were performed on a calorimeter (Charbonnières-les-Bains, France). Samples of 2–10 mg of lyophilised PECS were sealed in aluminium pans and heated from 20 to 450°C at a rate of 10°C/min, under constant purging argon flow rate of 25 mL min^–1.

TG-DTG analyses were carried out using a Setaram TGA. 10–20 mg of samples were put in a ceramic pot and heated at a typical heating rate of 10°C/min from room temperature up to 800°C in air atmosphere (flow rate: 20 mL/min).

Microscopy (SEM) images were obtained using a Hitachi S-4800 microscope at 5 kV. A droplet of 0.01% (v/v) nanoparticles dispersion was deposited on a sample holder, air-dried at room temperature (12 h), and coated with palladium in a cathode evaporator (Technics Hummer II) under an argon atmosphere. The SEM was carried out at the ‘Centre Technologique des Microstructures’, Université Claude Bernard Lyon 1.

Encapsulation efficacy

The content of tenofovir was calculated from the difference between the total amount of drug added in the nanoparticle preparation and the amount of free drug in the supernatants. The amount of free or unencapsulated drug was measured by UV spectrophotometer Model 680 Microplate Reader (BioRad, France) at a wavelength of 259 nm, using a calibration curve established in the same buffer. The lowest detectable concentration of tenofovir was 0.1 μg/mL.

Antibody sorption onto colloidal PECs

The sorption process consisted in mixing equal volumes of particle dispersion (CH-DS) and antibody solution (IgA) with moderate end-overhead stirring. Various solid contents and antibody concentrations were obtained by dilution of the initial particles dispersion and antibody solution with the same buffer. IgA/CH-DS particles were centrifuged 10 min at 14,000 g to remove potentially residual particles. The supernatant was separated; the pellet was resuspended in an identical buffer volume. Sorbed IgA was deduced from free IgA in the supernatant obtained by BCA assay titration, according to the manufacturer’s instructions, calibrated via serial dilutions in the same experimental buffer.

The sorption yield was calculated as follows:

where [IgA] input is the IgA concentration titrated in the control sample (i.e., all the reactants were present but not the particle suspension) of the original IgA solution for each independent experiment; [IgA] residual is obtained from titration of the supernatant, taking into account the background signal from a blank experiment in which all the reactants were present but the protein.

ELISA assays

The presence of either anti hCEA or anti-α4β7 of antibodies on the CH-DS particle surface was assayed by ELISA, as described in Figure 1. Ninety-six-well plates were coated overnight at room temperature with 100 μL of a 1 μg/mL antigen solution in PBS buffer (CEA antigen and Recombinant Human Integrin α4β7, from respectively Eurobio-Abcys and R&D System, France). The plates were then post-coated for 1 h at 37°C with 200 μL of 10% non-fat dry milk PBS and washed three times with 0.05% Tween20-PBS (PBST). Serial dilutions of IgA/CH-DS particles were prepared in 1% bovine serum albumin (BSA)-PBS; 100 μL were added to the plates in duplicates and incubated for 1 h at 37°C. After three washes with PBST buffer, peroxidase-conjugated IgA anti-human IgA-HRP (Invivogen) at a concentration of 0.2 μg/mL in 1% BSA-PBS was added and the plate was incubated for 1 h at 37°C. After washing, plates were developed with tetramethylbenzidine (TMB) (BD Pharmingen, France) prepared according to the manufacturer’s instructions, for 1min in the dark and the reaction was quenched with 100 μL of 1 N sulphuric acid. The absorbance at 450 nm was measured with a Model 680 Microplate Reader.

Figure 1

ELISA model [tetramethylbenzidine (TMB), enzyme substrate].

Nanoparticles elaboration and stability in the presence of different buffers

Particles were prepared by polyelectrolytes complexation of oppositely charged, CH and DS as described in the Methods Section. The PECs were elaborated in different buffers and characterised the nanoparticles by different technics (TGA, DSC and SEM).

The TGA curves of PECs and precursors (CH, DS and CH-DS) are shown in Figure 2. Two mass losses were observed. The first one, ranging from 25 to 200°C, was attributed to the evaporation of weakly bound water 11~14%. CH-DS presented the highest water content, CH an intermediate quantity and DS the smallest value. The most important water loss was founded for the nanoparticles, probably resulting from the entrapment of water in the void volume after centrifugation. The degradation of CH, DS and CH-DS occurred on a second weight loss stage as clearly shown in the TGA curves (Figure 2). The exact value of samples degradation temperature was obtained through derivatives (DTG) of respective TGA curve. The degradation temperature of the complex was about 250°C, between DS (227°C) and CH (307°C), confirming that the polymers were effectively associated and not simply physically blended. As a result, the CH-DS PECs particles were more thermally stable than DS, but less than CH [30], which is consistent with others works using chondroitine sulphate [31], peptide [32] or alginates [33] as polyanions. The DSC thermograms of CH-DS nanoparticles and precursors are depicted in Figure 3. All curves exhibited an endothermic peak between 50 and 150°C attributed to the loss of water and/or a possible chain relaxation [34]. Exothermic peaks at 307°C and 213°C respectively showed CH and DS degradation. PECs exothermic peak at 222°C can be attributed to the cleavage of electrostatic interactions between both polyelectrolytes [33]. Similar electrostatic interactions were referred to explain chitosan/chondroitin sulphate interactions [35]. As shown by these TGA and DSC results, the stability of the polyelectrolyte complex was lower than chitosan because the formation of strong electrostatic interactions between CH and DS charged groups induced the loss of the crystalline structure of chitosan, as shown by Denuzière [30].

Figure 2

TGA profiles of chitosan (A) CH-DS complex (B) and dextran sulphate (C).

Figure 3

DSC profiles of dextran sulphate (A), CH-DS complex (B) and chitosan (C).

The smooth and spherical morphology of CH-DS nanoparticles observed by SEM after drying (Figure 4) was consistent with the mean size measured by DLS, Table 1. The Zeta potentials in various media were positive, as seen in Table 2. A lower zeta potential of CH-DS was observed in PBS comparing to 0.1 M citric acid, illustrating the pH dependence of the charge density of the chitosan shell.

Figure 4

(A) SEM CH-DS 0,001% in water and (B) zoom of picture (A) (chitosan: DA 48%, Mw 130,000 g/mol; DS Mw 740,000 g/mol).

The dispersion colloidal stability was investigated by storing the particle dispersions at different solid contents for one month at room temperature in the following buffers: 150 mM sodium chloride, PBS and 0.1 M citric acid pH 5. The variations in time of the average particle diameters were monitored by DLS. As seen in Table 1, samples stored in 150 mM sodium chloride aqueous solution remained stable for 1 month for all concentrations with average particle diameter lower than in citric acid and PBS buffers. At lower concentrations, both in PBS and citric acid buffers, the aggregation of the colloids started essentially after 14 days.

Encapsulation of tenofovir

Tenofovir encapsulation in CH-DS nanoparticles was only possible in PBS whereas in 0.1 M citric acid the presence of TF induced the nanoparticle flocculation. The drug encapsulation ratio (w/w), the molar mixing ratio of drug/chitosan size, PI, EE% and zeta potential of CH-DS loaded with tenofovir (TF) are listed in Table 3. After production, all CH-DS nanoparticles were stored at 4°C and remained stable for at least for 2 weeks.

Table 3

Physicochemical characteristics of CH-DS nanoparticles loaded with tenofovir (TF).

TF/CH Ratio, w/w	Ratio	EE%	TF loaded, μg	Zeta potential, mV	Size, nm	PI
0.1	24:1	100	3	12	538	0.2
0.3	8:1	99	9	10	529	0.2
0.6	3:1	96	17	4	560	0.3

EE%, encapsulation efficiency.

IgA Sorption on the CH-DS nanoparticles

The sorption of the model IgAs were investigated in PBS (pH 7.4) to mimic a physiological environment and citric acid (0.1 M, pH 5) to mimic an acidic environment. The results are expressed by the amounts of particle-associated antibody (w/w), calculated as the antibody ratio content by weight units of nanoparticles. A series of experimental conditions were screened (particle concentration, buffers). We observed that, at constant IgA concentration of 10 μg/mL, the amount of IgA immobilised increased as the PECs solids decreased (Table 4). On the colloidal stability standpoint, an irreversible flocculation was observed at high particle concentration and only solid contents of 0.1% and lower maintained the colloidal stability.

The sorption kinetics of anti-hCEA and anti-α4β7 IgAs are shown in Figures 5 and 6 respectively. The sorption process was faster for both IgAs in acidic medium (0.1 M citric acid pH 5) than in PBS at pH 7.4, probably due to the higher charge density at lower pH (Table 2) of CH-DS nanoparticles. Around 80%–98% of the initial IgA input was bound in the first 2 h and 100% of IgA was immobilized after 6 h of incubation in citric acid medium. In PBS, the sorption process started with a fast initial step after 4 h or less, in which 40%–70% of the initial IgA input was bound. Then, the process slowed down to reach completion in 24–72 h depending on the IgA nature and concentration. Interestingly, the sorption kinetics in PBS decreased with increasing the antibody concentration and, for the two highest concentrations; there was a change in adsorption kinetics after 10 h of incubation. These changes in sorption kinetics observed in PBS could be related to different modes of interactions with the colloids involving rearrangement of the already bound IgAs at the surface (or within the chitosan shell) of the colloid, to allow more antibodies to bind till saturation.

Figure 5

Sorption kinetics of anti-hCEA IgA onto CH-DS particles, IgA adsorbed (%): (blue) in citric acid medium; (red) in PBS medium. The data are the average of three independent experiments ± standard deviation.

Figure 6

Sorption kinetics of anti-α4β7 IgA onto CH-DS particles, IgA adsorbed (%): black in citric acid medium (green) in PBS medium. The data are the average of three independent experiments ± standard deviation

During IgA sorption, particle size increased to 650–700 nm and remained stable in buffers for one week at room temperature, under moderate end-overhead stirring. The zeta potential of the bionanoparticles varied from +10 mV to +7 mV in citric acid medium and from +4 mV to +0 mV in PBS. After, antibody sorption onto the PECs, removal of the unbound proteins by centrifugation and redispersion of the pellet, a second centrifugation step induced only desorption of 1% of the total amount of bound proteins, proving the robustness of the binding.

Nanoparticles loaded with tenofovir drug and functionalised with IgA

The sorption was carried out only in PBS since the nanoparticles loaded with TF in acidic medium flocculate. IgAs sorption onto nanoparticles TF loaded was studied at the ratio (w/w): 0.1, 0.3 and 0.6 different concentrations both type of IgA. In Figure 7 are represented the sorption yields of anti-α4β7 IgA after 72 h of incubation; similar results were obtained anti-hCEA IgA sorption (data not shown).

Figure 7

Sorption of different concentration of IgA anti-α4β7 on the CH-DS nanoparticles tenefovir loaded at different ratio. The data are the average of three independent experiments ± standard deviation

Antibody bioactivity

The pharmaceutical property of therapeutic proteins closely depends on the retention of their biological activity after immobilization on the carrier [36]. Consequently, the recognition capacity of the particle bound immunoglobulins was assayed using a specific solid phase enzyme-linked immunosorbent assay (ELISA) as showed in Figure 1. For this test, the antigen was immobilised on a solid phase, then the particles were added and the detection was achieved with an enzyme conjugated anti-specie antibody. The antibody bioactivities reported in Figure 8 resulted from the ratio of the absorbance measured with the particle-immobilised IgAs vs the same concentration of free antibody. After one week of storage in PBS under stirring, 72% and 60% of anti-hCEA and anti-α4β7 IgA, respectively, immobilised onto CH-DS were still bioactive.

Figure 8

The ELISA relative bioactivity of: A) IgA anti-hCEA adsorbed onto CH-DS nanoparticles in PBS. (blue bars) and citric acid medium (yellow bars). B) IgA anti- α4β7 adsorbed onto CH-DS nanoparticles in PBS (black bars) and citric acid medium (orange bars). The data are the average of three independent experiments ±standard deviation.