In this study, genipin, a natural crosslinking agent, was used to crosslink chitosan microspheres to evaluate its effect on drug release. Genipin reacts with compounds containing primary amine groups, such as chitosan and some peptides and polypeptides, to form covalently crosslinked networks [7–10]. In the case of chitosan, genipin reacts with the free amino groups present in the glucosamine units. It has been reported that two separate reactions lead to the crosslinking between these two compounds [5]. The reaction between chitosan hydrochloride and genipin in solution was characterized by UV spectra analysis before spray-drying. The influence of reaction time and genipin concentration on the crosslinking reaction was studied.

Genipin dissolved in water displayed an absorption peak at 240 nm, which was higher with increased genipin concentrations. After the addition of chitosan to the genipin solution, the absorption at 240 nm decreased and a new absorption peak at 290 could be seen. This absorption peak increased with reaction time and genipin concentration (Figure 2). Mi and co-workers [7] described that this decrease at 240 nm was due to the conversion of the ester group of genipin into an amide linkage formed after the nucleophilic attack by chitosan amino groups. On the other hand, the increase at 290 nm was attributed to the formation of the heterocyclic genipin-chitosan compound [11]. The presence of this peak and its increase with reaction time and genipin concentration is therefore an indication of the extent of the crosslinking between genipin and chitosan.

The effect of increasing the concentration of genipin on the crosslinking reaction at fixed temperature, 50 ºC (based on previous studies carried out in our laboratory) and time (5 h), was evaluated. In the same way, different reaction times (0–30 min) at 50 ºC and fixed genipin concentration (5 mM) were studied in relation to the crosslinking process. As shown in Figure 2, both increasing genipin concentration and reaction time led to a higher degree of crosslinking. For biomedical applications, the adequate crosslinking conditions should be studied in each case depending on the drug, location of release within the human body, and pharmacokinetic parameters, to achieve optimal controlled release.

Figure 3 shows the morphology of chitosan microspheres crosslinked with genipin and loaded with clarithromycin (Figure 3A), tramadol hydrochloride (Figure 3B) or LMWH (Figure 3C) obtained by SEM. As can be seen, <10 μm microspheres were obtained in all cases. The obtained microspheres presented a smooth surface with indentations as a result of rapid particle shrinking during the drying process [12].

The zeta potential of the microspheres is shown in Table 1 (microspheres loaded with clarithromycin), Table 2 (microspheres loaded with tramadol hydrochloride) and Table 3 (microspheres loaded with LMWH).

As can be seen with these three cases, the zeta potential values decreased when the microspheres were crosslinked with genipin, proving the genipin-polymer interaction. Nevertheless, all the microspheres showed positive charge, so all of them maintained the chitosan mucoadhesive and absorption enhancement properties and can be considered as potential drug release vehicles [13].

The X-Ray diffraction patterns of genipin, chitosan hydrochloride, assayed drugs, and the combination of all microsphere components are shown in Figures 4–6. Genipin itself, as well as clarithromycin and tramadol hydrochloride, showed high crystallinity. On the other hand, the chitosan hydrochloride X-ray pattern showed its amorphous structure. Genipin, as well as clarithromycin (Figure 4) and tramadol hydrochloride (Figure 5) are included in the amorphous structure of chitosan when incorporated in microspheres. As can be seen in Figure 6, the LMWH X-ray pattern shows it as an amorphous compound. LMWH loaded microspheres crosslinked with genipin show an X-ray diffraction pattern with no marked peaks, which indicates that the crosslinking agent is molecularly dispersed in the polymer matrix [14,15], which should help to retard the delivery of the drug [16].

As mentioned in the experimental section, preliminary studies using different times, temperatures, and genipin concentrations were carried out for assayed drugs. 5 h and 50 ºC were revealed as adequate time and temperature conditions to achieve the desired degree of crosslinking. The genipin concentration range used for each drug is also a consequence of these previous results (data not shown).

The encapsulation efficiency was >85% for clarithromycin, >90% for tramadol, and 15–30% depending on the crosslinking degree (the higher the crosslinking degree the lesser the encapsulation efficiency) in the case of LMWH.

A 5 mg/mL chitosan solution was prepared by disolving it in deionized water. Genipin solutions of different concentrations (0–5 mM) were added and the resulting mixture was incubated for different time intervals (30 min–7 h) to allow the crosslinking to occur, at 50 ºC. The chitosan-genipin crosslinking reaction was characterized by UV/Vis (GBC UV/Vis 920, GBC Scientific Equipment Pty. Ltd., Dandenong, Australia). During the crosslinking reaction, spectra were obtained at certain time intervals in a wavelength range from 200 to 700 nm, at 2 nm resolution.

The required amount of the model drug with respect to polymer weight (50% w/w of clarithromycin, 30% w/w of tramadol hydrochloride, 10% w/w of LMWH) was added to the chitosan solution (5 mg/mL in the case of clarithromycin and tramadol hydrochloride and 1mg/mL for LMWH) with constant stirring. Genipin solution was added to the chitosan solution in different concentrations (0–25 mM) and the resulting mixture was incubated at 50 ºC during different time periods (0 and 15 h). Drug/polymer concentration, reaction temperature, and genipin concentration ranges used in each case are based on previous investigations carried out in our laboratory (data not shown). Chitosan-genipin-drug solutions were spray-dried using a Büchi Mini Spray Dryer B-290 (Switzerland) with a standard 0.5 mm nozzle. Briefly, spray-drying conditions were inlet temperatures of 150 ºC for LMWH and 160 ºC for clarithromycin and tramadol hydrochloride, spray flow rate of 473 NL/h, aspirator air flow rate of 32 m³/h and sample flow rate of 2.5 mL/min.

To determine shape and surface of the microspheres scanning electron microscopy (SEM) was used. The samples were sputter coated with Au/Pd using a vacuum evaporator (Balzers SDC 004 Sputter coater, Oerlikon Corporate Pfäffikon, Switzerland), and examined using a scanning electron microscope (JEOL JSM-6400 (JEOL, Tokyo, Japan) at 10 kV accelerating voltage.

Microspheres were suspended in ethanol. A solution of 10⁻³ M KCl was prepared. 1 mL of microsphere suspension was diluted into 99 mL KCl solution. Zeta potential measurements were made by microelectrophoresis using a Malvern Zetasizer Nanoseries Nano ZS (Malvern Instruments, Herrenberg, Germany). Measurements were performed at an effective voltage of 150 V and 25 ºC. Electrophoretic mobility data were automatically converted into zeta potential using the Henry equation and the approximation of Helmholtz-Smoluchowski [25].

X-ray diffraction patterns were obtained using an X-ray diffractometer (PHILIPS X’PERT SW) with a copper anode. The samples were scanned continuously from 0º to 50º (2θ) at 45 kV and 40 mA.

Microspheres inside a cellulose dialysis bag (dialysis tubing, Mw cut off 12,000 Da, Sigma Aldrich, Madrid, Spain) to avoid the loss of microspheres during the experiment, were suspended in HCl 1N pH 1.2 as simulated gastric fluid, SGF (clarithromycin and tramadol hydrochloride) or PBS pH 7.4 (LMWH) without enzyme. The dialysis bag was placed in a glass with 50 mL (clarithromycin), 200 mL (tramadol hydrochloride) or 100 mL (LMWH) of the release medium at 37 ºC with orbital agitation of 100 rpm (Rotabit horizontal shaker, Selecta, Barcelona, Spain). The volume of the release recipient was established on the basis of previous studies for each drug, so that the release medium would not be saturated. At predetermined interval times, 2 mL of sample were withdrawn. The withdrawn volume was replaced with fresh medium at the same temperature. The drug released was quantified by HPLC in the case of clarithromycin and UV/Vis in the case of tramadol hydrochloride and LMWH. All the experiments were carried out in triplicate.