Hydrogels as platforms for vaccine delivery and mucosal restoring

Resumo

Mucosal administration of active molecules, such as proteins, is an attractive and advantageous approach to induce local or systemic therapeutic and/or immunostimulatory responses. Critical aspects, such as rapid protein drainage or enzyme- or pH-mediated degradation, limit the residence time of therapeutic molecules on mucosal surfaces and condition the passage of substances through the mucosa and, therefore, their effectiveness. Hydrogels are 3D polymeric networks with characteristics that can be modulated to achieve highly mucoadhesive formulations, capable of interacting with mucous membranes and increasing the residence time of drugs at the target site, enhancing their absorption and the desired therapeutic effect. In this sense, thermosensitive and mucoadhesive hydrogels are of particular interest to improve the efficacy or bioavailability of proteins when administered on mucous membranes. In addition, these hydrogels are excellent candidates to overcome the limitations of transmucosal delivery due to their biocompatibility, mucoadhesion, thermosensitivity and drug loading capacity. However, the design of such dosage forms involves the preparation of multicomponent systems, which is a challenging task. In recent years, artificial intelligence (AI) tools such as neural networks, fuzzy logic or genetic algorithms (GA) have proven to be very useful in studying the effects of composition and operational variables on the characteristics of different pharmaceutical systems. Therefore, this thesis work addresses the usefulness of AI tools in the development of multicomponent hydrogels with advanced performance. Thus, we explore the potential of ternary combinations of selected polymers to generate thermosensitive and highly bioadhesive hydrogels suitable for protein loading and delivery with two applications: sublingual vaccination and local treatment of colorectal ulcers in inflammatory bowel disease (IBD). For hydrogels design, after the initial selection of components, a reduced experimental design was established using a balanced density method. This made possible to delimit both the knowledge space and the design space of ternary systems using a limited number of formulations. Texturometric and rheological properties and release profile of hydrogels were carried out. After hydrogel characterization, artificial neural networks (ANN) combined with fuzzy logic techniques were used to model the hydrogel characteristics according to their composition and, to analyze the effect of each material on their properties. Then, the combination of ANN and GA was used to model and select, in each case, an optimal formulation that gels at physiological temperature and shows the best characteristics for application and maintenance on the corresponding mucosa (resistance to washout or dilution and high mucoadhesion). Finally, mice experiments were conducted. In one hand, the sublingual optimized hydrogel was loaded with CTH522 Chlamydia trachomatis vaccine and tested in terms of immunostimulatory capacity. In vivo studies demonstrated that the most promising strategy for eliciting local immunity in the genital tract was parenteral-sublingual boost immunization, using CAF01 and the optimized hydrogel as carriers of CTH522 vaccine. This approach was the most effective in inducing cell-mediated immune responses (Th1/Th17) and significantly increasing IFN-γ production by cervical lymph nodes when compared with parenteral vaccination alone. Conversely, the regenerative and anti-inflammatory capacity of the rectal optimized hydrogel, loaded with human uterine cervical stem cells conditioned medium (H-hUCESC-CM), was evaluated in an experimental mouse model of acute colitis. Experimental data indicate that treatment with the loaded hydrogel significantly decreases body weight loss, colon shortening, colonic mucosal degeneration and TNF-α, IFN-γ and IL-6 mRNA levels. Our data indicate H-hUCESC-CM effectively alleviates DSS-induced colitis in mice, suggesting that H-hUCESC-CM may represent a very attractive cell-free therapy for the local treatment of IBD. In general, we demonstrate that the development of complex multicomponent systems is greatly facilitated by AI approaches, since their use allows a significant reduction of experiments, saving costs and time. In addition, the deep knowledge of pharmaceutical systems that derives from their use allows the establishment of the adequate design space for achieving advanced dosage forms.