DEVELOPMENT OF A NEXT-GENERATION IN VITRO PLATFORM FOR IN-DEPTH EVALUATION OF ALTERNATIVE SUSTAINABLE FISH FEED FORMULATIONS

Intensive farming of commonly consumed fish species has continuously increased to meet the growing global demand and to relieve marine fish stock that has reached its production plateau . However, many fish intended for human consumption are carnivorous species and therefore fed using fishmeal and fish oil obtained by processing small oily fish species caught for non-food purposes, affecting aquatic wildlife. In this perspective, academia and feed industries are dedicating their efforts to the identification of novel feed formulations based on alternative ingredients, able to ensure animal welfare while preserving their growth performances. Among others, vegetable-based raw materials have been proposed to be suitable candidates. However, high inclusion levels of plant-based ingredients in carnivorous fish diets not only lead to gastrointestinal side effects (such as diarrhea, intestinal inflammation, and gastric bloating), which affect feed intake and apparent digestibility; but they also impact the beneficial characteristics of fish meat, such as the high-quality protein content and fatty acid profile. For these reasons, the search for healthier and more sustainable alternatives is still open. To predict and assess the final effect of the new compounds on the digestive system, extensive in vivo raw materials evaluation programs are necessary. Nevertheless, in vivo feeding trials require many animals, are expensive, time consuming, and are subjected to highly restricted rules. The objective of this PhD thesis is to develop an intestinal in vitro model that responds physiologically to the effects of selected raw materials and novel formulations, providing reliable predictions of in vivo results. The rainbow trout (RT) was selected as the experimental model since its wide occurrence in the local and European aquaculture. The study utilized RT intestinal epithelial cell lines (RTpiMI and RTdiMI) cultured in a bi-cameral system that simulate the intestinal lumen and bloodstream, promoting cell differentiation. The first goal was to adapt feed pellets for cell culture by developing an in vitro digestion (IVD) protocol using RT gastric and intestinal enzymes to obtain a bio accessible fraction (BAF), mimicking chyme. After defining biocompatible conditions (50% BAF in L-15 medium), the model was exposed to different diets: a vegetable-based protein diet (Soy Diet, SD), known to negatively affect intestinal health in vivo, an experimental diet based on feather meal (FH), promising as a sustainable ingredient, and a reference diet rich in fish meal (reference diet, RD). The RTpiMI cells showed clear responses to each diet, with SD causing significant barrier damage. The FH triggered two notable stress responses: cellular edema and increased cell proliferation, indicating a certain degree of stress. In contrast, the RD induced only a cell proliferation. This mild stress response, observed in both RD and HF, may be attributed to the absence of the protective mucus layer typically present in vivo. Despite it, RD is less stressful overall compared to the other diets. Although these results were promising, the model did not fully replicate the complexity of the intestinal environment, particularly in terms of mucus production. In vivo, the stromal layer is crucial for supporting cellular differentiation, including the development of secretory lineages such as the mucus-secreting goblet cells, and for maintaining overall homeostasis. Adding this layer to the bicameral system could enhance the model’s similarity to native tissue, resulting in more physiologically relevant responses. To achieve this, the bicameral system was coated with a solubilized basal membrane, both alone and enriched with fibroblasts, or alternatively, a 3D scaffold populated with fibroblasts that produced their own extracellular matrix was used. The latter, though more complex and time-consuming, proved to be the most effective in promoting epithelial cell differentiation and barrier formation. In the final experiment, where the reference diet was tested across all platforms, from the simplest to the most advanced, the 3D platform emerged as the most predictive, closely mimicking the complexity of the intestinal mucosa and responding in a more physiological, dose-dependent manner during prolonged exposure to digested feed.