• gut microbiota
• in vitro validation
• ingestible capsule
• magnetic actuation
• material jetting
• mucosal brushing

Recent advances in the field of gut microbiota research have highlighted the need for minimally invasive, spatially selective, and functionally robust tools for in vivo sampling of the intestinal microbial ecosystem. The gut microbiota, composed of trillions of microorganisms residing in the small and large intestine, plays a central role in maintaining host health through metabolic, immunological, and neuroendocrine pathways. Its imbalance, or dysbiosis, has been linked to a wide spectrum of diseases, including inflammatory bowel disorders, colorectal cancer, metabolic syndrome, and neurodegenerative conditions. Understanding the spatial distribution, temporal dynamics, and compositional features of these microbial communities is therefore essential for advancing both diagnostic precision and therapeutic strategies. Traditional sampling techniques—such as stool collection, colonic lavage, endoscopic biopsies and brushings—each present considerable limitations in terms of invasiveness, spatial resolution, and sample integrity. These challenges motivate the development of novel ingestible devices that can reliably collect mucosal samples under physiological conditions and preserve them for downstream omics analyses.
This thesis addresses such a need through the redesign, optimisation and functional evaluation of the MECCA capsule (Microbiota Endoscopic Capsule with Compliant Activation), an innovative wireless, battery-free, magnetically actuated ingestible device for mucosal microbiota sampling. The work began with a comprehensive review of the anatomical, physiological and microbial landscape of the gastrointestinal tract, laying the foundation for defining the theoretical requirements of a robust sampling system. Key distinctions between luminal and mucus-associated microbiota were emphasised, including their ecological differences, immune interactions, and clinical relevance. The mucus layer, being closely associated with the epithelial barrier, hosts stable and immunologically active communities, making it a preferred target for precision sampling. Additionally, the thesis underscored the importance of preserving spatial and temporal fidelity in microbiota collection and discussed the biochemical constraints in preserving microbial DNA and RNA.
The initial MECCA prototype was reviewed in terms of design rationale, magnetic actuation principle, and mechanical structure. The capsule’s activation is entirely passive and relies on a series of internal NdFeB magnets arranged in concentric torsional springs, activated via proximity to an external permanent magnet (EPM). This configuration allows for precise triggering of anchorage, brush exposure, sample collection, and resealing, without any internal electronics. A central reservoir preloaded with a preservative solution ensures immediate stabilisation of the collected mucus.
However, early prototypes revealed a critical structural weakness: axial bursting occurred during magnetic actuation, compromising capsule integrity. The first major objective of this thesis was therefore to resolve this issue by redesigning the mechanical elements to absorb axial loads without impeding the rotational functionality of the system. The adopted solution provided for the enlargement of the reservoir and introduced a pair of threaded lateral nuts integrated with micro-bearings, allowing the central part to remain mobile while being structurally secured. These changes restored axial robustness and enabled the capsule to maintain concentricity and function throughout the entire actuation sequence.
A rigorous design process followed, combining CAD modelling, static magnetic simulations using a Boundary Element Method-based MATLAB code, and physical prototyping through high-resolution additive manufacturing (Material Jetting with VisiJet M3 Crystal). All structural elements were iteratively refined to reduce friction, improving tolerances, and optimise the bonding procedure. Validation included both theoretical estimation of magnetic thresholds for each activation phase and experimental in vitro tests using a transparent tube with precise manual positioning of the EPM. These confirmed the reproducibility and robustness of the anchorage, brush exposure, rotation of the reservoir and reclosure phases.
The second major goal concerned the integration of functional elements necessary for sampling. This included the fabrication of PDMS microbrushes using soft lithography techniques. These brush arrays, although significantly miniaturised compared to commercial cytology devices, demonstrated comparable mucus collection efficacy in a mechanical simulator, surpassing the 18 mg threshold required for transcriptomic analyses. Furthermore, internal and external caps were engineered to ensure sterility, prevent pre- and post-activation contamination, and allow irreversible locking of the reservoir after activation. A particular focus was placed on ensuring hydraulic sealing of the preservative reservoir, with attention to surface finishing, cap design, and the incorporation of PDMS sealing layers.
To meet clinical size constraints (maximum 33 × 12 mm, in accordance with PillCam COLON 2 standards), a miniaturisation phase was undertaken. All structural components were reduced in size while preserving relative magnet positions and mechanical functionality. Updated magnetic simulations validated the preservation of activation sequences within the new spatial configuration. Notably, in the reduced geometry, magnetic torques between adjacent magnets approached critical overlap, threatening phase discrimination. This was counteracted by leveraging the elastic friction of the newly designed internal caps, which served as mechanical thresholds selectively resisting rotation until specific torque levels were reached. This integration of mechanical resistance and magnetic control enabled distinct, reliable phase transitions even in the compact format.
The final prototype, fully functionalised and miniaturised, was successfully tested in vitro. All actuation phases occurred in the correct sequence, at predictable and safe distances, and without structural failure. These results validate the feasibility of MECCA as a clinically viable microbiota sampling capsule, opening the path to preclinical and eventually human trials.
Looking forward, further work will be directed toward improving hydraulic sealing, currently not fully achieved due to the delicate trade-off between increased compression and reduced friction between components. In addition, reconfiguration of the magnetic layout, introduction of miniaturised localisation systems, and integration with electromagnetic actuation platforms are among the proposed future developments. Moreover, in vivo validation in animal models is required to confirm sampling efficiency, peristaltic tolerance, and sample integrity across the full transit time. Ultimately, this work lays a solid foundation for the realisation of a non-invasive, spatially selective, and functionally autonomous capsule capable of revolutionising microbiota analysis and enabling a new generation of personalised diagnostic and therapeutic strategies.