Poster AA15 - Next-Generation Aligned Collagen Scaffolds: Harnessing Artificial Intelligence and Two-Photon Strategies

Introduction: Regenerative medicine has revolutionized the landscape of tissue repair and replacement. One focus in this field is on the development of aligned collagen scaffolds that are crucial for in vivo-like cell growth and tissue formation (Lee et al. 2021). The architectural complexity of many tissues is characterized by highly organized collagen fibril matrices, typically arranged into parallel fiber bundles. These assemblies predominantly consist of type I collagen, which is essential for the structural integrity and functionality of various tissues including intervertebral discs, tendons, ligaments, muscles, neurons, and the cornea. Despite the progress in manufacturing techniques like isoelectric focusing, electrospinning, microfluidics, freeze-drying, extrusion, cyclic stretching, and shear flow-induced crystallization (Dewle et al. 2020), clinical translation of these tissue-engineered devices frequently encounters challenges. These often stem from local structural disparities between the engineered scaffolds and the native tissue, which can compromise functionality and integration. To address these challenges, we have been developing a novel approach that involves an artificial intelligence (AI)-driven, laser-based technique for precise collagen alignment, which promises to facilitate the in-situ formation of scaffolds and significantly advance the generation of implantable biomedical devices.

Materials and

Methods: Our innovative method employs a tunable 80MHz femtosecond Ti:Sapphire laser, which utilizes a two-photon (2P) effect to induce controlled and localized protein assembly, known as fibrillogenesis. This results in the precise formation of collagen fibers and bundles. The scanning process involves the laser tracing predefined vector paths through a photoinitiator-mediated collagen solution contained in a custom-designed well plate. These paths are determined based on an AI model’s predictions of optimal collagen network configurations necessary for specific tissue repair, such as for those at enthesis sites of tendons and ligaments. The AI model articulates the laser creating linear, fan-like, or criss-cross patterns to mimic the native fiber alignment essential for functional tissue engineering. Following scaffold formation, these are subjected to biocompatibility and functionality tests involving embryonic rat fibroblasts (3T3 cells). The cells are seeded onto the collagen scaffolds after undergoing a rigorous cleaning process to ensure removal of any residual photochemicals that might affect cell health.

Results, Conclusions, and Discussions: Utilizing our 2P laser system, we have successfully generated various collagen structures, including intricate sinusoidal mesh patterns and single layers of collagen bundles arranged in square configurations. These structures support the attachment, proliferation, and alignment of 3T3 cells, demonstrating the biocompatibility and functional adequacy of the scaffolds. Our results indicate that the scaffolds reach confluence within 72 hours post-seeding, suggesting an effective removal of potentially cytotoxic chemicals through our scaffold treatment process. Moreover, further cytotoxicity testing and analysis of 3T3 cell behavior have provided deeper insights into fibril alignment within the generated constructs. The cells exhibited alignment along the fabricated fibers, orienting their long axis either parallel or perpendicular to these fibers, and showed pseudopodial extensions characteristic of active engagement in the compaction and integration processes within the collagen gels.

The laser-based collagen biofabrication process we have developed represents a substantial leap forward in the field of tissue engineering. Its ability to optically induce collagen fibrillogenesis with exceptional spatial and temporal precision introduces a novel, highly effective method for crafting customizable three-dimensional collagen structures. Unlike conventional methods that generally aim for bulk collagen alignment, our technique achieves highly localized collagen assembly, enabling the precise replication of intricate native fiber alignments found in various tissues. This method holds immense potential for advancing the field of regenerative medicine, particularly in developing highly tailored and mechanically robust collagen-based tissues suitable for diverse biomedical applications. Future work will focus on refining the laser parameters, exploring additional photochemicals for crosslinking, and broadening the scope of producible three-dimensional structures to further enhance the efficacy and applicability of this technology in clinical settings.

Acknowledgements (Optional): This research and the ADAPT in SC RII Track-1 project are supported by the National Science Foundation (OIA-2242812).