Back

Self-feeding living materials enabled by cell responsive glycogen nanoparticles as metabolic batteries

Gurian, M.; Willemen, N. N. G. A.; Porsul, I. I. R.; Bassous, N.; Hiemstra, J.; Gawlitta, D.; Shin, S.; Leijten, J.

2026-06-16 bioengineering
10.64898/2026.06.11.731644 bioRxiv
Show abstract

Scaling engineered living materials to clinically relevant dimensions is limited by diffusion-dependent depletion of oxygen and nutrients, which rapidly induces metabolic failure. We introduce glycogen as nutritional nanoparticle that provides cell-mediated, autonomous nutrient release to support long-term survival under extreme metabolic stress. We demonstrate that human mesenchymal stromal cells (hMSCs) survive for weeks in anoxia and serum deprivation when provided extracellular glycogen. Contrary to long-held assumptions, hMSCs secrete glycogen-degrading enzymes, enabling cell-density controlled extracellular glycogenolysis and sustained release of glucose and metabolic intermediates, positioning glycogen as the first-of-its-kind metabolic battery. This cell-responsive process maintains metabolic activity, limits glycolytic acidosis, and enhances pro-angiogenic signaling. To translate this mechanism into a versatile materials platform, we engineered core-shell dextran-tyramine microcapsules that stably encapsulate glycogen while permitting diffusion of enzymes and degradation products. Integrated into centimeter-scale GelMA constructs, these microcapsules maintained hMSC viability and function for at least one month under anoxia. In vivo, glycogen-loaded implants promote deep cellular infiltration, enhanced matrix remodeling, increased M2 macrophage polarization, and orchestrated accelerated vascularization. This work establishes the novel concept of glycogen-based nutritional nanoparticles as metabolic batteries to endow engineered tissues with autonomous self-feeding capacity, enabling scalable and functional living materials for regenerative medicine and related technologies.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.