Sequential multi-material 3D bioprinting of dual-stiffness peptide scaffolds for compartmentalized, multicellular bone marrow niche modeling
Three-dimensional (3D) bioprinting holds great potential for tissue engineering, yet reproducing the bone marrow (BM) microenvironment remains challenging: it requires biomaterials that support the diverse BM cell types while reproducing the mechanical heterogeneity of its two major niche compartments. Here we developed an in vitro BM model representing both the endosteal and central marrow niches, using ultrashort self-assembling peptide hydrogels as bioinks and scaffolding material. Rheological characterization showed that Ac-IIZK-NH₂ (IIZK) hydrogels were stiffer than Ac-IIFK-NH₂ (IIFK); IIZK was therefore used for the endosteal compartment and IIFK for the central marrow compartment. Both peptides showed excellent printability and shape fidelity using a robotic arm-based bioprinter with a customized coaxial nozzle. The compartments were fabricated sequentially: the stiffer IIZK compartment was printed first as a self-supporting structure, and the separately cellularized IIFK compartment was then printed into the cavity it defines. This order lets each compartment's composition be set by the tissue it represents rather than by the demands of extrusion, while in-line coaxial gelation preserves the interface. Cell viability exceeded 90% for mesenchymal stromal cells (MSCs), BM endothelial cells (BMECs) and the leukemia lines KG1a and HL-60. BMECs spontaneously formed capillary-like networks, and MSCs upregulated RUNX2, BGLAP and SP7 even without induction medium, depositing a calcium-rich mineralized matrix upon osteogenic differentiation. Tri-culture experiments demonstrated multicellular interaction and spatial organization. Looking ahead, the CD34-positive KG1a line used here as a hematopoietic surrogate can be replaced by primary HSCs without altering the scaffold design or the printing workflow, opening the way to studies of niche-driven hematopoiesis, leukemic drug resistance and patient-specific BM modeling.
