A Biomimetic 3D-Self-Forming Approach for Microvascular Scaffolds

Abstract

The development of science and technology often drew lessons from natural phenomena. Herein, inspired by drying-driven curling of apple peels, hydrogel-based micro-scaled hollow tubules (MHTs) are proposed for biomimicking microvessels, which promote microcirculation and improve the survival of random skin flaps. MHTs with various pipeline structures are fabricated using hydrogel in corresponding shapes, such as Y-branches, anastomosis rings, and triangle loops. Adjustable diameters can be achieved by altering the concentration and cross-linking time of the hydrogel. Based on this rationale, biomimetic microvessels with diameters of 50-500 µm are cultivated in vitro by coculture of MHTs and human umbilical vein endothelial cells. In vivo studies show their excellent performance to promote microcirculation and improve the survival of random skin flaps. In conclusion, the present work proposes and validifies a biomimetic 3D self-forming method for the fabrication of biomimetic vessels and microvascular scaffolds with high biocompatibility and stability based on hydrogel materials, such as gelatin and hyaluronic acid.

Keywords: bioinspired materials; biomimetic microvessels; self‐forming; vascular scaffolds.

Figure 1

Establishment of the MHT scaffolds. a) The process of drying‐driven curling like “apple peels”. b) The 3‐step forming of the hydrogel‐based MHTs. c,d) Macroscopic and microscopic images of the hydrogel‐based MHTs. The MHTs were stained with rhodamine and excited by a e–h) fluorescent microscopy or i–k) UV light. e,f) Morphology of single MHTs. g) The cross section of an MHT. h) Morphology of a branched MHT. i) Morphology of a single MHT. j) Morphology of an MHT circle. k) Morphology of a T‐shaped MHT. l,m) Scheme of the curling of the hydrogel flake driven by the heterogeneous internal strain after re‐swelling, caused by non‐uniform upper‐lower density after photocrosslinking and drying. p) SEM images of the lyophilized GelMA hydrogel after crosslinking, drying, and swelling. o,q) Detail view of the upper and lower layer in (p). n) Pore size in (o) 176 ± 23 µm and q) 292 ± 17 µm, **p < 0.01 (n = 6).

Figure 2

Physical characterization of the MHTs. a,b) SEM image of the “omelet”‐like MHTs. e,f) SEM images of the biaxial handscroll‐like MHTs. c,d,g,h) FEM analysis of the two types of the MHTs. FEM simulations of 3D‐shape‐morphing of hydrogel tubes. The gradient of material properties across the thickness of the hydrogel sheets was simplified by a bi‐layer denoted by “dense‐crosslinked” and “sparse‐crosslinked”, indicated by red and green respectively. Two different morphological types were achieved by controlling the thickness distribution of the materials along the width of the hydrogel sheets. c)“Omelet”‐like hydrogel tube with a spiral cross sectiond) was achieved by placing more materials at the left side than at the right side, g) biaxial handscroll‐like tube with a symmetric spiral cross section h) was achieved by placing more materials in the middle than at the sides. i) Detail view of the mesoporous structure of the wall of the MHT. j) Siphon effect of the MHT. k) Degradation profile of GelMA hydrogel. l) Stress–strain test for MHT. m) Stress–strain curve of MHTs fabricated with different hydrogel concentration. n,o,p) Young's modulus, tensile strength, and stretching capacity of the MHTs, derived from the stress–strain curve.

Figure 3

Biocompatibility of the GelMA hydrogel. a–d) Live/dead staining of HUVEC cultured on GelMA for 1, 3, 5, 7 days. Green: live cells. Red: dead cells. Yellow lines in (b): lumen spontaneously formed by the HUVECs. e) Proportion of the live cells. f) CCK‐8 results of the cells cultured on the pretreated (soaking or drying) hydrogels. g–j) Cytoskeleton staining of HUVECs, fibroblasts and coculture of both cells on the hydrogel. Red: cytoskeleton marked by phalloidin. Blue: nucleus marked by DAPI.

Figure 4

Confocal imaging and 3D reconstruction of the biomimetic vessels developed in vitro. a–d) Survival of the HUVECs incubated for 3 days on the MHTs. s) Proportion of live cells on the MHTs. e–h) ZO‐1 expression on the cytomembranes of HUVECs cultured on the MHTs for 3 days, marked by immunofluorescent staining. i–l) The extension of the HUVECs cultured on the MHTs for 3 days, marked by phalloidin staining. m,n) Live/dead staining of the HUVECs cultured on the omelet‐like MHTs for 3 days. o,p) Live/dead staining of the HUVECs cultured on the biaxial handscroll‐like MHTs for 3 days. q,r) The cross section of the HUVEC seeded MHTs incubated for 3 days, frozen sliced and stained by phalloidin/DAPI.

Figure 5

Random flap surgery and analysis on rat dorsal random flap (n = 6). a) Scheme of MHTs implantation beneath skin flap. b) Left: immediately and 5 min after placement of GelMA MHTs. Right: in situ suture of the random flap after surgery. c1,d1) necrotic area in the flaps 7 days after surgery of different groups. c2,d2) the laser speckle contrast imaging captured real time blood flow images of different groups. Blood perfusion density increases from blue to red signal. e) Necrotic area proportion in different groups. ****p < 0.0001. f) Average signal intensity of blood flow in different groups. **p < 0.01. g1,g2) H&E staining of the necrosis and survival junction area of different groups skin flaps. g3,g4) H&E staining of region near tube from different slice direction. h1–h4) Detail views of (g1–g4) in high magnification. i1–i4) Immunohistochemical images of groups above showing blood vessel CD31‐positive endothelial cells. Red lines in (i4): a new vessel extending longitudinally into the scaffold. j1–j4) Immunohistochemical images of groups above showing the CD68‐positive macrophage/monocytes. k) average density of the microvessels, ****p < 0.0001, l) average macrophage density of the control and experimental group, ***p < 0.001. m) Comparison of microvessel density in the area near the MHT and necrotic area in the experimental group, **p < 0.01. n) Comparison of inflammation in the area near the MHT and necrotic area in the experimental group, ****p < 0.0001.