Spatiotemporal blood vessel specification at the osteogenesis and angiogenesis interface of biomimetic nanofiber-enabled bone tissue engineering

Abstract

While extensive research has demonstrated an interdependent role of osteogenesis and angiogenesis in bone tissue engineering, little is known about how functional blood vessel networks are organized to initiate and facilitate bone tissue regeneration. Building upon the success of a biomimetic composite nanofibrous construct capable of supporting donor progenitor cell-dependent regeneration, we examined the angiogenic response and spatiotemporal blood vessel specification at the osteogenesis and angiogenesis interface of cranial bone defect repair utilizing high resolution multiphoton laser scanning microscopy (MPLSM) in conjunction with intravital imaging. We demonstrate here that the regenerative vasculature can be specified as arterial and venous capillary vessels based upon endothelial surface markers of CD31 and Endomucin (EMCN), with CD31⁺EMCN^- vessels exhibiting higher flowrate and higher oxygen tension (pO₂) than CD31⁺EMCN⁺ vessels. The donor osteoblast clusters are uniquely coupled to the sprouting CD31⁺EMCN⁺ vessels connecting to CD31⁺EMCN^- vessels. Further analyses reveal differential vascular response and vessel type distribution in healing and non-healing defects, associated with changes of gene sets that control sprouting and morphogenesis of blood vessels. Collectively, our study highlights the key role of spatiotemporal vessel type distribution in bone tissue engineering, offering new insights for devising more effective vascularization strategies for bone tissue engineering.

Keywords: Angiogenesis; Biomimetic nanofibers; Blood vessel specification; Bone tissue engineering; Intravital imaging; Oxygen tension pO(2).

Fig.1.. Analyses of nanofiber-enabled cranial bone defect repair.

Schematic to show layer-by-layer enabled tissue engineering strategy to assemble multilayered tissue construct for 2-mm cranial defect repair (A). MicroCT images (B) and H&E histology (C) of the defects treated with or without BMSC-seeded fibers at week 5. Quantitative MicroCT analyses of bone volume within defects at week 5 (D). Histomorphometric analyses of bone and fibrotic tissue within the defects (E). n=6. ****p<0.0001. The representative H&E histology of BMSC-fiber-mediated healing at week 3 (F) and 5 (G). The corresponding fluorescence images to demonstrate donor Col 1 (2.3) GFP⁺ osteoblasts and host OSX-cherry⁺ osteoblasts within the defect at week 3 (H) and 5 (I). Scale bar = 500μm.

Fig. 2.. Spatiotemporal vessel specification during nanofiber-mediated repair.

Representative 3D MPLSM images of the 2-mm cranial defects at week 3 and 5, reconstructed from different combinations of channels as indicated. Defect only at week 3 (A1-3). Boxed region in A2 is shown in A3. Defect treated with AF construct at week 3 (B1-3). Boxed region in B2 is shown in B3. Defect treated with BMSC-F construct at week 3 (C1-3). Images of the same BMSC-F treated defect were reconstructed at a depth below 150μm to show vessel types associated with Col 1 (2.3) GFP⁺ cells (cyan) (D1-3). Boxed regions in D1 are shown in D2 and 3. Defect only at week 5 (E1-2). Boxed region in E1 is shown in E2. Defect treated with AF construct at week 5 (F1-2). Boxed region in F1 is shown in F2. Defect treated with BMSC-F construct at week 5 (G1-2). Images of the same BMSC-F treated defect was reconstructed at 150μm below to show vessels in newly formed bone (H1-2). Boxed region in H1 is shown in H2. Scale bar = 500μm. Arrows indicate the connections between CD31⁺EMCN⁻ (red) and CD31⁺EMCN⁺ (yellow or yellowish green) vessels. Vessels were stained with CD31 and EMCN antibodies labeled by far red and red fluorescent dye. Tiling boundaries can be seen in some images.

Fig. 3.. Functional analyses of the different types of vessels at the defect repair site.

Intravital imaging was performed in a cranial window chamber model (A). Representative image of the defect perfused with vascular dye at 3-week post-transplantation (B). Boxed region in B is illustrated to show vessels with indicated pO₂ values (C). The same sample was stained with CD31 and EMCN antibodies to reveal vessel types (D). Flow rates were measured via line scans as indicated (insert) with the calculated flow rate indicated at each vessel segment (E). The same sample was stained with CD31 and EMCN antibodies to reveal vessel types (F). CD31⁺EMCN⁻ vessels shown as red, CD31⁺EMCN⁺ vessels shown as green. Measurements of vessel diameter (G), RBC velocity (H), and pO₂ (I) in CD31⁺EMCN⁻ vessels (red) and CD31⁺EMCN⁺ vessels (green). * p<0.05. n=150 vessels in 6 mice. * p<0.05. Scale bar = l00μm.

Fig 4.. Quantitative analyses of vessel types via MPLSM in cranial defects treated with or without BMSC-seeded nanofibrous constructs.

ROIs of the cranial defects were created and reconstructed to show donor GFP⁺ cells (cyan), new bone (white), CD31⁺EMCN⁺ (green) and CD31⁺EMCN⁻(red) vessels within the defects. Defect alone (A1-4, D1-4), defect with acellular fibers (B1-4, E1-4) or defect with BMSC fibers (C1-4, F1-4) at week 3 (panel A, B and C) and week 5 (panel D, E and F). Vol. Fract. of bone as evaluated via SHG (G), total CD31⁺ vessels (H), CD31⁺EMCN⁺ vessels (I), and Length Fract. of CD31⁺EMCN⁻ vessels (J) are shown. n=4 per group, * p<0.05. Scale bar = 200μm.

Fig. 5.. Vessel type distribution in bone-forming and non-bone-forming regions of the cranial defects.

Defects treated with fibers at week 3 (A1-4) and week 5 (B1-4) post-implantation were reconstructed to show vessels in the regions of bone forming (A1-2 and B1-2) or non-bone forming region (A3-4 and B3-4). Similarly, defects treated with BMSC-seeded fibers at week 3 (C1-6) and week 5 (D1-6) were reconstructed to show regions of bone forming (C1-4 and D1-4) and non-bone forming (C5-6 and D5-6). New bone (white), Col 1 (2.3) GFP⁺ cells (cyan), CD31⁺EMCN⁺ (green) and CD31⁺EMCN⁻ (red) vessels were reconstructed in the respective regions to show the spatial relationship of vessel types and bone/cells. Vol. Fract. Of CD31⁺EMCN⁺ (E) and Length Fract. of CD31⁺EMCN⁻ (F) in various regions of defects are shown. n=4; a, p<0.05 when comparing with or without cells in bone region at week 3; b, p<0.05 when comparing with or without cells in non-bone region at week 3; c, p<0.05 when comparing with or without cells at week 5 in bone region; d, p<0.05 when comparing between week 3 and 5 with cells in bone region; e, p<0.05 when comparing between week 3 and 5 with cells in non-bone region. Scale bar = 200μm.

Fig. 6.. Transcriptomic analyses of genes in healing and non-healing defects.

(A)The microarray volcano plot shows differentially up- and down-regulated genes in nanofibrous constructs seeded with or without BMSCs at week 3 post-implantation (n=3). (B) Enrichment plots of GO terms with indicated scores for NES, NOM, and FDR. (C) Gene expression heatmap shows an array of upregulated (top) and downregulated (bottom) angiogenesis-related genes in implants with or without BMSC treatment. (D) RT-PCR shows indicated gene expression in cranial samples. (E) RT-PCR shows progressive induction of Slit2, Slit3, and Sema3b,5a and 7a during BMSC differentiation in culture, along with the expression of osteogenic genes, namely Runx2, Sp7(OSX) and ALP. * indicates p<0.05.