Engineering of immunoinstructive extracellular matrices for enhanced osteoinductivity

Abstract

The increasing recognition of the contribution of the immune system to activate and prime regeneration implies that tissue engineering strategies and biomaterials design should target regulation of early immunological processes. We previously proposed the cell-based engineering and devitalization of extracellular matrices (ECMs) as a strategy to generate implant materials delivering custom-defined signals. Here, in the context of bone regeneration, we aimed at enhancing the osteoinductivity of such ECMs by enriching their immunomodulatory factors repertoire. Priming with IL1β a cell line overexpressing BMP-2 enabled engineering of ECMs preserving osteoinductive signals and containing larger amounts of angiogenic (VEGF) and pro-inflammatory molecules (IL6, IL8 and MCP1). Upon implantation, these IL1β-induced materials enhanced processes typical of the inflammatory phase (e.g., vascular invasion, osteoclast recruitment and differentiation), leading to 'regenerative' events (e.g., M2 macrophage polarization) and ultimately resulting in faster and more efficient bone formation. These results bear relevance towards the manufacturing of potent off-the-shelf osteoinductive materials and outline the broader paradigm of engineering immunoinstructive implants to enhance tissue regeneration.

Keywords: Bone tissue engineering; Hypertrophic cartilage matrices; Immunomodulation; Inflammation; Osteoinduction.

Immunoinstructed endochondral bone formation from engineered cartilage. Priming with IL1β cells overexpressing BMP-2 enables the generation of a hypertrophic cartilaginous matrix enriched in angiogenic and pro-inflammatory factors that promote in vivo vascular invasion, osteoclast activity and M2 macrophage polarization, enhancing ultimately endochondral bone formation.

Fig. 1

Engineering of an immunomodulatory cell-laid HyC matrix using IL1β as inducer. (A) Scheme illustrating the protocol followed to generate a HyC matrix enriched in BMP2 (thanks to its overexpression by MSOD-B cells) and in immunomodulatory factors (thanks to MSOD-B cell induction with IL1β). (B–F) Protein concentration of (B) IL6, (C) IL8, (D) MCP1, (E) VEGF and (F) BMP2 in engineered control and IL1β-induced HyC matrices. (G) Safranin-O, Collagen type II, Collagen type X and MMP13 stainings in control and and IL1β-induced matrices. Scale bar, 100 μm. (H) Glycosamynoglycans (GAG) quantification in digested control and and IL1β-induced matrices. (I–K) Gene expression of (I) COL2A1, (J) COL10A1 and (K) MMP13 in control and and IL1β-induced matrices. (B–F), (H–K) Data are plotted as means ± SD; n = 3–12. *P < 0.05, **P < 0.01. Unpaired two-tailed t-test.

Fig. 2

IL1β-induced HyC matrix exhibits accelerated remodeling into endochondral bone. (A–B) Micro-computed tomography (μCT) 3D reconstructions showing mineralization in ossicles derived from control and IL1β-induced matrices after 3 weeks of in vivo remodeling. (C–D) Ratio Bone Volume (BV) to Total Volume (TV) and percentage of trabecular (Tb) bone in ossicles derived from control and IL1β-induced matrices after 3 weeks of in vivo remodeling. (E–F) Trichrome of Masson and (G–H) Hematoxylin-Eosin stainings in 10 μm-sections of ossicles derived from control and IL1β-induced matrices after 3 weeks of in vivo remodeling. Dashed lines delineate the areas occupied by the remaining parts of the scaffold. Scale bar, 20 μm. (I–K) Percentage of (I) CD45⁺CD11b⁺Ly6G⁻ monocytes/macrophages, (J) CD45⁺CD11b⁺Ly6G⁺ neutrophils and (K) CD45⁺B220⁺ B cells in the femoral BM, and in ossicles derived from control and IL1β-induced matrices after 3 weeks of in vivo remodeling. Percentages of these populations in femoral BM were used as additional control. (L) Representative FACS plot illustrating the gating for monocytes/macrophages, neutrophils and B cells in engineered tissues. (C-D; I–K) Data are plotted as means ± SD; n = 3–9. *P < 0.05, ***P < 0.001. (C–D) Unpaired two-tailed t-test. (I–K) One-way ANOVA with Tukey's multiple comparison tests.

Fig. 3

IL1β-induced HyC promotes in vivo vascular recruitment, osteoclast activity and M2 macrophage polarization. (A–B) Immunofluorescent staining for CD31 (red) in 10 μm-sections of tissues derived from control and IL1β-induced matrices after 1 week of in vivo remodeling. MSOD-B-derived cells are labelled in green (GFP expression). Nuclei are labelled with DAPI (blue). The dashed lines delimitate the human cartilaginous tissue with MSOD-B cells from the surrounding mouse connective tissue. Scale bar, 100 μm. (C) Quantification of CD31⁺ area in remodeled tissues after 1 week in vivo. (D–E) Immunostaining for Osterix in 10 μm-sections of tissues derived from control and IL1β-induced matrices after 1 week of in vivo remodeling. Scale bar, 100 μm. (F) Quantification of Osterix ⁺ area in remodeled tissues after 1 week in vivo. (G–H) Tartrate-resistant acid phosphatase (TRAP) staining in tissues derived from control and IL1β-induced matrices after 1 week of in vivo remodeling. Scale bar, 100 μm. (I) Quantification of TRAP ⁺ area in remodeled tissues after 1 week in vivo. (J–N) Percentage of (J) CD11b⁺Ly6G⁻ monocytes, (K) CD11b⁺Ly6G⁺ granulocytes, (L) CD11b⁺F4/80⁺ macrophages, (M) CD11b⁺F4/80⁺CD206⁺ M2 macrophages and (N) CD11b^lowcKit ⁺ CD115⁺ osteoclast precursors. (C, F, I, J-N) Data are plotted as means ± SD; n = 5. *P < 0.05, **P < 0.01. Unpaired two-tailed t-test.

Fig. 4

IL1β-induced HyC upregulates IL6, TNFɑ and IL10 expression in vitro in a model of human M2 macrophage polarization. (A–D) Geometrical mean of (A) CD80, (B) CD86, (C) CD206 and (D) CD163 protein surface expression of human macrophages exposed to M1 control differentiating signals, M2 control differentiating signals, crushed control and IL1β-induced HyC matrices. (E–F) Representative flow cytometry plots for (E) CD80 and (F) CD206 stainings in the different conditions. (G–I) Gene expression of (G) IL6, (H) TNFɑ and (I) IL10 in human macrophages exposed to M1 control signals, M2 control signals, crushed control and IL1β-induced HyC matrices. (A-D, G-I) Data are plotted as means ± SD; n = 6. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. One-way ANOVA with Tukey's multiple comparison tests.

Fig. 5

IL1β-induced HyC enhances in vitro human osteoclast differentiation. (A–H) Brightfield images and TRAP staining in in vitro osteoclast cultures in presence of control (“C”) and IL1β-induced (“IL1β”) HyC matrices with (+) or without (−) RANKL. Scale bar, 500 μm. (I) TRAP quantification in osteoclast cultures in presence of control and IL1β-induced HyC matrices with/without RANKL. (J–K) Gene expression of (J) MMP9 and (K) CTSK in osteoclast cultures in presence of control and IL1β-induced HyC matrices with/without RANKL. (I–K) Data are plotted as means ± SD; n = 6. *P < 0.05, **P < 0.01, ****P < 0.0001. One-way ANOVA with Tukey's multiple comparison tests.