Recapitulation of in vivo angiogenesis and osteogenesis within an ex vivo muscle pouch-based coral-derived macroporous construct organoid model

Abstract

Background: Segmental bone defect is a challenging clinical problem that often requires autologous bone grafting, which has limitations such as donor site morbidity and insufficient supply. Bone tissue engineering aims to create functional bone substitutes that can mimic the properties and processes of native bone. However, the discrepancy between in vitro and in vivo conditions hinders the successful translation of bone tissue engineering from animal models to human applications. Organoids, such as muscle pouch-based models, are emerging as promising tools that can closely resemble the osteogenic niche and overcome some of the limitations of conventional in vitro models.

Methods: In this study, we explored two distinct muscle-biomaterial based bone induction models: an in vivo heterotopic implantation model and a novel ex vivo muscle pouch-based coral-derived macroporous construct organoid model. They both utilized the coral-derived constructs, specifically 13 % hydroxyapatite/calcium carbonate (13 % HA/CC) as the biomaterial. We implanted 72 coral-derived devices into rats' rectus abdominis muscle, divided equally between in vivo and ex vivo groups. Samples were harvested at 15, 30, and 60 days for molecular and histological analyses. We assessed the relative gene expression of angiogenesis markers (Vegfa and Col4a1) and osteogenesis signaling and structural markers (Runx2, Bmp2, Ocn and Alp) using qRT-PCR. We analyzed tissue morphogenesis, angiogenesis and induction of bone formation by H&E and modified Goldner's Trichrome staining. Immunostaining was further used to detect the expression and localization of OCN, VEGFA and CD31 in both in vivo and ex vivo models.

Results: We demonstrated that ex vivo muscle pouch-based coral-derived macroporous construct organoid model supported tissue survival up to 60 days with compromised tissue ingrowth compared to the in vivo model. Primary vascular structures formed at the tissue-scaffold interface in the organoid system with persistent up-regulation of Vegfa and Col4a1, while comprehensive angiogenesis took place with early up-regulation of Vegfa and Col4a1 in vivo. Proper bone formation was absent in both the ex vivo and in vivo models, but the in vivo models showed an up-regulation of Bmp2 and Alp in early phase and a delayed Ocn expression on day 30. The ex vivo model showed connective tissue formation, comprehensive OCN deposition, and gene expression patterns mimicking in vivo trends but with some distinctions.

Conclusions: The ex vivo muscle pouch-based coral-derived macroporous construct organoid model in this study can partially recapitulate angiogenesis and osteogenesis as compared to the in vivo model. However, key molecular signaling events that regulate these processes remained inactive. The study demonstrated that activating these events could enable the establishment of an ex vivo tissue-based vascularized model.

The translational potential of this article: This study partly elucidated the molecular signaling events involved in the development of an ex vivo tissue-based osteogenic organoid that closely resembled its in vivo counterpart. This would facilitate the development of well vascularized artificial bone grafts for treating segmental bone defects.

Keywords: Angiogenesis; Coral-derived construct; Muscle pouch-based; Organoids; Osteogenesis.

Graphical abstract

Fig. 1

Schematic diagram of the establishment of in vivo heterotopic implantation model and ex vivo muscle pouch-based coral-derived macroporous construct organoid model. 13 % HA/CC microporous devices were implanted into the rectus abdominis muscle pouch of rats and either maintained in vivo as the heterotopic implantation model or dissected and cultured in vitro as the ex vivo organoid model. Analysis was performed between the muscle-13 % HA/CC complex obtained from in vivo and ex vivo by histology/histomorphometry and gene expression/protein production assay. 13 % HA/CC, 13 % hydroxyapatite/calcium carbonate.

Fig. 2

Comparison of tissue survivability and growth between the in vivo and ex vivo muscle pouch-based coral-derived macroporous construct organoid models on day 0, 15, 30 and 60 (A–P).In vivo models (A–H) revealed that the invasion of connective tissue was broadly distributed within the macroporous devices on day 15 compared to day 0 (A, C). At higher magnification (D), a distinctive distribution of dense cellular deposition was observed at the interface (D, green arrow), in contrast to sparse cells found within the connective tissue matrix, whereas principal vasculature was found in the central zone (D, blue arrow). The cell density was much higher on day 30 compared to day 15, and mature vessels were broadly distributed throughout the tissue (E and F, blue arrows). Multinucleated giant cells (F, green arrow) formed near blood vessels in the concave space of the macroporous constructs. In vivo results by day 60 (G, H) depicted abundant vessels uniformly distributed within the firm connective tissue in the central region (H, green arrows) supporting the substantial formation of condensed fibers lining mainly at the interface of the constructs. In ex vivo organoid model, compared to Day 0 (I, J), connective tissue grew into the center of the construct on Day 15 (K, L), but failed to survive and develop over time with scattered live cells mainly deposited in the peripheral area (M, N). Nevertheless, tissue survival of in vitro culturing with viable cells deposited within the extracellular matrix can last up to 60 days (O, P). H&E staining. Scale bars: A, C, E, G, I, K, M and O = 1 mm; B, D, F, H, J, L, N, and P = 50 μm.

Fig. 3

Comparison of vasculogenesis between the in vivo and ex vivo muscle pouch-based coral-derived macroporous construct organoid models on day 15, 30 and 60 (A–K). On day 15 in the in vivo model, morphology showed capillaries formed within the pores of the coral-derived scaffold supported by areolar tissue (A, green arrow), indicating upregulated expression of Vegfa and Col4a1 (I). By day 30, the number of blood vessels increased (B, green arrow). Meanwhile, Vegfa and Col4a1 expression was dramatically downregulated (J); on day 60, the number of blood vessels increased even more and blood vessel bundles were formed (C, blue arrow), despite the continued downregulation of Vegfa and Col4a1 (K). A significant increase of the number and diameter of blood vessels was noted on day 60 compared to the early phase within in vivo models by quantitative histological analysis (G, H). Vascular structures were visible at the interface on day 15 (D) and survived by day 30 (E) in the ex vivo muscle pouch-based coral-derived macroporous construct organoid model, although there were no blood vessels within the 13 % HA/CC scaffold. Some blood vessel-like structures were observed on day 60 (F). Quantitative histological analysis of vessel numbers (G) and vessel diameters (H) showed very limited numbers of relatively small vessels within ex vivo organoid models without significance over time. Ex vivo organoid model demonstrated continuous upregulation of Vegfa and Col4a1 on day 15, 30, and 60, and in vivo results revealed early upregulation and later inhibition (I, J, K). H&E staining (A–F). Error bars are Mean ± SD. ∗, P < 0.05. Scale bars: A, B, and C = 50 μm; D, E and F = 50 μm. 13 % HA/CC, 13 % hydroxyapatite/calcium carbonate; CNRQs, calibrated normalized relative quantities.

Fig. 4

Immunohistological and gene expression analysis of vasculogenesis in ex vivo muscle pouch-based coral-derived macroporous construct organoid model. Immunohistological analysis of VEGFA protein production within the coral-derived constructs illustrated the presence of blood vessels in the ex vivo organoid model on day 60 (A–D). Meanwhile, the expression of angiogenic gene markers (Vegfa and Col4a1) was continuously upregulated over time (E, F). Decalcified sections cut at 2 μm underwent immunohistological analysis of VEGFA. Error bars are Mean ± SD. ∗∗∗, P < 0.001. Scale bars: A = 1 mm; B = 200 μm; C = 100 μm; D = 50 μm. CNRQs, calibrated normalized relative quantities.

Fig. 5

Immunofluorescence analysis of vasculogenesis within 13 % HA/CC over time in the ex vivo and in vivo models. Immunofluorescence staining of VEGFA and CD31 protein production within the coral-derived constructs identified limited VEGFA and CD31 production by different cells without vasculature formation ex vivo (A–C) compared to the early vessel formation (D) and mature vascular network formation over time (E, F) in vivo. Quantitative analysis (G–J) depicted the continuous production of VEGFA and CD31 at lower level ex vivo (G, H), while that of in vivo at a higher level with significant increase of CD31 production over time (I, J). Error bars are Mean ± SD. ∗, P < 0.05. Scale bars: A-F = 50 μm. 13 % HA/CC, 13 % hydroxyapatite/calcium carbonate; VEGFA, vascular endothelial growth factor A.

Fig. 6

Tissue morphogenesis within the in vivo heterotopic implantation model. Specific tissue morphogenesis patterns formed in the macroporous coral-derived devices within the heterotopic implantation model by day 60 (A–D). Mesenchymal tissue underwent condensation at the interface of the constructs (A and B, green arrows) with the support of abundant mature vessels (blue arrows). Transformation from condensed connective tissue to homogeneous tissue was observed in some concave spaces of the macro-porous constructs (C and D, red arrow). Bmp2 peaked on day 15 and exhibited a considerable decline over time (E); Alp gene expression assay revealed a similar regulation pattern (G). Ocn expression was upregulated on day 15 and peaked on day 30, following significant downregulation (F). By quantitative histological analysis of Goldner's trichrome stainings, significant increase of osteoid area was noted on day 30 and day 60 compared to day 15 within in vivo models while lower percentage of osteoid area was observed over time in ex vivo organoid models (H). H&E staining (A, B). Goldner's trichrome staining (C, D). Error bars are Mean ± SD. ∗, P < 0.05; ∗∗, P < 0.01. Scale bars: A, C = 200 μm; B, D = 100 μm. CNRQs, calibrated normalized relative quantities.

Fig. 7

Induction of osteogenesis within ex vivo muscle pouch-based coral-derived macroporous construct organoid model. On day 60 ex vivo, limited osteoid was formed at the interface of the macroporous spaces (A, B). Immunohistological assays depicted a complete OCN deposition within the 13 % HA/CC construct (C, D) and considerable histomorphometric increase over time (F). For gene expression analysis (E), Bmp2 was upregulated with a decreasing tendency; Alp was consistently suppressed; Ocn was significantly upregulated from day 30 and last to 60 days; and Runx2 was consistently upregulated at a high level. Goldner's trichrome staining (A, B). Decalcified sections cut at 2 μm underwent immunohistological analysis of OCN (C, D). Error bars are Mean ± SD. ∗, P < 0.05; ∗∗, P < 0.01. Scale bars: A = 1 mm; B = 100 μm. 13 % HA/CC, 13 % hydroxyapatite/calcium carbonate; CNRQs, calibrated normalized relative quantities.

Supplementary Figure 1

Quantitative immunohistological analysis of VEGFA protein production within ex vivo muscle pouch-based coral-derived macroporous construct organoid model. Immunohistological assays observed a comprehensive VEGFA deposition within the construct and quantitative analysis showed an obvious growth tendency over time but with no statistical significance. VEGFA, vascular endothelial growth factor A; MOD, mean optical density; IOD, integrated optical density; ROI, region of interest.