2017 Volume 58 Issue 2 Pages 266-270
Bone tissue has a highly anisotropic microstructure derived from the crystallographic orientation of apatite and the related collagen matrix alignment. Bone is also a highly vascularized tissue; intraosseous vascularization and bone formation are intimately coupled. Meanwhile, the structural relations between intraosseous vascular networks and bone microstructure are as yet unknown, partially due to technical difficulties in visualizing precise intraosseous vasculatures. The aim of this study is to develop a visualization method suitable for the structural analysis of intraosseous vascular networks and to reveal the relations between bone microstructure and the arrangement patterns of intraosseous vasculatures. Three-dimensional vascular networks were successfully visualized, and region-dependent arrangement patterns of blood vessels were clarified using fluorescent dye-conjugated lectin. Interestingly, the anisotropic structural correlation between bone matrix and the vascular system in a region-specific manner was clarified. The obtained results indicate the molecular interactions between the vascular system and bone tissue as a novel contributor for realization of anisotropic bone matrix construct.
Bone is a highly vascularized tissue, and the intraosseous blood supply plays a crucial role in bone modeling and remodeling. The blood supply modulates the microenvironment for bone cells by supplying requisite factors such as nutrients, oxygen and cytokines, and removing waste products such as carbon dioxide and acids.1) In addition to these conventional roles of blood vessels, increasing evidence suggests that the vascular system modulates organ homeostasis via crosstalk with other tissues or organs. In bone tissue, the vascular system plays significant roles in development and fracture healing process. The blood supply also helps bone cells to reach the appropriate site for bone modeling/remodeling; it transports osteoclast precursors to the bone surface where remodeling should occur and recruits circulating osteoblast precursor cells to the bone remodeling compartment.2,3) Because the mechanical and biological functions of bone tissue are governed by its anisotropic microstructure,4–6) it is essential to understand the structural relationship between bone tissue orientation and intraosseous vascularization.
To date, the involvement of intraosseous vascularization in bone maintenance and disease has been studied by only focusing on the regulation of bone mass. It was demonstrated that reduced intraosseous vascularization or blood flow is linked with reduced bone mass both in age-related7) and spinal cord injury-induced8) osteoporosis. It is also known that angiogenesis is stimulated in response to bone injury, resulting in increased transportation of osteoprogenitor cells and increased bone formation, whereas bone disuse leads to a reduction of the blood supply, resulting in decreased bone mass.9)
Evaluating intraosseous vascularization with regard to bone microstructure is important for treatments aimed at the recovery of bone tissue anisotropy and also the development of biomedical materials to induce oriented bone tissue regeneration. Nevertheless, even the evaluation methods for intraosseous vascular structure are not well established. The aim of this study is to develop a visualization method suitable for the structural analysis of intraosseous vascular networks and to analyze the relations between the arrangement patterns of intraosseous vasculatures and bone matrix microstructure in lamellar bone.
Some evaluation methods for intraosseous vasculatures in small animals are currently available. These include serial histology in longitudinal or transverse planes,10) intravenous injection of fluorescent dyes,11) X-ray micro-angiography with infusion of contrast agents,12,13) and scanning electron microscopy with corrosion casting.14,15) These techniques, however, still have respective problems related to sensitivity or manipulation difficulties when applied to the analysis of the vascular patterns of mouse bones.16)
Previously, Pazzaglia et al.17) reported a visualization method using black China ink in full-thickness specimens of rabbit femurs. They successfully confirmed the orientation patterns of the canal networks in Haversian lamellar bones. With this method, however, only a few planar images are obtained by focusing manually on different planes of thick cortical bone, and hence, it is difficult to discern the three-dimensional overlap geometry from vessel bifurcations in these images, as is often the case with projection images. In addition, the cortical bone sections are tightly pressed flat between two glass slides, inevitably resulting in specimen deformation and alterations in the relative position of canal networks.
In this study, the matrix anisotropy in non-Haversian lamellar bone of mouse femurs was analyzed by microbeam X-ray diffraction and birefringent measurement to elucidate the arrangement of both biological apatite (BAp) and collagen fibrils, the major constituents of bone matrix. We applied fluorescent dye-conjugated lectin that is highly specific to vascular endothelial cells to visualize the three-dimensional-vascular networks of mouse femurs and acquired serial image planes with confocal laser microscopy. The patterns of the vascular networks were qualitatively assessed both in mid-shaft regions and distal regions of femurs based on reconstructed image volumes. The structural relations between the intraosseous vasculatures and bone matrix microstructure were successfully revealed using the above novel procedures.
A schematic illustration of the procedures used for analyzing bone matrix orientation and visualizing intraosseous vascular networks is shown in Fig. 1. A crystallographic analysis of the BAp c-axis orientation was conducted on non-decalcified femurs using a microbeam X-ray diffractometer with a transmission optical system (R-Axis BQ, Rigaku, Tokyo, Japan) as previously described.18) Mo–Kα radiation was generated at a tube voltage of 50 kV and tube current of 90 mA, and collimated into a 300-μm circular spot by a double-pinhole metal collimator. Peaks for (002) and (310) reflections were identified in the profile, and the integrated intensity ratio of (002) to (310) was calculated.
A schematic illustration of procedures for analyzing biological apatite (BAp) c-axis orientation, collagen fibrils orientation, and vascular networks organization in mouse non-Haversian femurs.
To evaluate the collagen fibrils orientation, a WPA-micro birefringence measurement system (Photonic Lattice, Miyagi, Japan) attached to an upright microscope (Olympus) was used. Full-thickness decalcified specimens were imaged with a 20× objective lens by a transmission method. Data were acquired as the average of 50 images, with three settings of circularly polarized monochromatic light (laser wavelengths 523, 543, and 575 nm) for each image. The orientation of the polarization axis with the greater index of refraction (the slow axis) of each specimen was analyzed with WPA-VIEW software (version 18.104.22.168, Photonic Lattice). Owing to the nature of collagen fibrils as a positive birefringent material, we utilized the direction of the slow axis as the index of collagen fibrils orientation.
The visualization procedure for the vascular networks is as follows. C57BL6 mice (female, 13-weeks-old, Japan SLC) were anesthetized with an intraperitoneal administration of sodium pentobarbital (50 mg/kg) and intravenously injected in the tail vein with 150 μl of 1 mg/ml DyLight 594-conjugated Lycopersicon esculentum (tomato) lectin (Vector Labs, Burlingame, CA) in phosphate buffered saline (PBS). Fifteen minutes later, the mice were sacrificed with an additional administration of sodium pentobarbital (70 mg/kg) and perfused with saline containing heparin sodium (10 unit/ml, Mochida Pharmaceutical, Tokyo, Japan) followed by a 4% paraformaldehyde perfusion through the cardiac left ventricle. Femurs were resected and immersed in 4% paraformaldehyde at 4℃ for 24 hours. After washing with PBS, the femurs were decalcified with a 0.5 M EDTA-2Na solution (pH 7.4) for 7 days at 4℃. Then, by cutting longitudinally at a frontal plane and removing the metaphysis with a blade, sheets of diaphyseal cortical bone were obtained from the anterior side. The marrows were carefully removed.
The specimens were mounted onto microscope slides with 50% glycerol in PBS and observed with confocal laser microscopy (Olympus, Tokyo, Japan) at an excitation wavelength of 515 nm and an emission wavelength of 559 nm. Images were viewed with a 10× objective lens with field dimensions of 1280 × 1280 μm at the third trochanter, mid-shaft, and distal regions. Serial images were acquired at 2-μm intervals over a 100- to 120-μm depth to cover the full thickness of the specimens. Image processing was conducted using FluoRender (version 2.19.4, NIH, Bethesda, MD). All animal experimentation protocols were approved by the Animal Care and Use Committee of Osaka University.
The BAp c-axis orientation was measured by microbeam X-ray diffraction. Figure 2 shows diffraction patterns of whole bone specimens. The degree of BAp orientation was higher in the mid-shaft region (intensity ratio of (002)/(310) = 10.4) than in the distal region (6.0), indicating that the degree of BAp orientation in the longitudinal direction of the bone was higher in the mid-shaft region than in the distal region.
Microbeam X-ray diffraction patterns of a bone specimen. Left; the mid-shaft region, right; the distal region in mouse femurs. The values of (002)/(310) are indicated below the images. The degree of BAp orientation along the bone longitudinal axis in the mid-shaft region exhibited a higher value than in the distal region.
The collagen fibrils orientation of full-thickness cortical bone was determined by the quantitative birefringent measurement of decalcified bone specimen (Fig. 3(a)). Except for the region around osteocyte lacunae where matrix organization tends to be influenced by the geometry of lacunae, collagen fibrils were generally aligned along the bone longitudinal axis in the mid-shaft region. In contrast, collagen fibrils were less aligned along the bone longitudinal axis in the distal region. The distribution of collagen fibrils was broader in the distal region than in the mid-shaft region (Fig. 3(b)).
(a) Images of bright fields (top) and collagen orientation (bottom) of decalcified femurs analyzed by birefringence measurement. Left column; the mid-shaft region, right column; the distal region in mouse femurs. Arrows indicate the orientation of the slow axes that represent the collagen orientation in bone matrix. Bars = 50 μm. (b) Distribution of the collagen orientation in the fields indicated in (a).
The intraosseous vascular networks in mouse femurs were successfully visualized in three dimensions by the intravenous administration of a fluorescent dye to mice combined with confocal laser microscopy, even in irregularly shaped parts such as the third trochanter (Fig. 4). Figure 5(a) shows representative images of a single-plane observation, and Fig. 5(b) shows representative views of the surface rendering of reconstructed image volumes of diaphyseal cortical bone. We observed a pattern of the longitudinal alignment of the vasculatures along the bone longitudinal axis in the mid-shaft of femurs. On the contrary, the vasculatures was poorly aligned along the bone longitudinal axis in the distal region of the femurs. Vessel bifurcations and overlaps are clearly distinguishable in the rendered images (Fig. 5(c)).
Visualization of intraosseous vascular networks in mouse femurs. Left column; photograph of a mouse femur. Arrow indicates the third trochanter. Right column; vascular networks were successfully imaged even in the irregularly shaped regions such as the third trochanter. Bar = 100 μm.
Region-dependent arrangement patterns of blood vessels. Left column; the mid-shaft region, right column; the distal region in mouse femurs. (a) Representative single-plane images of intraosseous vascular networks in mouse femurs observed by confocal laser microscopy. Bars = 200 μm (b) Surface rendering of reconstructed image volumes. Bars = 50 μm. (c) Magnified images of cross-points in (b). Vessel bifurcations (arrows) and overlaps (arrowheads) are clearly distinguishable. Bars = 2 μm.
The significance of bone matrix alignment in long bone is now becoming increasingly clear. Bone matrix comprising collagen fibrils and biological apatite (BAp) crystals highly align along the bone longitudinal axis both in Haversian and non-Haversian lamellar structures in long bones, depending on stress distributions.4–6,19) More importantly, the alignment of collagen/BAp is an essential factor that contributes to bone mechanical function, as observed by the significant correlation between the degree of the BAp c-axis alignment and bone Young's modulus.6) Meanwhile, the association between the vascular patterns and bone microstructure is largely unknown, partially because of the technical difficulties associated with the vessels existing inside a highly mineralized matrix. In this study, the relationship between the region-dependent arrangement patterns of bone matrix and vascular networks was revealed using a combination of microbeam X-ray diffraction, birefringence measurement, and a newly developed visualization technique. Using the novel visualization method developed in the present study, we succeeded in imaging the intraosseous vascular networks at the third trochanter (Fig. 4) where three-dimensional visualization is hardly possible with preexisting techniques, in spite of its biomechanical significance.20)
The structural patterns of vascular networks and the manner of bone formation are intimately linked, as exemplified by Haversian remodeling systems seen in large mammalian animals such as rabbits, bovines, and humans. It has been well documented that in Haversian remodeling systems, the geometry of vasculatures directly determines the lamellar structure of secondary osteons; bone matrix is deposited in layers surrounding the Haversian canals that enclose vessels running along the bone longitudinal axis, and hence, lamellar structures are organized such that they encircle the vessels and extend parallel to the bone longitudinal axis.21,22) However, in small animals including mice and rats who lack Haversian remodeling systems, the association between bone matrix and vascular networks had been scarcely understood.
In this study, we analyzed mouse femurs and found a considerable correlation between the bone matrix anisotropy and structural patterns of vascular networks in non-Haversian lamellar bone. In the mid-shaft region, collagen fibrils, BAp c-axis, and vasculatures were highly aligned parallel to the bone longitudinal axis, whereas in the distal region, both the vasculatures and collagen fibrils were poorly aligned (Figs. 2, 3, and 5). Blood vessels not only mediate the transport of oxygen, nutrients, and circulating cells but also provide biological signals controlling organ homeostasis. The coupling of angiogenesis and osteogenesis was found to be regulated by specific molecules depending on the vessel type.23) The anisotropic microstructure of bone tissue is controlled by the mutual activities of bone cells involving osteoblasts, osteoclasts, and osteocytes. The obtained results indicate that the crosstalk between the vascular system and bone tissue also plays a key role in determining the crystallographic anisotropy of bone tissue. It is essential to clarify the molecular interaction between bone cells and vascular cells to understand the mechanism underlying anisotropic bone formation associated with organ crosstalk.
Recently, a high correlation between vascular patterning and bone formation in embryonic mouse bone was reported, indicating that the vascular patterns play a key role in determining bone shape by providing a template for collagen and mineral deposition during development phases.24) It was suggested that the vascular patterns are highly responsible for the bone microstructure during extensive stages of growth and that along with external factors, some inner factors that may be biologically programmed regulate the vascular patterns.4,9)
In addition to the bone matrix consisting of collagen fibrils and BAp, osteocyte networks are also an interacting element for the vascular patterns, because osteocyte networks have direct connections to vasculatures, enabling the transportation of molecules to and from osteocytes throughout the body.25) Osteocytes are known to act as a mechanosensor in bone tissue and transmit biomechanical signals to osteoclasts/osteoblasts through fluid flow inside the osteocyte networks.26) As the osteocyte networks display a regular geometry in lamellar bone and osteocytes align parallel to the lamellar plane and elongate canaliculi mainly perpendicular to the lamellar plane,27,28) some studies have suggested that the arrangement of the osteocyte networks is crucial for the efficient monitoring of mechanical strain and transportation of molecules.29,30) Therefore, studying the correlative geometry of vascular networks and osteocyte networks may be helpful to further understanding of bone function as a mechanosensitive and endocrine system.
In this study, the structural relationship between collagen/BAp orientation and intraosseous vascular networks was revealed by combining birefringence measurement, microbeam X-ray diffraction, and a newly developed visualization technique. The preferred orientation of the vascular networks was clearly correlated to the bone matrix orientation in non-Haversian lamellar bone of mouse femurs. The present findings indicate the interaction between bone cells and vascular cells as a novel trigger for anisotropic bone tissue construction.
This work was supported by JSPS KAKENHI Grant Number JP25220912, 15J01150.