Studying osteocytes within their environment

Abstract

It is widely hypothesized that osteocytes are the mechano-sensors residing in the bone's mineralized matrix which control load induced bone adaptation. Owing to their inaccessibility it has proved challenging to generate quantitative in vivo experimental data which supports this hypothesis. Recent advances in in situ imaging, both in non-living and living specimens, have provided new insights into the role of osteocytes in the skeleton. Combined with the retrieval of biochemical information from mechanically stimulated osteocytes using in vivo models, quantitative experimental data is now becoming available which is leading to a more accurate understanding of osteocyte function. With this in mind, here we review i) state of the art ex vivo imaging modalities which are able to precisely capture osteocyte structure in 3D, ii) live cell imaging techniques which are able to track structural morphology and cellular differentiation in both space and time, and iii) in vivo models which when combined with the latest biochemical assays and microfluidic imaging techniques can provide further insight on the biological function of osteocytes.

Figure 1

Osteocyte network and lacuno-canalicular network (LCN) retrieved from different high resolution ex vivo imaging techniques. (A) Corrosion casting replica of the LCN in cortical mouse bone. The data have been assessed by scanning electron microscopy (SEM). Figure reprinted from [5] with permission. (B) Osteocyte network of chick calvariae. The osteocyte cell bodies and their processes have been assessed from 3-μm-thick bone sections by transmission electron microscopy computed tomography (TEM CT) at a nominal resolution of 16 μm. Contrast of the osteocyte network surface for TEM CT imaging was imparted by silver staining, which allowed for the displayed three-dimensional reconstruction of the osteocyte and the cell processes’ surface. Figure reprinted from [24] with permission. (C) LCN of the femoral mid-diaphysis in the mouse assessed by serial focused ion beam/scanning electron microscopy (FIB/SEM). For serial FIB/SEM imaging, thin bone layers (down to a few nanometers) are milled away from the sample’s block face by an ion beam, alternating with SEM imaging of the block face. The acquired SEM sections are then registered and stacked together for following segmentation of the three-dimensional representation of the LCN, including osteocyte lacunae (yellow ellipsoid portions) and canaliculi (green tubes). The data have been assessed at a nominal resolution of 18.6 nm × 18.6 nm in-plane and at 29.5 nm between serial sections. Figure reprinted from [30] with permission.

Figure 2

Stained osteocyte network imaged by confocal laser scanning microscopy (CLSM). The images from parietal bone (a, c, e, g) and tibia (b, d, f, h) in the mouse show intensity z-projections of the original CLSM slices (a, b) and surface renderings of osteocyte cell bodies and processes (c, d and e, g), and of osteocyte nuclei. (e, f, g, h). Figure reprinted from [9] with permission.

Figure 3

Intracortical microstructure of femoral mouse bone. The canal network (red tubes) and osteocyte lacunae (yellow ellipsoids) have been obtained from synchrotron radiation-based micro-computed tomography (SR CT) at a voxel size of 700 nm. It is the bone’s X-ray absorption through which the intracortical microstructure has been segmented as a negative imprint of the mineralized bone matrix. Figure reprinted from [15] with permission.

Figure 4

Still frames from a time lapse imaging series of Dmp1-GFP-positive osteocytes in a 7 day old neonatal mouse calvarial explant. Images were acquired every 20 min for 12 h on a widefield fluorescence microscope using the green fluorescence channel for GFP (shown) and DIC for imaging of the bone explant (not shown). The osteocytes marked with an asterisk (*) show motions of their dendrites, which adopt various configurations at different times during the time lapse imaging period. Below each image, the outline of these osteocyte and their dendrites is traced to illustrate the dendrite conformations. Bar = 30 μm. Modified from [36] with permission.

Figure 5

a) Still frames from a time lapse imaging series of mineralizing primary osteoblast cultures from Dmp1-GFP transgenic mice in which alizarin red was used as a vital dye for mineral deposition. The movie was started when clusters of GFP positive cells had formed in the cultures and images were acquired every 20 min for 48 h in DIC and two fluorescent channels. Triple merged images of DIC, GFP and alizarin red are shown. Bar = 100 μm. b) Quantitation of mineralization dynamics from the time lapse sequence shown in (a). Mineralization was quantified using Image J software by thresholding of alizarin red image stacks followed by measurement of the mineralized area (red line). The number of GFP-positive cells were counted (green line). Note the deposition of mineral beginning 10 h after addition of 4 mM β-glycerophosphate and increasing until 40 h. Mineral deposition occurs specifically where there are clusters of GFP-positive cells and mineralization is accompanied by an increase in the number of GFP-positive cells. These data suggest that the cells responsible for mineral deposition are already transitioning towards the osteocyte phenotype. Figure reprinted from [36] with permission.

Figure 6

Time lapse imaging of the autonomic [Ca2+]i responses of osteocytes in intact calvarial explants (A) Fluorescence and pseudocolor images of living osteocytes in intact calvarial explants. The color scale represents the relative fluorescence intensity. The numbered cells in the fluorescent image correspond to the cells assessed in panel B. (B) representative normalized relative fluorescence intensity plots of individual osteocytes in intact calvarial explants. Each line represents the time course of the [Ca2+]i response of the individual osteocyte indicated by the corresponding number in panel A. The lower case letters (a-b) in (B) indicate the [Ca2+]i response during which the frames shown in (C) were taken. (C) serial pseudocolor images of the osteocytes shown in A taken at 0, 99, 162 and 252s after the initiation of monitoring. Consecutive images, which were collected from a single Z plane were taken at 3 second intervals for 5 min. Bar = 10mm. Figure reprinted from [49] with permission.

Figure 7

The Mechanical Systems Biology framework for investigating load-induced bone adaptation is a combined experimental and computational approach which can be separated into 3 different workflows. Figure reprinted from [64] with permission.