In Vivo 3D Histomorphometry Quantifies Bone Apposition and Skeletal Progenitor Cell Differentiation

Abstract

Histomorphometry and Micro-CT are commonly used to assess bone remodeling and bone microarchitecture. These approaches typically require separate cohorts of animals to analyze 3D morphological changes and involve time-consuming immunohistochemistry preparation. Intravital Microscopy (IVM) in combination with mouse genetics may represent an attractive option to obtain bone architectural measurements while performing longitudinal monitoring of dynamic cellular processes in vivo. In this study we utilized two-photon, multicolor fluorescence IVM together with a lineage tracing reporter mouse model to image skeletal stem cells (SSCs) in their calvarial suture niche and analyze their differentiation fate after stimulation with an agonist of the canonical Wnt pathway (recombinant Wnt3a). Our in vivo histomorphometry analyses of bone formation, suture volume, and cellular dynamics showed that recombinant Wnt3a induces new bone formation, differentiation and incorporation of SSCs progeny into newly forming bone. IVM technology can therefore provide additional dynamic 3D information to the traditional static 2D histomorphometry.

Figure 1

In vivo, multi-color two-photon imaging. (a) Injection regimens and imaging time points for in vivo histomorphometry. (b) The coronal suture space (S) between frontal (fb) and parietal (pb) bones. (c) Animals were treated with either rmWnt3a (in PBS, 44 ng/day) or just PBS for 14 days. Maximum intensity projection of a 25-μm layer was used for demonstration. Images on Day0 show Prx1 + cells (Green), their progeny (Red or yellow, co-expression of tdTomato and EGFP). In addition, on Day 28, the initial bone structure was identified by calcein blue (Blue) and the new bone fronts were demarcated by tetracycline (purple), as indicated by blue and purple arrows in Fig. 1(d). Spontaneous recombination (white arrowheads) and incomplete recombination (green arrows) may be present. It is noted that a higher number of osteocytes originated from Prx1 + cells were embedded in lacunae (the red arrows and the inset) from rmWnt3a treatment. (d) Quantifications of bone morphometry and cellular dynamics were performed over a 50-μm layer, as shown in 3D reconstructed images.

Figure 2

Image processing and quantitative analysis. Automatic cell counting adapted from Image J Plugins (a) Overlay of the identified seeds (yellow) with tdTomato + cells. Images were presented using maximum intensity projection. (b) Validation of automatic seed identification against 50 manually identified cells. (c) The counting accuracy of 5.7% ± 3.8% in the EGFP group, and 4.4% ± 3.5% in the tdTomato group was validated by 28 randomly picked regions (a total of 7127 and 9025 cells, respectively). Error bars represent standard deviation.

Figure 3

3D quantifications of bone morphometry. Volume measurement is demonstrated by a 3D reconstructed image. The bone fronts were manually segmented in each z-slice. The white dashed lines represent the suture area, and the green dashed lines represent the new bone area in the first slice. All the measured areas were summed over 50 μm in z dimension and converted to mm³.

Figure 4

Dynamic changes of bone histomorphometry in response to rmWnt3a. Quantitative measurement of dynamic changes through a period of 28 days in response to rmWnt3a treatment compared to the control group. The box plots demonstrate the mean, median, the interquartile range, the whiskers showing the total range of the measurements, and the individual data points greater or less than by 1.5 times the interquartile range. (a) tdTomato + osteocytes derived from Prx1 progeny (b) Bone growth (c) Cell density, tdTomato + osteocytes normalized to bone growth (d) tdTomato + osteocytes normalized to initial cell count at Day 0 (e) Frontal bone growth (f) Parietal bone growth (g,h) Osteocytes incorporated to frontal bone and parietal bone, respectively (i) Expansion of Prx1 progeny (j) Changes of Prx1 + population (k) Changes in suture volume. (N = 6 and 5 animals in rmWnt3a and PBS treated groups, respectively. 6 to 8 regions in the coronal suture of each animal were acquired).