Computational modeling based on confocal imaging predicts changes in osteocyte and dendrite shear stress due to canalicular loss with aging

Abstract

Exercise and physical activity exert mechanical loading on the bones which induces bone formation. However, the relationship between the osteocyte lacunar-canalicular morphology and mechanical stress experienced locally by osteocytes transducing signals for bone formation is not fully understood. In this study, we used computational modeling to predict the effect of canalicular density, the number of fluid inlets, and load direction on fluid flow shear stress (FFSS) and bone strains and how these might change following the microstructural deterioration of the lacunar-canalicular network that occurs with aging. Four distinct computational models were initially generated of osteocytes with either ten or eighteen dendrites using a fluid-structure interaction method with idealized geometries. Next, a young and a simulated aged osteocyte were developed from confocal images after FITC staining of the femur of a 4-month-old C57BL/6 mouse to estimate FFSS using a computational fluid dynamics approach. The models predicted higher fluid velocities in the canaliculi versus the lacunae. Comparison of idealized models with five versus one fluid inlet indicated that with four more inlets, one-half of the dendrites experienced FFSS greater than 0.8 Pa, which has been associated with osteogenic responses. Confocal image-based models of real osteocytes indicated a six times higher ratio of canalicular to lacunar surface area in the young osteocyte model than the simulated aged model and the average FFSS in the young model (FFSS = 0.46 Pa) was three times greater than the aged model (FFSS = 0.15 Pa). Interestingly, the surface area with FFSS values above 0.8 Pa was 23 times greater in the young versus the simulated aged model. These findings may explain the impaired mechano-responsiveness of osteocytes with aging.

Keywords: Aging; Computational fluid dynamics (CFD); Confocal imaging; Fluid–structure interaction (FSI); Mechanotransduction; Osteocyte.

Fig. 1

A 3D idealized osteocyte model components, depicting bone tissue (brown), lacunar-canalicular fluid (blue), and osteocyte cell body and dendrites (green). The magnified sections indicate a larger lacunar fluid space than canalicular fluidic space. B Four osteocyte models with ten or eighteen dendrites. The load was imposed from either the left side of the solid domains (models 1–3) or the top side (Model 4), as indicated by an orange arrow. All or one of the canaliculi (circled in red) on the loading face were designated as fluid inlets. The rest of the canaliculi were considered outlets. For instance, in model 1 with eighteen dendrites in which the load is applied from the left, all five canaliculi on that face were considered inlets, whereas in Model 2, with the same canaliculi number and load direction, there is only one fluid inlet

Fig. 2

A Single Z plane image from a confocal stack of an FITC stained 4mo old mouse femur (100 × 1.44NA oil objective, 1.7 digital zoom, step size 0.126 μm, 0.089 μm pixel resolution) showing the cropped ROI around the lacuna used for the CFD modeling, B young osteocyte 3D model generated using Materialise Innovation Suite. C Simulated aged osteocyte generated from the same confocal image stack using a thresholding process. D Cross-sectional 3D view of the young osteocyte model showing modeling of the osteocyte cell membrane, lacunar wall and lacunar fluid space. E Enlarged region from D showing modeling of the canalicular fluid space around the dendrite, where the fluid flows into or out of the cell

Fig. 3

A Comparison of the maximum FFSS on dendrites inside the canalicular fluid space of the four idealized models. The red dashed line shows the 0.8 Pa threshold for osteogenic response B FFSS contours of four osteocyte models with idealized geometries

Fig. 4

Strain contours in bone and lacunar-canalicular system for the four idealized models. The magnified parts of Model 3 and Model 4 show the strain magnitude and pattern in the top canaliculi

Fig. 5

Contour plots of FFSS in young and simulated aged osteocyte models based on confocal image stacks

Fig. 6

A Velocity streamlines in the realistic young osteocyte model inside the lacunar and canalicular fluid spaces surrounding the cell and dendrites. Alteration of fluid velocity profile in image-derived models due to tortuosity of the canaliculi is presented in B Canalicular branches, C Canalicular junction, D Narrow channels

Fig. 7

Cross-section of the osteocyte from the top view shows the velocity streamlines inside the lacunar space

Fig. 8

Velocity streamlines and pressure contours in idealized and confocal-based models shown in A & D and H & I, respectively. Idealized modeling indicates a constant velocity in single canaliculi (B), (C), while the confocal-based model shows variations in fluid velocity profiles in (E), (F), (G)