Vimentin contributes to changes in chondrocyte stiffness in osteoarthritis

Abstract

Actin and tubulin cytoskeletal components are studied extensively in chondrocytes, but less is known about vimentin intermediate filaments. In other cell types, vimentin is a determinant of cell stiffness and disruption of vimentin networks weakens the mechanical integrity of cells. Changes in vimentin organization correlate with osteoarthritis progression, but the functional consequences of these changes remain undetermined in chondrocytes. The objective of this study was to compare the contribution of vimentin to the mechanical stiffness of primary human chondrocytes isolated from normal versus osteoarthritic cartilage. Chondrocytes were embedded in alginate and vimentin networks disrupted with acrylamide. Constructs were imaged while subjected to 20% nominal strain on a confocal microscope stage, and the aspect ratios of approximately 1,900 cells were measured. Cytosolic stiffness was estimated with a finite element model, and live-cell imaging of GFP-vimentin was used to further analyze the nature of vimentin disruption. Vimentin in normal chondrocytes formed an inner cage-like network that was substantially stiffer than the rest of the cytosol and contributed significantly to overall cellular stiffness. Disruption of vimentin reduced stiffness approximately 2.8-fold in normal chondrocytes. In contrast, osteoarthritic chondrocytes were less stiff and less affected by vimentin disruption. This 3D experimental system revealed contributions of vimentin to chondrocyte stiffness previously not apparent, and correlated changes in vimentin-based chondrocyte stiffness with osteoarthritis.

Figure 1

A) Chondrocytes in alginate gel. These gels were cut in half radially for compression and confocal imaging. B) Schematic of the compression setup depicting the microscope objective, and the direction of compression (arrows) on the gel half. C) Typical images of uncompressed and compressed cells obtained with the confocal microscope. The green fluorescence is from the calcein-AM component of the live-dead stain. The yellow arrows indicate the measurements taken for analysis. D) GFP-Vimentin and cytosolic RFP staining of a transduced chondrocyte demonstrating the inner vimentin cage structure. E) Finite element mesh of a single alginate-embedded chondrocyte. The mesh (dark blue) has been sectioned to show the spherical cell (light blue) before and after 20% nominal strain is applied. F) Contour map of the displacement field of a cross-section of the cell-gel construct in the direction of compression, with the cell boundary outlined in green (units = micrometers).

Figure 2

Aspect ratio of compressed cells from healthy, TKA-normal, and TKA-OA cartilage, with and without disruption of vimentin with acrylamide treatment. Statistical analysis of ~1800 cell measurements are shown, * indicates p<0.05. Letters on the bottom of the bars indicate groups statistically different from each other by ANOVA with Tukey-Kramer’s post-hoc correction for multiple comparisons; bars with the same letter are not statistically different. Error bars represent 95% confidence interval of the mean.

Figure 3

Estimated cytosolic Young’s modulus based on cellular deformation in 20% nominal strain. Curve is derived from the finite element model; this assumed a Poisson’s ratios of 0.40 and 0.15 for the cytosol and the alginate, respectively. Horizontal error bars indicate 95% confidence interval of mean cellular aspect ratio. Measured X/Y ratios and estimated Young’s moduli are shown in table form below the graph.

Figure 4

GFP-Vimentin cytoskeleton forms a tight cage within the cell in alginate-embedded chondrocytes A) Single high-resolution confocal slice through the center of a chondrocyte. Arrow indicates superimposed cell outline, arrowhead indicates GFP-Vimentin. B) 3D reconstruction of high-resolution GFP-vimentin network after thresholding in DICOM software, with superimposed cell outline. C) 2D projection confocal image of a GFP-Vimentin-expressing chondrocyte fixed and stained with Phalloidin-Alexa-546 to identify cortical actin filaments (red) and inner GFP-vimentin network. Scale bar is 1um in A and B, 5um in C.

Figure 5

Slow turnover of vimentin cytoskeleton. A single alginate-embedded human chondrocyte expressing GFP-vimentin was photobleached, then observed from time t=0 to 9 minutes by confocal microscopy. There was little recovery of fluorescence in bleached area, suggesting that vimentin filaments form stable structures. Note that translation, rotation and distortion of the vimentin network occur in the living cell even in the absence of filament turnover. Scale 5um