Cyclic loading increases friction and changes cartilage surface integrity in lubricin-mutant mouse knees

Abstract

Objective: To investigate the effects of lubricin gene dosage and cyclic loading on whole joint coefficient of friction and articular cartilage surface integrity in mouse knee joints.

Methods: Joints from mice with 2 (Prg4(+/+)), 1 (Prg4(+/-)), or no (Prg4(-/-)) functioning lubricin alleles were subjected to 26 hours of cyclic loading using a custom-built pendulum. Coefficient of friction values were measured at multiple time points. Contralateral control joints were left unloaded. Following testing, joints were examined for histologic evidence of damage and cell viability.

Results: At baseline, the coefficient of friction values in Prg4(-/-) mice were significantly higher than those in Prg4(+/+) and Prg4(+/-) mice (P < 0.001). Cyclic loading continuously increased the coefficient of friction in Prg4(-/-) mouse joints. In contrast, Prg4(+/-) and Prg4(+/+) mouse joints had no coefficient of friction increases during the first 4 hours of loading. After 26 hours of loading, joints from all genotypes had increased coefficient of friction values compared to baseline and unloaded controls. Significantly greater increases occurred in Prg4(-/-) and Prg4(+/-) mouse joints compared to Prg4(+/+) mouse joints. The coefficient of friction values were not significantly associated with histologic evidence of damage or loss of cell viability.

Conclusion: Our findings indicate that mice lacking lubricin have increased baseline coefficient of friction values and are not protected against further increases caused by loading. Prg4(+/-) mice are indistinguishable from Prg4(+/+) mice at baseline, but have significantly greater coefficient of friction values following 26 hours of loading. Lubricin dosage affects joint properties during loading, and may have clinical implications in patients for whom injury or illness alters lubricin abundance.

Figure 1

Schematic diagrams of the 2 pendulum systems used in the experiment. A, Passive pendulum system used to measure the coefficient of friction. B, Active pendulum system used for cyclic loading. C, Illustration of how each mouse knee joint was positioned in the mounting block. Color figure can be viewed in the online issue, which is available at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)1529-0131.

Figure 2

Flow chart of the experimental protocol. Each experimental mouse limb underwent 26 cumulative hours of cyclic loading, 12 hours of unloading in cell culture medium, and oscillation data collection at 16 time points. The control mouse limbs remained in an unloaded state in culture medium for the duration of testing, and oscillation data were collected at 2 time points. Ctl = control; Exp = experimental; CoF = coefficient of friction.

Figure 3

A, Mean coefficient of friction values in experimental (Exp) and control (Ctl) mouse joints over the course of testing. The shaded bars at the bottom represent the time during which the experimental joint was cyclically loaded and unloaded. Note that the control coefficient of friction values did not change during the course of the experiment; however, the control coefficient of friction values in the Prg4^−/− mice were significantly higher than those in the other 2 genotypes (P < 0.0001). The experimental coefficient of friction values in the Prg4^−/− mice rose steadily during the 26 hours of cyclic loading. The coefficient of friction values in the Prg4^+/+ and Prg4^+/− mice were comparable at the onset of cyclic loading; however, at the conclusion of the cyclic loading experiment, the coefficient of friction values in the Prg4^+/− mice were closer to those in the Prg4^−/− mouse joints than to those in the Prg4^+/+ mouse joints. B, Coefficient of friction values in cyclically loaded joints from each genotype at 0, 16, 18, and 38 hours. At 0, 16, and 18 hours, the coefficient of friction values in the Prg4^−/− mice were significantly higher than those in the wild-type and heterozygous mice. However, at 38 hours, the coefficient of friction values in both the Prg4^−/− and Prg4^+/− mice were significantly higher than those in the Prg4^+/+ mice. Bars show the mean ± SD. * = P < 0.0001 versus Prg4^+/+ and Prg4^+/− mice; ** = P < 0.003; *** = P < 0.0001, versus Prg4^+/+ mice. Color figure can be viewed in the online issue, which is available at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)1529-0131.

Figure 4

Photomicrographs of mouse knee joints that had the median total histologic score for each Prg4 genotype and experimental group. A–C, Cyclically loaded Prg4^+/+, Prg4^+/−, and Prg4^−/− mouse joints with scores of 1.0, 2.5, and 5.0, respectively. D–F, Control (unloaded) Prg4^+/+, Prg4^+/−, and Prg4^−/− mouse joints with scores of 1.0, 2.0, and 5.0, respectively. Although the coefficient of friction increased significantly in all cyclically loaded joints, this was not accompanied by microscopic evidence of damage that could be attributed solely to loading. Bars = 100 μm; original magnification × 10.

Figure 5

Stacked fluorescence confocal microscopy images of sagittal sections (27.8 ± 12.1 μm) of fluorescein diacetate–stained mouse femoral condyle cartilage samples. Green fluorescence indicates the presence of living cells. A–C, Cyclically loaded Prg4^+/+, Prg4^+/−, and Prg4^−/− mouse cartilage. D–F, Control (unloaded) Prg4^+/+, Prg4^+/−, and Prg4^−/− mouse cartilage. Staining and imaging of each sample was performed at the conclusion of the experiment (38 hours). Note that live cells are present in all cartilage specimens and that the density of live cells did not decrease as a consequence of loading. Bars = 50 μm; original magnification × 20.