Association between friction and wear in diarthrodial joints lacking lubricin

Abstract

Objective: The glycoprotein lubricin (encoded by the gene Prg4) is secreted by surface chondrocytes and synovial cells, and has been shown to reduce friction in vitro. In contrast to man-made bearings, mammalian diarthrodial joints must endogenously produce friction-reducing agents. This study was undertaken to investigate whether friction is associated with wear.

Methods: The lubricating ability of synovial fluid (SF) samples from humans with genetic lubricin deficiency was tested in vitro. The coefficient of friction in the knee joints of normal and lubricin-null mice was measured ex vivo; these joints were also studied by light and electron microscopy. Atomic force microscopy was used to image and measure how lubricin reduces friction in vitro.

Results: SF lacking lubricin failed to reduce friction in the boundary mode. Joints of lubricin-null mice showed early wear and higher friction than joints from their wild-type counterparts. Lubricin self-organized and reduced the work of adhesion between apposing asperities.

Conclusion: These data show that friction is coupled with wear at the cartilage surface in vivo. They imply that acquired lubricin degradation occurring in inflammatory joint diseases predisposes the cartilage to damage. Lastly, they suggest that lubricin, or similar biomolecules, will have applications in man-made devices in which reducing friction is essential.

Figure 1

Schema of the pendulum apparatus used to measure whole joint coefficient of friction (μ), and representative data. A, Technique of whole-joint measurement using mouse knee joints in which muscles were removed and the joint capsule left intact. The tibia is fixed at a 45° angle so that the knee joint angle is 135° when the femur is held perpendicular. The joint is loaded with 20 gm weight, simulating the weight of an adult mouse standing on 1 limb, with the fulcrum of the pendulum located at the knee joint. The femur is manually deflected from the perpendicular by 30° and allowed to oscillate freely at the knee joint until joint motion stops. B, Representative raw data obtained using a modified Moiré encoder technique, measuring the oscillation of knee joints of Prg4^+/+ and Prg4^−/− mice. Deceleration of the pendulum was calculated from the pendulum decay and divided by G, the earth's gravitational constant, to determine the coefficient of friction. Note that the knee of the Prg4^+/+ mouse oscillated longer and for more cycles than that of the Prg4^−/− mouse. C, Box plots of μ values in 1-month-old and 2-month-old Prg4^−/− and Prg4^+/− mice, measured with the pendulum apparatus. Each box represents the 25th to 75th percentiles. Lines inside the boxes represent the median. For 1-month-old Prg4^−/− mice, lines outside the box represent the 10th and 90th percentiles, and circles indicate outliers.

Figure 2

Microscopy of cartilage from heterozygous (Prg4^+/−) and lubricin-null (Prg4^−/−) mice. A and B, Representative photomicrographs of toluidine blue–stained knee joint cartilage sections from 2-week-old Prg4^+/− (A) and Prg4^−/− (B) mice, demonstrating differences in surface smoothness. C and D, Representative electron micrographs of the cartilage surface in sections from 2-week-old Prg4^+/− mice, revealing a smooth, uniform surface and a clearly visualized lamina splendens. Note the parallel orientation of collagen fibrils adjacent to the cartilage surface. E and F, Representative electron micrographs of the cartilage surface in sections from 2-week-old Prg4^−/− mice, revealing roughening of the surface, lack of a recognizable lamina splendens, and disruption of the parallel orientation of the collagen fibrils. G and H, Toluidine blue–stained section (G) and electron micrograph (H) of the cartilage surface in a newborn Prg4^−/− mouse. I, Electron micrograph of the cartilage surface in a newborn Prg4^+/− mouse. Note that there is no difference in collagen fibril orientation between newborn Prg4^+/− and Prg4^−/− mice, but only the Prg4^+/− mouse has a detectable lamina splendens (arrows).

Figure 3

A–F, Atomic force microscopy of lubricin coated onto highly ordered pyrolytic graphite (HOPG). Topographic images (A, C, and E) and adhesion images (B, D, and F) of lubricin networks formed in experiments using lubricin at concentrations of 10 μg/ml (A and B), 50 μg/ml (C and D), or 300 μg/ml (E and F) are shown. In the topographic images, the networks of lubricin molecules are pseudo-colored to appear bright. In the adhesion images, regions of low adhesion are pseudocolored to appear dark. Note that at all concentrations, the lubricin network in the topographic images (bright areas in A, C, and E) corresponds to reduced adhesion (dark areas in B, D, and F). At 10 μg/ml, lubricin formed a loose, mesh-like network structure. At higher concentrations of lubricin (50 μg/ml and 300 μg/ml), the network remained mesh-like, although the mesh size decreased as the concentration increased. Arrows indicate an example of openings in the mesh-like network at each concentration. Boxed areas indicate the identical regions in a topographic and corresponding adhesive image, demonstrating the relationship between the presence of lubricin on the HOPG surface and the decrease in adhesion between the surface and the atomic force microscopy (AFM) probe. G, AFM force-versus-distance curves obtained in experiments using clean HOPG substrate and substrates coated with lubricin at various concentrations. Note the sigmoidal shape of the curve, which is compatible with lubricin undergoing a phase change when the concentration is ∼250 μg/ml. Values are the mean ± SD.

Figure 4

In vitro lubricating abilities of synovial fluid samples. With healthy human synovial fluid (HSF), friction (μ) was reduced to <0.02 and continued to decrease over time, likely due to more uniform covering of the in vitro bearing by the SF. In contrast, SF specimens from 6 individual patients with camptodactyly-arthropathy–coxa vara–pericarditis syndrome (CA1–6) did not reduce friction any more than was observed with normal saline (NS).