Home

Current Issue

Archives

Guidelines for Authors

Classified Ads

Links

Search PubMed

Subscriptions

Subscriber Registration

Guidelines for Website Users

E-mail Alert Service

Contact Info

Correspondence

Identity of the Joint Lubricant

To the Editor:

In a recent study, Jay and Cha¹ employed selective enzymatic destruction of joint lubricant to address the important issue of the identity of the vital "active ingredient" in synovial fluid (SF), which imparts effective boundary lubrication to the articular surface. Is it surface active phospholipid (SAPL)^2,3 or their preferred choice⁴ of lubricin, which is a macromolecular water soluble glycoprotein? Our study⁵ employing trypsin and phospholipase A₂ as the digestive enzymes favored SAPL as the boundary lubricant in bovine SF (BSF) based upon articular cartilage sliding upon glass, whereas Jay and Cha reach the opposite conclusion employing trypsin and phospholipase C (PLC) for glass sliding on rubber (i.e., hydrocarbon). This hydrophilic-hydrophobic combination would appear a strange selection of surfaces if the findings are to have any relevance to the situation in vivo.

Jay and Cha¹ conclude in favor of lubricin for two reasons: First, trypsin increases friction, but they then go on to castigate those favoring SAPL, namely ourselves^5,6, as "failing to address repeated reports of the removal of lubricating ability from SF by digestion with trypsin." This remark totally misrepresents the true issue. No one questions a major role for lubricin, but is lubricin the lubricant per se, or is lubricin the macromolecular water-soluble carrier for the otherwise highly insoluble SAPL, which is the true lubricant as we have advocated^6,7? Our analysis⁶ has shown how SAPL makes up 12% of lubricin, which would render this macromolecule an ideal carrier for SAPL, while phospholipids also bind to hyaluronic acid⁸, which has similar protein chains.

Thus, their trypsin results offer no means of differentiating between boundary lubricants, because by their theory trypsin destroys the lubricant per se, while by ours it destroys the carrier for the lubricant with a similar detrimental effect upon lubrication.

Second, Jay and Cha claim that lubricin per se is the boundary lubricant because PLC produces Dm values that do not differ significantly from straight BSF. However, these Dm values are derived by subtracting from the direct measurement of friction a mean value of saline controls that differ by 27% from each other, and thereby cancel out the difference in values for the primary measurement, i.e., BSF versus BSF + PLC, which they list in the first column of Table 1. However, the "controls" would appear to be a thin layer of saline, which, when sandwiched between two nonbiological surfaces such as glass and rubber, should surely give almost the same mean values. It also complicates the direct comparison of two boundary lubricants when they state "that only a thin layer of boundary fluid was present" at the interface, implicating hydrodynamic lubrication.

Table 1. Friction coefficients of bovine synovial fluid (BSF) and phosphatidylcholine (PC) preparations digested with phospholipase C. Data reproduced from Jay and Cha1

If we avoid these questionable "controls" and compare BSF with BSF digested with PLC, i.e., using one as a control for the other, then their results show that PLC increases friction 2.8-fold, i.e., from 0.028 to 0.095 (first column, Table 1). Surely this can only occur if it destroys phospholipid, demonstrating how SAPL is the boundary lubricant per se and not lubricin. However, the authors point out that their PLC was "contaminated with proteases" but, when adding a protease inhibitor, PLC still increases friction about 2-fold (m = 0.028 to 0.050). This comparison may not reach statistical significance, but surely it severely undermines any conclusion by Jay and Cha that lubricin per se is the boundary lubricant. Their results could even indicate the reverse, as did our study⁵ using higher numbers of runs and sliding surfaces, which were far more relevant physiologically, even if our friction apparatus was not as sophisticated.

We still contend that lubricin, and maybe other proteinaceous macromolecules in SF, plays an important role in the joint as the carrier for the otherwise highly insoluble SAPL, but is not the lubricant per se⁵.

This view is consistent with the fact that almost all commercial boundary lubricants deposited from adjacent fluids are surfactants, whereas the two major components of lubricin, namely proteins and carbohydrates, tend to be glues. Another factor emphasized elsewhere^2,7 is that boundary lubrication is imparted by the outermost layer of a surface, and so how could binding of such a hydrophilic, water-soluble substance as lubricin ever render the articular surface so hydrophobic, as reported by ourselves² and others quoted by Jay and Cha? The load-bearing boundary lubricant is surely surface-active phospholipid — the same lubricant found on other sliding surfaces in vivo⁷.

If this alternative conclusion is correct, it is fortunate clinically because exogenous phosphatidylcholine can be injected directly into the joint to replenish the deficiency of SAPL reported in osteoarthritis⁹, while preliminary human trials¹⁰ have proven most encouraging.

BRIAN A. HILLS, ScD (Cantab), Paediatric Research Centre, Mater Children's Hospital, South Brisbane, Queensland, Australia.

REFERENCES

1. Jay GD, Cha CJ. The effect of phospholipase digestion upon the boundary lubricating ability of synovial fluid. J.Rheumatol 1999; 26:2454-7.

2. Hills BA. Oligolamellar lubrication of joints by surface-active phospholipid. J Rheumatol 1989;16:82-91.

3. Hills BA.Oligolamellar nature of the articular surface. J Rheumatol 1990;17:349-56.

4. Jay GD, Cha CJ, Haberstroh K, Shaw R. Comparison of the boundary lubricating ability of bovine synovial fluid, lubricin and healon. J Biomed Mater Res 1998;40:414-8.

5. Hills BA, Monds MK. Enzymatic identification of the load-bearing boundary lubricant in the joint. Br J Rheumatol 1998;37:137-42.

6. Schwarz IM, Hills BA. Surface-active phospholipid as the lubricating component of lubricin. Br J Rheumatol 1998; 37:21-6.

7. Hills BA. Boundary lubrication in vivo. Proc Inst Mech Eng 2000; 214H:83-94.

8. Pasquali-Ronchetti L, Quaglino D, Mori D, Bacchelli B, Ghosh P. Hyaluronan-phospholipid interactions. J Struct Biol 1997;120:1-10.

9. Hills BA, Monds MK. Deficiency of lubricating surfactant lining the articular surfaces of replaced hips and knees. Br J Rheumatol 1998; 37:143-7.

10. Vecchio P, Thomas K, Hills BA. Surfactant treatment for osteoarthritis [letter]. Rheumatology 1999;38:1020-1.

Dr. Jay replies

To the Editor:

I respectfully submit that the evidence for phospholipid being the actual lubricating moiety of synovial fluid (SF) has been overdrawn. However, both Dr. Hills and I agree that lubrication occurs in the boundary mode, enabling deformable apposed surfaces to slide past one another at a very slow reciprocating speed¹. The consideration that lubricin is actually a carrier² for the lubricant deserves serious inquiry — necessitating our reconciling seemingly disparate observations^3,4. The conclusion that lubricin carries surface active phospholipid (SAPL) to articular cartilage, a function not unlike that of alveolar surfactant binding proteins (which have been comparatively better characterized) is based on the following evidence: (1) lipid is posited to occupy the 9.2-13% (w/w) "undetermined" proportion of bovine lubricin amino acid and glycosylation analyses², the purification of which closely followed the procedures of Swann⁵; (2) lipid is a boundary lubricant of both natural and synthetic surfaces in vitro, which is not in dispute; and (3) digestion of whole SF with phospholipase A₂(PLA₂) removes lubricating ability³. From these and the presence of lipid in SF, Dr. Hills has maintained that lipid is the boundary lubricant, transported to articular cartilage by lubricin. This theory was kindled² by the observation that 14% of radiolabelled purified bovine lubricin bound to articular cartilage⁶. The possibility that 14% of the radiolabel was delivered to articular cartilage by way of transported lipid is not valid, as the I¹²⁵ radiolabel in these earlier experiments was specifically linked to tyrosine residues.

Jay and Cha⁴ showed very clearly that the PLA₂ preparation used by Hills and Monds³ is contaminated with proteases. Not only was lubrication decreased or eliminated, but also digestion of Na-benzoyl-L-arginine ethyl ester occurred⁴, which is used in calibrating trypsin/protease solutions. Obviously this experimental approach cannot be used to support the notion that lubricin is a lipid carrying molecule. The presence of protease inhibitors (PI) leupeptin and aprotinin partially prevented the loss of lubrication when bovine SF (BSF) was digested with phospholipase C. If SAPL was the sole lubricant then m values on the order of 0.090 and greater would have been observed. This did not occur; the addition of PI prevented m from rising past 0.050 from 0.028, the m value for normal BSF. It is likely that our attempts in antiproteolysis, though diagnostic, were incomplete and typical of the need for multiple PI in state-of-the-art preparative biochemical efforts. Second, Schwarz and Hills² give no indication as to the lubricin purity. This would be assessed by means other than single chromatographic peaks arising from a replicated purification⁷. Third, no one has controlled for the quantity of lipid iatrogenically introduced into SF as a result of percutaneous aspiration.

The friction apparatus and bearing system of latex apposed to polished glass used in our study has been used by a number of other investigators^8-10 to study the lubricating ability of SF. In our study⁴, each data point of m for a sample has its own comparative normal saline control (NS), minimizing sample to sample variation of these rubbing surfaces. Data are not grouped and then subtracted en masse from the m for NS as suggested. The artificial test surfaces were chosen since m values are more reproducible than experimental cartilage containing bearings¹⁰. One cannot control for the weave of severed collagen fibrils and its resultant effect on surface features.

Trypsin and phospholipase digestion aside, lubricating ability can also be eliminated by galactosidase and neuraminidase digestion¹¹, removing the penultimate galactose from the a2,3NeuAc-b(1-3)Gal-GalNAc moiety on lubricin. It has been theorized that lubricin interacts with hydrophobic surfaces — articular cartilage and latex¹¹. Lubrication may be provided by apposed and pressurized hydrophilic moieties, and it is these mucinous glycoproteins that have amphipathic behavior. Synovial SAPL, if present naturally, may play some role in joint lubrication my rendering articular cartilage hydrophobic, through some as yet undiscovered means. This would enable hydrophobic-hydrophobic attraction (in an aqueous environment) of the lubricant onto its surface.

Resolution of whether lubricin is the lubricant or its carrier will depend on analyses for elemental phosphorus in purified lubricin, whose purity is assessed by liquid chromatography mass spectrometry.

GREGORY D. JAY MD, PhD, Department of Medicine, Brown University School of Medicine, Providence, RI 02903, USA.

REFERENCES

1. McCutchen CW. The frictional properties of animal joints. Wear 1962;5:412-5.

2. Schwarz IM, Hills BA. Surface-active phospholipid as the lubricating component of lubricin. Br J Rheumatol 1998;37:21-6.

3. Hills BA, Monds MK. Enzymatic identification of the load-bearing boundary lubricant in the joint. Br J Rheumatol 1998;37:137-42.

4. Jay GD, Cha CJ. The effect of phospholipase digestion upon the boundary lubricating ability of synovial fluid. J Rheumatol 1999; 26:2454-7.

5. Swann DA, Sotman S, Dixon M, Brooks C. The isolation and partial characterization of the major glycoprotein (LGP-1) from the articular lubricating fraction from bovine synovial fluid. Biochem J 1977;161:473-85.

6. Swann DA, Hendren RB, Radin EL, Sotman SL, Duda EA. The lubricating activity of synovial fluid glycoproteins. Arthritis Rheum 1981;24:22-30.

7. Jay GD, Britt D, Cha CJ. Lubricin is a product of megakaryocyte stimulating factor gene expression by human synovial fibroblasts. J Rheumatol 2000; 27:594-600.

8. Wilkins JF. Proteolytic destruction of synovial boundary lubrication. Nature 1968;219:1050-1.

9. Reimann I, Stougaard J, Northeved A, Johnsen SJ. Demonstration of boundary lubrication by synovial fluid. Acta Orthop Scand 1975; 46:1-10.

10. Davis WH, Lee SL, Sokoloff L. A proposed model of boundary lubrication by synovial fluid: structuring of boundary water. J Biomech Eng 1979;185-92.

11. Jay GD. Characterization of a bovine synovial fluid lubricating factor. I. Chemical, surface activity and lubricating properties. Connect Tissue Res 1992;28:71-88.