E. J . AMBROSE AND J . A . V. BUTLER 27 1 HYALURONIC ACID, CHONDROITIN SULPHATES AND THEIR PROTEIN COMPLEXES * BY KARL MEYER Department of Medicine, Columbia University College of Physicians and Presbyterian Hospital, New York Received 26th May, 1952 Surgeons, and the Edward Daniels Faulkner Arthritis Clinic of the In this paper the evidence for the existence of protein complexes of hyaluronic acid and also the chondroitin sulphates of hyaline cartilage and connective tissue is reviewed Most of the experimental data for hyaluronic acid are best explained by assumption of reversibly formed polar complexes. (The existence of secondary non-polar forces is not excluded.) The irreversible loss of viscosity observed during isolation, for example, may be a result of depolymerization of the carbohydrate chain by mechanisms not now understood, rather than the irreversible cleavage of a native protein complex.The data on chondroitin sulphate A in cartilage are also best explained by the assumption of polar linkages with protein. The chondroitin sulphates of connective tissue, however, are stably linked to protein in the native state. It has been assumed for many years that the mucopolysaccharides occur in nature as protein complexes which, although poorly defined chemically or physically, have been classified according to their origin in the older literature 1 *Supported in part by a grant from the Helen Hay Whitney Foundation, and the U.S. Public Health Service.272 HYALURONIC ACID COMPLEXES as vitreous muciii, syiiovial much, chondromucoid, osseomucoid, etc.The title of this section obviously presupposes the existence of mucoproteins. The author, some years ago, defined mucoproteins or mucoids as proteins containing mucopolysaccharides, i.e. hexosamine containing polysaccharides, in firm chemical union, that is, linked by a covalent bond to a peptide.2 Conjugated proteins of this type are ovomucoid, the various mucoproteins of serum, the acid muco- protein of submaxillary gland and the neutral mucoids of gastric mucosa. The separation of mucopolysaccharide and peptide has been achieved in some of these by action of hot alkali or by prolonged enzymatic hydrolysis. Whether the bonds (of unknown nature) between mucopolysaccharide and peptide have been broken after such treatment has not been established. From consideration of the various methods of isolation of hyaluronic acid and of the chondroitin sulphates, it is apparent that the stability of their protein complexes is of a different order of magnitude than that of the mucoproteins proper.It is the object of this paper to discuss the question of the nature of the inter- action of protein and hyaluronic acid and the chondroitin sulphates. It will be clear from the following that at the present time few definite answers can be given to the questions raised. HYALURON~C Acn-Hyaluronic acid occurs either in dissolved form as in vitreous humor, synovial fluid and some tumour fluids, or in form of gels as in umbilical cord, in certain mesodermal tumours and in dermis.3 Methods for the preparation of the acid, free of protein and of other accompanying substances (within reasonable limits) have been described.3 Hyaluronate prepared from various mammalian sources shows no significant differences in elementary analysis, in optical rotation, in the proportion of its component monosaccharides (glucos- amine and glucuronic acid), in the degree of hydrolysis by various enzymes or thus far in the nature and quantities of products which can be obtained on frac- tionation of enzymatic hydrolysis products.4 Physical properties, however, especially viscosity, differ among preparations, depending on the source and the method of purification of the polysaccharide.The viscosities are highest with samples prepared from umbilical cord and from group A haemolytic streptococcus, followed in order by tumour fluid, synovial fluid and vitreous humor.The latter appears to be also most polydisperse, although no detailed information of the molecular weight distribution of the acid from different sources has been published. Isolated hyaluronate has a lower viscosity in equivalent concentrations than that contained in the native fluid. Such fluids on dilution and acidification pre- cipitate as long fibres, the so-called mucin clots. Hyaluronate isolated by most procedures, especially from synovial and vitreous fluids, on addition of serum and acidification, precipitates in flocculent form. However, highly viscous hyaluronaie under the same conditions yields mucin clots.5 After incubation of native hyaluronate with 0.01 to 0.005 units of purified testicular hyaluronidase,3 the fluid no longer gives a mucin clot and precipitates in flocculent form.Isolated hyaluronate under these conditions shows no loss in turbidity formation and a negligible loss in viscosity. The " mucin " formed on acidification of synovial fluid has been shown to be a polar compound formed by the free amino groups of protein with the COOH groups of polysaccharide acid. In salt solutions or on alkalinization, these salts dissociate.6 From electrophoretic studies no evidence other than of dissociating complexes between hyaluronate and protein has been obtained. The mobility of isolated hyaluronate is identical with that of the fastest component in native fluids.7~ 8 Ogston and Stanier 9 9 10 on the basis of viscometry and sedimentation have concluded that synovial fluid contains a definite protein complex of hyaluronate which on isolation apparently breaks down irreversibly.The relatively highKARL MEYER 273 viscosity of the native fluids may perhaps equally well be explained by specific interaction of the native hyaluronate molecules, similar to that postulated for certain desoxypentose ribonucleic acids, or depolymerization by hydrolytic or oxidative processes of the carbohydrate chain during isolation, or, as on addition of enzyme, by cleavage of a few centrally located hexosaminidic bonds. CHONDROITIN suLPHAms.-We distinguish three different chondroitin sulphates, designated as A, B and C.11 The chemical structure of these compounds is at present unknown, although various structures have been proposed in recent years for chondroitin sulphate of hyaline cartilage.In contrast to hyaluronate which occurs both in fluids and in structural elements, the chondroitin sulphates do not occur in fluids." The physical properties of native chondroitin sulphates, especially their molecular size and the nature of the complexes in which they may occur in the matrix are very difficult to ascertain. In the older literature, chondroitin sulphate had been thought to occur as a protein complex, presumably as an ester which had to be hydrolyzed by alkali during isolation.1 Morner had, however, shown many years ago that a small fraction of chondroitin sulphate could be extracted by water.12 Following the study in our laboratory of artificial protein salts of hyalurinic acid, similar studies were carried out with chondroitin sulphate.13 On the basis of this work, a method for the preparation in excellent yield of chondroitin sulphate from finely ground defatted cartilage powder, based on extraction with 10 % CaC12, was presented.14 This procedure has been adopted by other investigators.15 The viscosity of isolated chondroitin sulphate varies greatly with the procedure adopted.A high pH or high temperature leads to loss of viscosity. The molecular weight of chondroitin sulphate (A?) has been reported 1s as 260,000 (birefringence of flow of carefully prepared chondroitin sulphate). Molecular weights of 15,000 to 43,000 obtained by an osmometric method have recently been reported.16 The lower figure corresponds to alkali-extracted, the highest to CaC12-extracted material.It appears quite clear that the method of treatment greatly influences the particle size of the isolated mucopolysaccharide. Isolated chondroitin sulphate was shown to react at low pH stoichiometrically as a dibasic acid with proteins.13 The quantity of chondroitin sulphate bound by the protein corresponded to the proportion of basic amino acids contained in the protein, and was also equivalent to the quantity of acid dyes bound by the protein. The ratio of hexosamine to total N of the artificial mucoids prepared from chondroitin sulphate and gelatin closely approximated the corresponding ratio for cartilage powder. It was suggested, for this and other reasons, that chondroitin sulphate in cartilage exists largely as a salt of the mucopolysaccharide with collagen.Electrophoretic studies of extracts of cartilage have been reported by Blix8 and by Partridge.17 Blix found no evidence for existence of protein complexes in aqueous extracts In Partridge's experiments with extract prepared at elevated temperatures, he observed the expected two components in the descending limb of the electrophoresis cell, corresponding to chondroitin sulphate and protein derived from degraded collagen. In the ascending limb he found, in addition, a new distinct zone, which he attributed to a dissociating complex between protein and carbohydrate. Costal cartilage has recently been shown18 to bind Na+. Ca2+ or Ba2+ in quantities equivalent to the chondroitin sulphate present its measured by the sulphate content.It was concluded that the chondroitin sulphate above pH 3.5 in this reaction behaved as a divalent acid. One may interpret this experiment to indicate that the acidic groups of chondroitin sulphate as present in cartilage are capable of dissociation. * Recently from the highly viscous contents of a cyst obtained on operation of a chondrosarconia there was isolated a sulphated polysaccharide ; from analysis and from its behaviour toward enzymes this appeared to be chondroitin sulphate A unaccompanied by other fractions.274 HYALURONIC ACID COMPLEXES Einbinder and Schubert19 have studied the binding by rat tail and bovine Achilles tendon of chondroitin sulphate (and hyaluronate) as compared to the binding of acid dyes. Chondroitin sulphate in these experiments had an equivalent weight of one (rather than one-half) repeating unit, and hyaluronate was bound only at a pH below 3.The authors called attention to the fact that collagen reacted with the anions in a manner very different from that of ~001.20 The explanation for the differences seems to lie in the fact that collagen at acid pH swells and is transformed irreversibly into degraded collagen. Further, the solubilities (and not impossibly the chemical natures) of the collagens of tendon differ markedly from those of cartilage. A comparison, therefore, of the behaviour of the two may not be justified. The difference in solubility may be caused by the large concentration of mucopolysaccharide in cartilage (ca. 20 to 40 % of the dry weight) while the amount of mucopolysaccharide in tendon is small (ca.1 %). It can be concluded that the nature of the complex of chondroitin sulphate and protein in hyaline cartilage remains to be elucidated. However, it may be said that none of the observations so far reported are in contradiction to the picture of a salt-like combination of polysaccharide and protein. The presence of primary covalent links between the two appears, at any rate, to be excluded. In cartilage, chondroitin sulphate B is absent. This substance has been isolated from skin,21 and later from tendon, heart valves and aorta.11 In skin, chondroitin sulphate is found together with hyaluronic acid, and in the other tissues the second mucopolysaccharide is chondroitin sulphate C. Chondroitin sulphates B and C are firmly bound to proteins which are distinct from collagen.It has been con- cluded from histological work that these protein + chondroitin sulphate complexes are contained in the interfibrillar ground substances. Extracts of bovine or porcine Achilles tendon with 0.02 N Ca(OH)2 (half-saturated lime water) on repeated precipitation yield mucoids from which the mucopolysaccharide cannot be separ- ated by either electrophoresis or treatment with 10 % CaCl2, in contrast with the behaviour of hyaline cartilage. A mucoid fraction prepared from heart valves at pH 8.5 in barbiturate buffer ( I = 0.1) had a mobility of 8-1 x 10-5, while the mobility of the chondroitin sulphates at the same condition is 12.5 x 10-5. The mucopolysaccharides are split off on treatment at 4" C with 0.33 N NaOH or on digestion with pepsin at pH 2 followed by tryptic digestion.The nature of the protein complexes, as well as the chemistry of the mucopolysaccharides has been studied very incompletely . On the biological level, the interaction of mucopolysaccharides with proteins may be of importance among other things in fibrillogenesis and in the organization of other structural elements, especially in growth and differentiation. The role of mucoids in the reconstitution of collagen has recently been described.22 It was shown in this work that soluble collagen gave, after dialysis, either long spac- ing (2000 A) or regular (640 A) axial striations, depending on the concentrations of serum mucoprotein fractions added (these mucoproteins contain neither sulphate nor uronic acid and are true neutral mucoids).In general, hyaluronate or the chondroitin sul phates on addition to dissolved collagen gave immediate precipit- ates devoid of cross striations. However, one of the samples of chondroitin sulphate A used gave the 2000A axial orientation.22 In contrast to the fibrils produced by mucoprotein, this cross striation appeared immediately, without prior dialysis. The difference between this sample and others appears to be only the higher molecular weight of the former. Anothcr exampIe of mucopolysaccharide 1- protein interaction is that de- scribed by Cohen.23 According to this author, macromolecules of very large size such as haemocyanine, liver particles and some viruses are precipitated by heparin, chondroitin sulphate or hyaluroiiate at neutral reaction.Some of the viruses in the presence of the mucopolysaccharides were transferred into para- crystals differing from their ordinary form. Hyaluronate of a high degree of polymerization was active in much lower concentration than either heparin orKARL MEYER 275 chondroitin sulphate (cp. ref. (3), p. 354). Enzymatically depolymerized hyal- uronate was inactive in this reaction. At present no satisfactory explanation can be given for these phenomena or for the orientation effects observed with soluble collagen. sUMMARY.--The evidence for the existence of reversible polar complexes of hyaluronic acid in native fluids is reviewed. Data on cartilage are reviewed and explained by the assumption of polar linkages with protein. In contrast, the chondroitin sulphates of connective tissue are stably Iinked to protein. 1 Levene, Hexosarnines and Mucoproteins (Longmans, Green & Co., London and 2 Meyer, Advances in Protein Chem., 1945, 2, 249. 3 Meyer, Physiol. Rev., 1947, 27, 335. 4 Rapport, Meyer and Linker, J. Amer. Chem. Soc., 1951, 73, 2416. 5 Meyer, J. Biol. Chem., 1948, 176, 993. 6 Meyer, Cold Spring Harbor Symposia, 1938, 6, 91. 7 Meyer and Chaffee, J. Biol. Chem., 1940, 133, 83. 8 Blix, Acta Physiol. Scand., 1940, 1, 29. 9 Ogston and Stanier, Biochem. J., 1950, 46, 364. 10 Ogston and Stanier, Biochern. J., 1951, 49, 585. 11 Meyer and Rapport, Science, 1951, 113, 596. 12 Marner, Skand. Arch. Physiol., 1889, 1, 210. 13 Meyer, Palmer and Smyth, J. Biol. Chem., 1937, 119, 501. 14 Meyer and Smyth, J. Biol. Chern., 1937, 119, 507. 15 Blix and Snellman, Ark. Kemi, Min. Geol. A, 1945,19. 16 Mathews, Fed. Proc., 1952, 11, 255. 17 Partridge, Biochern. J., 1948, 43, 387. 18 Boyd and Neuman, J. Biol. Chem., 1951,193,243. 19 Einbinder and Schubert, J. Biol. Chem., 1951, 188, 335. 20 Steinhardt, Fugitt and Harris, J . Res. Nat. Bur. Stand., 1942, 28, 201. 21 Meyer and Chaffee, J. Biol. Chem., 1941, 138,491. 22 Highberger, Gross and Schmitt, Proc. Nat. Acad. Sci., 1951, 37,286. 23 Cohen, J. Biol. Chem., 1942, 144, 353. York, 1925).