A Specific, Sensitive Method for the Determination of Hyaluronatel

Home , Sodium hyaluronate, Sodium periodate

ANALYTICAL BIOCHEMISTRY 96, 474-480 (1979)

GEORGE W. JOURDIAN, MELINDA WOLFMAN, ROBERT SARBER,* AND JACK DISTLER

Rackham Arthritis Research Unit, and the Departments of Biological Chemistry and Internal Medicine. The University of Michigan Medical School, Ann Arbor, Michigan 48109 Received November 29, 1978

A sensitive and specific assay for hyaluronate was devised. Hyaluronate contained in biological mixtures was digested with a commercially available microbial hyaluronate lyase. The P-A4,5-eneglucopyranuronic acid residues contained at the nonreducing termini of the resulting oligosaccharides were oxidized with periodic acid to yield, among other products, formyl pyruvic acid. The latter compound reacted with thiobarbituric acid to yield a chromophore with an absorption maximum at 549 nm. Optimal conditions for quantitative assay of hyaluronate are described.

At present there is no simple method for glycans by density gradient centrifugation, the rapid and specific detection and quanti- column chromatography on agarose, and tation of microamounts of hyaluronate. determination of the proteoglycan-hyaluro- Classically, the concentrations of hyaluronate complex elution profile by measure- nate in biological solutions have been deter- ment of uranic acid concentration (7). mined by turbidimetric methods (I), or by An alternative approach, utilized by a quantitative determination of its constituents number of investigators for the determina- after acid hydrolysis (2). However, these tion of hyaluronate, has been the application procedures are not specific for the deter- of microbial hyaluronate lyases. In contrast mination of hyaluronate. Other methods to animal hyaluronidases (hyaluronate depend upon isolation and chemical char- glycanohydrolase EC 3.2.1.35), the micro- acterization and involve ion-exchange bial hyaluronate lyases (hyaluronate lyase chromatography (3), cetylpyridinium chlo- EC 4.2.2.1) are eliminases and hydrolyze ride-cellulose microcolumns (49, and hyaluronate to oligosaccharides containing cellulose acetate electrophoresis (6). While terminal A4,Sunsaturated glucopyranuronic generally satisfactory, these methods are acid3 units (8); the unsaturated compounds time-consuming and are compromised if have a characteristic absorption maximum a large excess of other glycosaminoglycans at 232 nm. This property served as the basis are present in the sample. A highly sensitive for the development of direct spectrophoto- and specific proteoglycan-binding assay for metric means for the measurement of hy- hyaluronate contained in mixtures of gly- aluronate. However, biological fluids and cosaminoglycans has been reported but tissue extracts usually contain high levels requires prior isolation of cartilage proteo- of ultraviolet-absorbing compounds such as protein and nucleic acids which interfere ’ This work was supported by Grant HL 19685 with the quantitative measurement of the from the National Institute of Heart and Lung Diseases, National Institutes of Health, and in part by the Arthri- A4,5-unsaturated uranic acid residues by tis Foundation, Michigan Chapter. * Postdoctoral trainee, Training Grant AM 07080 a A4,SUnsaturated glucopyranuronic acid oligo- from the National Institute of Arthritis, Metabolic saccharides = (3-O-P-A4,5-ene-D-glucopyranuronic and Digestive Diseases, National Institutes of Health. acid-2-acetamido-2-deoxy-D-glucose),,.

OO03-2697/79/10O474-07$0200/O 474 Copyright 8 1979 by Academic Press. Inc. All rights of reproduction in any form reserved. SPECIFIC SENSITIVE METHOD FOR HYALURONATE DETERMINATION 475 this methodology (9). The latter problem standards and were generous gifts of Dr. has in part been circumvented by colori- M. B. Mathews, University of Chicago: metric measurement of liberated N-acetyl- hyaluronate (human umbilical cord), chon- glucosamine end groups by a modification droitin 4-sulfate (sturgeon cranial cartilage), of the Morgan-Elson reaction (10). How- chondroitin 6-sulfate (sturgeon notochord), ever, crude tissue extracts produce spurious dermatan sulfate (hog mucosa), keratan color and the calorimetric determination sulfate (bovine cornea), and heparin and is affected by the presence of even moderate heparan sulfate (beef lung). Hyaluronate protein concentrations (2). The enzymatic was also prepared from human synovial determination of hyaluronate in biological fluid by electrodeposition (15). Chondroitin fluids is further limited by the observation was prepared from chondroitin 4-sulfate that heparin, chondroitin 4-sulfate, and by solvolytic desulfation with dimethyl dermatan sulfate competitively inhibit some sulfoxide (16). hyaluronate lyases (2). Ohya and Kaneko Hyaluronate lyase. Hyaluronate lyase (11) have isolated a novel hyaluronate lyase purified from Streptomyces hyalurolyticus from Streptomyces hyalurolyticus which (specific activity 3000 turbidity reducing is reported not to be inhibited by glycos- units (TRU)/mg protein)4 was obtained aminoglycans. The enzyme is specific for from Calbiochem. hyaluronate, catalyzes the formation of Reagents. The following compounds oligosaccharides terminating in A4,5-unsat- were obtained from the indicated sources: urated glucuronic acid units, and is stable at sodium periodate and n-butyl alcohol, J. T. elevated temperatures. Baker Chemical Company; sodium arsenite Hascall et al. (12) observed that a perio- and p-dimethylaminobenzaldehyde, Baker date-thiobarbituric acid reagent reacted and Adamson; ammonium dihydrogen ortho- with A4,Sunsaturated uranic acid-terminat- phosphate, BDH Chemicals Ltd.; and thio- ing oligosaccharide chains (found on treat- barbituric acid, Eastman Chemical Com- ment of bovine nasal cartilage proteoglycan pany. All other reagents were of analytical with microbial chondroitinases) to yield a grade or of the highest purity commercially chromophore with an absorption at 549 nm. available. Dialysis tubing with a molecular This observation served as the basis for weight cutoff of 3500 was purchased from development of a convenient assay for the Arthur H. Thomas Company. degradation of bovine nasal proteoglycan by chondroitinase ABC and AC activities. METHODS Koseki et al. (13) applied a similar periodate- thiobarbituric acid assay procedure, de- Standard Assay Procedure scribed by Warren (14), to the microdeter- Incubation with hyaluronate lyase. Reac- mination of chondroitin 4- and 6-sulfate tions were performed in 15 x 125-mm Pyrex and dermatan sulfate after prior digestion tubes. Enzymatic digestions contained the with microbial chondroitinases. following components in a total volume of The above observations served as the 0.2 ml: sodium acetate buffer, pH 5.0, 4.0 basis for development of a specific and pmol; 1.66 pg hyaluronate Iyase (5 TRU); sensitive enzymatic-calorimetric assay and 0 to 100 pg of sodium hyaluronate. for hyaluronate described in this manuscript. Control tubes contained the same components but lacked either substrate or enzyme. MATERIALS The reaction mixtures were incubated at

Glycosaminoglycans. The following a Enzyme activity is expressed in turbidity units glycosaminoglycans served as reference (TRW as defined by Tolksdorf er al. (17). 416 JOURDIAN ET AL.

60°C for 5 h and enzyme action was termi- 1.6t 4 nated by the addition of 0.25 ml of 0.04 M NaIO, contained in 0.08 N H.&SO,. Measurement of eliminase products. The conditions used were similar to those described by Hascall et al. (12). After addition of the periodate-H,SO, reagent the reactants were mixed and the tubes placed in a 37°C water bath for 1 h to allow oxidation to proceed. Excess periodate was destroyed by addition of 0.5 ml of 3% (w/v) NaAsO, contained in 0.5 N HCl. The reactants were mixed, allowed to stand at room temperature for 30 min, and 4 ml of 0.3% HOURS thiobarbituric acid contained in 0.012 N FIG. 1. The formation of periodate-thiobarbituric HCl were added to each tube. The contents acid reactive compounds from hyaluronate as a func- were again mixed thoroughly and the tubes tion of time of incubation with Strepromyces hyaluro- capped with marbles, heated at 100°C for nate lyase: effect of substrate concentration. The 15 min, and cooled to room temperature. incubation conditions were those described for the standard assay procedure except the hyaluronate con- In contrast to the procedure of Hascall centrations (expressed in &incubation) were as et al. (12) we routinely extracted the chro- follows: (A) 10, (0) 20, (A) 35, (0) 50, (X) 75, and mophore with 4 ml of 5% HCl in n-butyl (cl) loo. alcohol to enhance the sensitivity of the assay procedure. Absorbance of the butanol hyaluronate lyase in each incubation. Lower layer was determined at 549 nm. amounts of hyaluronate lyase, i.e., less than 3 TRU per incubation, resulted in sig- RESULTS AND DISCUSSION nificantly lower initial rates and decreased Enzyme Incubation Conditions product formation (Fig. 2). Variation of absorbance values on a day-to-day basis Ohya and Kaneko (11) previously estab- using the same standard hyaluronate solu- lished that Streptomyces hyaluronate lyase tion was within 25%. is stable over a wide pH and temperature range; the presently described assay pro- Enzyme Specificity cedure was conducted at the established pH and temperature optima, pH 5.0 and Ohya and Kaneko (11) reported that 6O”C, respectively. Routinely, the enzyme Streptomyces hyaluronate lyase specifically was dissolved in 0.04 M sodium acetate catalyzes the hydrolysis of hyaluronate and, buffer, pH 5, at a level of 200 TRU/ml. No in contrast to other microbial hyaluronate appreciable loss in activity (less than 5%) lyases, is not inhibited by other glycosamino- was observed on storage of the diluted en- glycans. In view of the importance of these zyme at - 19°C for 1 month. considerations for the specific determina- In the standard assay procedure the sam- tion of hyaluronate in biological samples, ple is incubated with hyaluronate lyase for we reexamined the substrate specificity of a period of 5 h. As shown in Fig. 1, this time Streptomyces hyaluronate lyase and the of incubation resulted in complete product ability of other glycosaminoglycans to formation over the concentration range of inhibit Streptomyces hyaluronate lyase. hyaluronate employed. Maximal product The following glycosaminoglycans were formation was achieved with 5 TRU of studied: chondroitin 4- and 6-sulfate, chon- SPECIFIC SENSITIVE METHOD FOR HYALURONATE DETERMINATION 477

chromophore derived from sialic acid (18). Unextracted and n-butyl alcohol-extracted samples lost less than 8% of their color in 6 h and under the assay conditions described, the formation of n-butyl alcohol-extractable chromophore was linear with hyaluronate concentrations between 5 and 100 pg per reaction mixture. While the current studies were in prog- ress, a micromethod for the determination of chondroitin 4- and 6-sulfate and dermatan sulfate, after prior digestion with chon- 2 4 6 8 IO HOURS droitinase ABC and AC, was reported by Koseki et al. (13). The methodology applied FIG. 2. The formation of periodate-thiobarbituric a modification of the periodate-thiobarbi- acid-reactive compounds from hyaluronate as a func- tion of time of incubation with Sfrepromyces hyaluro- turic acid method of Warren (14) to the assay nate lyase; effect of enzyme concentration. The of A4,5-unsaturated uranic acid-containing incubation conditions were those described for the disaccharides derived from these glycos- standard assay except the hyaluronate lyase concen- aminoglycans. The authors reported their trations (expressed in TRUlincubation) were as fol- assay procedure was approximately three- lows: (A) 1, (0) 2, (A) 3, (0) 4, and (X) 5. fold more sensitive than that described by Hascall et al. (12). When the periodate- droitin, dermatan sulfate, heparin, heparan thiobarbituric acid conditions of the Jap- sulfate, and keratan sulfate. The standard anese investigators were applied to the assay conditions were employed. For spec- determination of hyafuronate (using the ificity studies 50 fig of a single glycosamino- incubation conditions described in the pres- glycan was added to a reaction mixture as a substrate and for inhibition studies each reaction mixture contained 50 pg of hyaluronate and 500 pg of another glycosamino- glycan. In agreement with the findings of Ohya and Kaneko no absorbance over the blank value was obtained with any glycos- aminoglycan other than hyaluronate. Fur- thermore, the absorbance obtained from hyaluronate was not altered when the other glycosaminoglycans were present during 0.2 the incubation. 0 IO 20 30 40 50 60 70 80 90 100 Factors That Affect Periodate- HYALURONIC ACID ( pg I Thiobarbituric Assay FIG. 3. Determination of the absorbance (color yield) generated by A4,5-unsaturated glucopyranuronic As shown in Fig. 3, the sensitivity of the acid-containing oligosaccharides in Streptomyces calorimetric procedure described by Hascall hyaluronate lyase digests of hyaluronate by the fol- et af. (12) was enhanced approximately lowing calorimetric and spectrophotometric proce- 2.5 times by extraction of the chromophore dures: (A) absorption at 232 nm (8); (0) 549 nm, Hascall et al. (12); (X) Koseki et al. (13); (0) 549 nm, into acidified n-butyl alcohol, a method determined by the methodology described in this previously described by Aminoff for en- manuscript; and (A) 585 nm, Morgan-Elson reagent hancement of the sensitivity of the same as modified by Reissig et al. (10). 478 JOURDIAN ET AL. ent manuscript), less than a 1% difference presence of hyaluronate is well documented in color yield was observed from that ob- and can be determined by classical methods. tained by the presently described method- Extracts of the soft tissue of lung are rich ology (Fig. 3). In our experience, the proce- in compounds that have the potential to dure described originally by Hascall et al. contribute to elevated blank values. These (12) coupled with extraction of the chromo- compounds include sialic acid-containing phore with acidified n-butyl alcohol is tech- glycoproteins (14,18), deoxyribose-con- nically easier, is less time-consuming, and taining nucleic acids (19), and lipids (20). requires fewer reagents than the procedure Neuraminidases present in lung extracts of Koseki ef al. (13). were found to release large amounts of free The absorbance values obtained at several sialic acid from sialic acid-containing lung hyaluronate concentrations with the modi- glycoproteins. Interference attributable fied Morgan-Elson procedure of Reissig to periodate oxidation of free sialic acid et al. (10) and direct spectrophotometric was avoided by dialyzing the extract to estimation at 232 nm are also presented in remove free sialic acid followed by heating Fig. 3. Nearly equivalent absorbance values at 100°C for 3 min to inactivate neuramini- were obtained with the Morgan-Elson and dase. As shown in Table 1, this heating step the periodate-thiobarbituric acid proce- produced tolerable blank values with the dures at equivalent hyaluronate concen- lung extracts. The results presented in Ta- trations; the absorbance values obtained ble 1 also show that hyaluronate was not by direct measurement of the unsaturated detected in lung extracts (less than 5 p.gl products at 232 nm were approximately 100 ~1). Analysis of the pooled lung extract fivefold lower than those obtained by the for uranic acid (after papain digestion by periodate-thiobarbituric acid or Morgan- the procedure of Schiller et al. (3)) showed Elson procedures. it contained 0.24 pmol uranic acid/100 ~1 extract. However, results described in Table 1 show that hyaluronate was not detected Assay of Hyaluronate in Biological Fluids by the coupled enzyme-calorimetric pro- and Extracts cedure. These results suggest that in our Accurate measurement in crude biologi- lung preparation, hyaluronate contributes to cal preparations of products arising from less than 6% of total uranic acid. This is the enzymatic degradation of hyaluronate surprising in view of the reported preva- by direct spectrophotometric measurement lence of hyaluronate in lung tissue (22) or by the Morgan-Elson procedure is re- and at present, no explanation can be of- stricted for the reasons cited in the introduc- fered for this apparent contradiction. When tion. The previous successful application exogenous hyaluronate was added to lung of the periodate-thiobarbituric acid reagents extracts prior to heating, more than 95% to the determination of free sialic acid in was recovered by the coupled enzyme- crude biological mixtures sugggested that calorimetric procedure (Table 1). the unsaturated oligosaccharides derived The residual color noted in the blanks from hyaluronate could also be measured from lung extracts has an absorption maxi- in such mixtures. While the presently de- mum at 532 nm and may arise from periodate scribed methodology has not been extended oxidation of deoxyribose or unsaturated to a large number of tissues, we are report- lipids. Oxidation of these compounds gives ing its application to the determination of malonaldehyde which in turn reacts with hyaluronate in lung extracts, as an example thiobarbituric acid to give a colored com- of a tissue which contains interfering sub- plex with an absorption maximum at 532 nm. stances, and in synovial tissue, where the The concentration of this chromophore SPECIFIC SENSITIVE METHOD FOR HYALURONATE DETERMINATION 479

TABLE 1

Assay OF HYALURONATE IN LUNG EXTRACT BY A Streptomyces HYALURONATE LYASE-PERIODATE-THIOBARBITURK ACID PROCEDURE

Absorbancy (549 nm)b Hyaluronate Hyaluronate Lung extracta added Without hyaluronate With 5 TRU found (/-a (fig) lyase hyaluronate lyase b.4

Heating step omitted 100 0.545 0.520 0 Heated at 100°C for 3 min 100 0.072 0.077 0 100 5 0.078 0.142 6.5 100 25 0.069 0.344 27.0 loo 50 0.065 0.636 55.5 100 100 0.069 1.056 95.8

a Soft tissue was scraped from six freshly excised, perfused guinea pig lungs. The tissue (3.7 g) was ground in a motor-driven Teflon-glass homogenizer with 10 ml 0.1 M sodium phosphate buffer, pH 6.0. The homogenate (20 mg protein/ml) was dialyzed at 4°C for 4 h against the buffer used for homogenization and 100~~1aliquots (with or without added hyaluronate) heated at 100°C for 3 min. b Samples were treated with hyaluronate lyase and assayed by the periodate-thiobarbituric acid procedure as described under Methods. obtained from lung extracts was not sig- the coupled enzyme-calorimetric proce- nificantly reduced by prior extraction with dure (Table 2) yielded values similar but not 2 vol of chloroform-methanol (2:l). The identical to those obtained from computa- residual blank value may therefore arise tions based on uranic acid content. These from the presence of deoxyribonucleic acid. tissues gave negligible blank values without Synovial fluid is a rich source of hyaluro- the heating and dialysis steps that were nate (15). Analysis of a limited number of required for analysis of the lung extracts. synovial fluid samples for hyaluronate by Analysis of two samples of synovial fluid

TABLE 2

ASSAY OF HYALURONATE IN SYNOVIAL FLUID BY A Srreptomyces HYALURONATE LYASE-PERIODATE-THIOBARBITURIC ACID PROCEDURE

Absorbancy at 549 nmc Hyaluronate Uranic acid Synovial Without With 5 TRU found foundd fluid” Diagnosis” hyaluronate lyase hyaluronate lyase (mg/ml) (~mol/ml)

S.Y. RA 0.025 0.116 0.37 1.6 M.W. RA 0.016 0.535 2.02 7.8 L.C. SLE 0.016 1.183 3.92 9.8

n Individual synovial fluid samples were received from the clinical laboratory of the Arthritis Division and were obtained as by-products of routine diagnostic procedures. * RA, rheumatoid arthritis; SLE, systemic lupus erythematosus. c Aliquots of the synovial fluids (25 ~1) were treated with hyaluronate lyase and assayed by the periodate- thiobarbituric acid procedure as described under Methods. d Uranic acid analyses were performed by the procedure of Bitter and Muir (21). 480 JOURDIAN ET AL. from rheumatoid arthritis patients indicated 5. Svejcar, J., and Robertson, W. V. B. (1967) Anal. that hyaluronate accounted for approxi- Biochem. 18, 333-350. 6. Breen, M., Weinstein, H. G., Andersen, M., mately 66% of the uranic acid content whereas and Veis, A. (1970)AnaI. Biochem. 35,146- 159. hyaluronate accounted for nearly all the Hardingham, T. E., and Adams, P. (1976) Biochem. uranic acid in a sample of synovial fluid J. 159, 143-147. from a patient with systemic lupus erythe- Linker, A., Meyer, K., and Hoffman, P. (1956) matosus. J. Biol. Chem. 219, 13-25. Greiling, H., Gunther, T., and Eberhard, T. (1960) Hoppe-Seyler’s Z. Physiol. Chem. 319, CONCLUSIONS 161- 166. 10. Reissig, J. L., Strominger, J. L., and Leloir, L. F. We believe the coupled enzyme-colori- (1955) J. Biol. Chem. 217, 959-966. metric assay described above will prove 11. Ohya, T., and Kaneko, Y. (1970) Biochim. Bio- particularly suitable for the determination phys. Acta 198, 607-609. of hyaluronate in partially purified mixtures 12. Hascall, V. C., Riolo, R. L., Hayward, J., Jr., and of glycosaminoglycans, and with proper Reynolds, C. C. (1972) J. Biol. Chem. 247, 4521-4528. precautions, in crude biological extracts. 13. Koseki, M., Kimura, A., and Tsurumi, K. (1978) The methodology is sensitive, specific, and J. Biochem. 83, 553-558. is not inhibited by the presence of other 14. Warren, L. (1959) J. Biol. Chem. 234, 1971-1975. glycosaminoglycans. 15. Roseman, S., Watson, D. R., Duff, I. F., and Robinson. W. D. (1955) Fed. Proc. 14, 312. 16. Nagasawa, K., Inoue, Y., and Kamata, T. (1977) REFERENCES Carbohyd. Res. 58, 47-55. 17. Tolksdorf, S., McCready, M. H., McCuJlagh, D. R., 1. Mathews, M. B. (1966) in Methods in Enzymology and Schwenk, E. (1949) J. Lab. C/in. Med. 34, (Neufeld, E. F., and Ginsburg, V., eds.), Vol. 8, 74-89. pp. 654-662, Academic Press, New York. 18. Aminoff, D. (1961) Biochem. J. 81, 384-392. 2. Greiling, H. (1965) in Methods of Enzymatic 19. Waravdekar, V. S., and Sastaw, L. D. (1957) Bio- Analysis (Bergmeyer, H. U., ed.), 2nd ed., chim. Biophys. Acra 24, 439. pp. 87-92, Academic Press, New York. 20. Bemheim, F., Bernheim, M. L. C., and Wilbur, 3. Schiller, S., Slover, G. A., and Dorfman, A. (l%l) K. M. (1948) J. Biol. Chem. 174, 257-264. J. Biol. Chem. 236, 983-987. 21. Bitter, T., and Muir, H. M. (1962) Anal. Biochem. 4. Antonopoulos, C. A., Gardell, S., Szirmai, J. A., 4, 330-334. and De Tyssonsk, E. R. (1964) Biochim. Biophys. 22. Horwitz, A. L., and Crystal, R. G. (1975) J. Clin. Acra 83, 1- 19. Invest. 56, 1312-1318.