Hum Genet (2003) 113 : 178–187 DOI 10.1007/s00439-003-0958-9
O R I G I N A L I N V E S T I G AT I O N
Leslie B. Gordon · Ingrid A. Harten · Anthony Calabro · Geetha Sugumaran · Antonei B. Csoka · W. Ted Brown · Vincent Hascall · Bryan P. Toole
Hyaluronan is not elevated in urine or serum in Hutchinson-Gilford Progeria Syndrome Received: 26 January 2003 / Accepted: 23 March 2003 / Published online: 1 May 2003 © Springer-Verlag 2003
Abstract Elevations in urinary hyaluronan have been used as the principal laboratory indicator for diagnosis of Hutchinson-Gilford Progeria Syndrome (HGPS). Previous reports have provided evidence suggesting that children with HGPS have altered hyaluronan metabolism as indicated by a mean 17-fold increase in urinary hyaluronan over normal values. In addition, adults with Werner’s syndrome have elevated urinary hyaluronan and even more prominent elevations in serum hyaluronan over age-matched controls. It is not known whether serum hyaluronan is elevated or whether serum hyaluronan levels correlate with urinary hyaluronan levels in children with HGPS. In a large cohort of 19 HGPS patients, we sought to confirm elevations in urinary hyaluronan concentration, to establish whether serum hyaluronan is elevated, to measure the size of urinary hyaluronan, and to determine whether serum or urine hyaluronidase levels are altered. We have
L. B. Gordon (✉) · I. A. Harten · B. P. Toole Department of Anatomy and Cellular Biology, Tufts University School of Medicine, 136 Harrison Avenue, Boston, MA 02111, USA e-mail:
[email protected] L. B. Gordon Department of Pediatrics, Rhode Island Hospital, Providence, R.I., USA A. Calabro · V. Hascall Department of Biomedical Engineering, Cleveland Clinic Foundation, Cleveland, Ohio, USA G. Sugumaran Department of Veterans Affairs, Connective Tissue Research Laboratory, Edith Nourse Rogers Memorial Veterans Hospital, Bedford, Ma., USA A. B. Csoka Department of Pathology, University of California, San Francisco, Calif., USA W. T. Brown Department of Human Genetics, Institute for Basic Research in Developmental Disabilities, Staten Island, N.Y., USA
analyzed urinary and serum hyaluronan levels in patients with HGPS and control patients (1) by using an enzymelinked immunosorbent assay (ELISA)-like method in which sample hyaluronan in solution and hyaluronan in solid phase compete for a solution of biotinylated hyaluronan-binding protein, and (2) by fluorophore-assisted carbohydrate electrophoresis. The size of urinary hyaluronan was measured by using Sepharose CL-6B size exclusion chromatography. Serum and urinary hyaluronidases were evaluated quantitatively, by using ELISA, and qualitatively, by using a gel detection method. HGPS patients did not show a significant elevation in either urinary or serum hyaluronan. We detected no difference in the size of urinary hyaluronan between HGPS children and age-matched controls. Serum and urinary hyaluronidase levels were not significantly different in normal and HGPS patients. These studies indicate that neither serum nor urinary hyaluronan concentration is a reliable diagnostic or prognostic marker for HGPS and underscore a difference between adult and childhood progerias.
Introduction Hutchinson-Gilford Progeria Syndrome (HGPS) is a rare and uniformly fatal genetic disease affecting approximately 1 in 8 million live births (DeBusk 1972). The extracellular matrix of mesodermally derived tissues is a target of the principal defects in HGPS, with increased collagen and elastin secretion, disorganized dermal collagen, and decreased decorin compared with normal controls (Beavan et al. 1993; Davidson et al. 1995; Giro and Davidson 1993; Sephel et al. 1988). Children experience normal fetal and post-natal development until around age 9 months, when severe failure to thrive ensues and is accompanied by extensive loss of subcutaneous fat and resulting prominance of scalp veins (for reviews, see Badame 1989; Brown 1990; Pesce and Rothe 1996). Cartilaginous and some distal bone resorption occurs, together with severely delayed dentition. Children reach a final height and weight of approximately 1 m and 15 kg,
179
respectively. Skin is thin with sclerodermatous areas and almost complete hair loss. Death occurs almost exclusively because of widespread atherosclerotic plaque formation at an average age of 13 years (DeBusk 1972). Werner’s syndrome (WS) is a rare autosomal recessive disease with onset in the third decade of life and shares several features with HGPS, including a shortened life span (average of 47 years) frequently because of cardiovascular disease, short stature, scleroderma-like skin with dermal fibrosis, and a tightened appearance with discrete areas of discoloration (for a review, see Reichel et al. 1971). HGPS and WS are termed “progeroid syndromes” because of a general aged appearance of the patients; patients with either disease develop severe premature atherosclerosis, a condition generally affecting the aging population. Because these two disorders share several prominent phenotypes, the finding of elevated urinary hyaluronan (formerly hyaluronic acid; hyaluronate) in WS patients in 1975 (Tokunaga et al. 1975) was quickly followed by examination of hyaluronan in HGPS. The first study found qualitative elevations in urinary hyaluronan of one patient with HGPS (Tokunaga et al. 1978). The most systematic report of altered hyaluronan metabolism in HGPS was carried out by using high-performance liquid chromatography (HPLC) of disaccharide products of chondroitinase digestion to measure urinary hyaluronan (Kieras et al. 1986). This study enrolled 11 children with HGPS and found a mean 17-fold increase in urinary hyaluronan over their normal age-matched controls. Although hyaluronan has not been directly linked to the development of the characteristic abnormalities in skin, bone, and cardiovascular system in either WS or HGPS, there is considerable circumstantial evidence that hyaluronan may play a role in these diseases. For example, hyaluronan has important structural and signaling properties that have been linked to the progression of atherosclerosis (Toole et al. 2002), to the inhibition of angiogenesis (Feinberg and Beebe 1983; Slevin et al. 1998; West et al. 1985), which could be involved in the bone resorption characteristic of HGPS, to the delay in hair follicle morphogenesis (Kaya et al. 1997), to the decrease in skin elasticity (Stern et al. 1998), and to sclerodermatous skin (Juhlin et al. 1986). However, several HGPS case studies have reported no elevation in urinary hyaluronan (Stables and Morley 1993; Wisuthsarewong and Viravan 1999; Yu and Zeng 1991). In addition, further analyses of urine by HPLC methods have given inconsistent results (W.T. Brown, unpublished), leading us to re-examine this issue. Thus, we have undertaken a detailed study of a large cadre of HGPS patients (19 of approximately 35 cases worldwide) and age-matched controls to determine whether there are significant differences in the characteristics of hyaluronan in urine and serum.
Materials and methods Patients and specimens All patients had been previously diagnosed with HGPS, based on the phenotypic expression of the disease. Nineteen subjects with HGPS and 53 pediatric control subjects donated blood and urine after either they and/or their parents gave informed consent. All subjects with HGPS received 2.5 g lidocaine/prilocaine cream (EMLA) onto the antecubital fossa of one arm for 60 min. Within 5 min of EMLA removal, all venipunctures were performed in a routine manner by using vacutainer equipment with a 21-gauge needle. Blood was centrifuged at 1700 g for 5 min; serum was removed in a sterile manner and frozen at –80°C until analysis. Urine samples were obtained periodically from each HGPS patient. Samples were obtained via clean catch, and each was divided into three sterile tubes: 0.1 mg/ml thymol was added to the first, 1% Triton X-100 was added to the second, and the third received no additive. Samples were frozen at –20°C and then mailed overnight to our laboratory where they were stored at –80°C until analysis. Creatinine measurement Urinary creatinine concentration was measured on a Synchron LX 20 multianalyzer (Beckman Coulter). The Synchron LX System utilizes the Jaffe rate method (Jaffe 1886) to determine creatinine concentration. Briefly, a solution consisting of 0.05 M picric acid and 0.188 M sodium hydroxide was mixed with urine at a ratio of 1:105, forming a red complex whose absorbance was measured at 520 nm. The rate of complex development is a direct measure of creatinine concentration. The Synchron LX System measures urinary creatinine concentration in the range of 10–400 mg/dl and has an inter-assay variability of 4.5%. Quantitative measurement of serum and urinary hyaluronan via competitive enzyme-linked immunosorbent assay-like method A competitive enzyme-linked immunosorbent assay (ELISA)-like method was adapted from that of Kongtawelert and Ghosh (1990) and used to measure hyaluronan concentrations in the urine and serum of study subjects. All samples were assayed in triplicate. Microtiter plates (96-well MaxiSorp plates; Nalge Nunc International, Denmark) were pre-coated by incubation with 100 µl/well of 0.1 mg/ml hyaluronan (Sigma, St.Louis, MO) in 0.1 M NaHCO3, pH 9.6, at 4°C for 16 h. Excess coating solution was removed, and the plate was gently dried with warm air to ensure adherence of hyaluronan to the wells. Wells were blocked by using 100 µl 1% bovine serum albumin (BSA; Sigma) in phosphate-buffered saline (PBS) at room temperature for 1 h, washed with PBS containing Tween-20 (Sigma), and dried with warm air. Samples and standards were diluted with 6% BSA in PBS and pre-incubated at room temperature for 1 h with an equal volume of 0.05 µg/ml biotinylated hyaluronan-binding protein (bHABP; Seikagaku America, Cape Cod, MA, USA) in 50 mM TRIS, pH 8.6. After the wells were blocked as described above, 100 µl samples and standards at multiple dilutions and blanks were added in triplicate to wells and incubated at room temperature for 1 h. Wells were then washed three times and dried as described. A solution of 0.4 mg/ml peroxidase-conjugated mouse monoclonal anti-biotin antibody (Zymed, San Francisco, CA) was diluted 1:3000 in PBS, and 100 µl was added per well and incubated at room temperature for 1 h. Wells were then washed three times and dried as described. A solution of 0.6 mg/ml o-phenylenediamine dihydrochloride (OPD; Sigma) and 0.0013% H2O2 in citrate/phosphate buffer, pH 5, was used for color development. The reaction was stopped by the addition of 2 M sulfuric acid. Absorbance was measured at 492 nm on an EL800 Universal Microplate reader (Bio-Tek Instruments, Winooski, VT). Concentrations of hyaluronan were calculated against a
180 standard curve of 5–300 ng/ml hyaluronan (Healon, Pharmacia Fine Chemicals, Uppsala, Sweden) by using GraphPad Prism V3.02 (GraphPad Software). Analysis of hyaluronan via fluorophore-assisted carbohydrate electrophoresis Fluorophore-assisted carbohydrate electrophoresis (FACE) analysis was carried out as previously published (Calabro et al. 2000). Briefly, samples were protease-digested, ethanol-precipitated, and underwent complete digestion of intact glycosaminoglycan (GAG) chains to characteristic saccharide structures by hyaluronidase SD and chondroitinase ABC (Seikagaku America). Saccharides were then fluorotagged by reductive amination with 2-aminoacridone HCL (Molecular Probes) and separated by FACE. The relative fluorescence in each band was quantified with a cooled charge-coupled device camera (Roper Scientific/Photometrics) and analyzed by using the Gel-Pro Analyzer program (Media Cybernetics).
Hyaluronan-substrate gel techniques for detection of hyaluronidase and hyaluronidase inhibitor Hyaluronidase and hyaluronidase inhibitor activity were measured by using the substrate gel techniques described previously (Guntenhoner et al. 1992; Mio and Stern 2000). Samples were electrophoretically separated on SDS acrylamide (30%/0.8% bis) gels impregnated with 0.17 mg/ml hyaluronan. After SDS removal, gels were incubated in assay buffers at either pH 3.7 or 7.4 and exposed to protease to remove protein that may interfere with the assay. After Alcian Blue staining, hyaluronidase activity was indicated as cleared bands on a light blue background. Counterstaining with Coomassie Blue revealed any residual protein as dark blue bands. Relative hyaluronidase activity in samples was compared with a standard of known activity run on the same gel. For detection of hyaluronidase inhibitor activity, gels prepared as above were then incubated in a solution containing 0.05 rTRU/ml bovine testicular hyaluronidase and stained with Alcian Blue for bands of hyaluronan, indicating the presence of inhibitors protecting hyaluronan from degradation.
Quantitation of urinary hyaluronan via HPLC HPLC analysis was generously performed by Dr. Fred Kieras using the protocol previously described (Zebrower et al. 1986a, 1986b). Briefly, total GAGs were precipitated from 30 ml urine with 0.4% cetyl pyridinium chloride (CPC) at 4°C overnight. After centrifugation, the CPC was removed via three successive precipitations with absolute EtOH/1% potassium acetate at 0°C. The cell pellet was dried and redissolved in deionized water. The GAGs were digested to disaccharides with chondroitinase ABC (Seikagaku), and proteins were removed via ethanol precipitation. The disaccharides were then further purified and separated by HPLC as described (Zebrower et al. 1986a, 1986b). Measurement of urinary hyaluronan size via gel-exclusion chromatography Urine (5 ml) from HGPS children or age-matched controls was lyophilized and resuspended in 1.6 ml deionized water. A concentrated sample (400 µl) was placed onto a Sepharose CL-6B (Pharmacia, Sweden) column measuring 1.8×98 cm; 75×2.2-ml fractions were collected via gravity drip. Hyaluronan concentration was measured in every third fraction of the included volume of the column by using the ELISA-like assay. In each sample assayed, total hyaluronan content was determined by the area under the curve and compared with total hyaluronan content determined by ELISAlike assay. Hyaluronan recovery was estimated at 85%–126% (mean=112±16%). Assays for hyaluronidase activity Hyaluronidase activity in urine and serum was determined in triplicate under both neutral (pH 7.4) and acidic (pH 3.7) conditions by using an ELISA-like assay as previously described (Frost and Stern 1997). Briefly, the free carboxyl groups of human umbilical cord hyaluronan (ICN, Irvine, CA) were partially biotinylated by using biotin hydrazide, and the hyaluronan was covalently bound to microtiter plate wells. Standards and unknown samples were incubated in the biotinylated hyaluronan (bHA)-coated wells, and residual bHA substrate was detected with an avidin-peroxidase color reaction (ABC kit, Vector Labs, Burlingame, CA) at 492 nm on an ELISA plate reader (Titertek Multiskan PLUS, ICN). Activity was expressed in relative turbidity units (rTRU). On each plate, bovine testicular hyaluronidase (Wydase, Wyeth-Ayerst, Philadelphia, PA) was used to produce a standard curve ranging from 0.015– 150 rTRU/ml. Positive and negative control wells (no enzyme and no avidin-biotin complex, respectively) were also included in triplicate.
Statistics Intra-assay and inter-assay variabilities of the ELISA-like assay were calculated from the mean of the coefficient of variation (Cv) for each triplicate assayed on a single assay plate and between two assay plates, respectively. All samples were assayed at least twice and data from the first two assays were averaged for statistical analyses. Two-factor analysis of variance (ANOVA) was used to analyze the influence of age on urinary and serum hyaluronan levels in HGPS and control subjects. Receiver-operator characteristic (ROC) curves were generated to analyze the clinical significance of using urinary and serum hyaluronan levels as diagnostic markers for HGPS. One-way ANOVA with repeated measurements was used to compare urinary hyaluronan from the same HGPS patients at different donation times. All other statistical comparisons of agematched samples were made by using two-tailed paired Student’s t-tests. Age-matched samples were chosen by pairing the control subject closest in age to each HGPS subject. The age of each control subject was within 6 months of its HGPS match (mean=2.4± 1.8 months). Urinary and serum hyaluronan values were recorded as mean±SEM, and urinary and serum hyaluronidase values were recorded as mean±SD. All statistical analyses were generated by using SPSS for Windows Version 11.5 software.
Results The ELISA-like assay for hyaluronan is both sensitive and specific Sensitivity Standard curves were generated with known quantities of hyaluronan in PBS, ranging from 5–300 ng/ml. The lower limit for reliable hyaluronan detection with the ELISAlike assay was 20 ng/ml. These standard curves were generated on each microtiter plate and were consistently linear from 20–80 ng/ml. Each sample in this study was analyzed at two dilutions within the 20–80 ng/ml range (in the linear portion of the standard curve). In addition, 100 or 200 ng/ml hyaluronan was added to urine and serum and gave increases in readings similar to those in PBS, demonstrating the absence of substances in urine or serum that interfere with the hyaluronan readings. Within each plate, each experimental sample was assayed together with its
181
age-matched control. Intra-assay variability was 12%, as determined from the mean of the Cv for each triplicate assayed. Inter-assay variability was 14%, as determined by calculating the mean Cv for each triplicate sample measured on two different plates. Negative controls consisted of both standard hyaluronan and urine samples, assayed in non-hyaluronan coated wells; these gave background absorbance readings. Positive controls consisted of samples from a single batch of adult urine, assayed on each microtiter plate. Specificity We found no detectable levels of hyaluronan in chondroitin sulfate proteoglycan (1 mg/ml), in urine treated with Streptomyces hyaluronidase, or in standard hyaluronan (Healon; 1 mg/ml) treated with Streptomyces hyaluronidase. Addition of 1 mg/ml chondroitin sulfate proteoglycan to urine, which initially contained 580 ng/ml hyaluronan, yielded no significant change in measured hyaluronan concentration. Urinary hyaluronan is not significantly elevated in children with HGPS regardless of age Urinary hyaluronan levels were analyzed by ELISA-like assay in 19 children with HGPS and 19 age-matched controls (Fig. 1). Both HGPS and control values followed a normal distribution. Age cohorts, number of subjects, and micrograms of hyaluronan per gram creatinine were as follows: 1–4 years, n=3 HGPS, mean=1300.29±465.6, and three controls, mean=1015.13±50.6; 5–8 years, n=5 HGPS, mean= 750.44±404.8, and five controls, mean=781.93±274.6;
Fig. 1 Urinary hyaluronan (HA) is not elevated in HGPS regardless of age (Cr creatinine). Urinary hyaluronan levels measured by ELISA-like assay from 19 children with HGPS and 19 agematched controls grouped by age cohorts. Two-factor ANOVA revealed no effect of age on urinary hyaluronan levels (P=0.198, f=1.652, df=3), and ROC curve generated without age weighting (not shown) revealed extremely low relevance of urinary hyaluronan as a clinical or biological hallmark of disease (area under curve =0.681; asymptotic significance =0.056; 95% confidence interval =0.5–0.86)
9–12 years, n=5 HGPS, mean=1089.10±153.3, and five controls, mean=796.50±397.5; 13–16 years, n=6 HGPS, mean=1075.31±540.8, and six controls, mean=665.94± 120.5. Only initial urine donations were compared in this analysis. Two-factor ANOVA revealed no effect of age on urinary hyaluronan levels (P=0.198, f=1.652, df=3). Therefore, we generated a ROC curve without age weighting and determined that the levels of urinary hyaluronan in control versus HGPS subjects were not useful as a clinically relevant marker for HGPS, regardless of age (area under curve =0.681; asymptotic significance =0.056, 95% confidence interval =0.5–0.86). We validated the ELISA-like assay results shown in Fig. 1 in several ways. First, samples were re-assayed in a laboratory that routinely measures hyaluronan levels in urine by using another established ELISA-like method (modified Fosang method; Fosang et al. 1990; Lokeshwar et al. 1997). These assays were kindly performed in a blinded fashion by Dr. Vinata Lokeshwar at the University of Miami School of Medicine, Miami, Fl. Three HGPS and three age-matched control urine samples, plus one adult positive control sample, were assayed in triplicate, yielding hyaluronan concentrations ranging from 600–1700 ng/ml urine (mean=1020.6±593.9) in HGPS and 390–1266 ng/ml urine (mean=926±532.8) in age-matched control samples. We compared these results with the modified Kongtawelert assay (Kongtawelert and Ghosh 1990), used for our routine measurements, in which hyaluronan concentrations ranged from 525–926 ng/ml urine (mean=677.6±217.5) in HGPS and 482–809 ng/ml urine (mean=607.8±176) in agematched control samples. Linear regression analysis for the two assay methods demonstrated a significant level of correlation (r=0.635, P=0.05). We then analyzed urinary hyaluronan levels by the FACE method, which has a completely different biochemical basis from the ELISA-like assay, being independent of hyaluronan interaction with a binding protein. It also has some similarity to the HPLC method in which the previously published elevation of hyaluronan in HGPS urine was determined (Kieras et al. 1986), since both assays employ the precipitation of GAGs and the enzymatic digestion of the GAGs to disaccharides, followed by the quantitative analysis of the disaccharides. However, a major difference in processing is the use of quaternary ammonium salts for GAG precipitation in the HPLC method; the FACE protocol uses only the ethanol precipitation of GAG. Fig. 2A shows a representative FACE gel displaying the separated enzymatic breakdown products of the GAGs, compared with standard disaccharide controls. The relevant disaccharide component for hyaluronan is outlined. In Fig. 2B, we compared urinary hyaluronan concentrations derived from the densitometric analysis of gel bands shown in Fig. 2A with those derived in the ELISAlike assay. As determined by using an age-matched paired two-tailed t-test, there was no significant difference in measured concentrations between these methods (P=0.77), and paired samples differed in concentration by 5%–38%. Finally, we determined whether values for urinary hyaluronan measured via the sample processing and HPLC meth-
182
Fig. 2A, B FACE and ELISA-like analyses give similar results. A FACE analysis was carried out on urine samples from three subjects with HGPS (HGPS) and their age-matched controls (Control). Bands corresponding to hyaluronan are located within the rectangular box. Band intensity corresponds to the amount of hyaluronan present in the sample and is not normalized for urinary creatinine. gal Nac N-acetylgalactosamine, ∆di HA fluorotagged unsaturated hyaluronan disaccharides [2-acetamido-2-deoxy-3-O(β-D-gluco-4-enepyranosyluronic acid)-D-glucose], ∆di OS fluorotagged unsaturated non-sulfated chondroitin sulfate disaccharides [2-acetamido-2-deoxy-3-O-(β-D-gluco-4-enepyranosyluronic acid)-D-galactose], ∆di 6S fluorotagged unsaturated chondroitin sulfate disaccharaides sulfated at position 6 [2-acetamido-2-deoxy3-O-(β-D-gluco-4-enepyranosyluronic acid)-6-O-sulfo-D-galactose], ∆di 4S fluorotagged unsaturated chondroitin sulfate disaccharaides sulfated at position 4 [2-acetamido-2-deoxy-3-O-(β-D-gluco-4-enepyranosyluronic acid)-4-O-sulfo-D-galactose], ∆di 2S fluorotagged unsaturated chondroitin sulfate disaccharaides sulfated at position 2 [2-acetamido-2-deoxy-3-O-(2-O-sulfo-β-D-gluco-4-enepyranosyluronic acid], ∆di 4,6S, ∆di 2,4,6S a mixture of fluortagged unsaterated chondroitin sulfate disaccharides sulfated at positions 4 and 6 or 2, 4, and 6 [2-acetamido-2-deoxy-3-O-(β-D-gluco-4enepyranosyluronic acid)-4,6-di-O-sulfo-D-galactose and 2-acetamido-2-deoxy-3-O-(2-O-sulfo-β-D-gluco-4-enepyranosyluronic acid)-4,6-di-O-sulfo-D-galactose]. B Comparison of values obtained from FACE by densitometric analysis and from the ELISAlike assay, after normalization for urinary creatinine (Cr) concentration (P=0.77)
ods used in the original publication (Kieras et al. 1986) differed from those obtained by the ELISA-like and FACE methods. Analysis of four urine samples (two HGPS and two age-matched controls) was kindly performed in a doubleblinded fashion by Dr. Fred Kieras and Gabe Radu. Results were inconsistent with previously published values, failing to yield significant values for hyaluronan concentration in any of the samples (values ranged from undetectable to 7.4 µg hyaluronan/g creatinine). Analysis of these samples by using the modified Kongtawelert ELISA-like assay (Kongtawelert and Ghosh 1990) yielded hyaluronan concentrations ranging from 414–1008 µg hyaluronan/g creatinine (mean=671.2±249.5). Thus, we confirmed that the similarity in hyaluronan levels that we found between HGPS children and controls was not caused by use of the ELISA-like assay. Urinary hyaluronan concentration is variable in both HGPS and control samples but does not change significantly with age or longitudinal analysis Overall, urinary hyaluronan levels were variable, ranging from 295 µg hyaluronan/g creatinine to 1898 µg hyaluronan/g creatinine. Variation in hyaluronan levels did not correlate with patient age. To determine whether urinary hyaluronan changed with the progression of the disease, we
183 Fig. 3 Urinary hyaluronan (HA) did not vary significantly over a 2-year period (Cr creatinine). Urinary hyaluronan levels as measured by ELISA-like assay in subjects with HGPS who gave multiple urine samples over a 20-month period. Samples were received every 4–6 months. The number of subjects providing samples declined with time because of lack of response or death of subject. First donation from control subjects shown for comparison of range of hyaluronan values. By single factor ANOVA with repeated measures, no significant change is seen between donations from HGPS subjects (P=0.20, f=1.63, df=3)
compared samples from each of 12 children with HGPS over a 20-month period. Samples were collected every 4–6 months but, primarily because of subject death, the number of repeat donations over the 20-month period diminished from 12 to 4 subjects. Figure 3 shows large variations in hyaluronan levels at each donation time in HGPS patients. However, no significant difference in hyaluronan levels was observed with repeated sampling over time, as determined by single factor ANOVA with repeated measures (P=0.204, f=1.613, df=3). The range of control values was also similar to HGPS values and did not vary significantly between longitudinal donations (first donation data only are shown in Fig. 3). Size of hyaluronan in urine is similar in HGPS and normal patients One explanation for the differences in values obtained in our study and those of previous investigators is the possible disparity in the size between hyaluronan in normal and in HGPS samples. For example, CPC precipitation, which was used in the previously published HPLC analysis, might differentially precipitate larger hyaluronan molecules and fail to precipitate smaller molecules. Thus, we evaluated urinary hyaluronan size in three HGPS and three agematched control subjects by gel exclusion chromatography, followed by measurement of hyaluronan concentration in each column fraction by our ELISA-like method. As demonstrated in Fig. 4, the size ranges for the HGPS and control samples corresponded almost precisely, indicating that there was no significant size difference in urinary hyaluronan between HGPS children and normal controls. The molecular weight range (3000–12,000) and peak (approximately 6000) were similar to previously published hyaluronan measurements in normal human urine (Laurent et al. 1987).
Fig. 4 Urinary hyaluronan (HA) from subjects with HGPS is similar in size to age-matched controls. Size exclusion chromatography profile of urine samples from HGPS and age-matched controls on a Sepharose CL-6B column (Pharmacia). Hyaluronan concentration was measured in every third fraction by using the ELISAlike assay. Total hyaluronan content was determined by calculating the area under the curve. Mean total hyaluronan recovery in each sample was 112±16%
Serum hyaluronan levels are not altered significantly in HGPS patients We analyzed the hyaluronan concentrations in serum samples from 15 children with HGPS and their age-matched controls. None of the samples displayed significantly elevated hyaluronan concentrations over control samples (Fig. 5). Hyaluronan concentrations ranged from 20–63 ng/ml serum in HGPS and 23–81 ng/ml serum in age-matched control samples (mean=40.1±2.0 and 45.7±3.8, respectively). An ROC curve was generated for the HGPS and control groups and demonstrates that serum hyaluronan is not a useful diagnostic biologic indicator of HGPS (area under curve=0.481; asymptotic significance=0.462; 95% confidence interval=0.191–0.646).
184
Urinary and serum hyaluronidase do not vary significantly between HGPS and normal patients
Fig. 5 Serum hyaluronan levels are not elevated in HGPS. A single serum sample from each HGPS and age-matched control subject was assayed in triplicate on two separate occasions by using the ELISA-like assay. Values indicate the means of the two assays. Clinical and biological relevance of serum hyaluronan as a disease marker for HGPS as determined by ROC curve (not shown) was extremely low (area under curve =0.481; asymptotic significance= 0.462; 95% confidence interval =0.191–0.646)
Fig. 6A–C Hyaluronidase activity is not elevated in HGPS. Urine and serum samples were assayed for hyaluronidase activity by using the ELISA-like assay. Activities were measured at both pH 3.7 and pH 7.4. Data from pH 3.7 is presented here as the log-transformed mean and SEM because of non-normal distribution of the data. Data for urine is normalized for creatinine content. A Urine (P=0.32). B Serum (P=0.79). C Hyaluronan gel zymography assay for serum acid-active hyaluronidase activity in subjects with HGPS (P) and agematched controls (C). Intensity of bands is proportional to activity (A). The bovine testicular hyaluronidase standard is a higher molecular weight molecule than the acid-active hyaluronidase in human serum
A potential source of variation in urinary hyaluronan levels could be differences in hyaluronidase levels between experimental and control samples. We assayed both the major (acid pH-active optimum) and minor (neutral pH-active optimum) forms of urinary hyaluronidase. Figure 6A shows the acid-active hyaluronidase activity in 11 paired urine samples from HGPS children and age-matched controls, measured by using an ELISA-like method (Nawy et al. 2001). Values for both groups varied widely, ranging between 0.05–12 rTRU/ml urine (mean=3.5±3) or 16– 2280 rTRU/g creatinine. Neutral-active hyaluronidase activity measured between 0.1–1.6 rTRU/ml (mean=0.3± 0.3 rTRU/ml) or 0–136 rTRU/g creatinine. In both acid-active and neutral-active analyses, there was no significant difference between the HGPS and control groups (P=0.33 and P=0.38, respectively). All values were within published ranges for normal adult urinary hyaluronidase (Csoka et al. 1997). Serum samples from 14 HGPS and 14 age-matched control samples were also assayed for both acid-active (via ELISA-like and gel zymography methods) and neutral active (via gel zymography method) hyaluronidase activity.
185
Figure 6B shows acid-active hyaluronidase activity in these samples measured from 0–12 rTRU/ml (mean=4.2± 2.6). No significant difference between the two groups was found (P=0.79). Adult serum was also assayed for acidactive hyaluronidase activity as a positive control and found to be 5–12 rTRU/ml (mean=.8.5±3.8), within range of published values (Afify et al. 1993; Podyma et al. 1997). We then analyzed serum hyaluronidase activity and size by gel zymography (Guntenhoner et al. 1992), where we could visualize protein size in HGPS patients compared with controls. This gel demonstrated that there was no difference in the size of the molecule between control and HGPS serum samples. Both samples were run at ~45 KDa, as previously published. Band intensity, which was proportional to enzyme activity, was similar in both groups (see Fig. 6C). No neutral-active hyaluronidase activity was detected in serum samples (data not shown). Previous experiments carried out by Mio et al. (2000) have shown the presence of inhibitors of hyaluronidases in serum. Alterations in inhibitor level may result in changes in hyaluronan turnover. Therefore, a parallel experiment was carried out to test for the presence of hyaluronidase inhibitors in three HGPS serum samples and four control samples by using a reverse gel zymography technique (Mio and Stern 2000). None was detected (data not shown). We also analyzed hyaluronidase inhibitor levels in three HGPS and three age-matched control urine samples, but again found none.
Discussion Previous reports have implied that children with HGPS have altered hyaluronan metabolism, as indicated by an increase in urinary hyaluronan over age-matched normal controls (Zebrower et al. 1986a, 1986b). This suggests a common basis of disease with WS, in which patients display elevated hyaluronan concentrations in urine and even more significant elevations in serum. Contrary to the previous findings in HGPS, we have established that these patients do not show a significant elevation in either urinary or serum hyaluronan. We have performed an extensive analysis of urinary and serum hyaluronan concentration, by using a competitive ELISA-like assay similar to that used by the investigators who have found increased hyaluronan in WS (Tanabe and Goto 2001), and by using quantitative FACE analysis of disaccharide components produced by chondroitinase digestion (Calabro et al. 2000). Similar values have also been obtained in a blinded fashion in an independent laboratory, by using another established ELISA-like protocol (Lokeshwar et al. 1997). However, independent analyses by HPLC of disaccharide components (Kieras et al. 1986) have given much lower values than those obtained by the other methods. Our control and experimental values for urinary hyaluronan, when normalized to creatinine concentration, fell within published ranges for middle-aged adult and pediatric (ages 1–18 years) normal individuals as measured by using ELISA-like assays (Erickson et al. 1998; Lokesh-
war et al. 2000; Maeda et al. 1999; Tanabe and Goto 2001). Creatinine is neither secreted nor absorbed by the kidney. Thus, creatinine levels have been used to normalize data to changes in urine concentration that occur with variable states of body hydration. FACE analysis of disaccharide components has yielded hyaluronan concentrations similar to the ELISA-like assay results and dissimilar to the results obtained by using HPLC of disaccharide components. This HPLC method has consistently yielded low values, as seen in the values given herein and in previous publications (Akiyama et al. 1991; Toyoda et al. 1991). HPLC analysis itself is an accurate means of measuring disaccharide products of enzymatic digestion. However, a major difference between the sample processing employed here for FACE analysis and that employed for HPLC analysis of urine has been the use of quaternary ammonium salts for GAG precipitation (Akiyama et al. 1991; Shum et al. 1999; Zebrower et al. 1986a), a method known to be inefficient for the precipitation of hyaluronan at low concentrations (Scott 1960). Therefore, the different methods of sample processing probably explain the low values previously obtained for hyaluronan levels in control urine in the HGPS studies. However, the difference found in these studies between HGPS and normal individuals remains unexplained. One possibility that we have considered is that urinary hyaluronan in HGPS patients is larger in size than that from normal children and consequently is more efficiently precipitated by quaternary ammonium salts. Thus, we have compared the size of hyaluronan in urine from normal and HGPS children but have found no significant difference. We have also explored differential hyaluronidase activity as a cause of the difference by analyzing acid-active and neutral-active hyaluronidase activity in serum and urine of HGPS children and age-matched controls. Again, no significant quantitative or qualitative differences have been observed. Our control and experimental values for serum hyaluronan levels are similar in average values and in range to those found in previously published studies of children aged 1–18 years (Andersson and Fasth 1994; Kumar et al. 1989; Shigemori et al. 2002) and similar to those found in normal middle-aged adults (Laurent et al. 1996), including the study of WS patients (Tanabe and Goto 2001). Examination of serum hyaluronan concentration yields values with much less variability than urinary hyaluronan and eliminates potential renal or bladder influences. Therefore, the absence of a significant elevation in serum hyaluronan is probably a better indication that total body hyaluronan turnover is not dramatically altered in patients with HGPS. Hyaluronan in serum and urine represents only a small fraction of total body hyaluronan. For example, human dermis hyaluronan concentration is three to four orders of magnitude greater than serum concentration (reviewed in Laurent and Fraser 1986). An examination of degradation and excretion pathways for hyaluronan in normal subjects has indicated that the kidneys excrete 1%–2% of total body hyaluronan into urine (Laurent and Fraser 1992). There is local hyaluronan degradation at the tissue site of at least 25%–33%, and the rest flows into the lymphatics,
186
so that hyaluronan levels in lymph are 10-fold higher than hyaluronan levels in blood or urine (Laurent and Laurent 1981). Therefore, although significant elevations in serum hyaluronan have been documented in systemic juvenile and adult rheumatoid arthritis (Shigemori et al. 2002), progressive systemic sclerosis (Freitas et al. 1996), and WS (Tanabe and Goto 2001; Tokunaga et al. 1975) and imply global or severe inflammatory activity, more subtle defects in hyaluronan metabolism and function would not be detected in this way. For example, in the course of the present study, we have found that cell-surface-associated, intracellular, and secreted hyaluronan are all significantly decreased in senescing fibroblasts from HGPS patients (data not shown), but this is not reflected in the levels of hyaluronan in serum or urine. WS and HGPS have often been compared as examples of premature aging syndromes with a potentially similar biological basis (Brown et al. 1985; Martin and Oshima 2000). WS affects individuals in their third decade of life and shares, with HGPS, the development of severe premature atherosclerosis (Epstein et al. 1966). However WS, but not HGPS, is marked by cataracts, frank diabetes, chronic skin ulcers, increases in neoplasias, severe osteoporosis, and graying and thinning of hair rather than total alopecia. The inheritance pattern for WS is autosomal recessive, and the primary defect lies in the gene encoding a DNA helicase (Yu et al. 1996). There is evidence that HGPS may be a disease of sporadic autosomal dominant inheritance (Brown et al. 1985), and the WS gene (WRN) is normal in HGPS (Oshima et al. 1996) and also that the gene defect in HGPS may be found on chromosome 1q (Brown 1990; Eriksson et al. 2002; Luengo et al. 2002), which does not harbor a helicase. The evidence presented here that hyaluronan is not elevated in HGPS sheds further doubt on a previously proposed similarity in the biological basis of disease and development of atherosclerosis in these two syndromes. Progeria is a disease in which some, but not all, of its manifestations represent a model of accelerated aging (for a review, see Sweeney and Weiss 1992). Clinical features common to HGPS and normal aging include alopecia (although the pattern of hair loss differs), sclerodermatosis, atherosclerosis, lipofuscin deposition, nail dystrophy, hypermelanosis, and decreased adipose tissue. Clinical differences between the two conditions include sequelae of maldevelopment in HGPS, with coxa valga, distal bone resorption, delayed dentition, facial disproportion, failure to thrive, and short stature. Features of aging that are absent in HGPS include neurosensory decline such as that in Alzheimer’s disease, dementia, hearing loss, and presbyopia. Both urinary and serum hyaluronan increase in normal individuals over the age of 80 years (Tanabe and Goto 2001) but, as we show here, are not increased in HGPS. Thus, our results add to a long list of differences between HGPS and the complex and multigenic aging process and suggest that HGPS is a disease that may shed light on only some of the genetic aspects of aging. Acknowledgements We thank the children with HGPS and their families for their tremendous efforts in participating in this re-
search endeavor. We also thank Drs. Frank Rothman, Paul Knopf, and Christine Harling-Berg for their contributions to the development and continued support of this project, Dr. Joan Lemire for her intellectual contributions to this project, Dr. Fred Kieras and Gabe Radu for hyaluronan analysis at the Institute for Basic Research in Staten Island, NY, Dr. Vinata Lokeshwar for hyaluronan analysis at the University of Miami School of Medicine, Dr. Robert Stern for assistance with the hyaluronidase assays, Dr. Prachya Kongtawelert for assistance with the ELISA-like method, and Dr. David Damassa for statistical assistance. This work was supported by The Progeria Research Foundation and by the American Heart Association (0030217 N)
References Afify AM, Stern M, Guntenhoner M, Stern R (1993) Purification and characterization of human serum hyaluronidase. Arch Biochem Biophys 305:434–441 Akiyama H, Toyoda H, Yamanashi S, Sagehashi Y, Toida T, Imanari T (1991) Microdetermination of hyaluronic acid in human urine by high performance liquid chromatography. Biomed Chromatogr 5:189–192 Andersson GB, Fasth A (1994) Serum concentration of hyaluronan, IgM and I rheumatoid factors in a population based study suvenile chronic arthritis. Scand J Rheumatol 23:183–190 Badame AJ (1989) Progeria. Arch Dermatol 125:540–544 Beavan LA, Quentin-Hoffmann E, Schonherr E, Snigula F, Leroy JG, Kresse H (1993) Deficient expression of decorin in infantile progeroid patients. J Biol Chem 268:9856–9862 Brown WT (1990) Genetic diseases of premature aging as models of senescence. Annu Rev Gerontol Geriatr 10:23–42 Brown WT, Kieras FJ, Houck GE, Dutkowski R, Jenkins EC (1985) A comparison of adult and childhood progerias: Werner syndrome and Hutchinson-Gilford progeria syndrome. Werner’s syndrome and human aging. Adv Exp Med Biol 190:229–244 Calabro A, Benavide sM, Tammi M, Hascall V, Midura R (2000) Microanalysis of enzyme digests of hyaluronan and chondroitin/dermatan sulfate by fluorophore-assisted carbohydrate electrophoresis (FACE). Glycobiology 10:273–281 Csoka TB, Frost G, Wong T, Stern R (1997) Purification and microsequencing of hyaluronidase isozymes from human urine. FEBS Lett 417:307–310 Davidson JM, Zang MC, Zoia O, Giro MG (1995) Regulation of elastin synthesis in pathological states. Ciba Found Symp 192: 81–99 DeBusk FL (1972) The Hutchinson-Gilford progeria syndrome. J Pediatr 80:697–794 Epstein CJ, Martin GM, Schultz AL, Motulsky AG (1966) Werner’s Syndrome: a review of its symptomatology, natural history, pathologic features, genetics and relationship to the natural aging process. Medicine 45:177–221 Erickson DR, Sheykhnazari M, Ordille S, Bhavanandan VP (1998) Increased urinary hyaluronic acid and interstitial cystitis. J Urol 160:1282–1284 Eriksson M, Singer J, Scott L, Dutra A, Boehnke M, Collins FS (2002) Uniparental isodisomy of chromosome 1q in a patient with Hutchinson-Gilford progeria syndrome (abstract). American Society of Human Genetics, 2002 Baltimore Meeting Archive, Abstract 176 http://www.faseb.org/genetics/ashg/annmeet/2002/menu-annmeet-2002.shtml Feinberg R, Beebe D (1983) Hyaluronate in vasculogenesis. Science 220:1177–1179 Fosang AJ, Hey NJ, Carney SL, Hardingham TE (1990) An ELISA plate based assay for hyluronan using biotinylated proteoglycan G1 domain (HA-binding) region. Matrix 10:306–313 Freitas J, Filipe P, Ermerit I, Meunier P, Manso C, Guerro FR (1996) Hyaluronic acid in progressive systemic sclerosis. Dermatology 192:46–49 Frost GI, Stern R (1997) A microriter-based assay for hyaluronidase activity not requiring specialized reagents. Anal Biochem 251:263–269
187 Giro M, Davidson JM (1993) Familial co-segregation of the elastin phenotype in skin fibroblasts from Hutchinson-Gilford progeria. Mech Ageing Dev 70:163–166 Guntenhoner MW, Pogrel MA, Stern R (1992) A substrate-gel assay for hyaluronidase activity. Matrix 12:388–396 Jaffe MZ (1886) Rate method to determine creatinine concentration. Physiol Chem 10:391 Juhlin L, Tengblad A, Ortonne J, Lacour J (1986) Hyaluronate in suction blisters from patients with scleroderma and various skin disorders. Acta Derm Venereol (Stockh) 66:409–413 Kaya G, Rodriguez I, Jorcano JL, Vassalli P, Stamenkovic I (1997) Selective suppression of CD44 in keratinocytes of mice bearing an antisense CD44 transgene driven by a tissue-specific promoter disrupts hyaluronate metabolism in the skin and impairs keratinocyte proliferation. Genes Dev 11:996–1007 Kieras FJ, Brown WT, George E. Houck J, Zebrower M (1986) Elevation of urinary hyaluronic acid in Werner’s syndrome and progeria. Biochem Med Metabol Biol 36:276–282 Kongtawelert P, Ghosh P (1990) A method for the quantitation of hyaluronan (hyaluronic acid) in biological fluids using a labeled avidin-biotin technique. Anal Biochem 185:313–318 Kumar S, West DC, Ponting JM, Gattamaneni HR (1989) Sera of children with renal tumours contain low-molecular-mass hyaluronic acid. Int J Cancer 44:445–448 Laurent T, Fraser JRE (1986) The properties and turnover of hyaluronan. Funct Proteoglycans 124:9–29 Laurent T, Fraser J (1992) Hyaluronan. J FASEB 6: 2397–2404 Laurent TC, Lilja K, Brunnberg L, Engstrom-Laurent A, Laurent UB, Lindqvist U, Murata K, D.Yitterberg (1987) Urinary excretion of hyaluronan in man. Scand J Clin Lab Invest 47:793– 799 Laurent TC, Laurent UBG, Fraser JR (1996) Serum hyaluronan as a disease marker. Ann Med 28:241–253 Laurent UBG, Laurent TC (1981) On the origin of hyaluronate in blood. Biochem Int 2:195–199 Lokeshwar VB, Obek C, Soloway MS, Block NL (1997) Tumorassociated hyaluronic acid: a new sensitive and specific urine marker for bladder cancer. Cancer Res 57:7373–7777 Lokeshwar V, Obek C, Pham, HT, Wei D, Young M, Duncan R, Soloway M, Block N (2000) Urinary hyaluronic acid and hyaluronidase: markers for bladder cancer detection and evaluation of grade. J Urol 163:348–356 Luengo WD, Martinez AR, Lopez RO, Basalo CM, Rojas-Atencio A, Quintero M, Borjas L, Morales-Machin A, Ferrer SG, Pineda Bernal L, Canizalez-Tarazona J, Pena J, Luengo JD, Hernandez JC, Chang JC (2002) Del(1)(q23) in a patient with Hutchinson-Gilford progeria. Am J Med Genet 113:298–301 Maeda H, Fujita H, Sakura Y, Miyazaki K, Goto M (1999) A competitive enzyme-linked immunosorbent assay-like method for the measurement of urinary hyaluronan. Biosci Biotechnol Biochem 63:892–895 Martin GM, Oshima J (2000) Lessons from human progeroid syndromes. Nature 408:263–266 Mio K, Stern R (2000) Reverse hyaluronan substrate gel zymography procedure for the detection of hyaluronidase inhibitors. Glycoconjugates J 17:761–766 Mio K, Carrette O, Maibach HI, Stern R (2000) Evidence that the serum inhibitor of hyaluronidase may be a member of the interalpha-inhibitor family. J Biol Chem 275:32413–32421 Nawy SS, Csoka AB, Mio K, Stern R (2001) Hyaluronidase activity and hyaluronidase inhibitors. Assay using a microtiterbased system. Methods Mol Biol 171:383–389 Oshima J, Brown WT, Martin GM (1996) No detectable mutations at Werner helicase locus in progeria. Lancet 348:10–19 Pesce K, Rothe MJ (1996) The premature aging syndromes. Clin Dermatol 14:161–170
Podyma KA, Yamagata S, Sakata K, Yamagata T (1997) Difference of hyaluronidase produced by human tumor cell lines with hyaluronidase present in human serum as revealed by zymography. Biochem Biophys Res Commun 241:446–452 Reichel W, Garcia-Bunuel R, Dilallo J (1971) Progeria and Werners syndrome as models for the study of normal human aging. J Am Geriat Soc 19:369–375 Scott JE (1960) Aliphatic ammonium salts in the assay of acidic polysaccharides from tissues. Methods Biochem Anal 8:145– 197 Sephel GC, Sturrock A, Giro MG, Davidson JM (1988) Increased elastin production by progeria skin fibroblasts is controlled by the steady-state levels of elastin mRNA. Soc Invest Dermatol 90:643–647 Shigemori M, Takei S, Imanaka H, Maeno N, Hokonohara M, Miyata K (2002) Diagnostic significance of increased serum hyaluronic acid in juvenile rheumatoid arthritis. Pediatr Int 44: 394–399 Shum DKY, Gobel MDI, Tam P (1999) Hyaluronans: Crystallization-promoting activity and HPLC analysis of urinary excretion. J Am Soc Nephrol 10 (Suppl 14):S397–S403 Slevin M, Krupinski J, Kumar S, Gaffney J (1998) Angiogenic oligosaccharides of hyaluronan induce protein tyrosine kinase activity in endothelial cells and activate a cytoplasmic signal transduction pathway resulting in proliferation. Lab Invest 78: 987–1003 Stables GI, Morley WN (1993) Hutchinson-Gilford syndrome. J R Soc Med 87:242–243 Stern R, Frost GI, Shuster S, Shuster V, Hall J, Wong T, Gakunga P, Csoka T (1998) Hyaluronic acid and skin. Cosmetics Toiletries 113:43–48 Sweeney KJ, Weiss AS (1992) Hyaluronic acid in progeria and the aged phenotype? Gerontology 38:139–152 Tanabe M, Goto M (2001) Elevation of serum hyaluronan level in Werner’s syndrome. Gerontology 47:77–81 Tokunaga M, Futami T, Wakamatsu E, Endo M, Yosizawa Z (1975) Werner’s syndrome as “hyaluronuria”. Clin Chim Acta 62:89–96 Tokunaga M, Wakamatsu E, Sato K, Satake S, Aoyama K, Saito K, Sugawara M, Yosizawa Z (1978) Hyaluronuria in a case of progeria. J Am Geriatr Soc XXVI:296–302 Toole BP, Wight TN, Tammi MI (2002) Hyaluronan-cell interactions in cancer and vascular disease. J Biol Chem 277:4593– 4596 Toyoda H, Motoki K, Tanikawa M, Shinomiya K, Akiyama H, Imanari T (1991) Determination of human urinary hyaluronic acid, chondroitin sulphate and dermatan sulphate as their unsaturated disaccharides by high-performance liquid chromatography. J Chromatogr 565:141–148 West DC, Hampson IN, Arnold F, Kumar S (1985) Angiogenesis induced by degradation products of hyaluronic acid. Science 228:11324–11326 Wisuthsarewong W, Viravan S (1999) Hutchinson-Gilford progeria syndrome. J Med Assoc Thai 82:96–102 Yu CE, Oshima J, Fu YH, Wijsman EM, Hisama F, Alisch R, Matthews S, Nakura J, Miki T, Ouais S, Martin GM, Mulligan J, Schellenberg GD (1996) Positional cloning of the Werner’s syndrome gene. Science 272:258–262 Yu QX, Zeng LH (1991) Progeria: report of a case and review of the literature. Pathol Med 20:86–88 Zebrower M, Kieras FJ, Brown WT (1986a) Urinary hyaluronic acid elevation in Hutchinson-Gilford progeria syndrome. Mech Ageing Dev 35:39–46 Zebrower ME, Kieras FJ, Brown WT (1986b) Analysis by highperformance liquid chromatography of hyaluronic acid and chondroitin sulfates. Anal Biochem 157:93–99
