cclm.1993.31.8.503

Gassier et al.: HPLC assay of glycosaminoglycans in human blood 503

Eur. J. Clin. Chem. Clin. Biochem.Vol.31, 1993, pp. 503-511© 1993 Walter de Gruyter & Co.

Berlin · New York

A High Performance Liquid Chromatography Methodfor the Determination of Glycosaminoglycans in Human Blood

By N. Gassier1, C. Rei ner*, N. Janzen\ Heike K hnen2 and K. Kleesiek2

1 Institut ar Klinische Chemie IIder Medizinischen Hochschule Hannover, Zentrallabor im Ostsladtkrankenhaus,Hannover

2 Institut fur Laboratoriums- und Transfusionsinedizin Herzzentrum Nordrhein-Westfalen, Bad Oeynhausen,Germany

(Received February 9/May 6, 1993)

Summary: A method is described for the determination of plasma and serum glycosaminoglycans, which canbe used in any laboratory equipped with an HPLC system. It is based on the sequential application ofchondroitinases AC and ABC and separation of the resulting disaccharides by high-performance liquidchromatography. All reagents are commercially available. This simple and rapid separation yields an accuratequantification and an exact distribution pattern. The determination of glycosaminoglycan disaccharides islinear between 7 and 7000 μιηοΐ/ΐ with coefficients of variation between 3.0 and 7.7% for serum and between2 and 14% for plasma. The recovery of the assay ranged from 93 to 106% for different concentrations ofglycosaminoglycan disaccharides. This HPLC method may therefore be considered as a candidate referencemethod.

Introduction Materials and Methods

Glycosaminoglycans, which are widely distributed in Reagentshuman tissues and fluids (1), are intimately related to Chondroitinase AC (EC 4.2.2.5)1) of Anhrobacter aurescensproteoglycan metabolism. In plasma, glycosamino- Ο01 Ε88301) and chondroitinase ABC (EC 4.2.2.4) of Proteus. , . - /. , i ι vulgar is (lot E88301) were purchased from Seikagaku Kogyo

glycans have a wide range of molecular masses, but (T(fky0j Japan) a.Mannosidase (EC 3.2.1.24) of Canavalia en-a large proportion have relative molecular masses of siformis Got 48F-9545) was from Sigma (St. Louis, USA) andless than 5000 (2). Various laboratories have demon- PaP*in <E£ 3.4.22.2) from Carica papayat (lot 12385425-75/May

x ^ , Al ~" - . 1 92) from Boehrmger Mannheim (Mannheim, Germany),strated the presence of glycosaminoglycans in humanplasma or serum (2-5). The results indicate that a Calibrator disaccharidesgreat quantitative variability may exist among differ- 2-acetamido-2-deoxy-3-O-( -jD-gluco-4-enepyranosyluronicent methods for determination of blood glycosami- acid)-D-ga!actose (Δ DiOS),noglycans. 2-acetamido-2-deoxy-3-O-( -D-gluco-4-enepyranosyluronic_ , j t ι - ι j. acid)-4-O-sulpho-Z>-galactose ( Δ Di4S),Recently we measured the glycosaminoglycan disac-charides from human blood donors, using two differ- 2-acetamido-2-deoxy-3-O-( -Z)-gluco;4-enepyranosyluronic

, , ' Γ . . c . . . . acid)-6-O-sulpho-Z>-galactose (Δ Di6S) andent methods for isolation of glycosaminoglycans inplasma and serum. The Chromatographie conditionsare identical for both methods of glycosaminoglycanisolation. l) Enzymes:

Significant sex differences in glycosaminog^can di- §£!Α£ igc^gsaccharide composition were found in the present oc-Mannosidase (EC 3.2.1.24)Study. Papain (EC 3.4.22.2)

Eur. J. Clin. Chem. Clin. Biochem. / Vol. 31,1993 / No. 8

504 Gassier et al.: HPLC assay of glycosaminoglycans in human blood

2-acetamido-2-deoxy-3-O-( -JD-gluco-4-enepyranosyl uronicacid)-£-glucose (Δ DiHA)

were available from Seikagaku Kogyo (Tokyo, Japan).

DEAE-Sephacel was from Pharmacia (Freiburg, Germany). Allother chemicals were of analytical grade (E. Merck, Darmstadt,Germany).

HPLC equ ipment

Separation and measurements were performed with a Jascoflow pump Model 880-PU (Techlab, Erkerode, Germany), aspectrophotometer SP-4 (Gynkotek, Gering, Germany), an au-tosampler WISP™ Model 710B (Millipore, K nigstein, Ger-many) and a Philips CM 9043 AT Computer (Philips, Hamburg,Germany) with an Axiom 717 chroraatography data system(LDC Analytical, Hasselroth, Germany). A Nucleosil SB 5 an-ion exhange column (4.6 χ 125 mm, 5 μιη particles) connectedwith a Hypersil ODS RP-C 18 reversed phase column (4.6χ 125 mm, 5 μπι particles) was used to separate sulphatedglycosaminoglycan disaccharides, and a Hypersil APS weakanion exchange column (4.6 χ 250 mm, 5 μπι particles) con-nected with a Hypersil ODS RP-C18 column (4.6 χ 15 mm,5 μιη particles) was used to separate non-sulphated glycosa-minoglycan disaccharides. Columns were from Shandon/VDS(Optilab Chromat. Technic, Techlab, Erkerode, Germany). De-tection was by UV absorbance at 232 nm.

Chromatographie conditions

A filtered (0.4 μπι, Millipore, Neu-Isenburg, Germany) andhelium-saturated mobile phase of 0.22 mol/1 sodium chloridewas used for separation of sulphated disaccharides. The flowrate was 0.8 ml/min. Non-sulphated components were separatedwith the same flow rate by the use of a filtered and oxygen-poor mobile phase of 2.5 mmol/1 disodium hydrogen phosphate,pH 3.0 containing 15 mmol/1 sodium chloride.

Disaccharides were injected in quantities of 0.07 to 70 nmoldissolved in ΙΟ μΙ water (7 — 7000 μπιοΐ/ΐ). Solutions of com-mercially available calibrators were used for quantification (ex-ternal calibrator method).

Sample preparat ion

Plasma samples were obtained from about 10 ml of whole bloodwhich were collected in a Monovette® (Sarstedt, Numbrecht,Germany) tube containing 0.05 mmol/1 EDTA, then centrifugedat 3000 min'1 (2000g) for 10 minutes at 4°C.

Serum samples were collected in Monovette® (Sarstedt, Num-brecht, Germany) tubes containing Kaolin. After clotting for!/2 to 1 hour at room temperature, samples wer centrifugedunder the same conditions. Specimens were stored at 4 °C foranalysis within one day. At -20 °C specimens could be storedfor several months.

Serum supernatants were digested for 16 hours at 60 °C with300 μΐ papain solution (3 MU/1 cysteine buffer) in 10 ml glasstubes containing 4.5 ml cysteine buffer (0.011 mol/1 cysteine,0.045 mol/1 disodium EDTA, 0.045 mol/1 potassium dihydrogenphosphate, 0.067 mol/1 dipotassium hydrogen phosphate, pH5.8). Papain (150 μΐ containing 3 MU/1 buffer) and cysteine(0.05 raol) were then added and digestion continued for afurther 4 h. The cholesterol ester content was diminished byvigorously extracting each sample with 3.0 ml chloroform.

Peptides were precipitated by storage overnight after the ad-dition of 2 ml trichloroacetic acid (0.61 mol/1). The solutionswere centrifuged (5000 min"1,20 min at 4 °C) and the recoveredsupernatants were washed three times with 5ml w-butanol.Remaining aqueous phases were evaporated for 5 minutes toremove the n-butanol. One millilitre equilibration buffer (0.15mol/1 lithium chloride in 0.02 mol/1 Tris/HCl, pH 8.6) wasadded and the solution applied to columns (4.0 χ 1.0 cm I. D.)packed with DEAE-Sephacel. After washing two times with5 ml equilibration buffer, glycosaminoglycans were eluted with8.0 ml of 2.0 mol/1 lithium chloride containing 0.02 mol/1 Tris/HC1, pH 8.6. The eluate was pooled, evaporated to near dry-ness, and glycosaminoglycans were precipitated overnight afteraddition of 5 μΐ of saturated sodium acetate solution and 12 mlethanol at 4 °C. The precipitates were centrifuged (5000 min"1,20 min at 4°C) and dryed. After dissolving in 1.0 ml doubledistilled water, the samples were ready for enzymatic digestion.

Plasma samples (5.0 ml) were applied to columns (4.0 χ 1.0cm I. D.) packed with DEAE-Sephacel and washed with 10 mlof a solution (0.15 mol/1 sodium chloride in 0.02 mol/1 Tris/HC1, pH 8.6). The glycosaminoglycans were eluted with 10 mlof 2.0 mol/1 lithium chloride containing 0.02 mol/1 Tris/HCl,pH 8.6, desalted and concentrated to 1.0 ml by ultrafiltration(membrane 1000K, 0.27 ml/min). Buffer (1.0 ml, 0.1 mol/1trisodium citrate, pH 4.5) and ΙΟ μΙ α-mannosidase (450 U/lbuffer) were added to the concentrates, which were digested for18 hours at 37 °C. The solutions were centrifuged (17 000 min"1,30 min at 8 °C) and the supernates again applied to the sameDEAE-Sephacel columns. After washing the columns with 2.5ml 0.1 mol/l sodium chloride in 0.02 mol/1 Tris/HCl, pH 8.6,the glycosaminoglycans were eluted with 2.0 ml of 2.0 mol/1lithium chloride containing 0.02 mol/1 Tris/HCl, pH 8.6. Theeluates were desalted by ultrafiltration (membrane 1000K, 0.16ml/min) and concentrated to 600—700 μΐ. After adjusting thevolume of each eluate to 1.0 ml by the addition of doubledistilled water, the samples were ready for enzymatic digestion.

The aqueous solutions of glycosaminoglycans were analysed bythe uronic acid method of Kosakai & Yosizawa (6), and by thedirect photometric method of Farndale (7) for sulphated gly-cosaminoglycans, using 1,9-dimethylrnethylene blue.

The method used for sequential digestion of glycosaminogly-cans with chondroitinase AC and chondroitinase ABC has beendescribed by Gurr et al. (8). Each enzymatic degradation stepwas followed by separation of products from the undigestedglycosaminoglycans by precipating the undigested polysaccha-rides with ethanol. The digested products, disaccharides ofhyaluronan, non-sulphated chondroitin, chondfoitin sulphate,non-sulphated dermatan and dermatan sulphate, were concen-trated to dryness to rotary evaporation and finally dissolved in200 μΐ double distilled water. Ten microlitres of sample wereinjected for the high performance liquid Chromatographie assay.

Specimens

Blood was obtained in the morning from 30 apparently healthyvolunteers and 78 healthy blood donors (53 men and 55 female,aged 18 to 60 years). All were in good nutritional state; nonetook drugs.

Calibration

Solutions of pure disaccharides were used for calibration (ex-ternal standard method). Losses of sample preparation werequantified with glycosaminoglycan Cpolysaccharide calibrators(internal standard method).

All analyses were performed in duplicate.

Eur. J. Clin. Chem. din. Biochem. / Vol. 31,1993 / No. 8

Gassier et al.: HPLC assay of glycosaminogJycans in human blood 505

Results and Discussion

High performance liquid chromatography ofthe disaccharides

Effect of ion exchanger and ionic strength on the elutionpatlern

Nucleosil SB is a potent ion exchanger bonded tosilica gel, with a functional quaternary ammoniumgroup. Aqueous sodium chloride solutions were usedto separate a calibrator solution of the pure sulphatedglycosaminoglycan disaccharides. Good separation ofsulphated disaccharides was obtained in serum sam-ples, but in plasma samples no satisfactory separationwas achieved. Further separation on a Hypersil ODSCIS reversed phase column allowed separation ofplasma components and of the disaccharides chon-droitin-6-sulphate, chondroitin-4-sulphate, dermatan-6-sulphate, and dermatan-4-sulphate. The best resultswere obtained with 0.22 mol/1 sodium chloride solu-tions and a flow rate of 0.8 ml/min (fig. 1).

-0.020

0.020-

0.01 δ-

Ι 0.010-

0.005-

0.000

Ι Ι Π Ι Π Γ I I4 || 12 20 28 36 Λ 44

-0.015

0.010

0.005

ι ι ι ι τ ι τ ι ι ι ι8 16 24 32 40

Retention time [min]48

Fig. 1. HPLC chromatograms of glycosaminoglycan disaccha-rides in human plasma and serum (same chromatogramsbut shorter retention times).Sulphated glycosaminoglycans in front: chondroitin-4-sulphate (II) and chondroitin-6-sulphate (I) disaccha-rides.Non-sulphated glycosaminoglycans (back): hyaluronan(IV) and chondroitin (III) disaccharides.

Hypersil APS is a weak potent ion exchanger basedon a primary ammonium salt. Buffered sodium phos-phate solutions (pH 4.5 to 6.0) were used in prelimi-nary experiments to separate a calibrator solution ofthe pure non-sulphated glycosaminoglycan disaccha-rides and serum samples. With an additional HypersilODS CIS reversed-phase column, plasma sampleswere separated satisfactorily. However, plasma solu-tions showed a double peak for the two anomericforms of the hyaluronan disaccharide, both in pure

calibrator solutions and in plasma solutions (for fur-ther details see Hjerpe et al. 1982). Variation of ionicstrength, concentration of hydrogen ions and addi-tionally salt concentrations of the eluent were studied.Chromatography was finally performed with a solu-tion containing disodium hydrogen phosphate (2.5mmol/1, pH 3.0) and sodium chloride (15 mmol/1)with a flow rate of 0.8 ml/min (fig. 1).

Linearity and recovery

In order to determine the range of linearity, chro-matography was performed using solutions contain-ing mixtures of the pure disaccharides in concentra-tions between 7 and 7000 μπιοΐ/ΐ (0.07 and 70 nmolper injection). A linear increase in the signal accom-panied the increase in sample concentration for theglycosaminoglycan disaccharides over the entire range(fig. 2). Linear regression analyses of the results gave

y = —0.102 + 0.934x with a correlation coefficientof 0.999 for 2-acetamido-2-deoxy-3-( -£>-gluco-4-enepyranosyluronic acid)-4-O-sulpho-Z)-galactose(Δ Di4S);

y = 0.958 + 0.924x with a correlation coefficientof 0.999 for 2-acetamido-2-deoxy-3-( -D-gluco-4-enepyranosyluronic acid)-6-O-sulpho-Z)-galactose(Δ Di6S);

y = 1.138 + 1.054x with a correlation coefficient of0.999 for 2-acetamido-2-deoxy-3-( -D-gluco-4-ene-pyranosyluronic acid)-D-galactose (Δ DiOS); and

y = 1.618 + 1.127x with a correlation coefficient of0.999 for 2-acetamido-2-deoxy-3-( -D-gluco-4-ene-pyranosyluronic acid)-Z)-glucose (Δ DiHA).

10000

οa? 1000

•S 100JS3sα>ιCOCO

10

10 100 1000Disaccharide injected (μηποΙ/Ι)

10000

Fig. 2. Linearity of detector response: peak area vs. sampleconcentration.Hyaluronan (o), chondroitin (o), chondroitin-4-sul-phate (x), chondroitin-6-sulphate (n) disaccharide.



Tab. 1. Recovery of glycosarainoglycan disaccharides in different concentrations.

Glycosaminoglycan disaccharide

2-Acetamido-2-deoxy-3-O-( -/)-gluco-4-enepyranosyluronic acid)-£>-galactose (Δ DiOS)

Expectedconcentrations[μιηοΐ/ΐ]

107170229352414436

Recoveredconcentration[μπιοΐ/ΐ] , f

166243343393434

Relativedifference[%]

-2.35+6.11-2.84-5.07-0.46

mean: —0.92

2-Acetamido-2-deoxy-3-O-( -£-gluco-4-enepyranosyluronic acid)-D-glucose (Δ DiHA)

66.6203342411475

__209331395460

+ 2.96-3.22-3.89-3.16

mean: —1.83

2-Acetamido-2-deoxy-3-O-( -D-gluco-4-enepyranosyluronic acid)-6-O-sulpho-Z)-galactose (Δ Di6S)

18.629.648.289.1

108

28.044.990.2

109

-5.41-6.85+ 1.24+0.93

mean: -2.52

2-Acetamido-2-deoxy-3-O-( -D-gluco-4-enepyranosyluronic acid)-4-O-sulpho-Z)-galactose (Δ Di4S)

18.627.248.975.7

103213

26.445.474.6

103211

-2.94-7,16-1.45+0.00-1.13

mean: -2.54

To investigate the recovery of the method, variousconcentrations of pure sulphated and non-sulphateddisaccharides were added to blood samples. Recoveryranged

from 92.8 to 100% for Δ Di4S,

from 93.1 to 101% for Δ Di6S,

from 96.1 to 103% for Δ DiHA and

from 94.9 to 106% for Δ DiOS (tab. 1, fig. 3).

Calibrator added

0.04-

0.02-

0.00-

^τ \ —

η, A*lfi|, ΐ η ι A

~l \ Τ 1 p-

+53%

+36%+23%+15% ι+ 5% l

0 4 8 12 16 20Retention time [min]

Fig. 3. Recovery of chondroitin disaccharide. Chromatogramsof increasing quantities of added calibrator are super-imposed.

For comparison, glucuronic acid added to blood sam-ples showed a recovery of 92.3% with the assay ofKosaki & Yosizawa, while added polymeric chondro-itin sulphate showed a recovery of 89.5% with theassay of Farndale et al. These findings imply that thelosses are accounted for in the preparation steps ofplasma/serum and not in the enzymatic digestionsteps.

The sensitivity of the method when defined as a signalto noise ratio > 3 is at least 3.0 μηιοΐ/ί serum orplasma (0.1 nmol per injection). This value is in ac-cordance with other methods (14, 15).

Precision

The total imprecision of the method was determinedfor pure calibrators, and for plasma and serum sam-ples. Coefficients of variation ranging from 1.3 to2.3% for non-sulphated calibrators (ADiHA andΔ DiOS) and from 1.5 to 2.1% for sulphated calibra-tors (Δ Di4S and Δ Di6S) were calculated within aseries of 18 and 20 chromatograms, respectively (con-



centrations between 67 and 238 μιηοΐ/ΐ; data notshown). For plasma/serum samples the coefficients ofvariation did not exceed 5.4% for sulphated glycos-aminoglycan disaccharides (n = 14, resp. 20; tab. 2).

For determination of the inter-assay coefficient ofvariation replicate chromatograms were run on 7 or10 consecutive days (tab. 3). The coefficients of var-iation ranged between 1.1 and 4.7% for non-sul-phated calibrators, up to 6.8% for sulphated calibra-tors, up to 7.6% for non-sulphated glycosaminogly-can disaccharides in plasma/serum samples and be-tween 7.1 and 17% for sulphated glycosaminoglycandisaccharides in blood samples. No recalibration wascarried out during these experiments, but the columnswere changed after approximately 150 runs. The highcoefficients of variation within a series for Δ DiHAand from day to day for Δ Di6S can be explained bythe low concentrations of these disaccharides inplasma samples. Apart from these exceptions, coef-ficients of variation do not exceed 7.9%.

The total imprecision of the method, within-run im-prescision and day-to-day variation are equal to orless than those reported for other HPLC assays inplasma/serum (15).

Determination of glycosaminoglycan distri-bution pattern

For the determination of glycosaminoglycan distri-bution patterns a technique has been described inwhich hyaluronan, chondroitin, chondroitin sulphate,dermatan and finally dermatan sulphate are digestedin sequential steps using papain/mannosidase, chon-droitinase AC and chondroitinase ABC, followed byhigh performance liquid Chromatographie analysis ofthe metabolites. Firstly, interfering plasma/serumcomponents must be removed by column chromato-graphy. Up to 30% of serum proteins were bound onDEAE-Sephacel, but more than 98% of plasma pro-teins were eluted together with plasma glycosamino-glycans. Samples were then desalted by ultrafiltrationor precipitated with ethanol. This step was necessaryfor accurate and quantitative determination of gly-cosaminoglycans. Four or five working days wererequired to determine the glycosaminoglycans andtheir distribution pattern. The proteolysis with pa-pain/mannosidase was time-consuming but necessary.After this enzymatic digestion, all glycosaminoglycanswere detected in the supernates, so that blood samplescould easily be processed in a few days and thenanalysed overnight.

Tab. 2. Reproducibility of high-performance liquid Chromatographie analysis in plasma and serum. Coefficients of variationwithin a series.

Glycosaminoglycan disaccharides

2-Acetamido-2-deoxy-3-O-( -/)-gluco-4-enepyranosyl-uronic acid)-Z)-glucose (Δ DiHA)

2-Acetamido-2-deoxy-3-O-( -/)-gluco-4-enepyranosyl-uronic acid)-Z)-galactose (Δ DiOS)

2-Acetamido-2-deoxy-3-O-( -/)-gluco-4-enepyranosyl-uronic acid)-4-O-sulpho-D-galactose (Δ Di4S)

2-Acetamido-2-deoxy-3-O-( -,D-gluco-4-enepyranosyl-uronic acid)-6-O-sulpho-jD-galactose (Δ Di6S)

Material

plasmaserum

plasmaserum

plasmaserum

plasmaserum

Number ofmeasurements

1820

1820

1420

1420

ConcentrationQimol/1]

7.00 ± 0.988.60 ± 0.27

110 +4.059.70 ± 0.29

169 ± 4.437.90 ± 0.30

5.00 + 0.309.32 ± 0.32

Coefficient ofvariation [%]

14.03.56

3.683.03

2.624.02

5.383.62

Tab. 3. Reproducibility of high-performance liquid Chromatographie analysis in plasma and serum. Coefficients of variationbetween the series (from day to day).

Glycosaminoglycan disaccharides

2-Acetamido-2-deoxy-3-O-( -£-gluco-4-enepyranosyl-uronic acid)-D-glucose (Δ DiHA)

2-Acetamido-2^deoxyr3-O-( ^/)-gluco-4-enepyranosyl-uronic acid)-D-galactose (Δ DiOS)

2-Aretarnido-2-deoxyr-3-O-( -Z>-gluco-4-enepyranosyl-uronic acid)-4-O-sulpho-D-galactose (Δ Di4S)

2-Acetamido-2-deoxy-3-O-( -Z)-gluco-4-enepyranosyl·uronic acidJ-o-O-sulpho-^galactose (Δ Di6S)

Material

plasmaserum

plasmaserum

plasmaserum

plasmaserum

Number ofmeasurements

710

710

710

710

Concentration[μιηοΐ/ΐ]

7.00 + 0.488.62 ± 0.61

110 + 7.309.73 ± 0.51

169 + 13.37.90 ± 0.56

5.00 ± 0.859.32 ± 0.59

Coefficient ofvariation [%]

6.867.61

6.645.06

7.877.07

17.06.97

Eur. J. Clin. Chem. Clin. Biochcm. / Vol. 31,1993 / No. 8


The chondroitin and chondroitin-4-sulphate disaccha-rides account for up to 92% of the total glycosami-noglycans in serum; only traces of dermatan-6-sul-

phate and dermatan were detectable (tab. 4). Differentmethods (2,3,5,10) yield similar distribution patternsof disaccharides.

Tab. 4. Glycosaminoglycan disaccharide concentrations, analysed in serum

a) males (n = 53)

Glycosaminoglycan disaccharides [μιηοΐ/ΐ]

Patients

123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748Af\4950C 4515253

χs

Hyaluronan

0.390.710.000.080.171.200.500.1901280.130.000.000.270.510.000.520.000.240.000.000.390.710.000.080.170.500.190.280.130.170.660.550.690.000.060.280.060.000.170.620.000.270.150.050.330.020.250.100.260.490.380.370.39

0.2630.251

Chondroitin

5.144.845.631.484.643.057.185.307.093.664.384.253.631.001.988.085.515.203.538.525.144.845.631.484.647.185.307.093.669.196.402.946.464.723.925.714.043.443.565.061.389.972.384.395.667.08

11.014.059.673.001.89

11.573.74

5.102.39

Chondroitin-4-sulphate

0.681.154.781.302.465.090.304.325.072.196.324.554.771.051.990.880.685.111.254.220.683.154.781.302.460.304.325.072.191.341.103.014.962.520.602.012.812.490.800.092.026.161.960.560.772.876.132.391.370.622.146.113.69

2.621.84


0.081.160.200.210.000.701.300.250.490.000.130.180.230.070.070.001.000.170.140.340.081.160.200.210.091.300.250.490.000.200.660.120.050.000.181.500.200.000.050.540.240.360.000.000.130.090.550.000.100.020.000.300.00

0.2980.380

Dermatan

0.100.110.000.000.000.190.380.000.370.000.000.000.000.000.000.090.110.000.000.000.100.110.000.000.000.380.000.370.000.110.270.290.000.000.200.220.000.000.220.190.000.000.000.220.050.000.260.000.150.150.000.540.00

0.0980.134

Dermatan-4-sulphate

0.510.611.611.150.830.520.941.901.260.630.000.000.000.000.000.820.891.271.420.860.510.611.611.150.830.941.901.260.630.700.720.721.281.391.070.861.340.730.670.751.331.180.891.060.701.391.510.790.860.901.101.540.67

0.9210.453

Dermatan-6-sulphate

0.000.000.020.220.000.000.090.000.000.000.000.000.000.000.000.000.000.000.000.020.000.000.020.220.000.090.000.000.000.000.000.010.000.000.000.000.010.000.000.070.000.000.000.000.000.180.210.000.010.000.000.120.00

0.0240.058

Sum

6.908.58

12.244.448.10

10.7510.6911.9614.566.61

10.838.988.902.634.04

10.398.19

11.996.34

13.966.908.58

12.244.448.19

10.6911.9614.566.61

11.719.817.64

13.448.636.03

10.588.466.665.477.324.97

17.945.386.287.64

11.6319.927.33

12.425.185.51

20.558.49

9.33



Our data showed significant differences (p < 0.005,Wilcoxon test for pair differences) between the mainglycosaminoglycan disaccharide values in men and

women. There was no detectable age-dependency, butsubjects younger than 18 and older than 60 were notincluded in the study.

b) females (n = 55)

Glycosaminoglycan disaccharides [μιηοΐ/ΐ]

Patients

123456789

10111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455

Xs

Hyaluronan

0.460.440.000.081.031.480.470.040.070.800.000.100.060.000.000.240.400.200.000.520.460.440.000.081.031.480.470.040.070.800.400.360.180.000.000.000.550.170.070.350.380.780.000.151.050.480.250.070.000.610.260.360.270.000.60

0.3380.364

Chondroitin

12.438.46

12.837.157.73

12.2610.1413.539.809.844.513.734.581.423.979.68

10.2312.5010.287.43

12.438.46

12.837.157.73

12.2610.1413.539.809.84

11.2611.3213.265.97

11.3710.3410.1114.6510.507.90

14.0510.753.406.309.03

14.348.108.68

12.2411.2514.645.908.326.269.80

9.523.09

Chondroitin-4-suLphate

3.421.78

11.397.852.562.651.738.619.423.107.352.673.901.355.022.422.10

12.7011.082.643.421.78

11.397.852.562.651.738.619.423.102.563.26

11.797.244.663.785.22

14.938.781.714.214.433.953.123.194.302.729.98

10.682.403.651.83

10.208.234.84

5.423.53


0.730.420.100.000.040.410.490.040.260.100.240.020.030.060.100.360.400.000.240.190.730.420.100.000.040.410.490.040.260.100.290.510.160.090.160.480.770.000.290.020.280.550.000.000.160.480.340.000.070.100.610.260.280.090.10

0.2350.211

Dermatan

0.000.000.000.000.160.000.000.000.130.000.000.000.000.000.000.000.000.000.090.000.000.000.000.000.160.000.000.000.130.000.000.000.000.000.000.000.000.050.000.000.000.000.000.000.000.000.000.000.370.000.000.000.160.000.00

0.0230.065

Derma tan -4-sulphate

0.250.001.291.121.380.000.381.060.861.310.000.000.000.000.000.000.001.381.330.850.250.001.291.121.380.000.381.060.861.310.170.431.301.061.170.000.281.331.161.190.000.411.310.951.360.360.271.221.281.220.000.001.101.370.95

0.7050.552

Dermatan-6-sulphate

0.000.000.000.100.000.000.000.000.000.000.000.000.000.000.000.000.000.000.000.000.000.000.000.100.000.000.000.000.000.000.000.000.000.000.000.000.000.050.090.100.000.000.000.000.080.000.000.000.100.080.000.000.160.170.21

0.0230.049

Sum

17.2911.1025.6116.3012.9016.8013.2123.2820.5415.1512.106.528.572.839.09

12.7013.1326.7823.0211.6317.2911.1025.6116.3012.9016.8013.2123.2820.5415.1514.6815.8826.6914.3617.3614.6016.9331.1820.8911.2718.9216.928.66

10.5214.8719.9611.6819.9524.7415.6619.168.35

20.4916.1216.50

16.31

Eur. J. Clin. Chera. Clin. Biochem. / Vol. 31,1993 / No. 8

510 Gassier et aL: HPLC assay of glycosaminoglycans in human blood

Dermatan-4-sulphate disaccharides were found inconcentrations up to 0.81 μπιοΐ/ΐ serum. Dupuoy etal. (16) determined 0 to 2 mg/1 for the reference valueof polymeric dermatan sulphate in human plasma.The principle of the assay is the measurement ofresidual amidolytic thrombin activity after incubationwith heparin cofactor II in defibrinated plasma; der-matan sulphate accelerates thrombin inhibition byheparin cofactor.For hyaluronan the reference value has been reportedbetween 10 and 100 μg/l serum (17,18). In the presentstudy, the mean serum concentrations of hyaluronandisaccharides were 0.26 μηιοΐ/ΐ (men) and 0.34 μπιοΐ/ΐ(women).When these results are converted from μηιοί disac-charides into μ§ polymeric glycosaminoglycans, theyare in accordance with the reference values.

Serum was preferred, although plasma could also beused (19, 20). In practice, the isolation of glycosami-noglycans from serum was shorter and less demand-ing. With both methods the isolation of glycosami-noglycans was quantitative.

Total glycosaminoglycan concentrations show goodagreement with the results of Larking (2) and Strapanset al. (5).

Conclusions

The presented method quantifies glycosaminoglycansin human serum or plasma as disaccharides. Applyingthis high specific and sensitivefmethod, quantitativeresults can be obtained with good precision, accuracyand performance. This HPLC method can thereforebe considered as a candidate reference method. Fur-thermore, it allows the detection of the distributionpatterns of the glycosaminoglycans, thereby providinga useful tool for elucidating aspects of the biosynthesisand degradation of proteoglycans.

Changes of plasma/serum glycosaminoglycan concen-trations have been demonstrated in physiological andpathological conditions (11 — 13). Therefore in clinicalpractice the method could lead to a better understand-ing of glycosaminoglycan involvement in disease.

AcknowledgementWe thank Miss M. Arlart for her skilful technical assistance.This work contains parts of the biochemistry graduation thesesof N. Janzen and C. Rei ner.

References1. Poulson, J. H. (1975) Urine and tissue glycosaminoglycans

and their interactions. Dan. Med. Bull. 33, 75-96.2. Larking, P. W. (1987) Quantitative gas Chromatographie

measurement of glycosaminoglycan hexosamines in urineand plasma. J. Chromatogr. 420, 231 —239.

3. Kodama, C., Ototani, N., Isemura, M., Aikawa, J. & Yos-izawa, Z. (1984) High-performance liquid chromatographyof pyridylamino derivatives of unsaturated disaccharidesproduced from chondroitin sulfated isomers by chondro-itinases. J. Biochem. (Tokyo) 96, 1283-1287.

4. Saito, H., Yamagata, T. & Susuzi, S. (1968) Enzymaticmethods for the determination of small quantites of iso-meric chondroitin sulfates. J. Biol. Chem. 243,1536-1542.

5. Staprans, I. & Felts, J. M. (1985) Isolation and character-ization of glycosaminoglycans in human plasma. J. Clin.Inv. Inc. 7(5, 1984-1991.

6. Kosakai, M. & Yosizawa, Z. (1979) A Partial Modificationof the Carbazole Method of Bitter and Muir for Quanti-tation of Hexuronic acids. Anal. Biochem. 93, 295-298.

7. Farndale, R. W., Sayers, C. A. & Barrett, A. J. (1986)Improved quantitation and discrimination of sulfated gly-cosaminoglycans by use of dimethylmethylene blue.Biochim. Biophys. Acta 883, 173-177.

8. Gurr, E., Pallasch, G., Tunn, S., Tamm, C. & Delbr ck, A.(1985) High performance liquid Chromatographie assay ofdisaccharides and oligosaccharides produced by the diges-tion of glycosaminoglycans with chondroitin sulphate ly-ases. J. Clin. Chem. Clin. Biochem. 23, 77-87.

9. Hjerpe, A., Antonopoulos, C. A. & Engfeldt, B. (1982)Determination of hyaluronic acid using high-performanceliquid chromatography of chondroitinase digests. J. Chro-matogr. 245, 365-368.

10. Larking, P. W. (1989) Total glycosaminoglycans in theplasma of adults: effects of age and gender, and relationshipto plasma lipids: A preliminary study. Biochem. Med. Me-tab. Biol. 42, 192-197.

11. Friman, C., Juvani, M. & Skrifvars, B. (1977) Acid glycos-aminoglycans in plasma It. Findings in rheumatoid arthri-tis. Scand. J. Rheumatol. 6, 177-182.

12. Friman, C., Nordstr m, D. & Eronen, I. (1987) Plasmaglycosaminoglycans in systemic lupus erythematodes. J.Rheumatol. 14, 1132 -1134.

13. Ferlazzo, A. M., Vinci, R., Muraca, U, Ferlazzo, A. &Calatroni, A. (1991) Suggestions arising from plasma GAGlevel changes in human and animal subjects. Ital. Biochem.Soc. Trans. 2, 242.

14. Gysen, P. & Franchimont, P. (1984) Radioimmunoassay ofproteoglycans. J. Immunoassay 5, 221 — 243.

15. Silvestro, L., Viano, I., Torri, G., Da Col, R. & Baiocchi,C. (1992) High-performance liquid chromatographic-massspectrometric analysis of oligosaccharides from enzymaticdigestion of glycosaminoglycans. J. Chromatogr. 591,225 —232.

, 16. Dupuoy, D., Sie, P., Dol, F. & Boneau, B. (1988) A simplemethod to measure derrnatan sulfate at sub-microgramconcentrations in plasma. Thromb. Haemost. 60, 236—239.

Bur. J. Clin. Chem. Clin. Biochem. / Vol. 31,1993 / No. 8


17. Engstroem, L. A., Laurent, U. B., Lilja, K. & Laurent, T. 19. Janzen, N. (1992) Glykosaminoglykan-Nachweis in verschie-C. (1985) Concentration of sodium hyaluronate in serum. denen Körperflüssigkeiten des Menschen — Pathobiochemi-Scand. J. Clin. Lab. Invest. 45, 497 — 504. sehe Bedeutung. Graduation Biochemistry, University of

18. Chichibu, K., Matsuura, T., Shichijb, S. & Yokoyama, M. Hannover.M. (1989) Assay of serum hyaluronic acid in clinical ap- 20. Reißner, C. (1992) Nachweis von Glykosaminoglykanen inplication. Clin. Chim. Acta /£/, 317—324. humanem Blut. Graduation Biochemistry, University of

Hannover.

Dr. N. GassierCentral LaboratoryCaritas Trägergesellschaft Trier e. V.Hospital St. ElisabethKoblenzer Straße 91D-55516 WittlichGermany


cclm.1993.31.8.503

Documents