Scandinavian Journal of Clinical and Laboratory Investigation
ISSN: 0036-5513 (Print) 1502-7686 (Online) Journal homepage: http://www.tandfonline.com/loi/iclb20
Quantitation of urinary hydroxypyridinium crosslinks from collagen by high-performance liquid chromatography G. Kollerup, G. Thamsborg, H. Bhatia & O. Helmer Sørensen To cite this article: G. Kollerup, G. Thamsborg, H. Bhatia & O. Helmer Sørensen (1992) Quantitation of urinary hydroxypyridinium cross-links from collagen by high-performance liquid chromatography, Scandinavian Journal of Clinical and Laboratory Investigation, 52:7, 657-662, DOI: 10.3109/00365519209115510 To link to this article: http://dx.doi.org/10.3109/00365519209115510
Published online: 08 Jul 2009.
Submit your article to this journal
Article views: 7
View related articles
Citing articles: 6 View citing articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=iclb20 Download by: [RMIT University Library]
Date: 18 April 2016, At: 19:15
Scand J Clin Lab Invest 1992; 52: 657-662
Quantitation of urinary hydroxypyridinium cross-links from collagen by high-performance liquid chromatography G. KOLLERUP,* G. THAMSBORG,* H. BHATlAt & 0. H E L M E R S Q ) R E N S E N *
Downloaded by [RMIT University Library] at 19:15 18 April 2016
*Department of Medicine and tDepartment of Clinical Chemistry, Sundby Hospital, Copenhagen, Denmark
Kollerup G, Thamsborg G, Bhatia H, Sorensen OH. Quantitation of urinary hydroxypyridinium cross-links from collagen by high performance chromatography. Scand J Clin Lab Invest 1992; 52: 657-662. Pyridinoline and deoxypyridinoline are intermolecular cross-links in mature collagen in bone and cartilage. The urinary excretion of the two compounds correlates well to bone turnover. A fast, sensitive, and accurate isocratic ionpairing reverse-phase high-performance liquid chromatography method for measurement of pyridinoline and deoxypyridinoline in urine has been established. Intra- and inter-assay precision were 5-7% and 12- 14%, respectively. Recovery for pyridinoline was 97.4% and for deoxypyridinoline 94.3%. The detection limit was 0.4 pmol. Pyridino1ine:creatinine and deoxypyridinoline: creatinine ratios in healthy subjects, were 38.8 nmol:mmol and 13.0 nmol: mmol, respectively. Increased values of both cross-links were observed in children, in the age group 20-29 in both sexes, and in post-menopausal women.
Key words: collagen degradation, deoxypyridinoline, HPLC, pyridinoline, urinary crosslinks. Gina Kollerup, MD, Department of Internal Medicine, Sundby Hospital, 2300 Copenhagen S , Denmark.
The 3-hydroxypyridinium crosslinks pyridinoline (Pyr) and deoxypyridinoline (Dpyr) are natural fluorescent intermolecular crosslink amino acids, in mature collagen type I and I1 in bone, cartilage, tendon and to a lesser extent in other connective tissues except skin [l]. Pyr is the major component in all these tissues, and is formed either by a reaction between two bifunctional keto-amino crosslinks [2], or by condensation of hydroxylysine aldehyde with an existing bifunctional ketoamine crosslink [3]. More recently Dpyr has been identified [4]. It is formed from the
reaction of lysine rather than hydroxylysine as the N-donor residue, and appears to be located mainly in bone and dentin [2, 51. Both crosslinks are excreted in the urine, so measurement may provide an estimate of the degradation of mature collagen from bone and cartilage [6]. Bone histomorphometric analysis and radioisotopic measurements of bone resorption indicate that urinary excretion of Pyr and Dpyr are sensitive markers of bone turnover [7, 81. Urinary excretion of Pyr and Dpyr were shown to be abnormally high in Paget’s disease [6, 91, hyperparathyroidism and 657
Downloaded by [RMIT University Library] at 19:15 18 April 2016
658
G. Kollerup et al.
hyperthyroidism IY, 101, in osteoarthritis and rheumatoid arthritis [6, 111. Unlike hydroxyproline, another biochemical marker of collagen degradation, Pyr and Dpyr are free from interference from degradation of newly synthesised collagen molecules, non-collagen proteins, or from dietary collagen intake 1121. The 3-hydroxypyridinium amino-acids have been quantitated in the urine after paper chromatography [ 131, by enzyme-linked immunoassay [14] or by fluorescense detection after high-performance liquid chromatography (HPLC) using a time-consuminggradient system or long column 16, 15, 161. The purpose of this study was to apply a more rapid quantitative analysis of Pyr and Dpyr in the urine, using a isocratic ion-paired reversephase HPLC. Normal levels were determined in healthy children and adults. PATIENT AND METHODS Overnight fasting second void urine samples were collected from healthy adults of both sexes (62 women and 27 men) and 8 healthy children, without any history of bone and/or articular disease, and no intake of medicine that could affect bone and cartilage metabolism. Preparation of' the pyridinium standard Purified preparations of Pyr and Dpyr were used in each assay as external standards. They were isolated from powdered bone from a hen, that was extracted with guanidine buffer (4M guanidine-HCI, 0,05M Tris-HCI, pH 7.5) at 4" C for 48 h to remove proteoglycans [2], followed by washing with demineralized water and refluxed for 48 h in 1.51 of 3~ HCI. The hydrolysate was evaporated to 150 ml, and 150 ml glacial acetic acid and 600 ml n-butanol were added. The 70 ml brown oil, that settled over 4 days, was separated and dissolved in 200 ml of demineralized water, and mixed again with 1 vol glacial acetic acid and 4 vol nbutanol. This precipitation step was carried out 4 times The oily precipitate was dried under vacuum and dissolved in 5 ml of 10'Yo acetic acid. Both hydroxypyridinium compounds were isolated by elution from a bio-gel (3-10 column ( l O O X 2 , 5 cm) and monitored for fluorescens with an UV-lamp. The central region of the
fluorescent peak were pooled and dried under vacuum, and dissolved in 5 ml 10%0acetic acid. The hydroxypyridinium compounds were further isolated by elution from a bio-gel P-2 (40x2,s cm), and the central regions of the fluorescent peak were pooled. Elution from this column was carried out twice, and the pooled fraction was kept in 0.01M HCI in -80" C until use. The pooled fractions, containing Pyr and Dpyr standards, were equilibrated with standards supplied by Tony Colwell, Medical School, University of Sheffield, England. Sample preparation Second void urine samples were acidified with acetic acid and stored at -20" C until measured. A 250 pl urine sample was hydrolysed with 250 pl 12M hydrochloric acid for 18 h at 107" C in glass tubes to release bound forms of the crosslinks. Pyr and Dpyr were extracted from the cold hydrolysates using cellulose CF1 (Whatman cellulose powder CF-1) partition column [6]. To the cold hydrolysated samples were added 0.5 ml acetic acid, 0.5 ml CFI slurry and 2 . 0 ml butan-1-01 (acetic acid glacial and I-butanol were HPLC grade from Merck), and the cellulose CF1 slurry was prepared from 5% CF1 in a mobil phase of butan-1-01: acetic acid: water, 4:l:l. A column was prepared by 5 ml CF1, washed with 5 ml mobile phase. The hydrolysate/slurry mix was applied to the column, and the hydrolysate tube was washed with 2 ml mobile phase. The column was then washed with 2x11 ml mobile phase, and the washings were discharged. Finally the crosslinks were eluted with 5 ml water, and the aqueous eluate was collected in conical centrifuge tube, concentrated in a vacuum rotor, and stored at -20" C until use. Prior to chromatography samples were redissolved in 250 pl loading buffer made of 10/0 sequanal grade n-heptafluorobuturic acid (HFBA) (Pierce Chemical Co) in lO0mM amoniumcloride pH 3 3 , and centrifugated 5 minutes at 1500 rpm. HPLC assay This HPLC assay is a modification of the method described by Eyre et al. [2] and Black et al. [6]. Pyr and Dpyr were quantitated in the urine extract by isocratic HPLC separation using a short 3.3 cm reversed phase column (Supelco LC-18-DB 3.3 cm x 4.6 mm i.d. pore
Downloaded by [RMIT University Library] at 19:15 18 April 2016
Urinury hydroxypyridinium cross-links from collagen by H P L C size lOOA), on a Jasco HPLC-system, equipped with two pumps, an automatic injector (Jasco 851 AS Intelligent sampler, Fasco International, Tokyo, Japan), and a system controller (802-SC). Fluorescens of the eluted peaks was monitored by a Jasco 821 FP detector with exitation at 295 nm and emmision at 395 nm. The chromatograms were analysed on a computerized unit (Carlo Erba Baseline 810, Dynamic Solution, Millipore, Ventura, CA, USA). A 20 pl sample loop was used for application to the column, which was protected by an C-18 guard catridge (Supelquard LC-WDB, Sudelco Inc., Bellafonte, USA). Flow rate was 1 ml min-I. Solvent A was 0 . 0 1 ~HFBA in 2mM NH&I solution. Solvent B was 7.5% acetonitrile and 25% solvent A. HFBA concentration was adjusted to 0 . 0 1 ~ and pH to 2.26 in both solvents. The column was equilibrated with 88% A and 12% B, and samples were eluted within 5 min. After every 10 samples the column was stripped with 1OO'X B for 5 min and reequilibrated at 12% B for 10 min before the next injection. All samples were analysed at least in duplicate. The results of Pyr and Dpyr were given according to a comparison with known external standards, and expressed as nmol Pyr mmol-' creatinine or nmol Dpry m m o P creatinine. Urinary creatinine concentration was analysed by Jaffe chromogen reaction (Astra, Beckman Instruments, Palo Alto, CA, USA). RESULTS With our system Pyr and Dpyr were separated without any interfering artifacts and eluted at 2.5 min and 3 min, respectively. A typical chromatogram of a standard mixture, consisting of 5.44 pmol of Pyr and 2.68 pmol of Dpyr, and a chromatogram of a normal urine sample, are illustrated in Figure 1. Calibration curves for Pyr in the range of 0-58 pmol and Dpyr in the range of 0-30 pmol are linear, such as shown in Figure 2. In validating our method we have determined the recovery, when increasing amount of Pyr and Dpyr were added to non-hydrolysed urine samples. Recovery of Pyr and Dpyr (mean? SD) were 97,Yf6,2% (n=7) and Y5,4+3,6%, (n=7), respectively, irrespective of the amount of standards added.
659
The intra-assay precision determined by 10 repeated measurements of three different urine samples with increasing amount of Pyr and Dpyr, were 5-7% for both Pyr and Dpyr independent on the amount of Pyr and Dpyr in the urine. The inter-assay precision determined in control non-hydrolysed urine samples, extracted and measured in 13 assays over a period of 3.5 months, were 13.6% for Pyr (mean 692.7 nmol I-' and 12.9% (mean 187.3 nmol I-') for Dpyr. The detection limit of Pyr and Dpyr is approximately 0.4 pmol. The mean values+SD for Pyr:- and Dpyr: crea value in 89 adults (62 women and 27 men, mean agekSD 40.0k15.0) were 38.8k10.8 nmol:mmol (range 16.0-61.7) and 13.0k4.6 nmol: mmol (range 4.5 - 32.3), respectively. The mean values for Pyr:- and Dpyr:crea value in 47 premenopausal (mean a g e f S D 35.3k9.1) and 15 post-menopausal women (aged 62.1 f10.1) and 27 men (aged 35.7?15.0) are shown in
1.0
1.0
2.0
Pyr
3.0
2.'0 3.0 Elution time (rnin)
4.0
4'.0
FIG.I . A typical chromatogram 0 1 the separation of a standard mixture containing pyridinoline and deoxypyridinoline (upper panel). A typical chromatogram of a fractionated hydrolysate of urine from a normal individual (lower panel).
660
G. Kollerup et al.
Figure 3. The mean valueskSD for Pyr:- and Dpyrxrea in 8 children (aged 12.0k4.7) were 181.0583.0 nmol mmol-' and 53.5k23.7 nmol mmol-I, respectively. There was a signi-
ficant increase in both crosslinks in post-menopausal compared to pre-menopausal women (p
ISSN: 0036-5513 (Print) 1502-7686 (Online) Journal homepage: http://www.tandfonline.com/loi/iclb20
Quantitation of urinary hydroxypyridinium crosslinks from collagen by high-performance liquid chromatography G. Kollerup, G. Thamsborg, H. Bhatia & O. Helmer Sørensen To cite this article: G. Kollerup, G. Thamsborg, H. Bhatia & O. Helmer Sørensen (1992) Quantitation of urinary hydroxypyridinium cross-links from collagen by high-performance liquid chromatography, Scandinavian Journal of Clinical and Laboratory Investigation, 52:7, 657-662, DOI: 10.3109/00365519209115510 To link to this article: http://dx.doi.org/10.3109/00365519209115510
Published online: 08 Jul 2009.
Submit your article to this journal
Article views: 7
View related articles
Citing articles: 6 View citing articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=iclb20 Download by: [RMIT University Library]
Date: 18 April 2016, At: 19:15
Scand J Clin Lab Invest 1992; 52: 657-662
Quantitation of urinary hydroxypyridinium cross-links from collagen by high-performance liquid chromatography G. KOLLERUP,* G. THAMSBORG,* H. BHATlAt & 0. H E L M E R S Q ) R E N S E N *
Downloaded by [RMIT University Library] at 19:15 18 April 2016
*Department of Medicine and tDepartment of Clinical Chemistry, Sundby Hospital, Copenhagen, Denmark
Kollerup G, Thamsborg G, Bhatia H, Sorensen OH. Quantitation of urinary hydroxypyridinium cross-links from collagen by high performance chromatography. Scand J Clin Lab Invest 1992; 52: 657-662. Pyridinoline and deoxypyridinoline are intermolecular cross-links in mature collagen in bone and cartilage. The urinary excretion of the two compounds correlates well to bone turnover. A fast, sensitive, and accurate isocratic ionpairing reverse-phase high-performance liquid chromatography method for measurement of pyridinoline and deoxypyridinoline in urine has been established. Intra- and inter-assay precision were 5-7% and 12- 14%, respectively. Recovery for pyridinoline was 97.4% and for deoxypyridinoline 94.3%. The detection limit was 0.4 pmol. Pyridino1ine:creatinine and deoxypyridinoline: creatinine ratios in healthy subjects, were 38.8 nmol:mmol and 13.0 nmol: mmol, respectively. Increased values of both cross-links were observed in children, in the age group 20-29 in both sexes, and in post-menopausal women.
Key words: collagen degradation, deoxypyridinoline, HPLC, pyridinoline, urinary crosslinks. Gina Kollerup, MD, Department of Internal Medicine, Sundby Hospital, 2300 Copenhagen S , Denmark.
The 3-hydroxypyridinium crosslinks pyridinoline (Pyr) and deoxypyridinoline (Dpyr) are natural fluorescent intermolecular crosslink amino acids, in mature collagen type I and I1 in bone, cartilage, tendon and to a lesser extent in other connective tissues except skin [l]. Pyr is the major component in all these tissues, and is formed either by a reaction between two bifunctional keto-amino crosslinks [2], or by condensation of hydroxylysine aldehyde with an existing bifunctional ketoamine crosslink [3]. More recently Dpyr has been identified [4]. It is formed from the
reaction of lysine rather than hydroxylysine as the N-donor residue, and appears to be located mainly in bone and dentin [2, 51. Both crosslinks are excreted in the urine, so measurement may provide an estimate of the degradation of mature collagen from bone and cartilage [6]. Bone histomorphometric analysis and radioisotopic measurements of bone resorption indicate that urinary excretion of Pyr and Dpyr are sensitive markers of bone turnover [7, 81. Urinary excretion of Pyr and Dpyr were shown to be abnormally high in Paget’s disease [6, 91, hyperparathyroidism and 657
Downloaded by [RMIT University Library] at 19:15 18 April 2016
658
G. Kollerup et al.
hyperthyroidism IY, 101, in osteoarthritis and rheumatoid arthritis [6, 111. Unlike hydroxyproline, another biochemical marker of collagen degradation, Pyr and Dpyr are free from interference from degradation of newly synthesised collagen molecules, non-collagen proteins, or from dietary collagen intake 1121. The 3-hydroxypyridinium amino-acids have been quantitated in the urine after paper chromatography [ 131, by enzyme-linked immunoassay [14] or by fluorescense detection after high-performance liquid chromatography (HPLC) using a time-consuminggradient system or long column 16, 15, 161. The purpose of this study was to apply a more rapid quantitative analysis of Pyr and Dpyr in the urine, using a isocratic ion-paired reversephase HPLC. Normal levels were determined in healthy children and adults. PATIENT AND METHODS Overnight fasting second void urine samples were collected from healthy adults of both sexes (62 women and 27 men) and 8 healthy children, without any history of bone and/or articular disease, and no intake of medicine that could affect bone and cartilage metabolism. Preparation of' the pyridinium standard Purified preparations of Pyr and Dpyr were used in each assay as external standards. They were isolated from powdered bone from a hen, that was extracted with guanidine buffer (4M guanidine-HCI, 0,05M Tris-HCI, pH 7.5) at 4" C for 48 h to remove proteoglycans [2], followed by washing with demineralized water and refluxed for 48 h in 1.51 of 3~ HCI. The hydrolysate was evaporated to 150 ml, and 150 ml glacial acetic acid and 600 ml n-butanol were added. The 70 ml brown oil, that settled over 4 days, was separated and dissolved in 200 ml of demineralized water, and mixed again with 1 vol glacial acetic acid and 4 vol nbutanol. This precipitation step was carried out 4 times The oily precipitate was dried under vacuum and dissolved in 5 ml of 10'Yo acetic acid. Both hydroxypyridinium compounds were isolated by elution from a bio-gel (3-10 column ( l O O X 2 , 5 cm) and monitored for fluorescens with an UV-lamp. The central region of the
fluorescent peak were pooled and dried under vacuum, and dissolved in 5 ml 10%0acetic acid. The hydroxypyridinium compounds were further isolated by elution from a bio-gel P-2 (40x2,s cm), and the central regions of the fluorescent peak were pooled. Elution from this column was carried out twice, and the pooled fraction was kept in 0.01M HCI in -80" C until use. The pooled fractions, containing Pyr and Dpyr standards, were equilibrated with standards supplied by Tony Colwell, Medical School, University of Sheffield, England. Sample preparation Second void urine samples were acidified with acetic acid and stored at -20" C until measured. A 250 pl urine sample was hydrolysed with 250 pl 12M hydrochloric acid for 18 h at 107" C in glass tubes to release bound forms of the crosslinks. Pyr and Dpyr were extracted from the cold hydrolysates using cellulose CF1 (Whatman cellulose powder CF-1) partition column [6]. To the cold hydrolysated samples were added 0.5 ml acetic acid, 0.5 ml CFI slurry and 2 . 0 ml butan-1-01 (acetic acid glacial and I-butanol were HPLC grade from Merck), and the cellulose CF1 slurry was prepared from 5% CF1 in a mobil phase of butan-1-01: acetic acid: water, 4:l:l. A column was prepared by 5 ml CF1, washed with 5 ml mobile phase. The hydrolysate/slurry mix was applied to the column, and the hydrolysate tube was washed with 2 ml mobile phase. The column was then washed with 2x11 ml mobile phase, and the washings were discharged. Finally the crosslinks were eluted with 5 ml water, and the aqueous eluate was collected in conical centrifuge tube, concentrated in a vacuum rotor, and stored at -20" C until use. Prior to chromatography samples were redissolved in 250 pl loading buffer made of 10/0 sequanal grade n-heptafluorobuturic acid (HFBA) (Pierce Chemical Co) in lO0mM amoniumcloride pH 3 3 , and centrifugated 5 minutes at 1500 rpm. HPLC assay This HPLC assay is a modification of the method described by Eyre et al. [2] and Black et al. [6]. Pyr and Dpyr were quantitated in the urine extract by isocratic HPLC separation using a short 3.3 cm reversed phase column (Supelco LC-18-DB 3.3 cm x 4.6 mm i.d. pore
Downloaded by [RMIT University Library] at 19:15 18 April 2016
Urinury hydroxypyridinium cross-links from collagen by H P L C size lOOA), on a Jasco HPLC-system, equipped with two pumps, an automatic injector (Jasco 851 AS Intelligent sampler, Fasco International, Tokyo, Japan), and a system controller (802-SC). Fluorescens of the eluted peaks was monitored by a Jasco 821 FP detector with exitation at 295 nm and emmision at 395 nm. The chromatograms were analysed on a computerized unit (Carlo Erba Baseline 810, Dynamic Solution, Millipore, Ventura, CA, USA). A 20 pl sample loop was used for application to the column, which was protected by an C-18 guard catridge (Supelquard LC-WDB, Sudelco Inc., Bellafonte, USA). Flow rate was 1 ml min-I. Solvent A was 0 . 0 1 ~HFBA in 2mM NH&I solution. Solvent B was 7.5% acetonitrile and 25% solvent A. HFBA concentration was adjusted to 0 . 0 1 ~ and pH to 2.26 in both solvents. The column was equilibrated with 88% A and 12% B, and samples were eluted within 5 min. After every 10 samples the column was stripped with 1OO'X B for 5 min and reequilibrated at 12% B for 10 min before the next injection. All samples were analysed at least in duplicate. The results of Pyr and Dpyr were given according to a comparison with known external standards, and expressed as nmol Pyr mmol-' creatinine or nmol Dpry m m o P creatinine. Urinary creatinine concentration was analysed by Jaffe chromogen reaction (Astra, Beckman Instruments, Palo Alto, CA, USA). RESULTS With our system Pyr and Dpyr were separated without any interfering artifacts and eluted at 2.5 min and 3 min, respectively. A typical chromatogram of a standard mixture, consisting of 5.44 pmol of Pyr and 2.68 pmol of Dpyr, and a chromatogram of a normal urine sample, are illustrated in Figure 1. Calibration curves for Pyr in the range of 0-58 pmol and Dpyr in the range of 0-30 pmol are linear, such as shown in Figure 2. In validating our method we have determined the recovery, when increasing amount of Pyr and Dpyr were added to non-hydrolysed urine samples. Recovery of Pyr and Dpyr (mean? SD) were 97,Yf6,2% (n=7) and Y5,4+3,6%, (n=7), respectively, irrespective of the amount of standards added.
659
The intra-assay precision determined by 10 repeated measurements of three different urine samples with increasing amount of Pyr and Dpyr, were 5-7% for both Pyr and Dpyr independent on the amount of Pyr and Dpyr in the urine. The inter-assay precision determined in control non-hydrolysed urine samples, extracted and measured in 13 assays over a period of 3.5 months, were 13.6% for Pyr (mean 692.7 nmol I-' and 12.9% (mean 187.3 nmol I-') for Dpyr. The detection limit of Pyr and Dpyr is approximately 0.4 pmol. The mean values+SD for Pyr:- and Dpyr: crea value in 89 adults (62 women and 27 men, mean agekSD 40.0k15.0) were 38.8k10.8 nmol:mmol (range 16.0-61.7) and 13.0k4.6 nmol: mmol (range 4.5 - 32.3), respectively. The mean values for Pyr:- and Dpyr:crea value in 47 premenopausal (mean a g e f S D 35.3k9.1) and 15 post-menopausal women (aged 62.1 f10.1) and 27 men (aged 35.7?15.0) are shown in
1.0
1.0
2.0
Pyr
3.0
2.'0 3.0 Elution time (rnin)
4.0
4'.0
FIG.I . A typical chromatogram 0 1 the separation of a standard mixture containing pyridinoline and deoxypyridinoline (upper panel). A typical chromatogram of a fractionated hydrolysate of urine from a normal individual (lower panel).
660
G. Kollerup et al.
Figure 3. The mean valueskSD for Pyr:- and Dpyrxrea in 8 children (aged 12.0k4.7) were 181.0583.0 nmol mmol-' and 53.5k23.7 nmol mmol-I, respectively. There was a signi-
ficant increase in both crosslinks in post-menopausal compared to pre-menopausal women (p
