Urinary excretion of pyridinium crosslinks: a new marker of bone resorption in metabolic bone disease

81

Eon~~dMinrrul,8(199)57-96 Elsevier

Urinary excretion of pyridinium crosslinks: a new marker of bone resorption in metabolic bone disease

Daniel Uebelhart, Evelyne Gineyts, Mrie-Claire Pierre D. Delmas INSEAM

(Inif

Chapuy and

& Service de Rhumntologic cl de Parhologie Orseuse, H6piral E. Herriot, Lyon, France (Received 7 October 1989) (Aaepted

7 October 1989)

Summary The pyridinium derivatives hydroxylysylpyridinoline (HP) and lysylpyridinoline (LP) are intermolecular crosslinking compounds of collagen which em only present in its mature form. Contrasting to tbe wide distribution of type 1 and II rollcgens, HP and LP me r&em

from skin, ligament and fascia, and

their major socrces are bane and cartilage. Using a specific HPLC assay, we have determined the 24-h excretion

ofHP and LP crmslinks

in normal adultsof both scxs, in patientswith primary hyperparathy-

midism and in patients with Paget’s disease of bone before and after intravenous treamcnt wilh amino,pmpylidene bisphosphonate (APE). HPand 6.3 f 3.4 pmal/~molcreatinine

Mean adult normal vab~cswere 33 * 13 pmollflmol creatinine for 1arLP. Inwomen,mcnapauscindvceda2-3.loldincrcaseafHP

end LY reflecting the well documented postmenopausal increase ot bone turnover. In the urine of patients with primary byperparathymidism and of patients with active Paget’s disease cd hone. urinary crosslinks were signilcandy htgher than in age-matched controls. with a mean 3- and 12.fold increase, respectively. Urinary excretion of hydraryproline is a well recognized bitt pxwrly sensitive marker of bone tumover, re&cring mswpuon. In the scmc patiems, the effect of menopause and disease state on hydmxypmline excretion we.?much les dramatic than on HP end LP. During intravr nous APB treatment of pagetic patients. there was P!I early decrease of HP and LP, which was signiiiciwt after 24 hand reached 62% at 4 days, comrasdng with a late and milder decrease ofurbtmy hydmxypmline. Because APB is c potent inlubitorofresorption

which does not have adirect short-term cflcct on bone formation,

these data alw indicate thct urinary excretion of HP and LP reflect only collagen degradation occurring during osteoclastic resorption end nor the degradation of ncwty synthesized collcgen. We conclude that uriuwy HP and LP excretion represents the lin.sensitive and rprcifi: markerofbone should be valuable in the clinical investigation

Key wards: Hydmxylysylpyridinoline;

resorption. IN use

ofmetabolic hone diseases, especially osteoporosis.

L,sylpyridinoline;

Bone resorption; Metabolic bone disease

-Ccrrespondence to: Pr. P.D. Delmas. Hd!qital E. Herriot, Pav. F. 69437 LyonCcdcx 03.. France.

0169.N09/9iYS03.50@

t99OElsevicrScience PublishersRV.

(Biomedical Division)

88 lntrnduction A crucial step in the clinical evaluation of metabolic bone disease and to monitor the effects of treatment is to have an accurate and sensitive marke, of bone resorption. Urinary hydmxypmhne is a widely used index of bone resorption [l] but it has a number of limitations. Because hydroxyproline is rapidly oxydized by the liver, its excretion represents only 10% of that produced by tissue catabolism. Although bone is the major source of urinary hydroxyproline, there are several other sources (other tissues, circulating molecules, diet). For instance, the elastin and the acetylcholinestecase metabolism also contribute to hydroxyproline excretion, although in small amounts. The other important source of hydroxyprohne is the Clq component of complement system. For normal individuals, the OHP released from Clq breakdown could represent lS-50% of the total daily excretion [2,3]. In addition, urinary hydroxyproline is not specific for bone collagen resorption, as it is also derived from collagen synthesis during the breakdown of the procollagen N-terminal extension peptides and of neosynthesized collagen molecules [4], Urinary excretion of the two hydroxylysine glycosides, galactosyl-hydroxylysyl and glucosyl-galactosyl-hydroxylysyl, were postulated to be valid markers of bone resorption, since the galactosyl fraction is not contributed by the Clq [S], but recently a specific a-glucosidase was found in rat kidney cortex which converts the glucosyl fraction in galactosyl[6]. If this enzyme is widespread in other tissues and species, it could also account for the urinary excretion of galactosyl-hydroxylysyl in humans and therefore this marker would not be specific of bone tissue breakdown. Plasma Ievels of tartrate resistant acid phosphatase have been proposed as a markcr of bone resorption as it originates in bone tissue from the osteoclasts [7] and because high levels have been measured in Paget’s disease and primary hyperparathyroidism [a], but its assay has limitations and its specificity needs to be further evaluated. Very recently, y-carboxyglutamic acid (Gla), which is present in at least two major non collagenous proteins of bone matrix (osteocalcin and matrix Gla-protein), has been shown to circulate in its free form and to be elevated in the serum of patients with primary hyperparathyroidism. Therefore, serum free Gla could also be a new marker of bone resorption, but most of it seems to result from the metabolism of the vitamin K-dependent clotting factors [9]. Recently, a new class of potential markers of bone turnover has been studied with the discovery of the intermolecular crosslinking amino-acids of the collagen molecule. The extracellular matrix is stabilized by the formation of covalent crosslinks between adjacent collagen chains and it is thought that the reducible aldimine crosslinks initially formed are converted to mature non-reducible compounds [lO]. There are two mature crosslinking amino acids of the 3-hydroxypyridinium family that have been identified by their natural fluorescent properties in many connective tissues, such as bone, cartilage and, to a lesser extent, in other connective tissues except skin [ll]. Hydroxylysylpyridinoline (HP) is the major mature component of these tissues, since lysylpyridinoline (LP) is significantly present only in bone tissue and dentine where it accounts for 21% of the total mature crosslinks

89 [K!]. The. pyridinium amino-acids have been measured in urine after paper chmmatography [13], by fluorescence detection after high pressure liquid chromatography (HPLC) ]14-161 or by enzyme-linked imm!monssay 1171.Its urinaryexcretion level has been reported in a few normals [17,18] and has been found to be elevated in five patients with osteoarthritis [17] and in five patients with rheumatoid arthritis [18,19], reflecting the increased cartilage breakdown of these diseases. To our knowledge. no comprehensive study of uricary pyridinoline has been performed in a large number of normal adults and in patients with bone disease. In this study, we have apphed a new sensitive and specific assay for pyridinium crosslinks to a welldefined population of normal adults of various age and to patients with primary hyperparathyroidism and Paget’s disease of bone, two conditions characterized by a typical increase of bone turnover. The specificity of HP and LP for resorption was demonstrated by looking at the short-term effect of aminopropylidene bisphospho“ate (APB) in pagetic patients.

Patients and Methods Healthy adults of both sexes (15 men and 39 women) without any history of bone and/or articular disease were studied. Male controls were 41 f 15 years old (range, 23-68). Female controls were 43 + 13 years old (range, 23-60) and were subdivided into two groups according to their menopausal status: 20 pre-menopausal women, 31 f 6 years old (23-46) and 19 post-menopausal women, 55 It 3 years old (52 -60). We further studied a group of 15 patients (65 & 10 years old) with well characterized active Paget’s disease of bone. All of them were analysed before treatment and ten of them were also studied during parenteral treatment with APB. A continuous intravenous infusion of 180 mg of APB was administered over 72 h. Urine was collected the day before starting the treatment and during the next 4 days. Seven patients (60 f 11 years old) with primary hyperparathyroidism were studied prior to the surgical removal of a parathyroid adenoma. Urine was collected over 24 h on a gelatine free diet for measurement of HP and LP crosslinks, total hydroxyproline and creatinine.

Before hydrolysis, urinary creatinine was measured on an aliquot according to the method of Folin and Wu [20]. A urine sample of 3 ml was hydrolysed in 6 M HCI for 3 h in glass tubes having a screw cap with a teflon liner in a pressure cooker (125 “C), a procedure which has been shown to give the same results as the classical 24-h hydrolysis at 105-108 “C in sealed glass tubes. The cold hydmlysates were first centrifuged at 1000 x g for 10 mitt, then the supernatant was carefully removed and applied to a CFl-cellulose column for fractionation according to the method originally described by Skinner for the purification of the elastin crusslinks [21]. According to the expected relative content of mature crosslinking amino-acids in 24-h

90

urine, 0.5-l ml of urine hydrolysate were mixed with acetic acid, water and n-butano1 (1:1:4; v/v) and applied to the CFl-cellulose column (Whatman Co. Ltd., Maidstone. England). After washing the column with 3 column volumes of buffer, the crosslinking amino acids wcrc eluted from the column with distilled water (5 ml) into a glass tube, evaporated to dryness on a Savant Speed-l/a@ overnight, and the dry residue was stored at 4 “C prior to analysis. For the HPLC analysis, the dry residues were redissolved in a 1% sequanal grade n-heptafiuorobutyric acid (HBFA) solution (Pierce Chemical Co.). Urinary hydroxyproline measurements were made on hydrolysed urine samples according to the calorimetric method ol^Firschein and Shill 1221,corrected for urinary creatinine and expressed as mp/p creatinine. Preparation ofpyridinoline standard

A bone hydrolysate preparation was used regularly as an external standard prior to the application of urine samples to the chromatogaphic procedure. The HP and LP standards were prepared from powdered human cortical bone after demineralization by sequential extraction with 0.5 M EDTA, pH 7.6. The collagenous residue was further washed with water, then hydrolysed with 6 M HCI for 24 h at 108 “C. Hydrolysate was evaporated to dryness on a Biichler Evapomix and the cooled hydrolysate was applied to a Biogel P2 column (2.5 x 90 cm), 200-400 mesh (BioRad), equilibrated in 10% acetic acid (v/v) at room temperature according to Eyre’s method [15] and eluted with 10% acetic acid at a flow rate of 30 ml/h. The fluorescence of the eluate was monitored with a Shimadzu RF 530 spectromonitor (Shimadzu Co., Kyoto, Japan). Fluorescence was determined by excitation at 297 nm and emission at 3110nm and each 5 ml fraction collected in glass tubes. Pooling a tairly broad region of eluant (50 ml) allowed recovery of all crosslinking amino acids. The pooled fractions were then diluted twice and dried in a Biichler Evapomix. Standards of HP and LP were purified by HPLC (see below), calibrated with a 1 nmol standard of HP generously supplied by D. Eyre, and stored at -30 “C in 500 gl aliquots until used. HPLC assay HP and LP were assayed in the urine extract by HPLC with a modification of the

method described by Eyre et al. [!5] on a Kontron HPLC System 400 (Kontron Instruments Co., Zurich, Switzerland) equipped with a serial dual-piston two-pump system and an automatic injector (Autosampler 460) controlled by a central multitask computerized unit (Data System 450). The reversed phase column used was an Altex Ultrasphere ODS (5 pm; 25 cm x 4.6 mm) protected by a Brownlee Cl8 guard cartridge (5pm; 3 cm x 4.6 mm). Eluant A was 0.01 M n-heptafluombutyric acid (HBFA) in 5% (v/v) acetonittile. Eluant B was 100% acetonitrile. The column was equilibrated in 88% A/12% B and the samples were eluted with the same isocratic gradient run at a flow rate of 1 ml/mitt for 20 min. The column was then stripped with 100% B for 5 mitt, reequilibrated at 12% B for 5 min before the next injection. For each urine specimen, two samples of 15 and 3041 were injected. Fluorescence of the eluted peaks was monitored with a Kontron spectrofluorimeter SFM-25with excitation at 297 nm and emission at 380 nm. The results of HP !d LP

were given according to a comparison with an external standard injected at two different amounts and expressed as pmoll~mol creatinine. Other methods Total alkalinephosphatase The total alkaline phosphatase activity was determined in the serum of pagetic pa-

tients according to the automated procedure based upon the enzymatic hydrolysis ofp-nitrophenyl phosphate during incubation at 37.5 “C [23]. Results are expressed in NJ/I.

Comparisons between groups were assessed with an unpaired Studegr’s r-!tet. In order to test the differences between tbc values at thr variao- days of treatment of pagetic patients, a Wilcoxon paired test was used and the % decrease was assessed with a single Student’s t-test.

With our system, HP and LP were clearly separated without any interfering artefact, eluting consistently at i3 and 14 mitt, respectively (Fig. 1). The accuracy of the

Fig. 1. Typical chromatogramof the assayof HP and LP in urine of a normaladult. The samplewasextractedandanalysedby HPLC asdescribedin Methods.

Fig. i. urirury excretionof HP (left) and LP (right) in 20 pre-menopausaland 19 post-menopausal healthy women.

assay assessed by determination of the recovery of a known standard spiked into non-hydrolysed urine samples ?om both a control and a pagetic patient wah 90% for HP and 91% for LP. The intra.,assay variation, determined by repeated measurements of the same urine extract was 10 and 13%, respectively, for HP and LP. The interassay variation of non-hydrolysed samples extracted and measured on different days was 9.6% for HP and 14.9% for LP. In the 54 healthy adults of both sexes, urinary HP was 24 + 13 pmol/flmol creatisine and urinary LP was 4.4 f 3.3 pmol@mol creatinine. Values were not significantly different in males and in females (20 + 5 vs. 25 f 15 for HP and 3.7 f 0.4 vs. 4.7 ? 3.8 for LP). When normal women were analysed according to their age, postmenopausal women showed HP and LP that were significantly higher (2-3-times) than premenopausal women (Fig. 2). The LP/HP ratio was also significantly increased after the menopause (0.21 + 0.12 vs. 0.13 + 0.07, P < 0.05). In the same women, urinary hydroxyproline was slightly but not significantly increased after the menopause (Table 1). There was a significant 3-fold increase in HP and LP in patients with primary hyperparathyroidism compared to age-matched controls (Table 2). In patients with active untreated Paget’s disease of bone, there was a dramatic ll-fold increase of HP and LP over controls, contrasting with a milder 6.fold increase of hydroxyproline. In pagetic patients treated with APB, there was a linear decline of HP and LP which was significant as early as day 1 for HP and day 2 for LP. Urinary HP before

Table 1 Urinary excretion of pyridinium crosslinks (HP and LP) and of hydroxyproline in normal women (data are expressed as mean & 1 SD) Normalwomenn

Ace ~ycars) HP (pmoi/pmat ----

Premenopaural 20 Postmenopausal 19

3116 55 ?;3

LP

LPiHP(%)

15.6+ 10.3 2.3 i2.3 35.6 + 13.4b6.9 2 3.Q

vJ<0.05 “I. p,emenopause:~P
Hydroxyproline (mg/gcreatinine)

creatinine)

!3 211

37.8 f 12.3 46.6 k 16.6

93 Table 2 excretion of pyridinium crosslinks (HP and LP) and of bydroxyproline in disease of bone and in primary hyperparathyroidism (data are expressed as mean f 1 SD) Urinary

Paget’s

Groups

P&s discam Prim. hyperpara. Controls

n

15 7 25

A$e

61?10 &I+ 11 56 ?: 4

HP

LP

Hydroxyproline

~pmoll~molcreatinine~

(mglgcreatinine)

405 f 381b 98+6# 33+13

271 i: 223b 107 * 71’ 4** 16

7S.O rt 80.Zb 15.8 f 9.gb 6.3 * 3.4

*PC 0.01 vs. controls: hP 4 0.001vs. mntmlr.

and after treatment was 408 + 373 pmoU,umol creatinine and 145 f 127 pmol/:lmol creatinine (-59%) and urinary LP 73 f 77 and 16 + 19 pmol@mol creatinine (-62%), respectively. The decrease of hydroxyproline occurred later (day 3) and was of lesser magnitude (259 i: 249 before vs. 202 f 199 mglg creatinine after, -30%). Meanwhile, serum alkaline phosphatase did not change significantly throughout the study time (Fig. 3). HP and LP values were highly correlated in pagetic patients and in primary hyperparathyroidism 0. = 0.94 and 0.96, P < 0.001) but not in controls (I = 0.40 in men and 0.74 in women). Hydroxyproline was correlated with HP and LP in disease states, with an t value ranging from 0.56 to 0.86, but not in controls.

A%

Fig. 3. Urinary excretion of HP, LP, hydraxypmiine (OHP) and serum alkaline phosphatase in 10 patientswith Paget’sdiseaseofbone during treatmentwith parenteralAPB. (Data are expressed asthe % variationfrom the initial valuerderivedfram eachpatient.)

94

The pyridinimn crosslinks arc markers of type I and II collagen breakdown, and their two main components, HP and I.? are essentially present in cartilage (HP) and bone (HP and LP). Their relative content in spongy and cortical bone increases during growth a.ld remains stable throughout life [24]. The LP fraction is found in bone and dentine, but the latter tissue does not contribute significantly to its level, so bone can be considered as the unique source of it, which makes it a specific bone marker. In urine, HP and LP are present both in the free form (40%) and in pep tide-bound form (60%) [lS,W], and the hydrolysis of urine prior to the extraction allows measurement of the total excretion of the crosslinks with various techniques. T:!e great stability of these compounds when urine is frozen offers a storing possibility for further measurements [I9]. The crosslinking assay used in our study is accurate, repioducible and it is suitable for longitudinal studies. To our knowledge, this is the first comprehensive study of urinary HP and LP excretion in normal adults and in patients with bone diseases. The values found in normal adults were similar to the one reported in a small cohort 1181. Mean values of HP and LP were similar in men and in premenopausal women. In the postmenopausal women, HP values were found to be twice as high, and LP three times as high as in the premenopausal group. This increase is further assessed by a significant increase of the LP/HP ratio (P C 0.05). This pattern reflects the marked increase of bone turnover at this period which is well documented by kinetic, histomorphometric and biochemical studies [26-291 and which is related to the postmenopausal oestrogen deficiency. These data confirm a recent animal study of the urinary excretion of HP and LP in ovariectomized rats, showing significantly higher values in the ovariectomized group as compared with the sham opcrated group [30]. The authors found that the increased excretion was mainly due to the contribution of the free crosslinks with no significant change in the peptidebound form of HP and LP. In our study, the postmenopausal increase of the small free fractior. of urinary HP and LP was not larger than the increase of the total excretion of HP and LP (data not shown). Contrasting with the marked changes of HP and LP after the menopause, urinary hydroxyproline was only slightly increased indicating that the latter is a less sensitive marker of bone turnover. Urinary excretion of HP and LP was significantly increased in two conditions characterized by a typical increase of bone turnover-primary byperparathyroidism and Paget’s disease of bone. The highest values were found in patients with active Paget’s disease, with a mean 12-fold increase of HP and LP. The fact that the increase of urinary hydroxyproline in these patients over the controls was of lesser magnitude confirms its poor sensitivity. In order to confirm that urinary crowlinks are derived from the breakdown of bone collagen matrix during the process of osteoclastic resorption and not from the breakdown of neosynthesized collagen molecules, we looked at the short-term effect of parenteral APB administration in pagetic patients. Like all bisphosphonates, APB acts primarily on the osteoclasts, inducing a rapid and prolonged inhibition of bone resorption [31]. Because bone resorption and formation are tightly

9.5

coupled, the decreased bone resorption is responsible for a delayed decline of bone formation in those patients, both events reaching a maximum 6 months after the infusion [32]. Because APB has no direct effect on the osteoblast, the early effect of parenteral APB is to decrease resorption without significant change of bone formation, as shown by the stable levels of serum alkaline phosphatase and bone Gla-protein [33]. In our patients under APB, the decline of’HP and LP was parallel to the urinary hydroxyproline excretion and contrasted with unchanged levels of serum alkaline phosphatase, demonstrating the spec;.ftcityof HP and LP to reflect bone resorption. Again, HP and LP measurunents were more sensitive than hy droxyprolinc, with an earlier and more pronounced suppression during treatment. Hydroxyproline assay can be improved by an HPLC method involving pre-colum derivatizatioll [34], but our data suggest that HP and LP are more suitable for detecting mild changes of bone resorption in disease states and to monitor the effects of specific treatments. This assay could be of special interest in osteoporosis where a moderate increase in resorption could account for the increased bone loss.

We thank Ciba-Geigy, France and Ciba-Geigy Ltd., Switzerland for providing us with APB.

References 1 f’rockop

DJ. Isotopic studies on collagen degradation end the urine excretion of hydroxypmline. J Chn Invest 1964:43:453-460. 2 Krane SM. Kantmwitz FG. Byrne M, Pinnell SR. Singer FR. Utincry cxctetion of hydroxylyrine and its glycerides as an index of collagen degradation. J Clin Invest 3977$9:819-827. 3 Nimni ME. Collagen: its structute and function in normal and pathological connective tissues.Sem Arthr Rheutn 1974;4:95-1%. 4 HBrlein D, Fietzek PP. Ftthn K. Pm-ght: the procollagen peptidase cleavagesite in the alpha 1 (I) chain of demtatosparactic calf skin procollagen. FEES Lett 1978;81):279-282. 5 Kelleher PC. Urinary excretion of hydtoxypmline, hydrorlyrine and hydmxylysyl glycosidesby patients with Paget’s disease of bone and carcinoma with tnetestasesitt bone. Clin Chim Acta 1979:92:373-379. 6 Sternberg M, Spiro RG. Studies on the catabolism of the hydroxylysine-linked disaccharideunits of basement tnembmne and collagens. Isolation and characterization of a tet kidney alpha.glucosidase of high speci6city. J Viol Chem 1979;254:10329-10336. 7 Minkin C. Bone acid phosphatase:tartrate-resistant acid phosphetaseesB marker of osteaclastfunc. tion. CalclfTisrue fnt 1982:34:285-290, 8 Lau KHW, Kraenzlin ME, Onishi T, Baylink DJ. Development of an improved assayof tartrete-resistant acid phasphatese(TrACP) in serwtt: potential use to essessbane rerorptin. In: Christiansen C, lohanaen JS. Riis BJ, eds. Osteoporosis 1987.2.Cownha8en: Osteopns~ APS, 1987:682-687. 9 Fournier B. Gineyts E, Delmas PD. Evidence that free 8s~tnaarb&ygb&mic acid circulates in serum. Clin Chim Acta 1989;182:173-182. 10 Eyre DR. Crosslink maturation in bone cottegen. In: Veis A, ed. ‘tote chemistry and biology of tnineralizcdconnectivetissues. Amsterdsm: Etsevier, 1981;51-55. 11 Eyre DR, Pee MA, Galop PM. Crowlinking in collagen end elastin. Ann Rev Bioehem

.

96 1934;53:717-748. 12 Eve DR. Koob TJ. Van Ness KP. Quantitation of hydroxypyridiniom crosslinks in collagen by highpcrformancc liquid chromatography. Anel Biochem 19&%;1373380-388. 13 Gunja-Smith

Z.

Boucek RJ. Collagen cross-linking compounds in human urine. Biochem J

1981:197:759-762. 14 Fujimom D. Aging and cross-linking in human aorta. Biwhcm

Biophys Res Commun 1982;lCtZ

1264-1269. 15 Eyre DR. Collagen crass-linking amino.acidr.

In: Cwmiogham LW, cd. Methods in enzymology,

Vol. 144. New-York: AcademicPress Inc., 1987;ll5-139. 16 Elack D. Duncan A. Robins SP. Qeanlitalive urine using ion-paire

analysis of the pyridinium crosslinks of collagen in

reversed-phase high-performance

liquid chromatography.

Anal Biochem

,988;,69:197-203. 17 Robins SP. An enzyme-linked immunoassay for the collagen crosslink pyridinoline.

Biochem J

l982:*cl7:617-6*0. 18 Fujimoto D. Suzuki M, Ucbiyama A. Miyamoto S. lnoue T. Analysis of pyridinoline, a crosslinking compound uf colldgee fibers, in human urine. J Biocbem 1963;94:1133-1136. 19 Robins SP. Stewart P, Astbury C. Bird HA. line in urine as an index

Measurement of the crosslinking compound pyridino-

ofcollagen degradation

in joint disease. Ann Rheum Dis 1986;45:969-973.

20 Foalin A, Wo T. In: Hawk HB. Oser BL. Summerrsan WH, eds. Practical physiological chemistry, l2rh Edn. New-York: McGraw-Hill,

l965;506.

21 Skinner SJM. Rapid method for the purification of the elastin cross-links, desmoaine md isodermo. sine. J Chromatogr 1982:229X%3-204. 22 Finchein HE, Shill JP. The determination of tolal hydroxypmlinc in urine and bane extraets. Anal Biochem 1966:143296-3114. 23 Mrrgenstem

S. Kessler G, Auerbach J. Flor RV, Klein B. An automated p-nitrophcnylphosphate

serum alkaline phosphatase procedure for the auto-analyzer. Clin Chem 1965;11:876-X88. 24 Eyre DR.

tnckson iR, Van Nes K. Collagen cross-linking in human bone and articular cartilage.

Ap-r&ted

changes

in

the

content

of

mature

hydroxypyridinium

residues.

Biochem

J

1988;252:495-500. 25 Eyre DR. Ericsson L, Simon L, Krane S. ldcntification of urinary peptides derived from c:u&inking sites in hone collagen in Paget’s disease. I Bone Mineral Rer 1988;3:E210. 26 Lindsay R. Aitken JM. Anderson JB, Her1 DM,

Mac Donald EB. Clarke AC. Long-term preven-

tion of postmenopausal osteopororis by oestrogeen.Evidence for an increased hone mass after delayed onset ofoestrogen treatment. Lancet l976:i:1038-1041. 27 Heaoey RP. Reeker RR, Saville PD. Menopausal changes in bone remodelling. J Lab Clin Med 1978:92:964-970. 28 Johansen JS, Riis BJ. Debnar PD. Christiansen C. Plasma BOP: en indicswr of spontaneous bone lo=

end of the effect of oertrogen

treatment in postmenopzeral

women. Eur J Clin Invest

l988:18:191-195. 29 Parfit, AM. Bane remodelling: rclationsbip to the amount and ~tr”ct”re of bone, and the pathogenesis and p~rvention of fractures. In: Riggs BL. Melton W, eds. Osteoporosis. New York: Raven Press, 1988.45-94. 30 Black D. Farquhanon

C. Robins SP. Excretion of pyridinium cross-links of collagen in avariecto-

mired rats as urinary markers for increased bone rcsorplion. C&if 31 Frijlink WB, Bijvoet OLM, 32 DelmasPD.

Tissue Int 1989,44:343-347.

Te Vclde J, Heynen G. Treatmettt of Paget’s disease with (3.amino-l-

hydmxy-propylidene)-I-I-bisphosphunate

(APB).

CesezJP, ArlotMA,ChapuyMC.

Lanc
HeilmannMO,MeunierPJ.

trwenous aminopmpylidene diphosphonate (APD) bone. CalcifTirrue Int 1989:44:S104. 33 Papapaulos SE, Frolich M. Muddc AH,

Long:~meffectsofin-

in patients with refractory Paget’s disessc of

Harink HIJ. Berg H, Bijvoet OLM.

Serum osteocalcin in

Paget’s disease of bone: basal concentrations and response 10 bisphosphonate treatmeet. I Clin Endwrimd Melab 1987:G5:89-94. 34 Damon

CD, Jewel1 S, Driskell WI. Liquid-chromaiographic

inurine. CiinChcm

1988;34:1572-1574.

determination of rotal hydroxyproline