Sb

T

V

BcOpboicvo

FN

SM

RMGm

©

0

d

oluble cathepsin-L: A marker of bone resorption andone density?

HOMAS LANG, ULRIKE WILLINGER, and GEROLD HOLZER

IENNA, AUSTRIA

We sought to evaluate cathepsin L serum levels in the peripheral blood of patientswith low bone density. Blood samples from 60 patients (32 osteoporotic, 28 os-teopenic) and 16 healthy controls were taken and quantitative measurements ofcathepsin L were performed with the use of a commercially available enzyme-linked immunosorbent assay. Dual x-ray absorptionometry measurements and se-rum levels of alkaline phosphatase, bone-specific alkaline phosphatase, osteocal-cin, parathyroid hormone, sexual hormones, and N-terminal crosslinks of type Icollagen were examined. Group comparisons between patients with osteoporosisand controls showed significant differences with respect to cathepsin L (t � �2.839;df � 29 ; P � .008). Osteoporosis treatment decreased the serum level of cathespsinL in a statistically significant fashion (P � .002). These results suggest that the serumlevel of cathepsin L can serve as a marker of bone resorption and bone density. (JLab Clin Med 2004;144:163-6)

Abbreviations: ALT � alanine aminotransferase; AST � aspartate aminotransferase; BMD �bone-mineral density; DXA � dual x-ray absorptionometry; ELISA � enzyme-linked immunosor-

bent assay; NTX � N-terminal crosslinks of type I collagen

rrpccszt

bffosbrbmmt

one mass in adults is maintained locally by thebalance between osteoclastic bone resorptionand osteoblastic bone formation. Resorption is

arried out by hematopoietically derived osteoclasts.steoclasts and other specialized cells such as macro-hages secrete the precursor of cathepsin L, which cane processed into some form of active enzyme by acidr surface activation or limited proteolysis. Cathepsin Ls a lysosomal endopeptidase belonging to the papainysteine protease superfamily.1 Subsequently it is in-olved in connective-tissue degradation and the turn-ver of extracellular matrix proteins.

rom the Department of Orthopedics and the University Ear-Throat-ose Clinic, University of Vienna, Vienna General Hospital.

ubmitted for publication September 25, 2003; revision submitteday 19, 2004; accepted for publication June 2, 2004.

eprint requests: Gerold Holzer, MD, Department of Orthopedics,edical University of Vienna, Vienna General Hospital, Waehringeruertel 18-20, A-1090 Vienna, Austria; e-mail: gerold.holzer@eduniwien.ac.at.

2004 Elsevier Inc. All rights reserved.

022-2143/$ – see front matter

aoi:10.1016/j.lab.2004.06.001

Cathepsin L, a cysteine protease, plays an importantole in osteoclast-mediated bone resorption. Immuno-eactivity for cathepsin B, C, H, and L, as well as acidhosphatase, has been demonstrated in large multinu-leated osteoclasts attached to the bone surface andhondroclasts in the proximal growth plate. These cellshow intense immunoreactivity for these lysosomal en-ymes, suggesting that they play a role in the degrada-ion of organic constituents of the bone matrix.2

Like other lysosomal proteinases, procathepsin L haseen found to be secreted by many malignantly trans-ormed cells. Increased levels of cathepsin L have beenound in primary-cell cultures from samples of cancer-us prostate gland,3 in invasive breast-tumor cells4

erves as a prognostic factor for tumor recurrence inreast cancer,5 and human melanoma cells.6 The up-egulation of cathepsin L and B by the ras oncogene haseen reported for primary human colorectal carcino-as.7 Higher levels of cathepsin L were found in rheu-atoid arthritis synovial lining cells8 than in osteoar-

hritis.Several serum markers for bone metabolism are

vailable, among them alkaline phosphatase, osteocal-

163

cLbdsst

M

S(aiaiestf�t

w6mft1�

w1t4hbw

(ya6ebw

i(bD

ccIpwom

vwfws�

2(fbfo

tBi�

bvarzarmLl3lscrft

tbh3tOfipcBt

hoPctsoCL

J Lab Clin Med164 Lang, Willinger, and Holzer September 2004

in, and NTX. We evaluated serum levels of cathepsin, which is secreted by osteoclasts maintaining a mis-alance in remodeling, to show correlation with boneensity and serum level of cathepsin L. The aim of thetudy was to investigate whether cathepsin L mighterve as a marker for osteoclast-mediated bone resorp-ion and bone density.

ETHODS

eventy-six individuals—32 patients with osteoporosis42%), 28 patients with osteopenia (37%), and 16 healthyge-matched controls (21%)—were recruited and exam-ned at our institution. None of the patients or controls had

disease or was undergoing a treatment that could haventerfered with bone metabolism—including, in women,strogen-replacement therapy. All peripheral-vein bloodamples were collected before the start of osteoporosisherapy. After sampling, blood was immediately centri-uged at 4000g for 10 minutes, and sera were frozen at80°C until they could be examined. Clinical data were

aken from the patients’ files.The group of patients with osteoporosis comprised 24

omen (75%) and 8 men (25%) with a mean � SD age of8.5 � 13.7 years. All of the female patients had enteredenopause, doing so at a mean age of 47.6 � 8.9 years. Six

emale patients (25%) had undergone hysterectomy, 3 pa-ients (13%) ovariectomy. The mean number of births was.61 � 1.56, and the mean number of miscarriages was 0.52

0.79.The group of patients with osteopenia consisted of 21

omen (75%) and 7 men (25%) with a mean age of 60.9 �3.8 years. Three female patients were taking oral contracep-ives, and 17 (61%) had entered menopause, at a mean age of7.9 � 6.3 years. Seven female patients (37%) had had aysterectomy, 4 (21%) an ovariectomy. The mean number ofirths was 1.3 � 0.89, and the mean number of miscarriagesas 0.13 � 0.56.The healthy age-matched controls comprised 14 women

88%) and 2 men (12%) with a mean age of 57.5 � 13.3ears. None of these women was taking oral contraceptives,nd 8 (50%) had entered menopause, at a mean age of 49.8 �.2 years. Two female patients (20%) had undergone hyster-ctomy, 2 patients (20%) ovariectomy. The mean number ofirths was 1.3 � 1.01, and the mean number of miscarriagesas 0.55 � 1.04.Serum levels of cathepsin L in 12 patients were exam-

ned longitudinally on an average of 12 � 14.6 monthsrange 4 – 48) after the start of osteoporosis treatment (oralisphosphonate and a combination of calcium and vitamin).All laboratory examinations—blood counts, chemistry (eg,

reatinine, blood urea nitrogen, ALT, AST), and, more spe-ifically, serum alkaline phosphatase (normal range 98–279U/L), bone-specific alkaline phosphatase (reference interval:remenopausal females 11.6–26.9 U/L, postmenopausalomen 14.2–42.7 U/L, men � 25 years 15.0–41.3 U/L),steocalcin (standard range 2–95 ng/mL), parathyroid hor-

one (normal range 100–600 pg/mL), calcitonin (normal W

alue � 70pg/mL), estradiol (normal range: premenopausalomen 11– 526ng/L, postmenopausal women � 37 ng/L),

ollicle-stimulating hormone (normal range: premenopausalomen 2–20 U/L, postmenopausal women � 20U/L), andro-

tenedione (normal range: women 30–50 years 0.51–2.54g/L, women �50 years 0.2–0.80�g/L, men 31–50 0.91–.96 �g/L, men �50 years 0.83–1.48 �g/L), testosteronenormal range: postpubertal male 8–35 nmol/L, postpubertalemale � 4.0 nmol/L), and NTX (reference values nanomolesone collagen equivalents per liter [nM BCE]: premenopausalemale 6.2–19.0, men 5.4–24.2) were routinely performed inur laboratory.We used a commercially available ELISA for the quanti-

ative measurement of cathepsin L in serum (BMS257;ender MedSystems, Vienna, Austria). In brief, samples were

ncubated at room temperature for 2 hours in duplicate (100L) in microtiter plates precoated with a monoclonal anti-ody specific for human cathepsin L. After washing, strepta-idin–horseradish peroxidase, which binds to the biotinylatednti–cathepsin L antibody, was added. After incubation atoom temperature for 1 hour and washing, Tetramethylben-idin as Chromogen and Peroxide substrate solution wasdded. Color development was stopped after 10 minutes atoom temperature, and intensity was read at 450 nm within 30inutes (microwell ELISA reader Anthos ht-II; Anthosabtec Instruments, Salzburg, Austria). Results were calcu-

ated from a standard curve (Cathepsin L standard, range.12–50 ng mL) generated with the use of a 4-parameterogistic curve fit and expressed in nanograms per milliliter oferum. The assay has been reported to recognize both humanathepsin L and recombinant cathepsin L and to not cross-eact with a series of cytokines and growth factors. The manu-acturer says the sensitivity of the assay is 2.6 ng/mL�1 and thathe an intra- and interassay variations are less than 10%.

All patients underwent DXA of the lumbar spine (in somehe test was also performed on the proximal femur). Total-ody DXA was not possible because several of the patientsad some sort of metal implant. Patients were classified intogroups on the basis of the T-score of the lumbar spine and

he proximal femur as recommended in the World Healthrganization guidelines.9 The overall BMD score was de-ned as the lowest lumbar-spine or proximal-femur score. Aatient with an overall BMD score of less than �2.5 isonsidered osteoporotic. Osteopenia is defined as an overallMD score of �2.5 to �1.0. An overall BMD score greater

han �1 is regarded as normal.9

We assessed correlations between cathepsin L on oneand and NTX, the T-score of the lumbar spine, the neckf femur, and the 2 combined on the other with the use ofearson correlation. Group differences between patients andontrols were analyzed for significance with the use of unpairedtests. P values of less than .05 (2-tailed) were considered

tatistically significant. Data handling and analyses were carriedut with the use of SPSS for Windows, version 11.5 (SPSS Inc,hicago, Ill). Differences between the serum levels of cathepsinbefore and after treatment were analyzed with the use of

ilcoxon’s test.

R

pB

rc�dncdw�NP.ct.

ana

wna2

(hSa��4ptw���a.

D

sswCslcwrstTndtadrtamd

anscfiacFa

eb

Tit

LNB

D*

Tcfs

LFN

J Lab Clin MedVolume 144, Number 3 Lang, Willinger, and Holzer 165

ESULTS

We detected higher serum levels of cathepsin L inatients with osteoporosis than in subjects with normalMD (Table I).Group comparisons between patients with osteopo-

osis and controls showed significant differences inathepsin L levels with regard to the lumbar spine (t �2.839; df � 29; P � .008) and a nearly significant

ifference with regard to the lumbar spine and femoraleck combined (t � �1.98; df � 52; P � .053); in bothases patients had higher scores than controls. Weetected no significant differences in cathepsin L levelsith regard to the femoral neck (t � �.463; df � 28; P.647). We obtained significant correlations between

TX and the T-score for the lumbar spine (r � �.36;� .008), NTX and the femoral neck (r � �.338; P �

03), and NTX and the lumbar spine and femoral neckombined (r � �.313; P � .009). Significant correla-ions were found between NTX and cathepsin L (r �434; P � .0001) (Table II).

Serum cathepsin L levels examined longitudinallyfter patients started osteoporosis treatment were sig-ificantly different (z � �3.061; P � .002) before andfter treatment.All parameters of blood count and chemistry wereithin the normal ranges in patients and controls. Sig-ificant differences between patients with osteoporosisnd controls were found with respect to estradiol (t �

able I. Cathepsin L levels with respect to T-scoresn the lumbar spine, the neck of the femur, andhe combination of the 2

Test siteCathepsin in patients

(ng/mL)Cathepsin in controls

(ng/mL)

umbar spine 3.20 � 3.71 1.10 � 0.68*eck of femur 2.36 � 2.53 1.92 � 1.88oth 2.89 � 3.44 1.58 � 1.46

ata expressed as mean � SD.P � .005.

able II. Correlation coefficients betweenathepsin L and BMD (lumbar-spine T-score,

emoral-neck T-score), NTX, and BMD (lumbar-pine T-score, femoral-neck T-score)

T-score Cathepsin L NTX

umbar spine �.19 �.36emoral neck �.11 �.34TX .434 —

.223; df � 13.731; P � .044), and the sum of fractures N

�3.094; df � 36.325; P � .004), with older patientsaving lower levels of estradiol and more fractures.tatistical trends between patients with osteoporosisnd controls were found in progesterone (t � �1.73; df

29; P � .094), alkaline phosphatase (t � � 1.822; df42; P � .076), and osteocalcin (t � � 1.921; df �

4; P � .061), with patients having higher scores inrogesterone, alkaline phosphatase, and osteocalcinhan the controls. We found no significant differencesith respect to androstenedione (t � �.894; df � 39; P.377), follicle-stimulating hormone (t � �1.269; df27; P � .215), testosterone (t � �.532; df � 40; P.598), calcitonin (t � �1.291; df � 42; P � .204),

nd Parathormone (PTH) (t � �1.669; df � 44; P �102).

ISCUSSION

Ours is the first study to investigate the correlation oferum cathepsin L levels and bone density. We ob-erved higher serum levels of cathepsin L in patientsith osteoporosis than in subjects with normal BMD.omparisons showed significant differences in cathep-

in L with regard to the lumbar spine, as well as theumbar spine and femoral neck combined; in bothases, patients had higher scores than did controls. Weould like to emphasize that we found significant cor-

elations between cathepsin L levels and the lumbarpine but were unable to show any significant correla-ion between cathepsin L level and the femoral neck.he authors of a study of treatment with bisphospho-ates (alendronate, risedronate)10 have also reportedifferences in the percent change in bone density be-ween these sites. Different mechanisms of osteoclasticctivity and bone resorption seem to be at work inifferent areas. Among patients who received osteopo-osis treatment, we found significantly decreased ca-hepsin L levels. We found that this assay exhibits goodnalytical performance, cathepsin L being a sensitivearker with which to assess bone resorption and bone

ensity.Osteoporosis is diagnosed clinically through DXA

ssessment of BMD in the lumbar spine and femoraleck. BMD is a surrogate parameter, and DXA mea-urements hold some disadvantages. First, degenerativehanges in the lumbar spine can lead to inaccuratendings as a result of osteophyte formation. Second,ccording to Genant et al,11 BMD of the lumbar spineannot predict fracture risk in the proximal femur.inally, the effects of osteoporosis treatment can bessessed only after a minimum of 1 year.Several serum markers are available with which to

valuate bone turnover, including alkaline phosphatase,one-specific alkaline phosphatase, osteocalcin, and

TX. These markers are products of bone cells or

dafo

mbgteeusatropma

R

1

1

J Lab Clin Med166 Lang, Willinger, and Holzer September 2004

egradation products of matrix and respond immedi-tely to therapy. Markers are used to measure boneormation or bone resorption and to monitor the coursef therapy.Currently, patients are classified on the basis of DXAeasurements into 3 groups according to BMD (normal

one density, osteopenic, and osteoporotic), as sug-ested by the World Health Organization.9 The aim ofhis study was to determine whether serum markers,specially cathepsin L, correlate with BMD on initialxamination (without therapy) and whether they can besed as a tool to diagnose osteoporosis. The use oferum parameters would be obviate the need for radi-tion in DXA, although the levels are low. Also, pa-ients who have undergone spinal fusion or total jointeplacement in one or both hips could be evaluated. Ifur findings are confirmed by further studies in largeratients cohorts, the cysteine peptidase cathepsin Lay be accepted as a new marker for bone resorption

nd bone density.

EFERENCES

1. Leung-Toung R, Li W, Tam TF, Kaarimian K. Thiol-dependentenzymes and their inhibitors: a review. Curr Med Chem2002979–1002.

2. Ohsawa Y, Nitatori T, Higuchi S, Kominami E, Uchiyama Y.Lysosomal cysteine and aspartic proteinases, acid phosphatase,and an endogenous cysteine proteinase inhibitor, cystatin-beta, in

rat osteoclasts. J Histochem Cytochem 1993;41:1075–83.

3. Friedrich B, Jung K, Lein M, Turk I, Rudolph B, Hampel G, etal. Cathepsins B, H, L and cysteine prothease inhibitors inmalignant prostatae cell lines, primary cultured prostatic cellsand prostatic tissue. Eur J Cancer 199935:138–44.

4. Zajc I, Sever N, Bervar A, Lah TT. Expression of cysteinepeptidase cathepsin L and its inhibitors stefin A and B in relationto tumorigenicity of breast cell lines. Cancer Lett 2002;187:185-90.

5. Thomssen C, Schmitt M, Goretzki L, Oppelt P, Pache L, DettmarP, et al. Prognostic value of cysteine peptidase cathepsin B andcathepsin L in human breast cancer. Clin Cancer Res 1995;1:741–6.

6. Rozhin J, Wade RL, Honn KV, Sloane BF. Membrane-associ-ated cathepsin L: a role in metastasis of melanomas. BiochemBiophys Res Commun 1989;164:556-61.

7. Kim K, Cai J, Shuja S, Kuo T, Murnane MJ. Presence ofactivated ras correlates with increased cysteine proteinase activ-ities in human colorectal carcinomas. Int J Cancer 1998;79:324-33.

8. Ikeda Y, Ikata T, Mishiro T, Nakano S, Ikebe M, Yasuoka S.Cathepsins B and L in synovial fluids from patients with rheu-matoid arthritis and the effect of cathepsin B on the activation ofpro-urokinase. J Med Invest 2000;47:61-75.

9. Kanis JA, Devogelaer JP, Genari C. Practical guide for the use ofbone mineral measurements in the assessment of treatment ofosteoporosis: a position paper of the European Foundation forOsteoporosis and Bone Disease. Osteoporosis Int 1996;6:256-61.

0. Cranney A, Wells G, Willan A, Griffith L, Zytaruk N, RobinsonV, et al. Meta-analyses of therapies for postmenopausal osteo-porosis. II. Meta-analysis of alendronate for the treatment ofpostmenopausal woman. Endocrinol Rev 2002;23:508–16.

1. Genant HK, Lu Y, Mathur AK, Fuerst TP, Cummings SR.Classification basedon DXA measurements for assessing the risk

for hip fractures. J Bone Mineral Res 1996;11(suppl 1):120.