Annals of the Rheumatic Diseases 1995; 54: 281-288

Cathepsin B in osteoarthritis: zonal variation ofenzyme activity in human femoral head cartilage

Antonio Baici, Dorothy Horler, Angela Lang, Claude Merlin, Rudolf Kissling

AbstractObjectives-To determine the quantita-tive topographical distribution of cathep-sin B in human femoral head cartilage bymeasuring the zonal variation of enzymeactivity in specimens taken from variousanatomical regions of normal and osteo-arthritic (OA) tissues, and to correlatethis parameter with the severity ofthe OAlesions.Methods-OA articular cartilage was ob-tained at surgery for total hip replacementand control cartilage obtained at post-mortem. Cylinders of full thickness carti-lage with underlying bone were retrievedwith a biopsy trephine. Sections ofcartilage were produced by cryocuttingthe tissue as slices parallel to the articularsurface and assayed for cathepsin B witha specific, highly sensitive fluorogenicsubstrate. The severity of the OA lesionswas graded according to the histopatho-logical-histochemical method ofMankin.Results-Zonal cathepsin B activity ofnormal cartilage was uniform and low inall regions of the femoral head. Inapparently intact OA cartilage and inseverely degraded tissue the zonaldistribution and the amounts of enzymewere similar to control values. At siteswith active disease, cathepsin B activitywas much greater than in controls and itsirregular zonal distribution correlatedwith tissue degeneration, hypercellularity,or cloning of chondrocytes as determinedhistochemically. Particularly high en-zyme levels were observed at sites withregenerating cartilage, where some zonalpeaks attained 20-fold activity withrespect to controls.Conclusion-Cathepsin B may play a rolein sustainiing the chronicity of OA, not asan initiator, but rather as a perpetuator ofthe disease and as an antagonist ofregeneration.

(Ann Rheum Dis 1995; 54: 281-288)

Chondrocyte mediated breakdown of cartilagehas classically been studied in the contextof cell stimulation by cytokines.1 2 Matrixdegradation has been attributed to metallo-proteinases stimulated by cytokines fromsynovium,3 monocyte-macrophage lineagecells,4 or polymorphonuclear leucocytes,5 or ofautocrine origin.6 A rich literature and modemtrends in the pathophysiology of articu-lar cartilage have been comprehensively

reviewed.7 8 Enzymes other than metallo-proteinases have received less attention, butthe discovery that endogenous hydrolases areresponsible for cartilage autodegradation9 10stimulated research on the possible patho-logical significance of this phenomenon." 12Cathepsin B and cathepsin D are able todegrade both collagen and aggrecan invitro,'3-'" but cathepsin D, although present atextracellular sites in rheumatoid synovium,16 isnot responsible for the resorption of cartilagein organ culture.'7 Despite higher levels ofcathepsin D in osteoarthritis (OA) comparedwith normal cartilage,18 19 its action is re-stricted to intracellular catabolism of matrixmacromolecules within chondrocytes.20

Increased concentrations of cathepsin Bhave been demonstrated in rheumatoidsynovium1621 22 and synovial fluid,2' 2325whilemuch smaller amounts were found in OAsynovial fluid.24 25 Increased concentrations ofcathepsin B in human OA cartilage18 26 27 wereaccompanied by a decrease in cysteineendopeptidase inhibiting activity in severelydamaged tissue.27 Chondrocytes can bestimulated to increase dramatically the syn-thesis, storage, and secretion of cathepsin Bthrough modulation of their differentiatedphenotype to a more primitive one.28 Thisproperty is reversible: cathepsin B levels revertto normal after the phenotype is modulatedback to its original state.28 29 After stimulationwith interleukin-1l, cathepsin B biosynthesisand storage are slightly increased, while secre-tion is not.30We describe below zonal variations of

cathepsin B activity in normal and OA humanarticular cartilage and suggest that this enzymemay play a pathological role by sustaining andperpetuating the OA process.

Materials and methodsCARTILAGEFemoral heads were obtained from patientswho underwent total hip replacement for bothprimary and secondary OA. After resection,the specimens were immediately immersedin sterile Hank's balanced salt solutionsupplemented with penicillin 50 IU/ml andstreptomycin 50 ,ug/ml, and processed withintwo hours. Control femoral heads wereobtained at postmortem, usually four to fivehours, but never in excess of eight hours, afterdeath, immediately placed in Hank's balancedsalt solution with penicillin-streptomycin (asabove) and amphotericin B 2-5 ,ug/ml, andprocessed within 30 minutes. Only femoralheads of individuals with no history of joint

Department ofRheumatology,BiochemicalLaboratory,University Hospital,Zurich, SwitzerlandA BaiciA LangD HorlerPhysical Medicine andRheumatologyDivision,Department ofOrthopaedic SurgeryBalgrist,University ofZurich,SwitzerlandR KisslingC MerlinCorrespondence to:Dr Antonio Baici,University Hospital,Department ofRheumatology,CH-8091 Zurich,Switzerland.

Accepted for publication14 November 1994

281

on 9 January 2019 by guest. Protected by copyright.

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.54.4.281 on 1 A

pril 1995. Dow

nloaded from

Baici, Hdrler, Lang, Merlin, Kissling

Superior

Figure 1 Superior extremity ofa right human femurshowing thefour regions from which articular cartilage wassampled. A: Anterior view ofan osteoarthritic head showingbone eburnation of the superior region (dashed area),'shelving' cartilage (dotted area beneath the dashed area)and marginal osteophytes. P: Posterior view of a normalhead.

disease were considered as controls. Oc-casionally, femoral heads obtained at post-mortem presented clear signs of OA. Thediagnosis was confirmed by an experiencedpathologist and, accordingly, the specimenswere attributed to the OA group. (To avoidconfusion, normal cartilage specimens were

identified by the prefix 'control' before theirnumber, while OA specimens were assigned a

three digit number.) The anatomical position(fig 1) from which cartilage was obtained was

documented by photographing the femoralheads from all sides before and after sampling,and designated anterior, posterior, superior, or

inferior.3' Usually the superior region of OAfemoral heads was levigated, with exposure ofan eburnated bone surface, and the inferiorregion, mainly towards the margin, was

covered by osteophytes.The cartilage type'8 was determined for each

specimen: normal articular cartilage fromrelatively young subjects without history ofjoint diseases (type I); slightly soft or fibrillatedcartilage from normal femoral heads obtainedfrom elderly subjects and different from OAcartilage (type III); residual (degenerating)cartilage obtained from OA femoral heads aftertotal hip replacement (type IV, typically havinga yellow colour); and nascent (regenerating)cartilage from OA femoral heads (type V, witha white colour). Often, type V cartilage was

clearly found growing above and in closecontact with residual (type IV) cartilage; thesespecimens were designated type IV + V. TypeII cartilage (from fractures of the neck of thefemur) was not available to us and osteophyticcartilage (type VI) could not be investigatedwith the technique used in this study becausethe cartilaginous layer was very thin. Cylindersof full thickness cartilage and underlying bonewith a diameter of 4 0 mm were obtained usinga hollow biopsy trephine (Straumann Ltd,Waldenburg, Switzerland). With this methodthe cartilage remained intact, and local tissuedamaging compression forces could beavoided. The samples were immediately im-mersed in physiological saline, thoroughlyrinsed, trimmed to leave 5-7 mm of boneunderneath the cartilage, and blotted dry. Thecylindrical samples were then rapidly attached

at the bone side to a cork disk, using a smalldrop of Tissue-Tek® OCT compound (Miles,IN, USA), frozen in a quick freeze carbondioxide chamber and stored at -70°C. Thespecimens were allowed to reach the cabinettemperature of -25°C before being cryo-sectioned as 20 ,um thick slices parallel to thearticular surface. This is a modification of apreviously published method.26 Five consecu-tive sections were collected in conical, 1 mlEppendorfer vials until the subchondral bonewas reached, and left at -25°C. Five sectionsobtained as described have a volume of 1-26mm3 and an area of 126&9 mm2, whereas asingle, 100 pum thick section would have thesame volume, but an area of only 26-4 mm2.The presence of the subchondral bone waseasily perceived by the resistance offered to theknife and the colour of the sections, so that oneor two sections, at the most, with mixed boneand calcified cartilage tissue were eventuallypresent in the deepest sample analysed in eachcylinder. Cartilage slices included superficial,mid and deep layers-more precisely, zonesI_V.32

HISTOLOGY

Measurement of cathepsin B in a cartilage plugdid not allow simultaneous histological analysiswithin the same specimen. Nevertheless, theseverity of the OA lesions was estimated at sitesimmediately adjacent to the sampled cylindersaccording to the histological-histochemicalgrading method of Mankin et al.33 To avoidartificial loss of proteoglycans during prepara-tion for paraffin embedding, cartilage samples(either cylinders obtained as described above,or slices with underlying bone resected fromthe area between two adjacent holes producedfor cylinder extraction) were snap frozen andcryosectioned at a thickness of 14-18 ,um.After 15 minutes fixation with 4% formalde-hyde in 0-1 mol/ sodium phosphate buffer, pH7 4, sections were rinsed with this buffer,brought to 50% and then to 80% ethanol (fiveminutes each), and sequentially stained withWeigert's iron chloride haematoxylin, fastgreen FCF, and safranin 0 before being per-manently mounted.34 Under these conditionsglycosaminoglycans were stained by safranin 0in the orthochromatic form35 and their loss wasmade visible by the presence of the greencounterstain. The only disadvantage of thismethod is that the subchondral bone may bedamaged in some sections and the tidemarkintegrity could not be evaluated.

CATHEPSIN B ASSAY

The frozen cartilage sections were thawedrapidly to disrupt cells by suspending them in0 50 ml of 0-10 mmol/l Z-Arg-Arg-NMec(abbreviations of amino acids and substitu-ents according to the Joint Commissionon Biochemical Nomenclature;36 -NMec = 7-(4-methyl)coumarylamide) (Bachem Ltd,Bubendorf, Switzerland) in 0-1 mol/l Na+/K+phosphate buffer containing 2 mmolI EDTAand 2 mmol/l dithiothreitol (DT), pH 6-0.

282

on 9 January 2019 by guest. Protected by copyright.

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.54.4.281 on 1 A

pril 1995. Dow

nloaded from

Cathepsin B in osteoarthritis: zonal variation ofenzyme activity in humanfemoral head cartilage

After incubation at 37°C under gentle shakingfor exactly four hours the reaction was stoppedwith 10 iil of 0-1 molIl iodoacetamide in water.The fluorescence of liberated 7-amino-4-methylcoumarin was read in an AmincoSPF-500 fluorimeter operating in the ratiomode with excitation and emission wave-lengths set at 383 and 455 nm, respectively.The recorder scale was calibrated with7-amino-4-methylcoumarin solutions ofknown concentration determined using anabsorption coefficient of 16 000 mol-1 1 cm-' at342 nm for this substance. The described assaywas optimised in preliminary experimentsusing cartilage specimens with high and lowcathepsin B activities, and using purifiedhuman spleen cathepsin B.37 In order for thereaction to be linear with time and enzymeconcentration, an excess of substrate wasensured.Z-Arg-Arg-NMec is a substrate with high

specificity for cathepsin B; it does not reactwith cathepsin H38 and has only trace activitywith cathepsin L.39 Possible interference bythese cysteine endopeptidases or otherenzymes was excluded by control studies. Forthis purpose, four adjacent plugs of cartilagewere taken from a region of an OA femoralhead with uniform thickness. Two plugs wereused to determine the uniformity of the zonalvariation of cathepsin B activity and two plugswere used to determine the effect of inhibitorsincluded in the assay solution in alternatesamples. The inhibitors used were 3,4-di-chloroisocoumarin, L-trans-epoxysuccinyl-leucylamido-(4-guanidino)-butane (E-64)(Sigma, St Louis, MO, USA), Z-Phe-Ala-CHN2, Z-Phe-Phe-CHN2, Pro-Phe-Har-CH2 Cl (a gift of Dr E Shaw, Basle,Switzerland), and leupeptin (Bachem Ltd).One unit of cathepsin B activity was defined

as the amount of enzyme that hydrolysed1 nmole of substrate in four hours at pH 6-0and 37°C. The activities shown in the figuresare expressed as units released by a 'sample'-that is, by five consecutive sections of cartilageobtained as described above. The activitiescould not be normalised to the cellularity of thesamples because the fluorimetric assay did notallow simultaneous determination of the DNAcontent.

STATISTICAL METHODS

Statistical comparisons were performed with atwo tailed Student's t test, taking into accountequality or inequality of variances as de-termined with Fisher's F test. Probabilityvalues (p) were explicitly calculated; p < 0-05was considered significant.

five consecutive sections were pooled foranalysis. Actual concentrations of 7-amino-4-methylcoumarin released in the assay werein the range 0-2-80 ,umo1l/, thus allowingaccurate readings with any commercialfluorimeter. This method allowed the topo-graphical mapping of cathepsin B in thecartilage of a whole joint to show itsquantitative distribution in the various zones.Both thiol activation (by DTT) and the

presence of EDTA were necessary to ensuremeasurement of cathepsin B activity. In thecontrol studies for assay specificity, the activitywas completely abolished by Pro-Phe-Har-CH2 Cl, a potent inhibitor of cathepsin B,40 andby leupeptin. Good inhibition was obtainedwith Z-Phe-Ala-CHN2, and E-64, whereas noinhibition was observed with Z-Phe-Phe-CHN2 or with the general serine endopeptidaseinhibitor 3,4-dichloroisocoumarin.41 Thesedata (fig 2) confirmed the cysteine endo-peptidase character of the measured enzymeand are highly characteristic of cathepsin B.38

Cathepsin B activity in zones of normalcartilage (type I) never exceeded seven units/sample (fig 3). Activity as a function of thedistance from the subchondral bone was quiteregular, generally greater towards the articularsurface (zone I) and lower in the mid zones (IIand III). In the deeper zones (IV and V),catheptic activity was either the same asobserved in the central zones, or slightlygreater. Cell density in zones of articularcartilage is greater in the superficial layer thanin the mid and deep zones, where it is quiteuniform.42 Zonal variations of cathepsin Bactivity in normal cartilage can thus beattributed to cell heterogeneity, differentmetabolism, and different density. As alreadysuggested,26 greater activity in the superficiallayer is a property of the fibrocyte typechondrocytes of this zone. Type III cartilage,obtained from elderly subjects (controls 6 and8 in figure 3) was not fibrillated and was stillconsidered to be 'normal' on the basis ofmacroscopic morphology, the only difference

20 1 E

D-a,

E

15 -

>~10-

c0

V) D

C)

0

ResultsThe method for measuring the zonaldistribution of cathepsin B activity in articularcartilage was highly sensitive and could beperformed by using just one section of tissue4 mm in diameter and 20 ,um thick. However,to save time and reduce costs (100 sections areobtained from a 2 mm deep plug of cartilage),

E2.

(~jI?

E2.

7

E:. 0

E 7E2. awH

z z C

E co @ I

c I° a,

o a,-

a, ;:

0 0-5 1 0 1-5

Distance from the subchondral bone (mm)Figure 2 Zonal variation of cysteine endopeptidaseactivity in two plugs (continuous and dotted lines) ofOAfemoral head cartilage with uniform thickness showing theeffect ofselected substances, added to alternate samples asshown, in inhibiting enzyme activity. Two other plugs wereused to show the continuity ofcathepsin B activity in thesame area (not shown). A sharply decreased activity in agiven sample indicates inhibition. DTT= Dithiothreitol.

iI

283

on 9 January 2019 by guest. Protected by copyright.

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.54.4.281 on 1 A

pril 1995. Dow

nloaded from

Baici, Horler, Lang, Merlin, Kissling

Control-2 d 37 yr, R

A ..AI.. .--I zi

0-5 1-0 15 2-0 2-5

Control-6 9 85 yr, R

p,-.--P1-S

-..". J::,-~~~~...*S1

0 0-5 1-0 1-5 2-0

Control-3 9 35 yr, L

0-5 1-0 1-5 2-0 2-5

Control-B 9 78 yr, L and R

A(R) P(

0 06 1-0 1-5 2-0 25

Control-5 9 30 yr, L

0 0-5 10 1 5 2-0 25

Control-9 d 54 yr, L and R

p( P(L) A(R)

-n ~~~~~ji';s< A(L)- * -S(R)

0 05 1.0 15 2-0 2-5

Distance from the subchondral bone (mm)Figure 3 Zonal variation of cathepsin B activity in normal (control) femoral head cartilage. A, P, S, and I indicate theanterior, posterior, superior, and inferior regions, respectively. R = right, L = leftfemoral head.

from young cartilage being a slightly softerconsistency.The zonal variation of cathepsin B activity in

OA cartilages was less regular (figs 4-6). Whilesome OA specimens had cathepsin B profilessimilar to those of the controls, others hadconsiderably greater activities both towards thesurface (examples: fig 4, 011; fig 5, 019 and023; fig 6, 015) and in the deep zones(examples: fig 4, 032 and 028; fig 5, 026;fig 6, 018). Large variations in the activityprofiles were sometimes observed in samplesfrom the same anatomical site, as shown infigure 6, where the axes of the two cartilageplugs from the posterior region of patient 029,

and those of the anterior region of patient 018,were about 9 mm apart from each other. In thisstudy we examined cartilages from 20 OAfemoral heads (54 samples) and from both theleft and right femoral heads of nine controls(42 samples). (For reasons of space not all ofthe specimens analysed are shown in figures3-6, but all the available data have been usedin figure 7 and 8 for purposes of statisticalanalysis.) Although the primary indication forsurgery of patient 011 was hip necrosis, thecartilage had typical OA lesions.

Statistical evaluation of zonal cathepsin Bactivity in cartilage (figs 3-6) poses a practicalproblem. While comparison of activity between

011 d 41 yr,hip necrosis A

L

....

0 0-5 1.0 1-5 2-0

027 9 79 yr, PrOA, L

0 05 1-0 15 2-0

014 9 64 yr, SOA(dysplasia). L

032 d 75 yr, PrOA, R ..................

A

.....AS

0 0-5 1-0 1-5 2-0

028 d 43 yr, SeOA (dysplasia), LA A

.f*t-P

0 p

O 0-5 1-0 1-5 2 0

017 d 69 yr, PrOA, L

A

0 0-5 1.0 15 2-0

031 d 46 yr, PrOA. L

A

A

0 05 1-0 1-5 20

Distance from the subchondral bone (mm)Figure 4 Zonal variation ofcathepsin B acivity in osteoarthriticfemoral head cartilage from regions A (anterior) and P(posterior). A short diagnosis is given: Pr = primary idiopathic, Se = secondary, OA = osteoarthritis. R = right, L = leftfemoral head.

20

15

0

£ 10E0 5.

0

2-6°

10

0m

5 -

20

15

0

a 10E'uU' 54w._

0

.--20

m

.C 15U,

0.C 10

5

0

u I I

284

I

I

on 9 January 2019 by guest. Protected by copyright.

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.54.4.281 on 1 A

pril 1995. Dow

nloaded from

Cathepsin B in osteoarthritis: zonal variation ofenzyme activity in humanfemoral head cartilage

ILi-r-p

0-5 1-0 1-5 2-0 2-54

019 77 yr, SeOA (traumatic) --AL

AA

05 10. 1i 2-0 2-5

p021 74 yr, PrOA, R 023 d 53 yr, PrOA,L

-PP

0 0-5 1-0 1.5 i0 2A 3-0 3 5 0 0-5 1-0 1-5 2-0 2-5 3-0 3-5

024 d 84 yr, PrOA, L

..

;I.......A A-, ..., ;

.....I..

0 0-5 1-0 1-5 2-0 2-5

026 d 73 yr, SaOA(protrusio acetabuli), R

L. ,- A

; . ;~~~~~~~~~~~~...!:..........O 0-5 1-0 1-5 2-0 2-5 3-0 3-5

Distance from the subchondral bone (mm)Figure 5 Zonal variation ofcathepsin B activity in osteoarthriticfemoral head cartilagefrom regionsA (anterior), P(posterior), and I (inferior). Other abbreviations as in figure 4.

zones of diseased and control tissue soundsmathematically correct, this comparison wouldresult in an anatomical nonsense. The 'zones'in articular cartilage are a descriptive concept,but there are no boundaries between them andtheir thickness varies within a joint and withinindividuals. Moreover, the original thickness ofpartly degraded OA cartilage, where somelayers have been worn away, is unknown.Therefore, statistical comparison of cathepsinB activities in normal and OA cartilages wasbased on a normalised (fractional) activity ofeach specimen, namely the ratio of the areaunder the curves shown in figures 3-6 to thecartilage depth in mm.

The dependence of the normalised activitiesupon the anatomical site within the femoralhead is shown in figure 7A, together with thecartilage type. In normal subjects cathepsin Bactivity was similar in the four regions of thefemoral head: the mean values for theposterior, anterior and inferior regions werestatistically indistinguishable from one another,while the superior region had significantly lessactivity than the others (p = 0-0041 comparedwith posterior). The posterior, anterior andinferior regions of OA cartilage had statisticallyindistinguishable mean values of cathepsin Bactivity when compared with one another.Comparison of the activities from femoral head

40-P015 d 59yr p

PrOA, L30

Remaining cartiagecovered by a thick,

20 invading pannus.Only cartilaginouslayer shown here.

10 -

0 00 0-5 10o 1.5 2-0 2-5

20

15

10

5

0

I ...~~~~~~~~~~~~:_018 d66 yr, PrOA,RA.

.. ~~~~~A...... .. *-. ,.-,.

1- -i. 5 2

022 g 53 yr, SeOA(dysplasia), L

05 1-0 1-5 2-0 2-5

030 d 80 yr, PrOA, R

.P

.-p W

0 0-5 1-0 1-5 2-0 2-5 3-0 36 0

029 d 71 yr, PrOA, R

0-5 1-0 1-5 2-0 2-5 3.0

035 9 85 yr, PrOA, L

A

A

0-5 1-0 1-5 2-0

Distance from the subchondral bone (mm)Figure 6 Zonal variation ofcathepsin B activity in osteoarthriticfemoral head cartilagefrom regions A (anterior), P(posterior), and I (inferior). Other abbreviations as in figure 4.

20 1016 d 65 yr, PrOA,L

15 -

10

5

0I

20

15

10

a

Em0

2-4,)

*-

0to

r

._

._

OLu0

5

0

0'aEU)-0

._m

0.

0

-t

._

0

._

5el

*

285

on 9 January 2019 by guest. Protected by copyright.

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.54.4.281 on 1 A

pril 1995. Dow

nloaded from

Baici, Hdrler, Lang, Merlin, Kissling

regions in OA with those in normal cartilagerevealed significant differences between pos-terior (p = 0-0040) and anterior (p = 0-0014)sites. A p value of 0-0618 for the two-tailedcomparison at the inferior regions became sig-nificant when subjected to a one tailed t test.The superior region of all OA femoral headswas eburnated, with too little residual'shelving' cartilage to allow analysis for zonalvariations of enzyme activity. The mean of all42 normal (control) samples (2-8 (SD 0-8,SEM 0-13)) and the mean of all 54 OA samples(5-5 (SD 3-1, SEM 0-4)) were significantlydifferent (p = 9 x 10-8).The data of figure 7A are plotted as a

function of age in figure 7B, in which the lackof correlation between cathepsin B activity anddonor's age in the cartilage types I and III isevident. OA cartilages from relatively youngsubjects had the greatest activities and a largeproportion of typical OA samples from elderlysubiects had activities comoarable to those ofthe controls. The slope of thInificantly different from zero,

15-

10-E -

E -

a4-ra,-oQ)0)(U0

(U0

,

-

>1.4_

mC

a)Q

0

0._L

E0z

5-

0

15 -

10-

5-

0

Cartilage types0 1

A Ill

* IV* V* IV + V

A o

Mean ofall= 5-5

Mean of80 all = 2-8

)8 4 -

0b

P A SNormal c

*

_ 12- P

E

m 8-c -X

X C p=000024

0)

~ 0- 17T

z 04 35 7 24v'

11 I

= 00112

p < 0 0001

9

V

T

IV + VCartilage types

Figure 8 Data offigure 7plotted according to cartilagetype. The number ofspecimens in each group is shownwithin the boxes, bars indicate SD, and p values show thestatistical significance of differences of the means withrespect to type I cartilage.

e curve, sig- inverse correlation between enzyme activitysuggests an and age in the OA group.

Figure 8 shows the dependence of cathepsinB activity on cartilage type using all pooleddata. With cartilage type I as the reference, the

** * other types showed significantly higherA activities, although the difference between the

means of types I (2-7 (SD 0-8)) and III (3-5(0-8)) was small.The histological-histochemical grade of the

OA lesions33 was determined for all of thepatients and controls represented in figures

4 1 * 3-6, in at least one cartilage specimenimmediately adjacent to a plug analysed forcathepsin B activity (fig 9). Many moresamples were analysed than are shown here; for

q ,e all of them, structure, cellularity, and safranin0 staining could be determined and these datawere used for evaluation. For reasons ex-

P A plained in the Methods section, it was notDsteoarthritic always possible to judge tidemark integrity

of the samples; figure 9 shows the samples

B

30 40 50 60 70 80 90Age (years)

Figure 7 Normalised cathepsin B activity in femoral headcartilage. A: All data availablefor normal andosteoarthritic specimens showing regional variation(P= posterior, A = anterior, S = superior, I= inferior). B:Data from A plotted as a function of the donor's age. Thelinear regression lines (solid) are surrounded by theirrespective 95% confidence interval envelopes, shown asdashed lines for the cartilage types IV, Vand IV+ V (OAgroup) and as dotted linesfor cartilage types I and III(controls). Control group: correlation = 0-267, t = 1 -751,p = 0-0876; OA group: correlation = -0-388, t = -3-039,p = 0-0037.

E;:E

0.C a

ciaa

to' 0).--00=

-=

a) XUenUEn

"co D~z

15 -

10 -

5 -

0 -

Cartilage types01*A Ill

* IV* V ** IV+V -

.--4

+- -_,1

A * s

U *0

I II ;I

0 1 2 3 4 5 6 7 8 9 10 12 14Mankin's histopathological score

Figure 9 Dependence of cathepsin B activity upon theseverity of the lesions in femoral head cartilage. The pointfor type I cartilage was averagedfrom six specimens whilethe others represent single observations. The Mankinhistopathological score was independently determined by twoobservers and the mean value is shown (accountingfor theoccasionalfractional values). The slope of the regression line(correlation = 0-283, t = 1 -319, p = 0-217) surrounded bythe 95% confidence interval envelope, is not significantlydifferentfrom zero.

286

on 9 January 2019 by guest. Protected by copyright.

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.54.4.281 on 1 A

pril 1995. Dow

nloaded from

Cathepsin B in osteoarthritis: zonal variation ofenzyme activity in human femoral head cartilage

for which all four parameters could be un-equivocally determined. The slope of the curvewas statistically indistinguishable from zeroand no correlation was found betweencathepsin B activity and the Mankin score.

DiscussionOur results agree with and extend previousstudies on cathepsin B in human articularcartilage that were performed with tissuepooled from femoral heads of differentsubjects'8 26 or with cartilage pooled fromvarious regions and zones of tibial plateaus andfemoral condyles.27 Pathological specimensobtained at surgery represent just oneparticular moment of a disease the origins ofwhich are remote and which has a past historythat can probably not be traced back in itsparticular pathogenetic traits. The verydifferent cathepsin B activities observed in OAcartilages indicated great disease hetero-geneity, not only between different patients,but also between regions and zones from thesame femoral head, thus confirming the focalcharacter of the OA process, which may showa wide spectrum of activity in closely juxta-posed anatomical positions.43 We exploited thisproperty for interpreting the possible patho-genetic role of cathepsin B in various phases ofthe disease. Each specimen of OA cartilagerepresented a case that could be understoodand interpreted after a close examination ofthe medical history, and the macroscopic,enzymatic, and histological-histochemicalproperties. Thus the lack of correlation be-tween cathepsin B activity and the Mankinscore33 is not surprising and makes sense whenconsideration is given to the individual contri-butions of the four parameters that make upMankin's grading system. Although thepresent data do not demonstrate absolute facts,we observed some significant trends, asdescribed below.The mean cathepsin B activity in all pooled

data of residual cartilage (type IV) was onlyslightly greater than that of normal (type I)cartilage because the activity of a largeproportion of the type IV specimens was in therange of the controls, despite clear signs of OA.While this is an effect of data pooling, we stressthe importance of individual evaluation ofzonal cathepsin B distribution for identifyingfocal sites of high activity. Type IV cartilagerepresents the old, residual tissue that mayshow all possible stages of degradation, froman almost intact structure to completedisorganisation. Type IV samples with lowgrade destruction, almost normal cellularity,slightly impaired safranin 0 staining in thesuperficial or deeper zones, and undamagedtidemark, had low Mankin scores and lowcathepsin B levels. This cartilage can beconsidered to represent a tissue in an initialphase of OA. Other type IV specimens showedclear signs of degeneration, hypercellularity orcloning, impaired safranin staining, tidemarkdestruction without duplication, and highcathepsin B activity. We consider this cartilageas representative of active OA in a destructive

phase, but not yet with a reparative reaction.The type IV cartilage specimens with clearsigns of degeneration and low levels ofcathepsin B were either hypocellular orhypercellular, but with many remnants ofchondrons and cell clusters containing highlydegenerated, atrophied, and possibly deadcells. The overall structure of these specimenswas from irregular to completely disorganised,with severely reduced safranin 0 staining andthe tidemark duplicated, triplicated, ordamaged. This cartilage, representing a highlydegenerated tissue with a past history of activedisease, thus had high histological scores andlow cathepsin B levels.The greatest enzyme activities were found in

type V cartilage and in those specimens havingcartilage type V overgrowing type IV. Ihesesamples had zonal cathepsin B activities thatwere up to 20-fold greater than those found incorresponding zones of control cartilage.Histological analysis suggested that particu-larly high activity was present in zones wherehyperactive chondrocytes were involved in thereparative processes by synthesising cartilage ofa 'juvenile' type.26 At these sites the chondro-cytes were in clusters within a completely dis-organised interterritorial matrix showingstrong pericellular safranin 0 staining. Thesecartilage specimens covered a broad spectrumof cathepsin B activities that were alwaysgreater than controls, and a broad range ofhistological scores.The distribution of enzyme activity in OA

cartilages with various degrees of degenerationsuggests a pathological role for cathepsin B.The relatively low activities in apparently intactand severely degraded OA cartilages, incontrast with high levels in tissues with activedisease and at regenerating sites, suggest thatcathepsin B may act as a perpetuator ratherthan an initiator of cartilage destruction in OAand may be at least partly responsible forunsuccessful tissue repair. This concept issupported by cytochemical and histochemicalresults discussed in the accompanying paper.44

We are grateful to Professor A Schreiber and his staff at theBalgrist Orthopaedic University-Clinic, Zurich, for supplyingthe surgical specimens, and to Professor J Schneider and Mr NAlder, Department of Pathology, University Hospital, Zurich,for their invaluable help in obtaining postmortem specimens.The study was supported by grants from the Albert Boni-Stiftung fur Rheumaforschung and the Ciba-Geigy-Jubilaumsstiftung.

1 Dingle J T, Saklatvala J, Hembry R, Tyler J, Fell H B, JubbR. A cartilage catabolic factor from synovium. Biochem J1979; 184: 177-80.

2 Tyler J A. Chondrocyte-mediated depletion of articularcartilage proteoglycans in vitro. Biochem J 1985; 225:493-507.

3 Meats J E, McGuireM B, Russell R G G. Human synoviumreleases a factor which stimulates chondrocyte productionof PGE and plasminogen activator. Nature 1980; 286:891-2.

4 Deshmukh-Phadke K, Lawrence M, Nanda S. Synthesis ofcollagenase and neutral protease by articular chondro-cytes. Stimulation by a macrophage derived factor. Bio-chem Biophys Res Commun 1978; 85: 490-6.

5 Watanabe S, Georgescu H I, Kuhns D B, Evans C H.Chondrocyte activation by a putative interleukin-Iderived from lapine polymorphonuclear leukocytes. ArchBiochem Biophys 1989; 270: 69-76.

6 Ollivierre F, Gubler U, Towle C A, Laurencin C, TreadwellB V. Expression of IL-I genes in human and bovinechondrocytes: a mechanism for autocrine control ofcartilage matrix degradation. Biochem Biophys Res Com-mun 1986; 141: 904-i1.

287

on 9 January 2019 by guest. Protected by copyright.

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.54.4.281 on 1 A

pril 1995. Dow

nloaded from

Baici, Hdrler, Lang, Merlin, Kissling

7 Woessner J F Jr, Howell D S, eds. Joint cartilage degradation.Basic and clinical aspects. New York: Marcel Dekker,1993.

8 Kuettner K E, Schleyerbach R, Peyron J G, Hascall V C,eds. Articular cartilage and osteoarthritis. New York: RavenPress, 1992.

9 Lucy J A, Dingle J T, Fell H B. Studies on the mode ofaction of excess vitamin A. 2. A possible role ofintracellular proteases in the degradation of cartilagematrix. BiochemJ3 1961; 79: 500-8.

10 Ali S Y. The degradation of cartilage matrix by anintracellular protease. Biochem

_1964; 93: 611-8.

11 Ali S Y. The presence of cathepsin B in cartilage. Biochem_J 1967; 102: 10-ll C.

12 Ali S Y, Evans L, Stainthorpe E, Lack C H. Characterizationof cathepsins in cartilage. Biochem_J 1967; 105: 549-57.

13 Morrison R I G, Barrett A J, Dingle J T, Prior D. CathepsinsB 1 and D. Action on human cartilage proteoglycans.Biochim Biophys Acta 1973; 302: 411-9.

14 Burleigh M C, Barrett A J, Lazarus G S. A lysosomalenzyme that degrades native collagen. Biochem J 1974;137: 387-98.

15 Etherington D J. The purification of bovine cathepsin B1and its mode of action on bovine collagens. Biochem J1974; 137: 547-57.

16 Mort J S, Recklies A D, Poole A R. Extracellular presenceof the lysosomal proteinase cathepsin B in rheumatoidsynovium and its activity at neutral pH. Arthritis Rheum1984;27: 509-15.

17 Hembry R M, Knight C G, Dingle J T, Barrett A J. Evidencethat extracellular cathepsin D is not responsible for theresorption of cartilage matrix in culture. Biochim BiophysActa 1982; 714: 307-12.

18 Ali S Y, Bayliss M T. Enzymes involved in degradation ofcartilage in osteoarthrosis. Ann Rheum Dis 1975; 34(Suppl2): 65-6.

19 Ali S Y, Evans L. Enzymic degradation of cartilage inosteoarthritis. Fed Proc 1973; 32: 1494-8.

20 Vittorio N, Crissman J D, Hopson C N, Herman J H.Histologic assessment of cathepsin D in osteoarthriticcartilage. Clin Exp Rheumatol 1986; 4: 221-30.

21 Codorean E, Buzila L, Gabrielescu E. Immunocyto-chemical investigation of cathepsin B in the rheumatoidsynovial membrane and fluid. Rev Roum Biochim 1981;18: 241-7.

22 Trabandt A, Gay R E, Fassbender H G, Gay S. CathepsinB in synovial cells at the site of joint destruction inrheumatoid arthritis. Arthritis Rheum 1991; 34: 1444-5 1.

23 Lenarcic B, Gabrijelcic D, Rozman B, Drobnic-Kosorok M,Turk V. Human cathepsin B and cysteine proteinaseinhibitors (CPIs) in inflammatory and metabolic jointdiseases. Biol Chem Hoppe-Seyler 1988; 369: 257-61.

24 Gabrijelcic D, Annan-Prah A, Rodit B, Rozman B, CotitV, Turk V. Determination of cathepsins B and H in seraand synovial fluids ofpatients with different joint diseases._J Clin Chem Clin Biochem 1990; 28: 149-53.

25 Huet G, Flipo R M, Richet C, et al. Measurement of elastaseand cysteine proteinases in synovial fluid of patients withrheumatoid arthritis, sero-negative spondylarthropathies,and osteoarthritis. Clin Chem 1992; 38: 1694-7.

26 Bayliss M T, Ali S Y. Studies on cathepsin B in humanarticular cartilage. Biochem3r 1978; 171: 149-54.

27 Martel-Pelletier J, Cloutier J M, Pelletier J P. Cathepsin Band cysteine protease inhibitors in human osteoarthritis.J Orthop Res 1990; 8: 336-44.

28 Baici A, Lang A, Horler D, Knopfel M. Cathepsin B as amarker of the dedifferentiated chondrocyte phenotype.Ann Rheum Dis 1988; 47: 684-91.

29 Baici A, Lang A. Cathepsin B secretion by rabbit articularchondrocytes: modulation by cycloheximide and glycos-aminoglycans. Cell Tissue Res 1990; 259: 567-73.

30 Baici A, Lang A. Effect of interleukin- Il on the productionof cathepsin B by rabbit articular chondrocytes. FEBSLett 1990; 277: 93-6.

31 Grushko G, Schneiderman R, Maroudas A. Some bio-chemical and biophysical parameters for the study ofthe pathogenesis of osteoarthritis: a comparison be-tween the processes of ageing and degeneration inhuman hip cartilage. Connect Tissue Res 1989: 19:149-76.

32 Gardner D L. Pathological basis ofthe connective tissue diseases.London: Arnold, 1992.

33 Mankin H J, Dorfman H, Lippiello L, Zarins A.Biochemical and metabolic abnormalities in articularcartilage from osteo-arthritic human hips. II. Correlationof morphology with biochemical and metabolic data. 7Bone J'oint SurgAm 1971; 53: 523-37.

34 Lillie R D. Histopathologic technic and practical histochemistry,3rd edn. New York: McGraw-Hill, 1965.

35 Rosenberg L. Chemical basis for the histological use ofsafranin 0 in the study of articular cartilage. 7 Bone J7ointSurgAm 1971; 53: 69-82.

36 IUPAC IUB Joint Commission on Biochemical Nomen-clature (JCBN). Nomenclature and symbolism for aminoacids and peptides. Recommendations 1983. Biochem J1984;219: 345-73.

37 Baici A, Gyger-Marazzi M. The slow, tight-binding in-hibition of cathepsin B by leupeptin. A hysteretic effect.EurJBiochem 1982; 129: 33-41.

38 Barrett A J, Kirschke H. Cathepsin B, cathepsin H, andcathepsin L. Meth Enzymol 1981; 80: 535-61.

39 Kirschke H, Kembhavi A A, Bohley P, Barrett A J. Actionof rat liver cathepsin L on collagen and other substrates.BiochemJ 1982; 201: 367-72.

40 Shaw E, Kettner C. The specificity of cathepsin B. Acta BiolMed Germ 1981; 40: 1503-11.

41 Harper J W, Hemmi K, Powers J C. Reaction of serineproteases with substituted isocoumarins: discovery of3,4-dichloroisocoumarin, a new general mechanismbased serine protease inhibitor. Biochemistry 1985; 24:1831-41.

42 Stockwell R A, Meachim G. The chondrocytes. In:Freeman M A R, ed. Adult articular cartilage. London:Pitman Medical, 1973; 51-99.

43 Weiss C. Ultrastructural characteristics of osteoarthritis. FedProc 1973; 32: 1459-66.

44 Baici A, Lang A, Horler D, Kissling R, Merlin C. CathepsinB in osteoarthritis: cytochemical and histochemicalanalysis of human femoral head cartilage. Ann Rheum Dis1995; 54: 289-297.

288

on 9 January 2019 by guest. Protected by copyright.

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.54.4.281 on 1 A

pril 1995. Dow

nloaded from