• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of amjpatholAmerican Journal of Pathology For AuthorsAmerican Journal of Pathology SubscribeAmerican Journal of Pathology SearchAmerican Journal of Pathology Current IssueAmerican Journal of Pathology About the JournalAmerican Journal of Pathology
Am J Pathol. Dec 2001; 159(6): 2167–2177.
PMCID: PMC1850593

Cathepsin K Is a Critical Protease in Synovial Fibroblast-Mediated Collagen Degradation


Synovial fibroblasts (SFs) play a critical role in the pathogenesis of rheumatoid arthritis (RA) and are directly involved in joint destruction. Both SF-resident matrix metalloproteases and cathepsins have been implicated in cartilage degradation although their identities and individual contributions remain unclear. The aims of this study were to investigate the expression of cathepsin K in SFs, the correlation between cathepsin K expression and disease severity, and the contribution of cathepsin K to fibroblast-mediated collagen degradation. Immunostaining of joint specimens of 21 patients revealed high expression of cathepsin K in SFs in the synovial lining and the stroma of synovial villi, and to a lesser extent in CD68-positive cells of the synovial lining. Cathepsin K-positive SFs were consistently observed at sites of cartilage and bone degradation. Expression levels of cathepsin K in the sublining and vascularized areas of inflamed synovia showed a highly significant negative correlation with results derived from the Hannover Functional Capacity Questionnaire (r = 0.78, P = 0.003; and r = 0.70, P = 0.012, respectively) as a measure of the severity of RA in individual patients. For comparison, there was no correlation between Hannover Functional Capacity Questionnaire and cathepsin S whose expression is limited to CD-68-positive macrophage-like synoviocytes. The expression of cathepsin K was also demonstrated in primary cell cultures of RA-SFs. Co-cultures of SFs on cartilage disks revealed the ability of fibroblast-like cells to phagocytose collagen fibrils whose intralysosomal hydrolysis was prevented in the presence of a potent cathepsin K inhibitor but not by an inhibitor effective against cathepsins L, B, and S. The selective and critical role of cathepsin K in articular cartilage and subchondral bone erosion was further corroborated by the finding that cathepsin K has a potent aggrecan-degrading activity and that cathepsin K-generated aggrecan cleavage products specifically potentiate the collagenolytic activity of cathepsin K toward type I and II collagens. This study demonstrates for the first time a critical role of cathepsin K in cartilage degradation by SFs in RA that is comparable to its well-known activity in osteoclasts.

Rheumatoid arthritis (RA) is an inflammatory joint disorder characterized by a progressive destruction of articular cartilage and subchondral bone that eventually leads to loss of joint function. RA cartilage erosion is a three-directional process caused by activated chondrocytes, a cartilage-overlaying synovial pannus, and a subchondral cellular infiltration of the hyaline cartilage. Subchondral cartilage is primarily degraded by osteo/chondroclasts, multinucleated giant cells, and mononuclear precursor cells, although invading synovial fibroblasts (SFs) are considered the main culprit cells in pannus-mediated joint destruction. 1-4

Although the degradation of the extracellular matrix in joints is clearly mediated by proteolytic activities, the nature of the individual proteases remains unknown in most cases. To date, two protease families have been implicated in cartilage degradation: metalloproteinases of the MMP and ADAMs families, and cysteine proteases. 5-7 Traditionally, metalloproteinases have been favored as potential culprit enzymes over cysteine proteases, although inhibitors of both protease classes have proved to be equally effective in reducing inflammation and cartilage erosion in animal models of RA. 8,9 Recent advances in the identification and characterization of novel cysteine proteases have directed increased attention to the cathepsins as potential drug targets to treat tissue degenerative and inflammatory processes. Current interest is focused on the roles of cathepsin K in bone resorption and cathepsin S in antigen presentation. 10 Cathepsin K has been identified as the predominant osteoclastic protease with a unique and potent collagenolytic activity. 11-14 The critical involvement of cathepsin K in bone remodeling is best supported by the finding that cathepsin K deficiency causes the bone-sclerosing disorder pycnodysostosis, 15 that, on the molecular level, is characterized by insufficient degradation of type I collagen during bone remodeling. 16 In contrast, the nature of the proteases responsible for the cartilage erosion by SFs remains elusive although several matrix metalloproteinases and cathepsins L and B have been prime suspects. 6,17 Here, we report the specific expression of cathepsin K in the inflamed RA synovium and discuss the potential involvement of this protease in SF-mediated bone and cartilage degradation and synovial remodeling.

Materials and Methods

Patient Characteristics

All 21 patients fulfilled the American College of Rheumatology-revised criteria for the classification of RA. 18 All were positive for rheumatoid factor and had radiographic erosions. Parameters of disease activity were recorded as follows: 1) the swollen joint count evaluated 28 joints (shoulders, elbows, wrists, metacarpophalangeal-, interphalangeal-, proximal interphalangeal-, and knee-joints) as described in Fuchs and Pincus. 19 2) The modified Lansbury index reflects the number of swollen joints adjusted for their relative size. It is comparable to the area weighted swollen joint index described in van Leeuwen and colleagues. 20 This parameter is distinct from the Lansbury index in that it does not refer to “pain on motion.” In our opinion, the latter can also be influenced by degenerative changes without inflammation. 3) The Keitel functional index 21 is based on the scoring of limitations of joint movements observed by the physician. This test evaluates the function of hands, wrists, shoulders, and lower limbs by grading defined joint movements. 4) The Hannover Functional Capacity Questionnaire (HFCQ) was performed on 12 of 21 patients. This questionnaire is the most widely used questionnaire in Germany. It has been validated by several studies and shows good correlation (r = 0.87) with the health assessment questionnaire (HAQ). 22 Severe impairment of joint function is defined as a reduction of the HFCQ below 50% 22 ) Visual analog scale of joint pain was assessed using a scale ranging from 0 (no pain) to 10 (most severe pain). 6) The laboratory parameters included the erythrocyte sedimentation rate that was determined by the Westergren method. C-reactive protein was measured by nephelometry (monarch 2000; Instrumentation Laboratory GmbH, Kirchheim Heimstetten, Germany) using specific antiserum (Biokit GmbH; Kirchheim Heimstetten, Germany). Patient characteristics are given in Table 1 [triangle] .

Table 1.
Patient Characteristics

Synovial Specimen

Synovial tissue samples were obtained at the time of synovectomy (n = 10), or reconstructive surgery of joints, mostly resection of metatarsal heads (n = 11). All 21 tissue samples were obtained from the Department of Rheumatology, Rostock, Germany. The respective joints were metatarsophalangeal (n = 13), wrist (n = 5), metacarpophalageal (n = 2), and knee joint (n = 1). The samples were fixed in 5% buffered formalin and embedded in paraffin wax. The paraffin blocks were labeled with a numerical code only and the staining and the evaluation of the slides was performed without knowledge of any activity parameters of the patients.


Paraffin sections (5 μmol/L) were mounted onto Vectabond slides (Vector Laboratories, Burlingame, CA), dried overnight, and dewaxed with xylene. Sections were incubated with mouse monoclonal anti-human cathepsin K antibody at a dilution of 1:200 as previously described 23 for 1 hour at room temperature in a humidified chamber, followed by incubation with biotinylated goat anti-mouse secondary antibody for 20 minutes. Peroxidase-labeled avidin was used to localize the secondary antibody with oxidized diaminobenzidine as chromogen (Biogenix, San Ramon, CA). Counterstaining was performed with Meyer’s hematoxylin and eosin. All slides were evaluated using a departmental Nikon Eclipse E800 microscope.

For double-immunofluorescence staining, we used a rabbit polyclonal anti-human cathepsin K antibody as previously described. 23 Secondary antibodies were replaced by anti-mouse antibody conjugated with tetramethylrhodamine isothiocyanate and anti-rabbit antibody conjugated with fluorescein isothiocyanate (Sigma, St. Louis, MO).

Monoclonal anti cathepsin K antibodies were generated by immunization of A/J mice with purified bacterially-expressed cathepsin K. After an initial intraperitoneal boost of 50 μg of cathepsin K in complete Freund’s adjuvant, the mice were allowed to rest for 8 weeks. They were then boosted intraperitoneally at 2-week intervals; first with 25 μg of cathepsin K in incomplete Freund’s adjuvant; then with 10 μg of cathepsin K in 10 mmol/L phosphate, 150 mmol/L sodium chloride, pH 7.2; and finally with 5 μg of cathepsin K in 10 mmol/L phosphate, 150 mmol/L sodium chloride, pH 7.2, intravenously. The spleens of two mice were then fused in accordance with standard procedures. Supernatants were screened with a solution-phased enzyme-linked immunosorbent assay in which yeast-expressed (activated) and bacterially expressed cathepsin K bound to a microplate well was incubated with supernatant to be tested. Sequential incubations with anti-mouse IgG-horseradish peroxidase conjugate and then substrate allowed detection of anti-cathepsin K antibodies. Clone M21 was selected for further use and purified from hybridomas grown in HB Pro serum-free media (Irvine Scientific, Santa Ana, CA). Terminal culture supernatants were harvested by centrifugation when cell viability fell to <10 to 20%. Supernatants were purified over fast-flow Protein A-Sepharose (Amersham/Pharmacia, Piscataway, NJ) after 0.2-μm (pore size) filtration and addition of saturated sodium borate and sodium chloride to final concentration of 100 g/L and 3 mol/L, respectively. Bound antibody was eluted with 0.1 mol/L of glycine, pH 3.0, and fractions were neutralized by the addition of 10% (v/v) 1.2 mol/L of Tris, pH 8.5. Antibody-containing fractions were pooled and dialyzed into 10 mmol/L of phosphate and 150 mmol/L of sodium chloride, pH 7.2.

For control staining experiments, the antibody was preincubated for 1 hour with nitrocellulose paper strips containing recombinant human cathepsin K. Briefly, 100 μg of recombinant cathepsin K was applied to sodium dodecyl sulfate-polyacrylamide gel electrophoresis using 4 to 20% Tris/glycine gels and the protein was electroblotted onto nitrocellulose (Fisher Scientific, Pittsburgh, PA). The cathepsin K band with an apparent molecular weight of 29 was excised after Ponceau staining and the nitrocellulose strip was blocked with 9% nonfat milk powder in phosphate-buffered saline and subsequently used for affinity binding of anti-cathepsin K antibodies. Anti-cathepsin K antibody-depleted antibody solution did not reveal any detection of cathepsin K-positive material in tissue sections.

Semiquantitative Analysis

Tissue sections were coded and randomly analyzed by an investigator blinded to the origin and history of the samples. Tissue sections were scored separately for the presence of cathepsin K in the intimal lining layer, the interstitial stromal regions, and vascular-rich areas. Scores were assigned semiquantitatively on a 0 to 4 scale: 0, no staining; 1, rare positive staining or trace staining (1 to 5%); 2, scattered clusters of positive cells (5 to 20%); 3, moderate staining in a specific region (21 to 50%); and 4, extensive staining throughout a region (51 to 100%). Three sections per tissue sample were analyzed. 24

Statistical Analysis

Statistical analysis was performed by means of the SPSS software (SPSS Inc.). Because most of the data in the groups did not follow a Gaussian distribution, data are given as the median value and the interquartile range. The latter represents the range between the 25th and the 75th percentile. Correlation analysis was performed by Spearman’s rank correlation.

Primary Cell Culture of SFs

Two surgical specimens from joint replacement surgery were obtained from the Department of Orthopaedics, Mount Sinai School of Medicine. The patients fulfilled the American College of Rheumatology-revised criteria for the classification of RA. 18 Tissue samples were minced into ~1-mm 3 pieces and cultured in Dulbecco’s modified Eagle’s medium containing 10% fetal bovine serum, 100 U/ml of penicillin and100 mg/ml of streptomycin (Mediatech, Herndon, VA) was added. The cells were cultured at 37°C in a 5% CO2 incubator with medium exchanges twice a week. At cell confluence, cells were passaged and then analyzed between passages 3 and 8. Cartilage degradation by SFs was assessed by culturing SFs on pieces of bovine cartilage (4 × 4 × 1 mm) under the conditions described above. After 3 days, the culture media were supplemented by 10 μmol/L of the cysteine protease inhibitors Mu-Leu-hPh-VS-Ph or Mu-Np-hPh-VS-Np, respectively (both compounds were kindly provided by Axys Pharmaceuticals, South San Francisco, CA). The inhibitors were added to each media exchange for 10 days. Mu-Leu-hPh-VS-Ph is a potent cathepsin K inhibitor whereas Mu-Np-hPh-VS-Np is a very poor cathepsin K but a very potent inhibitor for cathepsins B, L, and S. 25 Primary SF cultures were cyto-stained for the fibroblast marker, 5B5 (proline-4-hydroxylase) and the macrophage marker, CD68 (Chemicon, Temecula, CA), and for cathepsin K using the monoclonal antibody M21 as described above. Fluorescent assays for intracellular cathepsin K activity were performed as previously described by Xia and colleagues 26 using Z-Gly-Pro-Arg-4-methoxy-β-naphthylamide (MβNA) and Z-Arg-Arg-MβNA as control for cathepsin B activity (Bachem Inc., Bubendorf, Switzerland).

Electron Microscopy

The cells cultured on the sliced cartilage specimens were immediately immersed in a solution containing 3% glutaraldehyde with 0.2 mol/L of sodium cacodylate at pH 7.4. After overnight fixation the fixative solution was removed and replaced with phosphate buffer followed by 1% osmium tetroxide buffered with sodium cacodylate. After 1 hour the osmium was replaced with increasing concentrations of ethanol through propylene oxide and flat-embedded in Embed 812 (EMS, Fort Washington, PA). One μm plastic sections were cut perpendicular to the plane of the cells grown on the cartilage surface, stained with methyl blue and azure II, and observed by light microscopy. Representative areas were chosen for ultrathin sectioning (50 nm) and observed with a JEM 100CX transmission electron microscope (JOEL, Ltd., Tokyo, Japan).

In Vitro Digestion of Aggrecan and Collagens

Bovine aggrecan was prepared as described previously 23 and type I (calf skin) and II (calf articular joints) collagens were purchased from United States Biochemical (Cleveland, OH). Recombinant human cathepsin K was prepared as described in Linnevers and colleagues. 27 Bovine aggrecan (200 μg/ml final concentration) was incubated in 100 mmol/L sodium acetate buffer, pH 5.5, or in 100 mmol/L Tris/HCl, pH 7.2, containing each 2.5 mmol/L of ethylenediaminetetraacetic acid and 2.5 mmol/L of dithiothreitol with 800 nmol/L of cathepsin K for 1 hour at 28°C. The digestion reactions were stopped by the addition of 10 μmol/L of E-64. The digestion samples were subjected to 0.6% agarose/1.2% polyacrylamide gel electrophoresis and the gels were stained with 0.1% toluidine blue (Sigma, St. Louis, MO). Collagen digests were performed in 100 mmol/L of sodium acetate buffer, pH 5.5, containing 2.5 mmol/L ethylenediaminetetraacetic acid and 2.5 mmol/L dithiothreitol with 800 nmol/L cathepsin K at 28°C. Aliquots were taken from the digest mixtures at 0, 2.5, 8, and 24 hours, inhibited with 10 μmol/L of E-64 and subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis using 4 to 20% Tris/glycine gels (Novex/Invitrogen, Carlsbad, CA). Gels were stained with Coommassie Blue. Collagen digests were performed in the absence and the presence of cathepsin K-predigested bovine aggrecan. The concentration of aggrecan in the digest mixtures was 0.1% based on the initial concentration of aggrecan in the predigest. Predigests of aggrecan were performed with 800 nmol/L of cathepsin K for 16 hours at 37°C. Remaining residual activity of cathepsin K in the predigest mixture was heat-inactivated at 65°C for 20 minutes. No Z-LR-MCA hydrolyzing activity (standard substrate for cathepsin K) was observed after the heat inactivation.


Characterization of Anti-Human Cathepsin K Antibodies

Purified monoclonal antibodies against human cathepsin K were shown by Western blotting to recognize cathepsin K specifically among a panel of six other human papain-like cathepsins (recombinant cathepsins B, L, S, F, V, and W) and as two protein bands in a cell extract of RA fibroblasts. The antibody recognized both the mature and the precursor form of cathepsin K (Figure 1) [triangle] . Moreover, immunohistochemical staining of cathepsin K in synovial tissue sections could be quenched by the preincubation of the antibody with nitrocellulose membrane-immobilized recombinant cathepsin K (Figure 2b) [triangle] .

Figure 1.
Western blot analysis of the specificity of monoclonal anti-human cathepsin K antibody. The specificity of monoclonal human cathepsin K antibody was tested by Western blot analysis with human cathepsins B, L, S, K, V, F, and W. Eight pmol of each enzyme ...
Figure 2.
Immunohistochemical analysis of cathepsin K in synovial RA specimens. A mouse monoclonal antibody against human cathepsin K was used to localize cathepsin K in in the synovium from the patients with RA. For fluorescence double staining, a rabbit polyclonal ...

Expression of Cathepsin K in RA Synovial Tissue

Archived paraffin-embedded specimens from 21 RA patients were analyzed by immunohistochemistry. Expression of cathepsin K was observed in all rheumatoid arthritic specimens examined. Five cell types were identified to express cathepsin K protein: fibroblast-like and macrophage-like synoviocytes, chondroclasts/osteoclasts, and giant multinucleated cells. The most dominant expression of cathepsin K was observed in the lining layer and sublining area of synovial villi. Low-power magnification (×40 to ×100) revealed either a characteristic ribbon-like distribution that extended from the stroma to the lining surface of synovial villi (Figure 2a) [triangle] or a more uniform expression pattern of cathepsin K-positive cells (Figure 2c) [triangle] . In contrast, normal synovium obtained from a cancer patient revealed only a very few isolated, individual spindle-shaped cells in the synovial lining (Figure 2d) [triangle] . Examination with high-power magnification suggested that the majority of cathepsin K-positive cells in these areas are fibroblast-like synoviocytes (Figure 2f) [triangle] . The intracellular staining was mostly vesicular implying a lysosomal localization (Figure 2, f and h) [triangle] . It should be noted that cathepsin K-containing fibroblast-like cells are frequently restricted to well-defined areas with cathepsin K-negative cells in their immediate vicinity. Figure 2f [triangle] shows cathepsin K-positive cells at the edges of a ribbon-like distribution of SFs adjacent to cathepsin K-negative spindle-like cells. Within the synovial stroma, cathepsin K expression is generally restricted to CD68-negative cells (not shown) whereas within the synovial lining CD68-positive cells also displayed immunostaining against cathepsin K (Figure 2e) [triangle] .

Cathepsin K-positive SFs were observed in large quantities in areas of high proliferation and vascularization within the RA synovium (Figure 2 [triangle] ; g to i). Cathepsin K-positive cells seemed to invade highly proliferating tissue areas and frequently formed ring-like cellular arrangements around developing blood vessels (Figure 2, g and i) [triangle] . Lateral sections of a necrotic blood vessel are characterized by surrounding cathepsin K-containing fibroblast-like cells. (Figure 2h) [triangle] . Around well-established blood vessels, cathepsin K-positive cells are located beyond the smooth muscle layer (Figure 2i) [triangle] . Endothelial cells are cathepsin K-negative and appear swollen and project into the lumen of the vessels.

Fibroblast-like cells were consistently observed at sites of synovium-mediated bone and cartilage destruction. Figure 2, j to l [triangle] , shows cathepsin K-positive cells in areas of synovial infiltration of bone and cartilage matrices where they seem to form a scalloped surface on the remaining cartilage (Figure 2k) [triangle] . These cells were CD68-negative (not shown) and in contrast to the spindle-like SFs displayed a rounder shape. Interestingly, SFs also seem to specifically attack bone debris particles deep in the synovial stroma (Figure 2m) [triangle] . Of note is the high accumulation of cathepsin K-positive multinucleated giant cells and mononuclear cells, including fibroblast-like cells at sites of deposits resembling amyloids (Figure 2, n and o) [triangle] . At the sites of cathepsin K-positive cells, the fibrous extracellular matrix is absent suggesting its complete degradation (Figure 2o) [triangle] .

Correlation between Cathepsin K Expression and Clinical RA Assessment

The staining intensity for cathepsin K did not correlate with the number of swollen joints, the modified Landbury index, or the Keitel functional index. On the other hand, there was a highly significant negative correlation of the results of the HFCQ index with the cathepsin K staining in the subsynovial (r = 0.78, P = 0.003) and perivascular regions (r = 070, P = 0.012) (Figure 3) [triangle] . No correlation was found for the erythrocyte sedimentation rate or the C-reactive protein. Cathepsin S expression did not reveal any correlation with any of the tested parameters of disease activity (data not shown). The immunohistochemical analyses of the tissue specimens was performed at the Mount Sinai School of Medicine in New York City, NY (by W-SH and DB) without the knowledge of the clinical patient parameters. The correlation between the immunostaining and the HFCQ data were performed at the Martin Luther-University-Halle, Germany (by GK).

Figure 3.
Cathepsin K expression in subsynovial lining and in vascular areas in correlation to the HFCQ. Cathepsin K expression was quantified using a 0 to 4 scale, with 0 indicating no staining to 4 indicating extensive staining throughout a region (51 to 100%) ...

Characterization of Primary RA-Synovial Fibroblast Cultures

Synovial fibroblast-like cells were derived from two patients undergoing joint replacement surgery. Expression of cathepsin K in primary SF cell cultures was demonstrated at the mRNA and protein level by Northern and Western blot analyses (data not shown). To determine the intracellular localization of cathepsin K, cells were cultured on tissue culture slides and stained with M21. Figure 4a [triangle] shows the perinuclear and vesicular distribution of cathepsin K polypeptide indicating a lysosomal localization of the protease. The fibroblast identity of the primary cell culture was verified by the positive staining against the fibroblast marker, proline 4-hydroxylase (5B5), and the negative staining against the macrophage marker CD68 (Figure 4, b and c) [triangle] . Using Z-Gly-Pro-Arg-MβNA for in situ substrate staining, fluorescent signals revealed the same distribution as the protein staining suggesting that cathepsin K is processed and fully active in lysosomes (Figure 4d) [triangle] . Z-Gly-Pro-Arg-MβNA is an efficient substrate for cathepsin K and to a lesser degree for cathepsin B. 26 Other cathepsins such as cathepsins L, S, H, F, and V do not hydrolyze this compound (D. Brömme, unpublished results). To exclude cathepsin B as a Z-Gly-Pro-Arg-MβNA-hydrolyzing activity, cells were incubated with the specific cathepsin B inhibitor, ibuNH-EPS-Leu-Pro (kindly provided by Dr. Gour, Adherex Technologies, Ottawa, Ontario, Canada). 26 The remaining Z-Gly-Pro-Arg-MβNA-hydrolyzing activity was concluded to be cathepsin K-specific (Figure 4e) [triangle] . Cathepsin K activity could be abolished by the addition of 1 μmol/L of Mu-Leu-hPh-VS-Ph, a potent and selective cysteine protease inhibitor (Figure 4f) [triangle] . In contrast, Mu-Np-hPh-VS-Np, a potent cathepsin B, L, and S inhibitor but an extremely poor cathepsin K inhibitor showed only a weak inhibition of the Z-Gly-Pro-Arg-MβNA-hydrolyzing activity comparable to that observed for the cathepsin B-specific ibuNH-EPS-Leu-Pro compound (Figure 4g) [triangle] . 25 The lysosomal uptake and intracellular activity of the selective cathepsin B inhibitor and Mu-Np-hPh-VS-Np were demonstrated by the inhibition of the hydrolysis of the cathepsin B substrate, Z-Arg-Arg-MβNA (not shown).

Figure 4.
Expression of cathepsin K in primary cultures of RA-derived FLS. All fibroblast-like cells derived from RA synovial specimens stained positive for cathepsin K protein and showed intracellular hydrolysis of a cathepsin K-specific substrate (a). Primary ...

In Vivo Cartilage Degradation

To determine whether SFs degrade cartilage, cells were cultured for 10 days on bovine cartilage pieces in the presence or absence of synthetic cysteine protease inhibitors. Cartilage-attached cells were analyzed by light and electron microscopy. Immunohistochemical analysis of cartilage attached SFs showed a vesicular distribution of cathepsin K (not shown). Although SFs seem to invade the cartilage matrix, the highest optical magnification achieved by light microscopy was not adequate for visualizing the cellular uptake of cartilage matrix components. Using electron microscopy, the engulfment andinternalization of individual type II collagen fibers became apparent (Figure 5a) [triangle] . Interestingly, there was no significant difference in the ability of RA-SFs in the absence or the presence of inhibitors to ingest collagen fibers. In the absence of inhibitors or in the presence of a non-cathepsin K but cathepsin B, L, and S inhibitor (Mu-Np-hPh-VS-Np), no or only a very slight accumulation of undigested collagen fibrils was observed in lysosomes (Figure 5, a and b) [triangle] . This suggested a rapid degradation of collagen fibers once they arrived in the lysosomal/endosomal compartment. In contrast, a potent cathepsin K inhibitor (Mu-Leu-hPh-VS-Np; this inhibitor has in addition a similar potency toward cathepsins L, B, and S as the Mu-Np-hPh-VS-Np compound) prevented degradation of intracellular collagen leading to an accumulation of undigested fibers within the cells. The comparison of the effects of both inhibitors strongly suggested that cathepsin K was the primary lysosomal activity responsible for the hydrolysis of type II collagen. The detection of very few undigested fibers in cells treated with 10 μmol/L of Mu-Np-hPh-VS-Ph suggested a partial inhibition of cathepsin K activity rather than the involvement of other redundant cysteine proteases. Because both compounds are irreversible inhibitors it can be anticipated that a long-term incubation with Mu-Np-hPh-VS-Ph leads to a partial inhibition of cathepsin K despite its extremely weak potency toward cathepsin K (second-order rate constant of inhibition of cathepsin K by Mu-Np-hPh-VS-Ph is more than 6 orders of magnitude lower when compared with Mu-Leu-hPh-VS-Np 25 ). The cellular uptake of both inhibitors by rat osteoclasts has been previously demonstrated by the inhibition of lysosomal cathepsin B activity using a cathepsin B-specific in situ substrate, Z-Arg-Arg-βMNA 26 (D. Brömme, unpublished results). The uptake of both compounds by human SFs is shown in Figure 4 [triangle] and discussed above.

Figure 5.
Electron microscopy images of RA-derived SFs growing on bovine cartilage disks. a: SFs were grown on bovine cartilage disks in the absence of cysteine protease inhibitors. The inset displays the engulfment of type II collagen fibrils by the SFs. No intracellular ...

In Vitro Aggrecan and Collagen Degradation by Cathepsins K, L, and S

In vivo collagen degradation experiments indicated that endocytosed collagen fibers in SFs are rapidly degraded by a lysosomal cysteine protease activity that is specifically inhibited by a cathepsin K inhibitor. In addition, the presence of SFs at sites of bone erosion (Figure 2 [triangle] ; j to m) suggested an active role of cathepsin K in bone matrix degradation as discussed for osteoclasts. 28 To demonstrate the ability of cathepsin K to degrade both major components of cartilage, aggrecan, and type II collagen as well as the major bone component, type I collagen, in vitro degradation experiments were performed. For comparison, digestion experiments with recombinant human cathepsins L and S were performed. Digestion experiments clearly showed the ability of the tested recombinant human cathepsins to degrade bovine aggrecan under acidic conditions typical for lysosomes (pH 5.0) and to a lesser degree at pH 7.2 reflecting the extracellular pH. Only cathepsin S had an equally potent activity at pH 7.2 and pH 5.0. Glycosaminoglycan staining with toluidine blue revealed a complete degradation of aggrecan to the size of free chondroitin sulfate polymers (Figure 6A) [triangle] at pH 5.0. This may indicate that cathepsins K, L, and S are able to cleave at multiple sites between the G2 and G3 domain of aggrecan that harbors the glycosaminoglycan side chains. Based on the recently identified potentiating effect of free chondroitin sulfate on the activity of cathepsin K toward collagens, 23 predigested aggrecan was added to the collagen degradation assays. The addition of predigested aggrecan increased the efficiency of triple helical collagen degradation by cathepsin K comparable to the addition of free chondroitin sulfate (Figure 6B) [triangle] . 23 Similarly to free chondroitin sulfate, the activity-enhancing effect of the aggrecan predigest was restricted to acidic pH conditions. No significant degradation of collagen was observed at neutral pH (data not shown).

Figure 6.
In vitro activities of recombinant cathepsin K. A: Degradation of bovine aggrecan by cathepsins K, l, and S at pH 5.5 and 7.2. Separation of aggrecan cleavage products using agarose gel electrophoresis and detection of glycosaminoglycan (GAG) products ...


Cathepsin K is a papain-like cysteine protease that constitutes the main proteolytic activity in bone-resorbing osteoclasts. 12,29 Whereas early studies suggested that cathepsin K is exclusively expressed in osteoclasts, recent research has revealed a wider distribution pattern. Besides its expression in osteoclasts and related chondroclasts and multinucleated giant cells, cathepsin K has also been described in mononuclear cells that may act as precursor cells for osteoclasts, 30 in macrophages, 31 and various epithelial cells. 32-35 In situ hybridization experiments have demonstrated cathepsin K expression in synoviocytes of RA patients. 36 However, no qualitative, quantitative, or functional data were available to characterize cathepsin K expression in SFs on the protein level. Here, we demonstrated high levels of cathepsin K expression in spindle-shaped fibroblast-like cells throughout the inflamed RA synovium. Although the expression of cathepsin K in SFs is widespread, it is not uniform. The finding that cathepsin K is only expressed in a subset of SFs within the inflamed synovium supports the hypothesis that the SF population in RA tissue is heterogeneous. 37 Considering the potent collagenolytic-, elastinolyic-, and proteoglycan-degrading activity of cathepsin K, it is tempting to speculate that cathepsin K-positive SFs in the deeper stroma are involved in synovial remodeling, whereas cells attached or in close vicinity to cartilage or bone surfaces directly contribute to joint erosion. These cells have been described as highly proliferating mesenchymal cells within the synovial stroma that may extend to the surface of synovial villi where they seem to replace the original lining cells. 38 Two types of cathepsin K-containing cells appear to be present in the synovial lining: CD68+ and CD68 cells. The data presently available do not allow us to classify the CD68+ cells as macrophage-like synoviocytes because the CD68 marker has also been described on activated SFs. 39 However, the dense population of cathepsin K-positive cells in the lining of the inflamed synovium suggests a role of the enzyme in cartilage- and bone-invasive processes. This role is further supported by the observation that cathepsin K-positive SFs accumulate at sites of bone and cartilage erosion.

Increased synovial remodeling including angiogenesis and vascular recession are common features in RA. 40 Cathepsin K-positive SFs were highly enriched at sites of vascularization and angiogenesis but also around necrotic vessels. It can be hypothesized that the protease plays a role in the loosening of the surrounding matrix to facilitate angiogenic growth on one side and contributes to the degradation of expired vessels on the other side. Similar to stromal cells in tumor tissues, stromal cells such as SFs could support the turnover of the synovial extracellular matrix in response to angiogenic signals. It is known that SFs are directly involved in angiogenesis by the expression and secretion of the angiogenic vascular endothelial growth factor whose secretion is in particular augmented under conditions of hypoxia. 41 Thus, the simultaneous activation/recruitment of endothelial cells by fibroblast-derived vascular endothelial growth factor and the degradation of the extracellular matrix by fibroblast-derived cathepsin K would be beneficial for the generation of new blood vessels.

Another hallmark of rheumatoid synovitis is the enormous influx of lymphocytes. Lymphocyte infiltration and migration through the synovium requires local matrix degradation. As described above, cathepsin K protein-containing SFs are concentrated in areas of lymphocytic infiltration. It can be speculated again, that cathepsin K activity facilitates the migration of these cells.

Synovial fibroblasts have also been directly linked to the degradation of articular cartilage and bone. Bone and cartilage erosions are considered an irreversible degenerative process leading to the loss of joint function. Various functional indices have been developed to address the clinical status and progression of joint destruction and the responsiveness to therapeutic treatments. The Health Assessment Questionnaire Disability Index (HAQ) is considered one of the most comprehensive tests presently available. In this study, we used the HFCQ, a modified version of this test applied to the German speaking population. The questionnaire comprises 19 questions about daily activities of life. 22 A value of 0% reflects maximal incapacitation whereas 100% indicates full functionality. HFCQ data were available for 12 RA patients and they correlated with the quantitative assessment of cathepsin K expression in perivascular and subsynovial areas of these patients. Although the number of patients available was rather small in this study, the r and P values of 0.70 (P = 0.012) and 0.78 (P = 0.003) were highly correlative and significant for the degree of the impairments. Interestingly, there was no correlation between HFCQ and the expression of cathepsin S that has been implicated in the inflammatory component of RA 42 (D Brömme, unpublished results).

It is well accepted that cathepsin K is the major osteoclastic proteolytic enzyme responsible for bulk collagen degradation during bone remodeling. This function of the protease is best documented in the phenotype of cathepsin K deficiency, which causes the autosomal-recessive bone-sclerosing dysplasia, pycnodysostosis. On the cellular level, the enzyme deficiency is characterized by an accumulation of undigested collagen fibrils in affected osteoclasts that was reminiscent of the collagen fiber accumulation first seen in E-64-treated osteoclast cultures. 43 Analogous to cathepsin K-deficient osteoclasts, SFs treated with a cathepsin K inhibitor displayed high amounts of collagen fibrils stored in intracellular vesicles. Neither one of the tested cysteine protease inhibitors prevented the phagocytosis of collagen fibers by the SFs that suggested that the collagen degradation primarily occurs intracellularly. This conclusion is supported by the previous observation that cells at the pannus-articular cartilage junction contained membrane-bound collagen fibrils, apparently in various stages of digestion. 44 It is possible, however, that cathepsin K is also involved in the extracellular degradation of matrix molecules. It has been demonstrated that cathepsin K is present in the extracellular resorption lacunae of osteoclasts and that the depth of the osteoclastic excavation pits on the bone surface depends on cathepsin K activity. 26,28 In contrast to the well-characterized polarized morphology of osteoclasts allowing the formation of an extracellular acidified resorption lacuna underneath the cell, SFs are not known to form such an extracellular matrix resorption compartment. In general, it is argued that because of the pH-neutral extracellular matrix secreted cysteine proteases, with the exception of the neutral pH stabile cathepsin S are unlikely to be proteolytically active. Parak and colleagues, 45 however, demonstrated that SFs from RA patients showed an enhanced secretion of acidic components that would acidify their pericellular microenvironment. Under these conditions, an extracellular activity of cathepsin K would be likely because primary cultures of SFs secrete mature cathepsin K as well as its precursor molecule that would be autoactivated at acidic pH (D. Brömme, unpublished results). 46

The inhibition of intracellular collagen fibril degradation by cathepsin-specific inhibitors in SFs raises the question about the involvement of matrix metalloproteinase in type II collagen degradation. Collagen-degrading matrix metalloproteinases have a neutral pH activity optimum and are expressed on the outer cell membrane or are secreted. In general they are regarded as the main collagenolytic activities in mammalian tissues and it is thought that collagen degradation is primarily an extracellular event. However, the finding reported here and previous observations about the intracellular accumulation of type I collagen fibrils in cathepsin K-deficient osteoclasts 42 suggests that 1) the bulk degradation of collagen fibrils occurs intralysosomally and 2) cathepsin K is a key protease for the intracellular collagen degradation in osteoclasts and SFs. Matrix metalloproteinase activities are likely to be involved in the extracellular predigestion of collagen fibrils or their loosening from the extracellular matrix structure thus allowing their phagocytosis.

It is of particular interest that the collagenolytic activity of cathepsin K is specifically activated by degradation products of aggrecan. Concentrations of chondroitin sulfate similar to those found in synovial fluids lead to an optimal stabilization of cathepsin K activity and subsequently to an increase of its collagenolytic activity. 23 In this report, we could demonstrate that chondroitin sulfate necessary for the increase of the collagenolytic activity can be generated by cathepsin K-catalyzed aggrecan degradation. In vitro aggrecan digestion assays with cathepsin K revealed a complete hydrolysis of the proteoglycan at acidic pH and at least partial degradation at physiological neutral pH. It is very likely that in SFs endocytosed aggrecan molecules are degraded intralysosomally by resident cathepsins including cathepsin K that will generate a chondroitin sulfate-rich microenvironment supportive to the collagenolytic activity of cathepsin K.

In conclusion, this report describes the expression of one of the most potent extracellular matrix-degrading mammalian activities, cathepsin K in SFs that are regarded as the pivotal cell type in RA-associated cartilage and bone erosion. The expression of cathepsin K in the subsynovium and in perivascular synovial areas correlates with the severity of the disease based on the HFCQ. The inhibition of cathepsin K in RA-derived SFs results in a lysosomal accumulation of undigested type II collagen fibrils comparable to that seen in cathepsin K-deficient osteoclasts in pycnodysostosis whereas the inhibition of cathepsins L, B, and S has no effect. The specific activation of the protease activity toward types I and II collagen by cathepsin K-digested aggrecan, corroborates the potential role of cathepsin K in RA-associated joint destruction. Taken as a whole, these findings identified cathepsin K as a potential target for anti-resorptive drugs in RA.


Address reprint requests to Dieter Brömme, Ph.D., Department of Human Genetics, Box 1498, The Mount Sinai School of Medicine, One Gustave L. Levy Place, New York, NY 10029-6574. E-mail: .ude.mssm@emmorb.reteid

Supported in part by National Institutes of Health grant AR46182 and by a Biomedical Science grant from the Arthritis Foundation.


1. Muller-Ladner U, Gay R, Gay S: Molecular biology of cartilage and bone destruction. Curr Opin Rheumatol 1998, 10:212-219 [PubMed]
2. Goldring S, Gravallese EM: Pathogenesis of bone erosions in rheumatoid arthritis. Curr Opin Rheumatol 2000, 12:195-199 [PubMed]
3. Buttle DJ, Handley CJ, Ilic MZ, Saklatvala J, Murata M, Barrett AJ: Inhibition of cartilage proteoglycan release by a specific inactivator of cathepsin B and an inhibitor of matrix metalloproteinases. Evidence for two converging pathways of chondrocyte-mediated proteoglycan degradation. Arthritis Rheum 1993, 36:1709-1717 [PubMed]
4. Firestein GS: Invasive fibroblast-like synoviocytes in rheumatoid arthritis. Passive responders or transformed aggressors? Arthritis Rheum 1996, 39:1781-1790 [PubMed]
5. Bresnihan B: Pathogenesis of joint damage in rheumatoid arthritis. J Rheumatol 1999, 26:717-719 [PubMed]
6. Brinckerhoff C: Joint destruction in arthritis: metalloproteinases in the spotlight. Arthritis Rheum 1991, 34:1073-1075 [PubMed]
7. Buttle DJ: Lysosomal cysteine endopeptidases in the degradation of cartilage and bone. Immunopharmacology of Joints and Connective Tissue. London, San Diego, New York, Boston, Sydney, Tokyo, Toronto, Academic Press, 1994, pp 225–243
8. Esser RE, Angelo RA, Murphey MD, Watts LM, Thornburg LP, Palmer JT, Talhouk JW, Smith RE: Cysteine proteinase inhibitors decrease articular cartilage and bone destruction in chronic inflammatory arthritis. Arthritis Rheum 1994, 37:236-247 [PubMed]
9. Conway JG, Wakefield JA, Brown RH, Marron BE, Sekut L, Stimpson SA, McElroy A, Menius JA, Jeffreys JJ, Clark RL: Inhibition of cartilage and bone destruction in adjuvant arthritis in the rat by a matrix metalloproteinase inhibitor. J Exp Med 1995, 182:449-457 [PMC free article] [PubMed]
10. Chapman HA, Riese RJ, Shi GP: Emerging roles for cysteine proteases in human biology. Annu Rev Physiol 1997, 59:63-88 [PubMed]
11. Brömme D, Okamoto K, Wang BB, Biroc S: Human cathepsin O2, a matrix protein-degrading cysteine protease expressed in osteoclasts. Functional expression of human cathepsin O2 in Spodoptera frugiperda and characterization of the enzyme. J Biol Chem 1996, 271:2126-2132 [PubMed]
12. Drake FH, Dodds RA, James IE, Connor JR, Debouck C, Richardson S, Lee-Rykaczewski E, Coleman L, Rieman D, Barthlow R, Hastings G, Gowen M: Cathepsin K, but not cathepsins B, L, or S, is abundantly expressed in human osteoclasts. J Biol Chem 1996, 271:12511-12516 [PubMed]
13. Garnero P, Borel O, Byrjalsen I, Ferreras M, Drake FH, McQueney MS, Foged NT, Delmas PD, Delaisse JM: The collagenolytic activity of cathepsin K is unique among mammalian proteinases. J Biol Chem 1998, 273:32347-32352 [PubMed]
14. Kafienah W, Bromme D, Buttle DJ, Croucher LJ, Hollander AP: Human cathepsin K cleaves native type I and II collagens at the N-terminal end of the triple helix. Biochem J 1998, 331:727-732 [PMC free article] [PubMed]
15. Gelb BD, Shi GP, Chapman HA, Desnick RJ: Pycnodysostosis, a lysosomal disease caused by cathepsin K deficiency. Science 1996, 273:1236-1238 [PubMed]
16. Hou W-S, Brömme D, Zhao Y, Mehler E, Dushey C, Weinstein H, Miranda CS, Fraga C, Greig F, Carey J, Rimoin DL, Desnick RJ, Gelb BD: Cathepsin K: characterization of novel mutations in the pro and mature polypeptide regions causing pycnodysostosis. J Clin Invest 1999, 103:731-738 [PMC free article] [PubMed]
17. Lemaire R, Huet G, Zerimech F, Grard G, Fontaine C, Duquesnoy B, Flipo RM: Selective induction of the secretion of cathepsins B and L by cytokines in synovial fibroblast-like cells. Br J Rheumatol 1997, 36:735-743 [PubMed]
18. Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF, Cooper NS, Healey LA, Kaplan SR, Liang MH, Luthra HS: The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1987, 31:315-324 [PubMed]
19. Fuchs HA, Pincus T: Reduced joint counts in controlled clinical trials in rheumatoid arthritis. Arthritis Rheum 1994, 37:470-475 [PubMed]
20. van Leeuwen MA, van der Heijde DM, van Rijswijk MH, Houtman PM, van Riel PL, van de Putte LB, Limburg PC: Interrelationship of outcome measures and process variables in early rheumatoid arthritis. A comparison of radiologic damage, physical disability, joint counts, and acute phase reactants. J Rheumatol 1994, 21:425-429 [PubMed]
21. Keitel W, Hoffmann H, Weber, Krieger U: Evaluation of the percentage of functional decrease of the joints using a motor function. Dtsch Gesundheitsw 1971, 26:1901–1903 [PubMed]
22. Lautenschlager J, Mau W, Kohlmann T, Raspe HH, Struve F, Bruckle W, Zeidler H: Comparative evaluation of a German version of the Health Assessment Questionnaire and the Hannover Functional Capacity Questionnaire. Z Rheumatol 1997, 56:144-155 [PubMed]
23. Li Z, Hou WS, Bromme D: Collagenolytic activity of cathepsin K is specifically modulated by cartilage-resident chondroitin sulfates. Biochemistry 2000, 39:529-536 [PubMed]
24. Cunnane G, Bjork L, Ulfgren AK, Lindblad S, FitzGerald O, Bresnihan B, Andersson U: Quantitative analysis of synovial membrane inflammation: a comparison between automated and conventional microscopic measurements. Ann Rheum Dis 1999, 58:493-499 [PMC free article] [PubMed]
25. Palmer JT, Rasnick D, Klaus JL, Brömme D: Vinyl sulfones as mechanism-based cysteine protease inhibitors. J Med Chem 1995, 38:3193-3196 [PubMed]
26. Xia L, Kilb J, Wex H, Lipyansky A, Breuil V, Stein L, Palmer JT, Dempster DW, Brömme D: Localization of rat cathepsin K in osteoclasts and resorption pits: inhibition of bone resorption cathepsin K-activity by peptidyl vinyl sulfones. Biol Chem 1999, 380:679-687 [PubMed]
27. Linnevers CJ, McGrath ME, Armstrong R, Mistry FR, Barnes M, Klaus JL, Palmer JT, Katz BA, Brömme D: Expression of human cathepsin K in Pichia pastoris and preliminary crystallographic studies of an inhibitor complex. Prot Sci 1997, 6:919-921 [PMC free article] [PubMed]
28. Saftig P, Wehmeyer O, Hunziker E, Jones S, Boyde A, Rommerskirch W, von Figura K: Impaired osteoclastic bone resorption leads to osteopetrosis in cathepsin K-deficient mice. Proc Natl Acad Sci USA 1998, 95:13453-13458 [PMC free article] [PubMed]
29. Brömme D, Okamoto K: Human cathepsin O2, a novel cysteine protease highly expressed in osteoclastomas and ovary. Molecular cloning, sequencing and tissue distribution. Biol Chem Hoppe Seyler 1995, 376:379-384 [PubMed]
30. Dodds RA, Connor JR, Drake FH, Gowen M: Expression of cathepsin K messenger RNA in giant cells and their precursors in human osteoarthritic synovial tissues. Arthritis Rheum 1999, 42:1588-1593 [PubMed]
31. Punturieri A, Filippov S, Allen E, Caras I, Murray R, Reddy V, Weiss SJ: Regulation of elastinolytic cysteine proteinase activity in normal and cathepsin K-deficient human macrophages. J Exp Med 2000, 192:789-800 [PMC free article] [PubMed]
32. Bühling F, Gerber A, Häckel C, Krüger S, Köhnlein T, Brömme D, Reinhold D, Ansorge F, Welte T: Expression of cathepsin K in lung epithelial cells. Am J Respir Crit Care Med 1999, 20:612-619 [PubMed]
33. Haeckel C, Krueger S, Buehling F, Broemme D, Franke K, Schuetze A, Roese I, Roessner A: Expression of cathepsin K in the human embryo and fetus. Dev Dyn 1999, 216:89-95 [PubMed]
34. Haeckel C, Krueger S, Kuester D, Ostertag H, Samii M, Buehling F, Broemme D, Czerniak B, Roessner A: Expression of cathepsin K in chordoma. Hum Pathol 2000, 31:834-840 [PubMed]
35. Tepel C, Bromme D, Herzog V, Brix K: Cathepsin K in thyroid epithelial cells: sequence, localization and possible function in extracellular proteolysis of thyroglobulin. J Cell Sci 2000, 113:4487-4498 [PubMed]
36. Hummel KM, Petrow PK, Franz JK, Muller-Ladner U, Aicher WK, Gay RE, Brömme D, Gay S: Cysteine proteinase cathepsin K mRNA is expressed in synovium of patients with rheumatoid arthritis and is detected at sites of synovial bone destruction. J Rheumatol 1998, 25:1887-1894 [PubMed]
37. Zvaifler NJ, Firestein GS: Pannus and pannocytes. Alternative models of joint destruction in rheumatoid arthritis. Arthritis Rheum 1994, 37:783-789 [PubMed]
38. Firestein GS: Etiology and pathogenesis of rheumatoid arthritis. Textbook of Rheumatology, ed 5. Edited by WN Kelley, ED Harris, S Ruddy, CB Sledge. Philadelphia, London, Toronto, Montreal, Sydney, Tokyo, W. B. Saunders Company; 1997, pp 851–897
39. Harris ED: Rheumatoid Arthritis. Philadelphia, London, Toronto, Montreal, Sydney, Tokyo, W. B. Saunders Company, 1997
40. Walsh DA: Angiogenesis and arthritis. Rheumatology 1999, 38:103-112 [PubMed]
41. Jackson JR, Minton JA, Ho ML, Wei N, Winkler JD: Expression of vascular endothelial growth factor in synovial fibroblasts is induced by hypoxia and interleukin 1beta. J Rheumatol 1997, 24:1253-1259 [PubMed]
42. Nakagawa TY, Brissetle WH, Lira PD, Griffiths RJ, Petrushova N, Stock J, McNeish JD, Eastman SE, Howard ED, Clarke SR, Rosloniec EF, Elliott EA, Rudensky AY: Impaired invariant chain degradation and antigen presentation and diminished collagen-induced arthritis in cathespin S null mice. Immunity 1999, 10:207-217 [PubMed]
43. Everts V, Aronson DC, Beertsen W: Phagocytosis of bone collagen by osteoclasts in two cases of pycnodysostosis. Calcif Tissue Int 1985, 37:25-31 [PubMed]
44. Harris EDJ, Glauert AM, Murley AH: Intracellular collagen fibers at the pannus-cartilage junction in rheumatoid arthritis. Arthritis Rheum 1977, 20:657-665 [PubMed]
45. Parak WJ, Dannohl S, George M, Schuler MK, Schaumburger J, Gaub HE, Muller O, Aicher WK: Metabolic activation stimulates acid production in synovial fibroblasts. J Rheumatol 2000, 27:2312-2322 [PubMed]
46. McQueney MS, Amegadzie BY, D’Alessio K, Hanning CR, McLaughlin MM, McNulty D, Carr SA, Ijames C, Kurdyla J, Jones CS: Autocatalytic activation of cathepsin K. J Biol Chem 1997, 272:13955-13960 [PubMed]

Articles from The American Journal of Pathology are provided here courtesy of American Society for Investigative Pathology
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...