Skip Navigation


Rheumatology Advance Access originally published online on January 17, 2006
Rheumatology 2006 45(7):808-814; doi:10.1093/rheumatology/kel003
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplementary data
Right arrow All Versions of this Article:
45/7/808    most recent
kel003v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Ma, G.-F.
Right arrow Articles by Salo, J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ma, G.-F.
Right arrow Articles by Salo, J.
Related Collections
Right arrow Osteoporosis and Metabolic Bone Disease
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2006. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

Expression of ADAM9 (meltrin-{gamma}) around aseptically loosened total hip replacement implants

G.-F. Ma1, M. Liljeström3, M. Ainola3, T. Chen1, V.-M. Tiainen4, R. Lappalainen5, Y. T. Konttinen1,4,6 and J. Salo2

1 Department of Medicine/Invärtes Medicin and 2 Department of Orthopaedics and Traumatology, Helsinki University Central Hospital, 3 Department of Anatomy, Institute of Biomedicine, University of Helsinki, 4 ORTON Orthopaedic Hospital of the Invalid Foundation, Helsinki, 6 COXA Hospital for Joint Replacement, Tampere and 5 Department of Applied Physics, University of Kuopio, Kuopio, Finland.

Correspondence to Y. T. Konttinen, Department of Medicine/Invärtes Medicin, P.O. Box 700, FIN-00029 Helsinki University Central Hospital, Finland. E-mail: yrjo.konttinen{at}helsinki.fi


   Abstract
Top
 Abstract
Introduction
 Materials and methods
 Results
 Discussion
 References
 
Objective. To investigate the involvement of a disintegrin and the metalloproteinase ADAM9 (meltrin-{gamma}) in the formation of multinuclear giant cells and osteoclasts in aseptic loosening of hip replacement implants.

Methods. We used in situ hybridization, immunohistochemical staining and western blotting of interface membrane surrounding loosened hip implants, macrophage-colony stimulating factor (M-CSF) and receptor activator of nuclear factor {kappa}B ligand (RANKL) costimulation and polymethyl methacrylate (PMMA) particle stimulation of human monocytes followed by immunofluorescence staining and flow cytometric analysis.

Results. Morphometric analysis revealed that the ADAM9+ area in the revision total hip replacement (THR) interface was larger than in primary THR samples (37.6±5.1 vs 5.2±0.8%, P=0.002). Double immunofluorescence staining showed that CD68+ interface tissue macrophages and multinuclear giant cells were ADAM9+. ADAM9 mRNA containing mononuclear and multinuclear cells was often seen in a close spatial relationship with other ADAM9+ cells. Western blotting disclosed a 50 kDa ADAM9 band in tissue extracts. Upon M-CSF and RANKL costimulation of human monocytes, the ADAM9 staining pattern changed over time and ADAM9+ cells formed bi- and multinuclear cells. Flow cytometry disclosed that cells of the monocyte/macrophage lineage changed from ADAM9-negative cells into strongly positive cells during a 3-day culture.

Conclusion. ADAM9 is expressed in interface tissues around aseptically loosened THR implants. ADAM9 may play a role as a fusion molecule in the formation of multinuclear giant cells and osteoclasts from mononuclear precursors in diseases characterized by bone tissue destruction.

KEY WORDS: ADAM9, Aseptic loosening, Total hip replacement, Multinucleated giant cells, Fusion molecule


   Introduction
Top
 Abstract
 Introduction
Materials and methods
 Results
 Discussion
 References
 
Aseptic loosening of prosthetic components is the most common mode of failure of total hip replacement (THR) arthroplasties. Aseptic loosening is characterized by hip pain and by radiolucent lines or polycyclic bone defects in the peri-implant bone. This is probably caused by wear particles, which are formed as a result of cyclic loading and which induce macrophage activation [1]. Activated macrophages are actively phagocytosing cells, which produce lysosomal proteinases to degrade the phagocytosed material. If these foreign bodies are too large to be phagocytosed or non-degradable, macrophages start to form foreign body giant cells trying to cope with a quantitatively and qualitatively difficult challenge [2, 3]. The site of this foreign body-driven chronic inflammation is the synovial membrane-like interface tissue, which always develops around loosening implants and apparently plays an important role in aseptic loosening [4]. This foreign body host response contributes inadvertently to peri-implant tissue destruction and accelerates implant loosening.

ADAMs (a disintegrin and a metalloproteinase) form a family of transmembrane glycoproteins, which all have a disintegrin and a metalloproteinase domain. Their aminoterminal end is formed of a propeptide (activation peptide) followed by a metalloproteinase domain, which may be involved in the degradation of extracellular matrix and necrotic tissue [5]. The disintegrin domain is a ligand for integrin [6]. ADAMs also contain a cysteine-rich domain, which acts as a ligand for the cell adhesion molecules and participates in cluster formation [7]. Most importantly, some cysteine-rich domains of ADAMs contain fusion peptide sequences involved in membrane fusion, which have been shown to be involved in the formation of multinuclear cells [8–10]. ADAM9, also known as meltrin-{gamma}, is involved in the formation of multinuclear cells, such as myotubes, macrophage-derived giant cells and osteoclasts, which are formed as the result of cell–cell fusion of mononuclear precursor cells. Expression of ADAM9 mRNA, enhanced by receptor activator of nuclear factor {kappa}B ligand (RANKL) and macrophage-colony stimulating factor (M-CSF) or by anti-CD98 antibody, stimulates monocyte fusion [11].

Based on these facts, this study was performed to study the expression of ADAM9 in the synovial membrane-like interface tissue in aseptic loosening of THR and in blood monocytes capable of forming multinuclear foreign body giant cell- and osteoclast-like cells. Aseptic loosening is characterized by chronic foreign body inflammation and periprosthetic osteolysis. Synovial membrane-like interface tissue from revision THR samples was compared with synovial membrane obtained from primary THR operations performed for osteoarthritis, which is normally not characterized by tissue-destructive osteolysis.


   Materials and methods
Top
 Abstract
 Introduction
 Materials and methods
Results
 Discussion
 References
 
Tissue samples
Ten samples of synovial membrane-like interface tissue were collected from the proximal bone–cement or bone–stem interfaces around aseptically loosened femoral stems during revision THR operations. Nine of the patients had cement fixation and one had an uncemented stem. Seven of the patients were women and three were men. Their mean age was 71 yr (range 57–80 yr). Seven arthroplasties had been performed for osteoarthritis, two for fractures and one for rheumatoid arthritis. The mean time from the primary to the revision THR was 7.7 yr (range 5–16 yr). Another set of ten samples was obtained from patients undergoing primary THR operations performed for osteoarthritis. Seven of these patients were women and three were men. Their mean age was 62.5 yr (range 47–67 yr). Samples were immediately immersed in saline. Half of the sample was immediately embedded in Tissue-Tek OCT compound (Lab-Tek Products, Miles Laboratories, Elkhart, IN, USA), snap-frozen in dry ice-precooled isopentane and stored at –70°C. The other half was processed to paraffin by fixation in 10% formalin followed by dehydration in ethanol, clearing in xylene and embedding in paraffin. All samples were analysed using immunohistochemistry, but five representative frozen tissue samples from each category were also used for in situ hybridization. In addition, tissue from three patients from revision and primary THR operations was extracted and analysed using western blotting. Sample collection was approved by the Ethics Committee of Helsinki University Central Hospital. Guidelines of the Declaration of Helsinki were followed.

Immunohistochemistry
Paraffin tissue sections were deparaffinized in xylene, rehydrated through a graded ethanol series and washed in 10 mM phosphate-buffered 150 mM saline, pH 7.4. Antigens were retrieved for 30 min using MicroMED T/T Mega Laboratory Microwave Systems (Milestone, Sorisole, Italy) in 0.01 M sodium citrate buffer, pH 6.0, using the AR98C S30M program. The intrinsic peroxidase activity was abolished by pretreatment of the tissue sections in 0.3% H2O2 in methanol for 30 min. The sections were incubated: (i) with normal goat serum (Jackson ImmunoResearch Laboratories, West Grove, PA, USA) in phosphate buffer containing 0.1% bovine serum albumin (Sigma, St Louis, MO, USA) for 30 min; (ii) 1 µg/ml polyclonal affinity-purified, rabbit anti-human ADAM9 IgG (Triple Point Biologics, Forest Grove, OR, USA) overnight at +4°C; (iii) 0.3 µg/ml biotinylated goat anti-rabbit IgG (Vector Laboratories, Burlingame, CA) for 1 h; (iv) avidin–biotin–peroxidase complex (1:100 in phosphate buffer, Vector Laboratories) for 1 h; (v) a mixture of 3,3-diaminobenzidine tetrahydrochloride (Sigma) and 0.3% H2O2 for 5 min. Finally, the slides were counterstained with haematoxylin for 1 min, dehydrated in ethanol, cleared in xylene and mounted. All incubations were performed at +22°C if not otherwise stated. Purified normal goat IgG (Jackson) at the same concentration as and instead of the specific primary IgG antibody was used as a negative staining control. The sections were inspected with Leitz Diaplan microscope coupled to a 12-bit digital image camera. Images were analysed with a semiautomatic Analysis Pro 3.0 image analysis and processing system.

Double immunofluorescence staining
After deparaffinization, tissue sections were incubated for 30 min in antigen retrieval 0.01 M sodium citrate buffer, pH 6.0, in MicroMED T/T Mega Laboratory Microwave Systems (Milestone) using the AR98C S30M program. The sections were washed in phosphate buffer followed by incubation for 30 min in normal donkey or goat serum (Jackson) for blocking. Blocked sections were incubated in a mixture of (i) 0.1 µg/ml monoclonal mouse anti-human CD68 IgG1 (NeoMarkers, Freemont, CA, USA) and 1 µg/ml affinity-purified polyclonal rabbit anti-human ADAM9 IgG (Triple Point Biologics) overnight at +4°C, and (ii) secondary tetramethyl rhodamine isothiocynate (TRITC)-conjugated donkey anti-mouse IgG (Jackson) and fluorescein isothiocyanate (FITC)-conjugated goat anti-rabbit IgG (Jackson) for 1 h. Purified normal mouse IgG1 or rabbit IgG (Jackson) at the same concentration as and instead of the specific primary antibodies was used as the negative staining control. All incubations were done at +22°C if not otherwise stated. Slides were air-dried and mounted in Vectashield (Vector) and kept in the dark at +4°C. The slides were analysed using an Olympus motorized revolving AX 70 system microscope coupled with a 12-bit digital image camera. Images were analysed using a semiautomatic Analysis Pro 3.0 image analysis and processing system.

ADAM9 primers and RNA probes
Total RNA was isolated from homogenized samples using TRIzol reagent (Invitrogen, Paisley, UK). mRNA was isolated from total RNA using magnetic (dT)25-polystyrene beads (Dynal, Oslo, Norway). The first-strand cDNA was synthesized from 100 ng of sample mRNA using the SuperScript Preamplification System (Invitrogen). For ADAM9 primer (Accession number, NM_003816 [GenBank] ), the sequence was sought from the National Centre for Biotechnology Information (NCBI) Entrez search system. Sequence similarity was sought using the NCBI Blastn 2 program. To obtain a polymerase chain reaction (PCR) amplification product, the following ADAM9 primers were designed: sense 5'-GCA TTT GTG GGA ACA GTG TG-3'; antisense 5'-CCA GCG TCC ACC AAC TTA TT-3'. The target specific primers were mixed with dNTP and AmpliTaq Gold DNA polymerase (PE Applied Biosystems, Foster City, CA, USA) in PCR buffer (15 mM Tris–HCl, pH 8.0, 1.5 mM MgCl2, 50 mM KCl). PCR amplification was performed in a thermal cycler (RoboCycler 40 Temperature Cycler, Stratagene, CA, USA) for 40 cycles of 1 min for denaturation at +95°C, 1 min for annealing at +60°C, 1 min for extension at +72°C, and 10 min for the last cycle. The amplified PCR products were cloned, using the TOPO TA cloning kit (Invitrogen), into a 3'-deoxythymidine residue overhang and topoisomerase I-activated pCRII-TOPO vector. Recombinant vector was purified using the NucleoSpin Plasmid Kit (Macherey-Nagel, Düren, Germany). The orientations of the acquired antisense and sense sequences were verified with an automated sequencer using M13 reverse or forward primers. Plasmid concentration was measured by spectrophotometer. The acquired sequence was verified with the NCBI Blastn 2 program. The plasmid was linearized using restriction enzymes NotI or HindIII. DNA templates were transcribed to a single-strand antisense RNA probe using SP6 RNA polymerase or a sense RNA control probe using T7 RNA polymerase. RNA was labelled with digoxigenin-conjugated UTP from the DIG RNA Labeling Kit (Roche Diagnostics, Mannheim, Germany). A series of dilutions of DIG-labelled RNA was applied to nitrocellulose membrane to verify the efficiency of the RNA labelling.

In situ hybridization
Frozen tissue sections (8 µm) were fixed in 4% paraformaldehyde at +4°C for 15 min, permeabilized for 30 min at +37°C in 100 mM Tris–HCl, 50 mM EDTA, pH 8.0, containing 5 µg/ml RNase-free proteinase K, and dipped in 0.1 M triethanolamine diluted 1:400 (v/v) in acetic anhydride, pH 8.0. To block non-specific hybridization, sections were prehybridized with hybridization buffer containing 100 µg/ml denatured salmon sperm DNA for 1 h at +57°C in humidified chambers. Blocked sections were incubated in hybridization buffer containing 1–2 ng/µl digoxigenin-labelled RNA probe in humidified chambers at +57°C overnight. Following stringent post-hybridization washes, single-stranded (unbound) RNA probe was degraded at +37°C for 30 min using 20 µg/ml RNase-A (Roche). After blocking in normal sheep serum (Jackson) at +4°C overnight, alkaline phosphatase-conjugated sheep anti-digoxigenin-Fab (Roche) was added and colour was developed using 5-bromo-4-chloro-3-indolylphosphate (BCIP; Promega, Madison, WI, USA) substrate and nitroblue tetrazolium (NBT; Promega) colour. Counterstaining was performed in 1% methylene green before mounting.

Western blotting
Tissue samples in RIPA buffer (RIPA buffer set; Boehringer Mannheim) were homogenized in an ice-bath, sonicated (Vibra-Cell 501; Sonics & Materials, Danbury, CT, USA) for 15 s and centrifuged at 21 000 g for 3 h at +4°C. The total protein of the sample was quantified using the Bio-Rad Protein Assay (Bio-Rad Laboratories, Richmond, CA, USA). Protein was adjusted to 100 µg per lane using sodium dodecyl sulphate (SDS) sample buffer (187.5 mM Tris HCl, pH 6.8, 6% SDS, 30% glycerol, 0.03% phenol red, and 125 mM dithiothreitol, diluted 1:3) (New England Biolabs, Beverly, MA, USA). After 10% SDS–polyacrylamide gel electrophoresis, the gels were equilibrated and blotted onto nitrocellulose membrane (Bio-Rad Laboratories). The membranes were incubated at room temperature in (i) 3% bovine serum albumin (Sigma) in Tris-buffered saline blocking solution overnight, (ii) 1 µg/ml polyclonal affinity-purified rabbit anti-human ADAM9 IgG (Triple Point Biologics) for 60 min, and (iii) alkaline phosphatase-conjugated goat anti-rabbit IgG (Bio-Rad Laboratories) for 60 min. Between the steps, the membranes were washed in 0.1% Tween 20 in Tris-buffered saline. Colour reaction was developed for 15 min using the Alkaline Phosphatase Conjugate Substrate Kit (Bio-Rad Laboratories).

Cell cultures
Monocytes were isolated from buffy coats of three healthy blood donors of the Finnish Red Cross using Ficoll–Paque PLUS density gradient isolation. Washed cells (1 x 107/well) were incubated in six-well plates with or without coverslips in 2 ml of MEM Alpha Medium with Glutamax-1 (Invitrogen) supplemented with 10% fetal calf serum, 1% penicillin and 1% streptomycin. After 1 h the non-adherent lymphocytes were washed away; the adherent monocytes were used in stimulation experiments before analysis using immunofluorescence staining and flow cytometry.

Immunofluorescence staining of stimulated cells
Cells cultured on coverslips for 1, 3, 7 or 14 days with or without 25 ng/ml rhM-CSF (R&D, Minneapolis, MN, USA) and 40 ng/ml rhRANKL (Alexis, Carlsbad, CA, USA) [12] were washed, fixed in 3% paraformaldehyde in phosphate buffer for 20 min and incubated in (i) 1 µg/ml polyclonal affinity-purified rabbit anti-human ADAM9 IgG (Triple Point Biologics) or, as a negative control, in purified normal rabbit IgG (Jackson) for 30 min, and (ii) Alexa Fluor 594-labelled goat anti-rabbit IgG (Molecular Probes, Leiden, The Netherlands) for 30 min. Before mounting, nuclei were stained for 5 min in 5 µg/ml 4',6-diamidino-2-phenylindole. Cells were observed using an Olympus motorized revolving AX 70 system microscope coupled with a 12-bit Sensicam digital image camera and an Analysis Pro 3.0 image analysis and processing system.

Polymethyl methacrylate uptake
Polymethyl methacrylate (PMMA, bone cement) particles containing BaSO4 were produced from a PMMA block using 500-grid SiC paper. The size of the particles produced was characterized using a NICOMPTM 370 (Particle Sizer Systems, Santa Barbara, CA, USA) submicron particle sizer based on dynamic light scattering of particles suspended in solvent. Particles of 0.6–2 µm were preserved in 99.5% ethanol solution as a 1.3 mg/ml stock solution until used in cell culture experiments. Before use, PMMA particles were ultrasound-sonicated for dispersal and cleaning with a Vibra CellTM processor (Sonics & Materials, Newtown, CT, USA) with 2-s intervals (amplitude 80, output 150 watts). Adherent monocytes were stimulated for 24 and 72 h using 25 ng/ml rhM-CSF (R&D) alone or together with 40 ng/ml rhRANKL (Alexis) in the presence or absence of 1 x 106/ml PMMA particles. Resting non-cultured cells were used as a control.

Flow cytometry of monocyte/macrophages
Adherent cells were gently scraped and washed followed by an incubation of 2 x 105 cells in 1 µg/ml polyclonal, affinity-purified rabbit anti-human ADAM9 IgG (Triple Point Biologics) at +4°C for 60 min. The cells were washed twice and incubated for 60 min in darkness with FITC-conjugated goat anti-rabbit IgG (Jackson). The stained cells were analysed with a FACScan analyser using CellQuest software (Becton Dickinson, San Jose, CA, USA). The level of non-specific binding was determined using isotype-matched antibodies of irrelevant specificity.

Statistical analysis
The results are expressed as mean and S.E.M. All data were checked for normality using Wilk's W test. Comparisons between groups were done using the t-test for normally distributed data and Wilcoxon's test for skewed data. Probability values less than 0.05 were considered significant. Statistical software BMDP-PC version 7.01 (BMDP Statistical Software, Cork, Ireland) was used for statistical calculations.


   Results
Top
 Abstract
 Introduction
 Materials and methods
 Results
Discussion
 References
 
ADAM9 in interface membrane
Under light microscopy, the paraffin sections of the interface membrane around aseptically loosened THR implants were found to contain: (i) small, black metal particles; (ii) clefts as PMMA voids; and (iii) birefringent polyethylene under polarized light. Some particles had been phagocytosed and were found intracellularly in macrophage-like cells and multinuclear giant cells. A synovial lining-like structure, usually one to three cell layers thick and consisting of fibroblast- and macrophage-like cells, covered the interstitial connective tissue stroma in all samples.

All synovial membrane-like peri-implant interface tissue samples around loosened implants showed ADAM9 immunoreactivity. The inflammatory cell-rich areas contained many ADAM9+ macrophage-like cells (Fig. 1A), whereas such cells were much more rare in the deep fibrotic tissue areas. In particular, all multinuclear foreign body giant cells were ADAM9+ (Fig. 1A). ADAM9 immunoreactive cells often occurred in small aggregates in intimate contact with one another or close by ADAM9+ multinuclear cells. The synovial lining-like layer facing the pseudosynovial fluid was always intensely stained (Fig. 1C). The primary THR control samples also exhibited immunoreactivity for ADAM9, although the staining was usually very weak and restricted to certain areas (Fig. 1B). Cellular ADAM9 staining was strong in revision THR compared with primary THR. As there was also some extracellular ADAM9 staining, it was difficult to count the number of positive cells. Instead, the percentage area of ADAM9 staining was measured. The percentage of ADAM9+ area in revision THR was higher than in primary THR samples (37.6±5.1 vs 5.2±0.8%, P = 0.002). Staining controls with purified normal rabbit IgG, used at the same concentration as and instead of specific ADAM9 IgG, were almost totally negative (Fig. 1D).


Figure 1
View larger version (126K):
[in this window]
[in a new window]
  		FIG. 1. Immunoperoxidase staining of ADAM9. In synovial membrane-like interface tissue around aseptically loosened THR implants, ADAM9 staining is seen in macrophage-like cells (A). Multinuclear foreign body giant cells seen in such samples were always ADAM9+ (A). ADAM9 staining was weak and sparse in the control samples obtained from primary THR arthroplasties (B). ADAM9 staining was also clearly seen in the synovial lining-like layer of interface membrane (arrow), which contains macrophage-like type A synovial lining cells, as well as in the stromal macrophage-like cells (C). Negative staining control was negative (D). Bar = 10 µm.

 
Double immunofluorescence staining
TRITC and FITC double staining showed that almost all CD68+ macrophages and multinuclear giant cells were also ADAM9+ in the interface tissue (Fig. 2A, B). Staining disclosed close spatial relationship between ADAM9+ mononuclear cells and ADAM9+ multinuclear giant cells and CD68+ mononuclear cells and CD68+ multinuclear giant cells (Fig. 2A, B). The synovial lining cell layer of the control osteoarthritic synovial membrane samples also showed ADAM9, although the staining was usually weak and restricted to certain areas (Fig. 2D). Almost all CD68+ macrophages (Fig. 2C) were ADAM9+ also in osteoarthritis.


Figure 2
View larger version (121K):
[in this window]
[in a new window]
  		FIG. 2. Immunofluorescence double labelling of macrophage marker CD68 and ADAM9 in interface membrane from revision THR (A and B) and in primary THR samples (C and D). CD68 is TRITC-labelled and red (A and C), whereas ADAM9 is FITC-labelled and green (B and D). All or almost all CD68+ multinuclear giant cells and macrophages were also ADAM9+ in the interface membrane. All or almost all CD68+ macrophages were also ADAM9+ in primary THR samples. Bar = 10 µm.

 
In situ hybridization
In situ hybridization disclosed ADAM9 mRNA in all synovial membrane-like interface tissue samples around loosened hip implants. ADAM9 mRNA was found in macrophage-like cells and multinuclear giant cells (Fig. 3A). Similarly to immunostaining, ADAM9 mRNA+ mononuclear cells were often seen in close contact with ADAM9 mRNA+ multinuclear cells. Macrophage-rich sublining stroma and the lining-like layer showed strong ADAM9 signal, whereas ADAM9 transcript-positive cells were more rarely found in the deep fibrotic stroma of the interface tissue samples. Negative labelling controls of the consecutive tissue sections, performed with the sense RNA probe, confirmed the specificity of the labelling (Fig. 3B). In the control samples obtained from primary THR operations, expression of ADAM9 mRNA was weak and was detected only in some regions of the synovial lining and sublining (Fig. 3C). The consecutive sections hybridized with the control sense ADAM9 RNA probe showed no staining in the same areas (Fig. 3D).


Figure 3
View larger version (127K):
[in this window]
[in a new window]
  		FIG. 3. In situ hybridization analysis of expression of ADAM9 mRNA. ADAM9 mRNA was detected in macrophage-like and foreign body giant cells in the synovial membrane-like interface tissue of revision THR samples (A). In the same area of the same section, the sense RNA control probe showed practically no staining (B). Compared with interface membranes, the ADAM9 mRNA signal was weaker and less extensive in the control synovial membranes obtained from primary THR arthroplasties (C). The sense RNA control probe showed only slight background methylene green counterstaining (D). Bar = 10 µm.

 
Western blotting
Western blots of tissue extracts of revision THR samples showed a 50 kDa ADAM9-immunoreactive band, which could not be found in samples from the primary THRs (Fig. 4).


Figure 4
View larger version (44K):
[in this window]
[in a new window]
  		FIG. 4. Western blot of ADAM9. Lane 1 shows molecular weight standards. The representative 50-kDa band seen in lane 2 is from an extract of synovial membrane-like interface tissue from revision THR (lane 2). Such bands were not seen in extracts prepared from control synovial membrane samples (lane 3).

 
ADAM9 in M-CSF- and RANKL-costimulated monocyte cultures
To analyse the cellular staining of human peripheral blood monocytes during different phases of fusion in a long-term stimulation, adherent cultured cells were labelled on coverslips. Resting cells, which had not been stimulated with M-CSF and RANKL, showed weak staining (data not shown). At culture day 1, M-CSF- and RANKL-costimulated monocytes stained relatively weakly (Fig. 5A). At culture day 3 all cells were still mononuclear but were larger and their cytoplasmic ADAM9 staining was stronger, widespread and slightly granular (Fig. 5B). At culture day 7 the cells were large and the culture contained many bi- or multinuclear cells with relatively strong cytoplasmic and granular ADAM9 staining (Fig. 5C). At culture day 14 the percentage of mononuclear ADAM9+ cells was already relatively low, as multinuclear giant cells with strong but non-homogeneous granular and patchy cytoplasmic staining had developed (Fig. 5D shows one cell with five nuclei).


Figure 5
View larger version (59K):
[in this window]
[in a new window]
  		FIG. 5. Immunofluorescence staining of ADAM9 of M-CSF- and RANKL-costimulated human monocyte cultures at culture days 1 (A), 3 (B), 7 (C) and 14 (D). ADAM9 staining is shown in red, together with the blue 4',6-diamidino-2-phenylindole nuclear counterstain (an overlay). Bar = 10 µm.

 
ADAM9 in PMMA particle stimulated monocyte cultures
According to flow cytometric analysis, non-stimulated cells were ADAM9– (Fig. 6A). Cells cultured in the presence but also in the absence of PMMA particles were ADAM9+ at day 1 (Fig. 6B) and the ADAM9 staining intensity increased further by culture day 3 (Fig. 6C), with a clear change in staining over time (Fig. 6D). M-CSF and RANKL did not affect the result of PMMA stimulation (data not shown).


Figure 6
View larger version (32K):
[in this window]
[in a new window]
  		FIG. 6. Flow cytometric analysis of ADAM9 expression in monocytes directly after isolation from the peripheral blood of a healthy donor (A) and after culture for 1 (B) and 3 days (C); results are summarized in D. ADAM9 expression was not affected by the presence or absence of PMMA particles, which seems to indicate maturation of monocytes to macrophages.

 

   Discussion
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
References
 
Host defence mechanisms lead to multinuclear giant cell formation from monocyte–macrophage precursors as part of the host response against foreign bodies. The particle disease in aseptic loosening represents a chronic inflammatory host response. Adverse responses of tissue to wear debris and mechanical failure lead to the formation of synovial membrane-like interface tissue characterized by a foreign body reaction [13]. The three principal cell types found in interface tissue, apart from those forming the blood vessels, are the macrophage, the giant cell and the fibroblast. A large amount of wear debris is always detected intra- and/or extracellularly in aseptic loosening. Particles, together with cyclic loading and implant-host interface micromovement, lead to or are at least associated with macrophage accumulation and activation. Cytokines are able to stimulate monocyte recruitment and osteoclast precursor proliferation and differentiation in the interface tissue. Both M-CSF [14] and RANKL [15, 16] have been described earlier in the interface tissue and are necessary and sufficient to stimulate osteoclastogenesis [17–19]. Multinuclear giant cells are phagocytic, but also express various destructive proteinases; whereas osteoclasts are terminally differentiated, multinucleated cells able to resorb bone. Both originate from the multipotent haematopoietic stem cells and belong to the monocyte/macrophage lineage [20]. The key step in polykaryon formation is the fusion of mononuclear precursor cells into multinucleated cells, and the present work suggests that ADAM9 may play a role in this process as a fusion molecule. This conclusion is based on the intense ADAM9 mRNA and protein expression in monocyte-macrophages in aseptic loosening in areas characterized by foreign bodies and inflammatory cell infiltrates compared with osteoarthritis synovial membrane. It was not possible to obtain corresponding control samples from well-fixed total hip implants for ethical reasons. The major limitation of this and other studies of synovial membrane-like interface obtained at revision THR is the absence of appropriate control samples. Following the general practice in this field of study, we used synovial membrane obtained from patients with osteoarthritis undergoing primary THR as control tissue. The histological features of synovial membrane from osteoarthritis patients are similar to those of the synovial membrane-like interface tissue obtained at revision THR for aseptic loosening. Immunofluorescence double staining showed that all or almost all CD68+ synovial tissue macrophages are also ADAM9+. ADAM9+ mononuclear cells in revision samples, apparently representing prefusion cells, were often seen in close contact with forming multinuclear cells. There was an apparent discrepancy between the lack of ADAM9 band in the western blot of primary THR control samples in spite of some immunohistochemical ADAM9 staining in such samples. This may relate to the relatively low intensity and extent of staining in the control samples, so that the ADAM9 concentration in the extracts probably remains below the detection level of the western blot method.

ADAM9, also known as meltrin-{gamma}, has a complex and multifunctional domain structure and seems to play multiple roles during cell fusion. Like ADAM1 (fertilin-{alpha}), it contains a fusogenic peptide in its cysteine-rich domain [21]. Synthetic peptides corresponding to fertilin fusion peptide, when reconstituted into liposomes or built into lipid bilayers, enhance fusion [22]. Structure–function analysis suggest that ADAM9, overexpressed in the interface membrane around loosening hip implants, may, via cell fusion, contribute to the formation of multinuclear foreign body giant cells and osteoclasts and thus have a role in loosening. RANKL, with the aid of M-CSF, has the ability to induce cell fusion and multinucleated giant cell formation from monocyte precursors. In our in vitro study, ADAM9 was enhanced in monocytes stimulated with RANKL and M-CSF [23]. However, this ADAM9 expression was not further enhanced by PMMA particles. According to our in vitro studies, ADAM9 seems be a useful marker of the maturation of monocytes to macrophages. Further studies are needed to clarify whether particles have any direct effect on ADAM9 expression.

The disintegrin domain of ADAM9 can interact with the {alpha}Vß5 integrin receptor, which is expressed in human monocytes [24, 25]. Such interactions regulate many cellular events, including adhesion and migration. The cytoplasmic tail of ADAM9 contains Src homology 3 (SH3) binding motifs [26], which interacts with actin cytoskeleton leading to changes in the shape and migration of the cell. Finally, ADAM9 has a metalloproteinase domain with several matrix metalloproteinase characteristics, such as a cysteine switch, a furin cleavage site, a catalytic site and a Met-turn [27–30]. Cell-membrane-bound ADAM9 metalloproteinase may cleave basement membrane components during transmigration of the peripheral blood monocytes from the vascular compartment into interface tissue. It may also mediate shedding of cytokines and lead to their activation, but may also function in receptor solubilization and contribute to negative feedback loops. Therefore, ADAM9 has several other potentially important roles in foreign body inflammation and aseptic loosening apart from its involvement in the formation of multinuclear giant cells. This study describes for the first time ADAM9 expression in interface tissue around aseptically loosened total hip implants as well as its potential involvement in multinuclear cell formation from blood monocytes stimulated with RANKL and M-CSF. ADAM9 may also contribute to interface tissue remodelling via its proteinase-like effect. It is thus of interest in this condition as well as in other diseases characterized by tissue destruction and osteoclast-driven bone resorption, such as the formation of erosions in rheumatoid arthritis. Further understanding of the regulation and role of ADAM9 in chronic inflammation may provide new, attractive therapeutic targets for the control of osteoclast formation and bone destruction.
Figure 7


   Acknowledgments
 
We thank Eija Kaila and Erkki Hänninen for technical help. This research was funded by the clinical EVO research grants, Finska Läkaresällskapet, an Orion Pharma grant, the Stockmann Foundation and the Centre of Excellence programme of the Academy of Finland and the Centre for Technological Advancement (TEKES), Invalid Foundation, University of Helsinki Group of Excellence scheme, and the PhD Graduate School on Biomaterials and Tissue Engineering of the Ministry of Education.

The authors have declared no conflicts of interest.


   References
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 

  1. Goodman SB, Lind M, Song Y, Smith RL. In vitro, in vivo, and tissue retrieval studies on particulate debris. Clin Orthop 1998;352:25–34.
  2. Lassus J, Salo J, Jiranek WA et al. Macrophage activation results in bone resorption. Clin Orthop 1998;352:7–15.
  3. Ma J, Chen T, Mandelin J et al. Regulation of macrophage activation. Cell Mol Life Sci 2003;60:2334–46.[CrossRef][ISI][Medline]
  4. Goldring SR, Jasty M, Roelke MS, Rourke CM, Bringhurst FR, Harris WH. Formation of a synovial-like membrane at the bone-cement interface. Its role in bone resorption and implant loosening after total hip replacement. Arthritis Rheum 1986;29:836–42.[ISI][Medline]
  5. Black RA, White JM. ADAMs: focus on the protease domain. Curr Opin Cell Biol 1998;10:654–9.[CrossRef][ISI][Medline]
  6. Zhu X, Evans JP. Analysis of the roles of RGD-binding integrins, alpha(4)/alpha(9) integrins, alpha(6) integrins, and CD9 in the interaction of the fertilin beta (ADAM2) disintegrin domain with the mouse egg membrane. Biol Reprod 2002;66:1193–202.[Abstract/Free Full Text]
  7. Iba K, Albrechtsen R, Gilpin B et al. The cysteine-rich domain of human ADAM 12 supports cell adhesion through syndecans and triggers signaling events that lead to beta1 integrin-dependent cell spreading. J Cell Biol 2000;149:1143–56.[Abstract/Free Full Text]
  8. Abe E, Mocharla H, Yamate T, Taguchi Y, Manolagas SC. Meltrin-alpha, a fusion protein involved in multinucleated giant cell and osteoclast formation. Calcif Tissue Int 1999;64:508–15.[CrossRef][ISI][Medline]
  9. Seals DF, Courtneidge SA. The ADAMs family of metalloproteases: multidomain proteins with multiple functions. Genes Dev 2003;17:7–30.[Free Full Text]
  10. Mandelin J, Li TF, Hukkanen MV et al. Increased expression of a novel osteoclast-stimulating factor, ADAM8, in interface tissue around loosened hip prostheses. J Rheumatol 2003;30:2033–8.[ISI][Medline]
  11. Namba K, Nishio M, Mori K et al. Involvement of ADAM9 in multinucleated giant cell formation of blood monocytes. Cell Immunol 2001;213:104–13.[CrossRef][ISI][Medline]
  12. Paloneva J, Mandelin J, Kiialainen A et al. DAP12/TREM2 deficiency results in impaired osteoclast differentiation and osteoporotic features. J Exp Med 2003;198:669–75.[Abstract/Free Full Text]
  13. Santavirta S, Konttinen YT, Hoikka V, Eskola A. Immunopathological response to loose cementless acetabular components. J Bone Joint Surg Br 1991;73:38–42.
  14. Xu JW, Konttinen YT, Waris V, Pätiälä H, Sorsa T, Santavirta S. Macrophage-colony stimulating factor (M-CSF) is increased in the synovial-like membrane of the periprosthetic tissues in the aseptic loosening of total hip replacement (THR). Clin Rheumatol 1997;16:243–48.[CrossRef][ISI][Medline]
  15. Mandelin J, Li TF, Liljeström M et al. Imbalance of RANKL/RANK/OPG system in interface tissue in loosening of total hip replacement. J Bone Joint Surg Br 2003;85-B:1196–201.
  16. Ritchlin CT, Schwarz EM, O'Keefe RJ, Looney RJ. RANK, RANKL and OPG in inflammatory arthritis and periprosthetic osteolysis. J Musculoskelet Neuronal Interact 2004;4:276–84.[Medline]
  17. Tanaka S, Takahashi N, Udagawa N et al. Macrophage colony-stimulating factor is indispensable for both proliferation and differentiation of osteoclast progenitors. J Clin Invest 1993;91:257–63.[ISI][Medline]
  18. Fuller K, Wong B, Fox S, Choi Y, Chambers TJ. TRANCE is necessary and sufficient for osteoblast-mediated activation of bone resorption in osteoclasts. J Exp Med 1998;188:997–1001.[Abstract/Free Full Text]
  19. Haynes DR, Crotti TN, Potter AE et al. The osteoclastogenic molecules RANKL and RANK are associated with periprosthetic osteolysis. J Bone Joint Surg Br 2001;83:902–11.
  20. Roodman GD. Cell biology of the osteoclast. Exp Hematol 1999;27:1229–41.[CrossRef][ISI][Medline]
  21. Blobel CP, White JM. Structure, function and evolutionary relationship of proteins containing a disintegrin domain. Curr Opin Cell Biol 1992;4:760–5.[CrossRef][Medline]
  22. Martin I, Epand RM, Ruysschaert JM. Structural properties of the putative fusion peptide of fertilin, a protein active in sperm-egg fusion, upon interaction with the lipid bilayer. Biochemistry 1998;37:17030–9.[CrossRef][Medline]
  23. Felix R, Cecchini MG, Fleisch H. Macrophage colony stimulating factor restores in vivo bone resorption in the op/op osteopetrotic mouse. Endocrinology 1990;127:2592–4.[Abstract]
  24. Huang S, Endo RI, Nemerow GR. Upregulation of integrins alpha v beta 3 and alpha v beta 5 on human monocytes and T lymphocytes facilitates adenovirus-mediated gene delivery. J Virol 1995;69:2257–63.[Abstract]
  25. Zhou M, Graham R, Russell G, Croucher PI. MDC-9 (ADAM-9/Meltrin gamma) functions as an adhesion molecule by binding the alpha(v)beta(5) integrin. Biochem Biophys Res Commun 2001;280:574–80.[CrossRef][ISI][Medline]
  26. Weskamp G, Kratzschmar J, Reid MS, Blobel CP. MDC9, a widely expressed cellular disintegrin containing cytoplasmic SH3 ligand domains. J Cell Biol 1996;132:717–26.[Abstract/Free Full Text]
  27. Birkedal-Hansen H. Proteolytic remodeling of extracellular matrix. Curr Opin Cell Biol 1995;7:728–35.[CrossRef][ISI][Medline]
  28. Pei D, Weiss SJ. Furin-dependent intracellular activation of the human stromelysin-3 zymogen. Nature 1995;375:244–7.[CrossRef][Medline]
  29. Mohan S, Thompson GR, Amaar YG, Hathaway G, Tschesche H, Baylink DJ. ADAM-9 is an insulin-like growth factor binding protein-5 protease produced and secreted by human osteoblasts. Biochemistry 2002;41:15394–403.[CrossRef][Medline]
  30. Wolfsberg TG, Bazan JF, Blobel CP, Myles DG, Primakoff P, White JM. The precursor region of a protein active in sperm-egg fusion contains a metalloprotease and a disintegrin domain: structural, functional, and evolutionary implications. Proc Natl Acad Sci USA 1993;90:10783–7.[Abstract/Free Full Text]
Submitted 31 July 2005; revised version accepted 6 December 2005.
Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplementary data
Right arrow All Versions of this Article:
45/7/808    most recent
kel003v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Ma, G.-F.
Right arrow Articles by Salo, J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ma, G.-F.
Right arrow Articles by Salo, J.
Related Collections
Right arrow Osteoporosis and Metabolic Bone Disease
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?