Gene Transfer of Tissue Inhibitor of Metalloproteinases-3 Reverses the Inhibitory Effects of TNF-α on Fas-Induced Apoptosis in Rheumatoid Arthritis Synovial Fibroblasts

Andreas Drynda, Paul H. A. Quax, Manfred Neumann, Willemijn H. van der Laan, Géza Pap, Susanne Drynda, Ingmar Meinecke, Joern Kekow, Wolfram Neumann, Tom W. J. Huizinga, Michael Naumann, Wolfgang König and Thomas Pap
J Immunol May 15, 2005, 174 (10) 6524-6531; DOI: https://doi.org/10.4049/jimmunol.174.10.6524

Abstract

Apart from counteracting matrix metalloproteinases, tissue inhibitor of metalloproteinases-3 (TIMP-3) has proapoptotic properties. These features have been attributed to the inhibition of metalloproteinases involved in the shedding of cell surface receptors such as the TNFR. However, little is known about effects of TIMP-3 in cells that are not susceptible to apoptosis by TNF-α. In this study, we report that gene transfer of TIMP-3 into human rheumatoid arthritis synovial fibroblasts and MRC-5 human fetal lung fibroblasts facilitates apoptosis and completely reverses the apoptosis-inhibiting effects of TNF-α. Although TNF-α inhibits Fas/CD95-induced apoptosis in untransfected and mock-transfected cells, fibroblasts ectopically expressing TIMP-3 are sensitized most strongly to Fas/CD95-mediated cell death by TNF-α. Neither synthetic MMP inhibitors nor glycosylated bioactive TIMP-3 are able to achieve these effects. Gene transfer of TIMP-3 inhibits the TNF-α-induced activation of NF-κB in rheumatoid arthritis synovial fibroblasts and reduces the up-regulation of soluble Fas/CD95 by TNF-α, but has no effects on the cell surface expression of Fas. Collectively, our data demonstrate that intracellularly produced TIMP-3 not only induces apoptosis, but also modulates the apoptosis-inhibiting effects of TNF-α in human rheumatoid arthritis synovial fibroblast-like cells. Thus, our findings may stimulate further studies on the therapeutic potential of gene transfer strategies with TIMP-3.

Tissue inhibitors of metalloproteinases (TIMPs)3 are involved critically in the control of extracellular proteolytic activity by counteracting matrix metalloproteinases (MMPs). To date, five members of the TIMP family have been described that share the potential to inhibit nearly every member of the MMP family (1). In addition to inhibiting MMPs, TIMP-3 has been associated with a number of features that are distinct from other TIMPs. It binds to extracellular matrix (2, 3) and also inhibits membrane type (MT)-MMPs (4) and ADAMs. The latter most likely explains the capability of TIMP-3 to prevent the shedding of cell membrane-anchored proteins such as TNFR (5), IL-6R (6), syndecan ectodomains (7), and the TNF-α-converting enzyme (TACE) (8). However, the most interesting feature of TIMP-3 is its ability to induce apoptosis in different cell types (9, 10, 11, 12). This distinguishes TIMP-3 clearly from other TIMPs, but to date neither the physiological relevance of this function nor its potential contribution to disease has been elucidated.

Rheumatoid arthritis (RA) is a prominent example for disorders that are characterized by an imbalance between MMPs and TIMPs. It is a chronic inflammatory disease that affects primarily the joints and results in their progressive destruction. Fibroblast-like cells (RA synovial fibroblasts (RA-SF)) have been assigned a key role in the pathogenesis of RA, and due to their specific properties have been termed transformed appearing, tumor-like, or simply activated (13). The role of specific proinflammatory cytokines in the activation of these cells is discussed controversially, but it is widely accepted that the inflammatory environment stimulates RA-SF and contributes to their aggressive behavior toward cartilage. TNF-α has been demonstrated to up-regulate the expression of MMPs (14, 15) and to stimulate the proliferation of these cells (16). In addition, some data suggest that TNF-α may modulate the apoptosis of RA-SF (17, 18, 19) most likely through pathways involving the transcription factor NF-κB (20) that constitutes an important link between inflammation and synovial hyperplasia (21). The overexpression of tissue-degrading MMPs and alterations in apoptosis are most prominent features of RA-SF that mediate the invasiveness of the hyperplastic synovium (13). The expression of MMPs and their inhibitors is well investigated in RA, and a number of studies have addressed the expression of apoptotic pathways in this disease (22). However, still little is known about mechanisms that initiate and maintain apoptotic changes in synovial fibroblasts and particularly the relation of MMP function and apoptosis therein.

In this study, we demonstrate that TIMP-3 in addition to its proapoptotic function sensitizes RA-SF to Fas ligand (FasL/CD95L)-induced apoptosis when expressed through adenoviral gene transfer. Moreover, adenoviral delivery of TIMP-3 reverses completely the apoptosis-inhibiting effects of TNF-α in RA-SF. Similar effects can be observed when TIMP-3 is delivered nonvirally into fetal human lung fibroblasts. Interestingly, neither synthetic inhibitors nor the addition of bioactive hTIMP-3 protein are able to mimic these effects suggesting intracellular mechanisms. Of note, gene transfer of TIMP-3 reduces the activation of the transcription factor NF-κB by TNF-α in RA-SF and inhibits TNF-α-mediated up-regulation of soluble Fas/CD95. These data shed new light on the regulation of apoptosis by TIMP-3 and extend our understanding of interactions between pathways involved in destruction of extracellular matrix and induction of apoptosis. In addition, they demonstrate effects of a secreted molecule that cannot be achieved by the application of the recombinant protein and may, therefore, stimulate gene transfer strategies delivering TIMP-3 to the rheumatoid synovium.

Materials and Methods

Isolation of fibroblast-like cells

Synovial tissue samples were obtained from 15 RA patients undergoing joint arthroplasty at the Clinic of Orthopedic Surgery, University of Magdeburg (Magdeburg, Germany). Tissue was obtained from these patients following informed consent and approval by the ethic committee of the university. The demographic and disease characteristics of these patients are shown in Table I⇓. Synovial fibroblasts were isolated, as described before (23). Briefly, tissue specimens were minced and digested enzymatically (Dispase I; Roche), and the released cells were grown in DMEM (Biochrom) with 10% FCS (Biochrom) in a humidified 5% CO2 atmosphere. After allowing the cells to adhere overnight, nonadherent cells were removed and the adherent cells were grown further over four passages. Human fetal lung fibroblasts (MRC-5 cell line) were used as control cells.

View this table:
Table I.

Demographic and clinical characteristics of the 15 patients with RA

Gene transfer of TIMP-3

For gene transfer of TIMP-3, a nonviral expression construct (CMV promoter) as well as replication-defective adenoviruses (E1 deleted, CMV promotor) encoding TIMP-3 (AdTIMP-3) were used (24). The respective constructs without insert (pCMV and AdCMV) as well as untransfected cells served as controls. Transductions were performed according to standard procedures (24). For adenoviral transduction of RA-SF, cells were seeded at 80% confluence. After adherence, cell culture medium was removed, and the cells were infected with the adenoviruses at concentrations of 100-1000 PFU/cells for 12 h. Nonviral transfection of MRC-5 cells was performed using lipofection (10–30 μg of DNA per 3 × 105 cells, Superfect; Qiagen), according to the manufacturer’s instructions.

Preparation of protein extracts, immunoblotting, and Abs

RA-SF were centrifuged, and the cell pellets were mixed with 20 μl of sample buffer. After incubation for 5 min at 95°C, proteins were separated on a SDS-PAGE (12% running, 5% stacking gel). Proteins were then transferred onto a (polyvinylidene difluoride) membrane using semidry blot. Membranes were blocked for 2 h with 5% dry defatted milk in PBS-T (PBS including 0.1% Tween 20) and incubated with anti-human TIMP-3 mAbs (R&D Systems) in PBS containing 5% milk. After washing with PBS-T, membranes were incubated with HRP-labeled anti-mouse Abs. Bound HRP was visualized using the ECL kit (Amersham), according to the manufacturer’s instructions.

Nuclear and cytoplasmic protein extracts from differently treated RA-SF were prepared according to Schreiber et al. (25), with slight modifications as described (26). Protein concentrations were determined by a Bradford assay. For the immunoblot assay, 5 μg of protein extracts was separated on 10% SDS polyacrylamide gels and electrophoretically transferred to nitrocellulose membranes. Equal loading was confirmed by PonceauS staining (Sigma-Aldrich) of the membrane. As another control for equal loading of cytoplasmic extracts, an immunodetection was performed using an ERK1/2-specific Ab.

The following polyclonal Abs were used for immunodetection: anti-p65 NF-κB (Santa Cruz Biotechnology; sc-109), anti-c-Rel (sc-070), anti-RelB (sc-226), anti-p50 NF-κB (sc-114), and anti-p52 NF-κB (Upstate Biotechnology; Ab 06-413).

Induction and detection of apoptosis

RA-SF and MRC-5 cells were stimulated with 100 ng/ml human rFasL (rhFasL) for 16 h, as described (27). Subsequently, apoptosis was determined using a histone fragmentation assay (Cell Death Detection ELISAPlus; Roche), according to the instructions of the manufacturer. It is based on a quantitative sandwich-enzyme immunoassay using mouse mAbs against DNA and histones that allow for the specific, quantitative determination of cytoplasmatic histone-associated DNA fragments (mono- and oligonucleosomes) in the cell lysates. The ELISA plates were read at 405 nm (490 nm reference).

The influence of TNF-α was studied by stimulation of RA-SF and MRC-5 cells with 10 or 100 ng/ml TNF-α alone or 24 h before the induction of apoptosis by FasL/CD95L. In control experiments, RA-SF were incubated with BB-94 (batimastat, 5 μg/ml), BB-2516 (marimastat, 50 μg/ml; both kind gifts from E. Bone, British Biotech, Oxford, U.K.), or bioactive rhTIMP-3 (10–1000 ng/ml; R&D Systems) (8) instead of transducing them with AdTIMP-3. All experiments were performed at least three times.

To study the responsiveness toward exogenous TIMP-3 produced by RA-SF, Transwell experiments were performed. Parental RA-SF (1 × 105/ml) were seeded into the lower compartment of 24-well cell transwell culture plate, while AdTIMP-3 or mock-transduced cells (5 × 104/ml) were seeded into the respective cell culture inserts (pore size: 4 μm) (Nunc). After 12 h, the cell culture medium was removed and the cells in the lower compartment were stimulated with TNF-α and/or rhFasL, according to the above protocol. Apoptosis of the monolayer cells was measured, as described.

EMSA and DNA-binding assay

EMSA was used to characterize the DNA-binding activities of NF-κB in RA-SF. Nuclear extracts were prepared according to standard procedures. The oligonucleotide probe used was a consensus NF-κB binding site derived from the κB element of the murine IL-2 promoter (TCEdA>C): 5′-AGCTTGACCAAGAGGGATTTCCCCTAAATC-3′.

The activation of NF-κB was also determined quantitatively by a recently developed DNA-binding assay (TransAM NF-κB Transcription Factor Assay Kit; Active Motif) (28). The assay is based on the use of multiwell plates coated with a cold oligonucleotide containing the consensus binding site for NF-κB. The kit was used according to the instructions of the manufacturer. Briefly, nuclear extracts of RA-SF were prepared according to standard procedures and added into the oligonucleotide-coated 96-well plate. Following incubation for 1 h at room temperature, the wells were washed and incubated with an NF-κB-specific Ab for 1 h. The samples were washed again, and the secondary HRP-conjugated Ab was added for 1 h. Finally, 3,3′,5,5′-tetramethylbenzidine substrate solution containing 1% DMSO was added for color development, and the plates were read in an ELISA reader at 450 nm. All experiments were performed in duplicates, and NF-κB wild-type as well as mutated consensus oligonucleotides were used to monitor the specificity of the assay.

Detection surface-bound and soluble Fas/CD95

Surface expression of Fas/CD95 was analyzed by flow cytometry. Briefly, 105 fibroblasts were seeded in 24-well cell culture plates. The cells were detached by mild trypsinization (0.3%) and washed once with serum-containing medium (DMEM, 10% FCS). For the reconstitution of surface-expressed proteins, fibroblasts were incubated with DMEM (with 10% FCS) in 1.5-ml tubes on a roller device. After washing with PBS, cells were stained with FITC-labeled mouse anti-human Fas/CD95 Abs (Serotec; clone LOB 3/11) for 30 min. After two washing steps with PBS, cell surface expression of Fas/CD95 was measured using a FACSCalibur flow cytometer (BD Biosciences). For the detection of soluble Fas (sFas) in the cell culture supernatants of RA-SF, a commercially available ELISA (Quantikine Assays; R&D Systems) was used according to the manufacturer’s instructions.

Statistical analysis

The difference between sample group means was tested for statistical significance using the Mann-Whitney U test. Means were considered significantly different at p < 0.05.

Results

Adenoviral gene transfer of TIMP-3 reverses the antiapoptotic effects of TNF-α and sensitizes fibroblasts to Fas/CD95-mediated apoptosis

As seen in Western blot analysis, transduction of RA-SF with increasing concentrations of AdTIMP-3 resulted in a dose-dependent increase in the production of TIMP-3 (Fig. 1⇓A). We detected both the nonglycosylated and glycosylated forms of TIMP-3, with higher levels of the nonglycosylated 24-kDa form. Transduction with 1000 PFU of AdTIMP-3 produced similar levels of glycosylated protein as 500 ng/ml bioactive rhTIMP-3 (Fig. 1⇓A). Similarly, nonviral gene transfer of MRC-5 control cells resulted in the production of TIMP-3 that was dependent on the amount of plasmid used for transfection. Again, TIMP-3 was detected both in the nonglycosylated and glycosylated forms, with higher levels of the nonglycosylated 24-kDa form (Fig. 1⇓B).

FIGURE 1.
FIGURE 1.

Gene transfer of TIMP-3 reverses the antiapoptotic effects of TNF-α. A, RA-SFs were transduced with different concentrations of adenoviral particles expressing human TIMP-3 (AdTIMP), and the expression of TIMP-3 was determined by Western blot. Both the nonglycosylated and glycosylated forms of TIMP-3 could be detected, with higher levels of the nonglycosylated 24-kDa form. rhTIMP-3 was used as control. B, Using lipofection, MRC-5 cells (3 × 105) were transfected with different concentrations of a nonviral expression construct of TIMP-3 (pTIMP-3). Western analysis revealed a dose-dependent increase in the production of both the nonglycosylated and glycosylated forms of TIMP-3, with higher levels of the nonglycosylated form. C, RA-SF were transduced with 300 PFU of AdTIMP-3, and apoptosis was determined using a histone fragmentation assay following stimulation with TNF-α, FasL/CD95L, or both, as described in Materials and Methods. As compared with untransduced and AdCMV (mock)-transduced RA-SFs, AdTIMP-3 reversed the antiapoptotic effects of TNF-α on FasL/CD95L-induced apoptosis (mean ± STD of 6 RA-SF; ∗, p < 0.05 vs untransduced). D, MRC-5 cells were transfected with pTIMP-3 (15 μg per 3 × 105 cells), and apoptosis was determined. As seen with RA-SF, pTIMP-3 reversed the antiapoptotic effects of TNF-α on FasL/CD95L-induced apoptosis of MRC-5 cells (mean ± STD of 5 independent experiments; ∗, p < 0.05).

Based on data that have demonstrated an apoptosis-promoting effect of adenoviral TIMP-3 (9), we investigated whether gene transfer with an adenoviral construct expressing human TIMP-3 (AdTIMP-3) altered the apoptosis of RA-SF. For these experiments, we used a concentration of 300 PFU of AdTIMP-3 that has been previously shown to inhibit the invasiveness of RA-SF (24). As expected from studies using other cells, transduction of RA-SF with AdTIMP-3 resulted in an increased rate of spontaneous apoptosis after 4 days. When compared with untransduced, AdMock-transduced, or even AdTIMP-1-transduced RA-SF, there were significantly more apoptotic cells among the AdTIMP-3-transduced cells after 96 h (data not shown).

More intriguingly, AdTIMP-3 completely reversed the inhibiting effects of TNF-α on FasL-induced apoptosis of RA-SF, and this effect was seen much earlier than the induction of spontaneous apoptosis. Although TNF-α inhibited FasL/CD95L-mediated apoptosis in untransduced and AdCMV (mock)-transduced RA-SF, stimulation with TNF-α clearly sensitized AdTIMP-3-transduced RA-SF to FasL/CD95L-induced cell death (Fig. 1⇑C). Addition of 100 ng/ml TNF-α over 24 h and subsequent stimulation with 100 ng/ml rhFasL for 16 h decreased the levels of histone-associated DNA fragments by 32% when compared with stimulation with FasL alone (p < 0.05; Fig. 1⇑C). In contrast, an 86% increase in the levels of cytoplasmic histone fragments was seen in AdTIMP-3-transduced RA-SF. Transduction of RA-SF with an empty adenoviral construct (AdCMV (mock)) had no effects (p < 0.05) (Fig. 1⇑C).

To further exclude that viral toxicity is responsible for the effects of AdTIMP-3 and to test whether these are specific for RA-SF, MRC-5 human fetal lung fibroblasts, which can be transfected more easily by nonviral methods, were transfected with a nonviral expression construct of TIMP-3 (15 μg of plasmid per 3 × 105 cells), and apoptosis was determined, as described above. As seen with RA-SF, gene transfer of TIMP-3 reversed the antiapoptotic effects of TNF-α (Fig. 1⇑D). Stimulation of untransfected MRC-5 cells with 100 ng/ml TNF-α over 24 h and subsequent induction of apoptosis with FasL/CD95L reduced the levels of cytoplasmic histone-associated DNA fragments by 7% as compared with MRC-5 cells that were stimulated with FasL/CD95L alone. In contrast, a strong increase of 43% in the levels of cytoplasmic histone fragments was seen in pTIMP-3-transfected MRC-5 cells. Again, transfection with the empty construct (pCMV) had no effects (Fig. 1⇑D).

Modulation of apoptosis by TNF-α is dose dependent and cannot be mimicked by synthetic inhibitors of metalloproteinases and bioactive rhTIMP-3

Next, we investigated whether the effects of TNF-α on apoptosis in RA-SF, both the inhibitory effects in untransduced cells and the apoptosis-facilitating effects in AdTIMP-3-transduced RA-SF, are TNF-α dose dependent. Indeed, stimulation of untransduced RA-SF with 10 ng/ml TNF-α reduced the levels of cytoplasmic histone fragments to lesser extent than 100 ng/ml (Fig. 2⇓). Conversely, preincubation of AdTIMP-3-transduced RA-SF with 100 ng/ml TNF-α facilitated subsequent FasL-induced apoptosis more pronouncedly than 10 ng/ml TNF-α (Fig. 2⇓). To test whether synthetic inhibitors of MMPs, in a similar way as AdTIMP-3, modulate the effects of TNF-α, we used the broad spectrum MMP inhibitor BB-94 (batimastat), which has been demonstrated to inhibit a variety of metalloproteinases, including TACE and ADAMTS, as well as BB-2516 (marimastat), which acts preferentially on MMPs. Interestingly, addition of BB-94 to the cell culture supernatants was unable to mimic the effects of AdTIMP-3, and the same was true for BB-2516 (Fig. 2⇓). RA-SF that were treated with either MMP inhibitor showed comparable rates of FasL/CD95L-induced apoptosis as unstimulated cells, and there were no differences between untreated RA-SF and cells treated with MMP inhibitors with respect to the effects of TNF-α (Fig. 2⇓).

FIGURE 2.
FIGURE 2.

The proapoptotic effects of AdTIMP-3 cannot be mimicked by synthetic MMP inhibitors. RA-SF were treated with MMP inhibitors (batimastat or marimastat) or transduced with 300 PFU of AdTIMP-3, and apoptosis was determined using a histone fragmentation assay following stimulation with increasing concentrations of TNF-α, FasL/CD95L, or both. TNF-α did not induce cell death, but inhibited FasL-induced apoptosis in a dose-dependent manner in untransduced RA-SF as well as in RA-SF treated with the MMP inhibitors. Conversely, facilitation of FasL-induced apoptosis by TNF-α in the AdTIMP-3-transduced RA-SF was also dose dependent (mean ± STD of 5 RA-SF; ∗, p < 0.05 vs untransfected).

Then we analyzed whether the effects of AdTIMP-3 are TIMP-3 dose dependent and whether bioactive TIMP-3 reverses the antiapoptotic effects of TNF-α in a similar way. The effects of AdTIMP-3 on the susceptibility of RA-SF to FasL/CD95L-induced apoptosis correlated clearly with the TIMP-3 concentrations. Although in untransduced RA-SF preincubation with 100 ng/ml TNF-α reduced the level of subsequent FasL/CD95L-induced apoptosis by 44%, 31% reduction was seen in RA-SF that were transduced with 100 PFU of AdTIMP-3. Starting with 300 PFU of AdTIMP-3, preincubation with TNF-α sensitized RA-SF to apoptosis and facilitated subsequent FasL/CD95L-induced cell death (an increase of 20% compared with stimulation with FasL/CD95L alone) (Fig. 3⇓A). Transduction of RA-SF with 1000 PFU of AdTIMP-3 had the strongest effects. Incubation of these cells with TNF-α before the induction of cell death by FasL/CD95L increased the levels of cytoplasmatic histone-associated DNA fragments by 76% as compared with RA-SF stimulated with FasL/CD95L alone (Fig. 3⇓A). Treatment of RA-SF with bioactive rhTIMP-3 at different concentrations was unable to achieve these effects (Fig. 3⇓B). This was true even for concentrations of rhTIMP-3 that exceeded clearly the amount of glycosylated protein produced by AdTIMP-3-transduced RA-SF (see Fig. 1⇑A). Preincubation of rhTIMP-3-treated RA-SF with TNF-α reduced the rate of subsequent FasL/CD95L-induced cell death between 39 and 58%, thus showing no differences to RA-SF that were not treated with rhTIMP-3 (Fig. 3⇓B). In additional experiments, parental human RA-SF were cocultured with AdTIMP-3 or mock-transduced RA-SF in a transwell cell culture system to analyze potential effects of adenovirally produced TIMP-3 on neighboring cells. Cells in the lower compartment were then treated with TNF-α/FasL according to the above protocol. The apoptotic rate of the monolayer cells in the lower compartment was identical in all experiments, indicating that the cells are nonresponsive to exogenous TIMP-3 even when produced by neighboring AdTIMP-3-transduced RA-SF (data not shown).

FIGURE 3.
FIGURE 3.

The proapoptotic effects of AdTIMP-3 cannot be mimicked by rhTIMP-3. A, RA-SFs were transduced with different concentrations of AdTIMP-3, and stimulated with 100 ng/ml TNF-α and then with 100 ng/ml rhFasL/CD95L. Apoptosis was determined using a histone fragmentation assay, as described in Materials and Methods. The effects of AdTIMP-3 on the susceptibility of RA-SF to FasL/CD95L-induced apoptosis correlated with the concentration of AdTIMP-3 used. In untransduced RA-SF, incubation with 100 ng/ml TNF-α clearly reduced the levels of FasL/CD95L-induced apoptosis, while AdTIMP-3 in a dose-dependent manner facilitated FasL/CD95L-induced cell death (mean ± STD of 3 RA-SF; ∗, p < 0.05 vs untransduced). B, RA-SFs were treated with different concentrations of rhTIMP-3 while stimulated with 100 ng/ml TNF-α and subsequently with 100 ng/ml rhFasL/CD95L. Apoptosis was determined using the histone fragmentation assay. rhTIMP-3 was unable to alter the susceptibility of TNF-α-treated RA-SF to FasL/CD95L-induced apoptosis (mean ± STD of 3 RA-SF).

Gene transfer of TIMP-3 reduces the TNF-α-mediated NF-κB activation in RA-SF

To investigate potential mechanisms that may be responsible for the apoptosis-modulating effects of TNF-α, we analyzed the nuclear binding of NF-κB in untransduced, mock-transduced, and TIMP-3-transduced RA-SF. Using EMSA, an increase in NF-κB DNA-binding activity was seen upon stimulation of untransduced and AdCMV (mock)-transduced RA-SF with 10 and 100 ng/ml TNF-α (Fig. 4⇓A) after 24 h. In AdTIMP-3-transduced RA-SF, however, NF-κB activation was drastically reduced, with the binding of the p50 homodimers being the same as in unstimulated cells (Fig. 4⇓A).

FIGURE 4.
FIGURE 4.

AdTIMP-3 inhibits TNF-α-induced activation of NF-κB in RA-SF. A, RA-SF were transduced with 300 PFU of AdTIMP-3 or AdCMV (mock), and TNF-α-mediated activation of NF-κB was determined by EMSA. TNF-α increased NF-κB DNA-binding activity in untransduced and AdCMV (mock)-transduced RA-SF after 24 h of stimulation. In AdTIMP-3-transduced RA-SF, NF-κB activation was reduced, with the binding of the p50 homodimers being the same as in unstimulated cells. B, RA-SF were stimulated with TNF-α, and the activation of NF-κB was determined quantitatively in their nuclear by an ELISA-based DNA-binding assay. TNF-α resulted in a dose-dependent manner increased the NF-κB activity, with no differences between untransduced and AdCMV (mock)-transduced cells. However, activation of NF-κB was reduced in AdTIMP-3-transduced RA-SF (mean ± STD of 2 RA-SF). C, Cytoplasmic (left panel) and nuclear (right panel) protein extracts were prepared from untransduced, mock (AdCMV)-, and AdTIMP-3-transduced RA-SF, as described in Materials and Methods. The cells were either left untreated or stimulated with TNF-α (100 ng/ml) for 24 h. Following SDS-PAGE and immunoblotting, the membranes were analyzed using the indicated Rel protein-specific Abs. With the exception of p52/NF-κB2, TNF-α increased the nuclear translocation of all Rel proteins. Intracellularly expressed TIMP-3 reduced this up-regulation.

To confirm these data, RA-SF were stimulated with TNF-α, and the activation of NF-κB was determined quantitatively in the nuclear extracts of these cells by an ELISA-based DNA-binding assay that has been demonstrated to be more sensitive than EMSA (26). Stimulation of RA-SF with 10 and 100 ng/ml TNF-α resulted in a dose-dependent increase in NF-κB activity, and there were no differences between untransduced and AdCMV (mock)-transduced cells. However, activation of NF-κB was reduced in AdTIMP-3-transduced RA-SF (Fig. 4⇑B).

The nuclear translocation is the critical parameter for the activation of NF-κB. In our study, this translocation into the nucleus is enhanced in the case of p65/RelA, c-Rel, RelB, and p50/NF-κB1 upon prolonged TNF-α stimulation (100 ng/ml) of RA fibroblasts. One notable exception was p52/NF-κB2, which is central to the recently discovered noncanonical, IκB-independent pathway of NF-κB activation. In all cases, the enhanced TNF-α-induced nuclear translocation of Rel proteins could be reduced by adenoviral gene transfer of TIMP-3 into the cells (Fig. 4⇑C). With the exception of c-Rel and p52/NF-κB2, the expression of all Rel proteins was increased in TNF-α-treated RA fibroblasts in addition to an enhanced nuclear translocation indicating that many Rel proteins such as RelB are themselves target genes of NF-κB (29). These higher expression levels were down-regulated by intracellular TIMP-3 (Fig. 4⇑C).

Intracellular TIMP-3 suppresses TNF-α-induced synthesis of sFas/CD95 in RA-SF

Finally, we examined the effects of TNF-α on the expression of Fas/CD95 in untransduced, AdCMV (mock)-transduced, and AdTIMP-3-transduced RA-SF. Consistent with other studies (30), stimulation of RA-SF with TNF-α over 24 h up-regulated Fas/CD95, as determined by quantitative real-time PCR (data not shown). Flow cytometry demonstrated an increase in the expression of cell surface Fas/CD95. In untransduced RA-SF, stimulation with 100 ng/ml TNF-α for 24 h increased the expression of Fas/CD95, and this increase was nearly identical in the AdTIMP-3-transduced cells (Fig. 5⇓A). As found by ELISA, TNF-α increased the production of sFas in the supernatants of RA-SF in a dose-dependent manner (Fig. 5⇓B). Following stimulation with 10 and 100 ng/ml TNF-α for 24 h, the levels of sFas increased equally in untransduced and AdCMV (mock)-transduced RA-SF. However, transduction of RA-SF with AdTIMP-3 significantly reduced the increase of sFas (Fig. 5⇓B).

FIGURE 5.
FIGURE 5.

Intracellular TIMP-3 suppresses TNF-α-induced synthesis of sFas in RA-SF. A, RA-SF were stimulated with TNF-α, and the expression of cell surface Fas/CD95 was determined by flow cytometry. In untransduced and AdTIMP-3-transduced RA-SF, stimulation with 100 ng/ml TNF-α for 24 h increased the expression of Fas/CD95 by ∼60%. B, As found by ELISA, TNF-α in a dose-dependent manner increased the production of sFas in the supernatants of untransduced and AdCMV (mock)-transduced RA-SF. However, transduction of RA-SF with AdTIMP-3 significantly reduced the increase of sFas (mean ± STD of 7 RA-SF, p < 0.05 vs untransduced).

Discussion

TIMP-3, like other TIMPs, is critically involved in the remodeling of extracellular matrix, but the inhibition of sheddases such as the TACE at low nanomolar concentrations (8) and the induction of apoptosis in a number of cancer cell lines (5, 9, 11), smooth muscle cells (10), and pigment epithelial cells (12) are unique features of TIMP-3 that distinguish it from other members of the TIMP family. It remains unclear how alterations of TIMP-3 expression may translate into distinct modifications of cell function, but the specific properties of TIMP-3 suggest that an imbalance between metalloproteinases and TIMP-3 may result in uncontrolled matrix destruction, increased levels of TNF-α, and altered cell death. Although expression of TIMP-3 has been demonstrated in the RA synovium (31), it is of interest that in RA chronic inflammation and other pathways of cellular activation result in aggressive fibroblasts that are characterized by an imbalanced production of MMPs and alterations in apoptosis (13). However, the factors that regulate the susceptibility of fibroblasts to apoptosis, the specific factors that regulate the susceptibility of RA synovial fibroblasts to apoptosis, as well as the role of matrix-degrading enzymes and their inhibitors therein are only incompletely understood. In this context, our present data suggest that TIMP-3 not only exerts proapoptotic effects, but also may act as a switch that regulates whether TNF-α inhibits or promotes Fas/CD95-induced apoptosis. This regulatory function of TIMP-3 appears to be related to mechanisms that cannot be mimicked by broad spectrum MMP inhibitors and even bioactive rTIMP-3. In this context, it is important to note that the MMP inhibitors used in our study are poorly soluble in water and that this poor solubility may have influenced the inhibitor experiments. However, we have used the inhibitors under the same conditions as in numerous other studies (e.g., 32 , 33), in which their capability to inhibit MMP activity has been demonstrated convincingly. As for potential effects of externally added TIMP-3, it may be speculated that nonglycosylated TIMP-3 is more effective and responsible for the observed effects. This notion is supported both by our experiments with glycosylated rhTIMP-3 (see Fig. 1⇑A) and our data from the transwell system. Although it has been shown that glycosylation is not essential for MMP inhibition of TIMP-3 (34), we were unable to test the effects of nonglycosylated hTIMP-3 on apoptosis. However, the conclusions, that intracellularly produced TIMP-3 reverses the antiapoptotic effects of TNF-α in RA-SF, would have been influenced by such experiments.

To date, the proapoptotic effects of TIMP-3 have been attributed mainly to the inhibition of metalloproteinases that are involved in the shedding of cell surface molecules. As reported by Smith et al. (5), expression of TIMP-3 in human colon carcinoma cells resulted in an increase of cell surface TNFRI/CD120a that was accompanied by a higher rate of cell death. Consequently, it was suggested that TIMP-3-mediated inhibition of receptor sheddases restores TNFRI signaling pathways, and carcinoma cells are killed by autocrine TNF-α. This notion was supported also by data from Bond et al. (35), who showed convincingly that the proapoptotic function of TIMP-3 was located to the N terminus and required functional metalloproteinase-inhibitory activity. Interestingly, synthetic inhibitors of MMPs were unable to mimic the effects of TIMP-3 (35), but the addition of the recombinant protein had comparable effects to gene transfer studies from this group (10, 11). It was hypothesized that the induction of apoptosis by TIMP-3 is related closely to the extracellular inhibition of proteases that cleave surface receptors involved in signaling apoptosis, specifically TNFRI/CD120a (36). However, such mechanisms would require that target cells are susceptible to TNF-α-mediated apoptosis or that TNF-α at least facilitates apoptosis.

This does not appear to be the case for RA-SF. In our experiments, TNF-α inhibited rather than stimulated apoptosis in these cells, as seen from a decrease in apoptosis following stimulation with TNF-α. Moreover, TNF-α in a clearly dose-dependent manner prevented both RA-SF from Fas/CD95-induced apoptosis that was paralleled by the up-regulation of both cell surface-bound and soluble Fas/CD95 receptor. This is consistent with data showing that in some fibroblast-like cells, apoptosis signals through members of the TNFR family are only executed when the survival signals are blocked (20, 37). Numerous data have demonstrated that the activation of RA-SF is accompanied by changes such as down-regulation of PTEN (38), increased phosphorylation of Akt kinase (37), and particularly the activation of NF-κB, which all act as survival signals and most likely contribute to the antiapoptotic effects of TNF-α. In this context, the activation of NF-κB appears to be of special importance, and this notion is supported not only by our demonstration that TNF-α in a dose-dependent manner induced the activation of NF-κB. As shown by Zhang et al. (20), at a functional level, inhibition of the nuclear translocation of NF-κB leads to TNF-α-induced apoptosis of synovial fibroblasts. In our experiments, gene transfer of TIMP-3 into RA-SF and MRC-5 cells resulted in proapoptotic effects of TNF-α. Specifically, TNF-α most significantly facilitated FasL/CD95L-induced cell death in AdTIMP-3-transduced RA-SF and pTIMP-3-transfected MRC-5 cells. In RA-SF, the extent of FasL/CD95L-induced apoptosis correlated positively with the concentration of TNF-α used for prior stimulation as well as with the levels of AdTIMP-3 expression. Most interestingly, a reduced activation of NF-κB was seen in the AdTIMP-3-transduced RA-SF, suggesting that a reduced activation of NF-κB by AdTIMP-3 may contribute to the proapoptotic effects of TNF-α in RA-SF that overexpress TIMP-3.

Another mechanism that may contribute to apoptosis-inhibiting effects of TNF-α in RA-SF is an increase of sFas by TNF-α. Several studies have shown that sFas is generated by alternative splicing of the Fas/CD95 mRNA (39, 40, 41) and acts as inhibitor of Fas/CD95-mediated apoptosis (39, 42, 43). As demonstrated in our study, stimulation of RA-SF with TNF-α increased significantly the levels of sFas in cell culture supernatants. Although the up-regulation of both Fas/CD95 mRNA and surface-bound Fas/CD95 receptor by TNF-α suggests that increased levels of sFas may reflect simply increased transcription, augmented shedding of Fas/CD95 receptor through up-regulation of putative Fas/CD95 sheddases may contribute also to this process. Although shedding of Fas/CD95 has been suggested to contribute to sFas levels (44), little is known about such shedding. In our study, AdTIMP-3 inhibited the TNF-α-mediated increase in sFas, which may also mediate the higher susceptibility of AdTIMP-3-transduced RA-SF to FasL/CD95L-induced cell death. The question of whether inhibition of a putative Fas/CD95 sheddase by TIMP-3 contributes to increased apoptosis of RA-SF is difficult to answer, but it is of interest that rhTIMP-3 even at high concentrations was unable to reverse the antiapoptotic effects of TNF-α. In addition, no differences in the expression of membrane-bound Fas/CD95 were observed between untransduced and AdTIMP-3-transduced RA-SF even after stimulation with TNF-α.

From these data, it may be speculated that intracellular mechanisms rather than mere inhibition of extracellular sheddases are responsible for important effects of TIMP-3. It may be linked to the glycosylation of TIMP-3, although data from Langton et al. (34) have demonstrated that glycosylation is not essential for either MMP inhibition or localization to the extracellular matrix. Of interest, Bond et al. (45) reported recently that overexpression of TIMP-3 induced apoptosis via Fas-associated death domain protein-dependent mechanisms that were independent of TNF-α, FasL/CD95L, or TRAIL, but involved the activation of caspases-8 and -9 as well as mitochondrial activation. In line with current concepts on signaling pathways in type I and type II cells (46), overexpression of bcl-2 was able to inhibit TIMP-3-mediated apoptosis in this study (45). Together with these findings, our data strongly support the concept that RA-SF are type II apoptotic cells, which are characterized by the expression of bcl-2 (47) and the induction of apoptosis by C2 ceramide (48). At the same time, our observations shed new light on the regulation of apoptosis in RA-SF cells by demonstrating that modulation of apoptotic pathways in type II cells alters their sensitivity to TNF-α. Although the exact mechanisms that mediate the effect of TIMP-3 require further investigations, these data have not only important implications for understanding the relation between aggressiveness of RA-SF and resistance to apoptosis, but also stimulate the development of gene transfer strategies delivering TIMP-3 as a novel approach for the treatment of RA and related disorders.

Acknowledgments

We thank Dr. Annelore Ittenson for her help with flow cytometry, and Bianca Henning, Sybille Pietzke, Desire Weber, and Jos Grimbergen for their expert technical assistance.

Disclosures

The authors have no financial conflict of interest.

Footnotes

  • The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

  • ↵1 This work was supported by the Deutsche Forschungsgemeinschaft (PA689-2 and PA689-3), The Netherlands Heart Foundation (M 93.001), and the Robert-Matthys Foundation.

  • ↵2 Address correspondence and reprint requests to Dr. Thomas Pap, Division of Molecular Medicine of Musculoskeletal Tissue, Department of Orthopedics, University Munster, Domagkstrasse 3, D-48149 Münster, Germany. E-mail address: thomas.pap{at}uni-muenster.de

  • ↵3 Abbreviations used in this paper: TIMP, tissue inhibitor of metalloproteinases; Ad, adenoviral; FasL, Fas ligand; h, human; MMP, matrix metalloproteinase; RA, rheumatoid arthritis; SF, synovial fibroblast; sFas, soluble Fas; TACE, TNF-α-converting enzyme.

  • Received August 12, 2004.
  • Accepted March 1, 2005.

References

  1. ↵
    Brew, K., D. Dinakarpandian, H. Nagase. 2000. Tissue inhibitors of metalloproteinases: evolution, structure and function. Biochim. Biophys. Acta 1477: 267-283.
    OpenUrlCrossRefPubMed
  2. ↵
    Blenis, J., S. P. Hawkes. 1984. Characterization of a transformation-sensitive protein in the extracellular matrix of chicken embryo fibroblasts. J. Biol. Chem. 259: 11563-11570.
    OpenUrlAbstract/FREE Full Text
  3. ↵
    Yu, W. H., S. Yu, Q. Meng, K. Brew, J. F. Woessner, Jr. 2000. TIMP-3 binds to sulfated glycosaminoglycans of the extracellular matrix. J. Biol. Chem. 275: 31226-31232.
    OpenUrlAbstract/FREE Full Text
  4. ↵
    Will, H., S. J. Atkinson, G. S. Butler, B. Smith, G. Murphy. 1996. The soluble catalytic domain of membrane type 1 matrix metalloproteinase cleaves the propeptide of progelatinase A and initiates autoproteolytic activation: regulation by TIMP-2 and TIMP-3. J. Biol. Chem. 271: 17119-17123.
    OpenUrlAbstract/FREE Full Text
  5. ↵
    Smith, M. R., H. Kung, S. K. Durum, N. H. Colburn, Y. Sun. 1997. TIMP-3 induces cell death by stabilizing TNF-α receptors on the surface of human colon carcinoma cells. Cytokine 9: 770-780.
    OpenUrlCrossRefPubMed
  6. ↵
    Hargreaves, P. G., F. Wang, J. Antcliff, G. Murphy, J. Lawry, R. G. Russell, P. I. Croucher. 1998. Human myeloma cells shed the interleukin-6 receptor: inhibition by tissue inhibitor of metalloproteinase-3 and a hydroxamate-based metalloproteinase inhibitor. Br. J. Haematol. 101: 694-702.
    OpenUrlCrossRefPubMed
  7. ↵
    Fitzgerald, M. L., Z. Wang, P. W. Park, G. Murphy, M. Bernfield. 2000. Shedding of syndecan-1 and -4 ectodomains is regulated by multiple signaling pathways and mediated by a TIMP-3-sensitive metalloproteinase. J. Cell Biol. 148: 811-824.
    OpenUrlAbstract/FREE Full Text
  8. ↵
    Amour, A., P. M. Slocombe, A. Webster, M. Butler, C. G. Knight, B. J. Smith, P. E. Stephens, C. Shelley, M. Hutton, V. Knauper, et al 1998. TNF-α converting enzyme (TACE) is inhibited by TIMP-3. FEBS Lett. 435: 39-44.
    OpenUrlCrossRefPubMed
  9. ↵
    Ahonen, M., A. H. Baker, V. M. Kahari. 1998. Adenovirus-mediated gene delivery of tissue inhibitor of metalloproteinases-3 inhibits invasion and induces apoptosis in melanoma cells. Cancer Res. 58: 2310-2315.
    OpenUrlAbstract/FREE Full Text
  10. ↵
    Baker, A. H., A. B. Zaltsman, S. J. George, A. C. Newby. 1998. Divergent effects of tissue inhibitor of metalloproteinase-1, -2, or -3 overexpression on rat vascular smooth muscle cell invasion, proliferation, and death in vitro: TIMP-3 promotes apoptosis. J. Clin. Invest. 101: 1478-1487.
    OpenUrlCrossRefPubMed
  11. ↵
    Baker, A. H., S. J. George, A. B. Zaltsman, G. Murphy, A. C. Newby. 1999. Inhibition of invasion and induction of apoptotic cell death of cancer cell lines by overexpression of TIMP-3. Br. J. Cancer 79: 1347-1355.
    OpenUrlCrossRefPubMed
  12. ↵
    Majid, M. A., V. A. Smith, D. L. Easty, A. H. Baker, A. C. Newby. 2002. Adenovirus mediated gene delivery of tissue inhibitor of metalloproteinases-3 induces death in retinal pigment epithelial cells. Br. J. Ophthalmol. 86: 97-101.
    OpenUrlAbstract/FREE Full Text
  13. ↵
    Pap, T., U. Muller-Ladner, R. E. Gay, S. Gay. 2000. Fibroblast biology: role of synovial fibroblasts in the pathogenesis of rheumatoid arthritis. Arthritis Res. 2: 361-367.
    OpenUrlCrossRefPubMed
  14. ↵
    Hanemaaijer, R., T. Sorsa, Y. T. Konttinen, Y. Ding, M. Sutinen, H. Visser, V. W. van Hinsbergh, T. Helaakoski, T. Kainulainen, H. Ronka, et al 1997. Matrix metalloproteinase-8 is expressed in rheumatoid synovial fibroblasts and endothelial cells: regulation by tumor necrosis factor-α and doxycycline. J. Biol. Chem. 272: 31504-31509.
    OpenUrlAbstract/FREE Full Text
  15. ↵
    Migita, K., K. Eguchi, Y. Kawabe, Y. Ichinose, T. Tsukada, T. Aoyagi, H. Nakamura, S. Nagataki. 1996. TNF-α-mediated expression of membrane-type matrix metalloproteinase in rheumatoid synovial fibroblasts. Immunology 89: 553-557.
    OpenUrlCrossRefPubMed
  16. ↵
    Grimbacher, B., W. K. Aicher, H. H. Peter, H. Eibel. 1998. TNF-α induces the transcription factor Egr-1, pro-inflammatory cytokines and cell proliferation in human skin fibroblasts and synovial lining cells. Rheumatol. Int. 17: 185-192.
    OpenUrlCrossRefPubMed
  17. ↵
    Wakisaka, S., N. Suzuki, Y. Takeba, Y. Shimoyama, H. Nagafuchi, M. Takeno, N. Saito, T. Yokoe, A. Kaneko, T. Asai, T. Sakane. 1998. Modulation by proinflammatory cytokines of Fas/Fas ligand-mediated apoptotic cell death of synovial cells in patients with rheumatoid arthritis (RA). Clin. Exp. Immunol. 114: 119-128.
    OpenUrlCrossRefPubMed
  18. ↵
    Mizushima, N., H. Kohsaka, N. Miyasaka. 1998. Ceramide, a mediator of interleukin 1, tumor necrosis factor α, as well as Fas receptor signalling, induces apoptosis of rheumatoid arthritis synovial cells. Ann. Rheum. Dis. 57: 495-499.
    OpenUrlAbstract/FREE Full Text
  19. ↵
    Firestein, G. S., M. Yeo, N. J. Zvaifler. 1995. Apoptosis in rheumatoid arthritis synovium. J. Clin. Invest. 96: 1631-1638.
    OpenUrlCrossRefPubMed
  20. ↵
    Zhang, H. G., N. Huang, D. Liu, L. Bilbao, X. Zhang, P. Yang, T. Zhou, D. T. Curiel, J. D. Mountz. 2000. Gene therapy that inhibits nuclear translocation of nuclear factor κB results in tumor necrosis factor α-induced apoptosis of human synovial fibroblasts. Arthritis Rheum. 43: 1094-1105.
    OpenUrlCrossRefPubMed
  21. ↵
    Miagkov, A. V., D. V. Kovalenko, C. E. Brown, J. R. Didsbury, J. P. Cogswell, S. A. Stimpson, A. S. Baldwin, S. S. Makarov. 1998. NF-κB activation provides the potential link between inflammation and hyperplasia in the arthritic joint. Proc. Natl. Acad. Sci. USA 95: 13859-13864.
    OpenUrlAbstract/FREE Full Text
  22. ↵
    Baier, A., I. Meineckel, S. Gay, T. Pap. 2003. Apoptosis in rheumatoid arthritis. Curr. Opin. Rheumatol. 15: 274-279.
    OpenUrlCrossRefPubMed
  23. ↵
    Pap, T., Y. Shigeyama, S. Kuchen, J. K. Fernihough, B. Simmen, R. E. Gay, M. Billingham, S. Gay. 2000. Differential expression pattern of membrane-type matrix metalloproteinases in rheumatoid arthritis. Arthritis Rheum. 43: 1226-1232.
    OpenUrlCrossRefPubMed
  24. ↵
    Van der Laan, W. H., P. H. Quax, C. A. Seemayer, L. G. Huisman, E. Pieterman, J. M. Grimbergen, J. H. Verheijen, F. C. Breedveld, R. E. Gay, S. Gay, et al 2003. Cartilage degradation and invasion by rheumatoid synovial fibroblasts is inhibited by gene transfer of TIMP-1 and TIMP-3. Gene Ther. 10: 234-242.
    OpenUrlCrossRefPubMed
  25. ↵
    Schreiber, E., P. Matthias, M. M. Müller, W. Schaffner. 1989. Rapid detection of octamer binding proteins with ‘mini-extracts,’ prepared from a small number of cells. Nucleic Acids Res. 17: 6419.
    OpenUrlFREE Full Text
  26. ↵
    Neumann, M., T. Grieshammer, S. Chuvpilo, B. Kneitz, M. Lohoff, A. Schimpl, B. R. Franza, E. Serfling. 1995. RelA/p65 is a molecular target for the immunosuppressive action of protein kinase A. EMBO J. 14: 1991-2004.
    OpenUrlPubMed
  27. ↵
    Peli, J., M. Schroter, C. Rudaz, M. Hahne, C. Meyer, E. Reichmann, J. Tschopp. 1999. Oncogenic Ras inhibits Fas ligand-mediated apoptosis by down-regulating the expression of Fas. EMBO J. 18: 1824-1831.
    OpenUrlAbstract
  28. ↵
    Renard, P., I. Ernest, A. Houbion, M. Art, H. Le Calvez, M. Raes, J. Remacle. 2001. Development of a sensitive multi-well colorimetric assay for active NFκB. Nucleic Acids Res. 29: E21.
    OpenUrlCrossRefPubMed
  29. ↵
    Bren, G. D., N. J. Solan, H. Miyoshi, K. N. Pennington, L. J. Pobst, C. V. Paya. 2001. Transcription of the RelB gene is regulated by NF-κB. Oncogene 20: 7722-7733.
    OpenUrlCrossRefPubMed
  30. ↵
    Zheng, Y., F. Ouaaz, P. Bruzzo, V. Singh, S. Gerondakis, A. A. Beg. 2001. NF-κB RelA (p65) is essential for TNF-α-induced fas expression but dispensable for both TCR-induced expression and activation-induced cell death. J. Immunol. 166: 4949-4957.
    OpenUrlAbstract/FREE Full Text
  31. ↵
    Takizawa, M., E. Ohuchi, H. Yamanaka, H. Nakamura, E. Ikeda, P. Ghosh, Y. Okada. 2000. Production of tissue inhibitor of metalloproteinases 3 is selectively enhanced by calcium pentosan polysulfate in human rheumatoid synovial fibroblasts. Arthritis Rheum. 43: 812-820.
    OpenUrlCrossRefPubMed
  32. ↵
    Yu, W. H., J. F. Woessner, J. D. McNeish, I. Stamenkovic. 2002. CD44 anchors the assembly of matrilysin/MMP-7 with heparin-binding epidermal growth factor precursor and ErbB4 and regulates female reproductive organ remodeling. Genes Dev. 16: 307-323.
    OpenUrlAbstract/FREE Full Text
  33. ↵
    Zhao, H., M. M. Bernardo, P. Osenkowski, A. Sohail, D. Pei, H. Nagase, M. Kashiwagi, P. D. Soloway, Y. A. DeClerck, R. Fridman. 2004. Differential inhibition of membrane type 3 (MT3)-matrix metalloproteinase (MMP) and MT1-MMP by tissue inhibitor of metalloproteinase (TIMP)-2 and TIMP-3 regulates pro-MMP-2 activation. J. Biol. Chem. 279: 8592-8601.
    OpenUrlAbstract/FREE Full Text
  34. ↵
    Langton, K. P., M. D. Barker, N. McKie. 1998. Localization of the functional domains of human tissue inhibitor of metalloproteinases-3 and the effects of a Sorsby’s fundus dystrophy mutation. J. Biol. Chem. 273: 16778-16781.
    OpenUrlAbstract/FREE Full Text
  35. ↵
    Bond, M., G. Murphy, M. R. Bennett, A. Amour, V. Knauper, A. C. Newby, A. H. Baker. 2000. Localization of the death domain of tissue inhibitor of metalloproteinase-3 to the N terminus: metalloproteinase inhibition is associated with proapoptotic activity. J. Biol. Chem. 275: 41358-41363.
    OpenUrlAbstract/FREE Full Text
  36. ↵
    Woessner, J. F., Jr. 2001. That impish TIMP: the tissue inhibitor of metalloproteinases-3. J. Clin. Invest. 108: 799-800.
    OpenUrlCrossRefPubMed
  37. ↵
    Zhang, H. G., Y. Wang, J. F. Xie, X. Liang, D. Liu, P. Yang, H. C. Hsu, R. B. Ray, J. D. Mountz. 2001. Regulation of tumor necrosis factor α-mediated apoptosis of rheumatoid arthritis synovial fibroblasts by the protein kinase Akt. Arthritis Rheum. 44: 1555-1567.
    OpenUrlCrossRefPubMed
  38. ↵
    Pap, T., J. K. Franz, K. M. Hummel, E. Jeisy, R. Gay, S. Gay. 2000. Activation of synovial fibroblasts in rheumatoid arthritis: lack of expression of the tumour suppressor PTEN at sites of invasive growth and destruction. Arthritis Res. 2: 59-64.
    OpenUrlCrossRefPubMed
  39. ↵
    Cheng, J., T. Zhou, C. Liu, J. P. Shapiro, M. J. Brauer, M. C. Kiefer, P. J. Barr, J. D. Mountz. 1994. Protection from Fas-mediated apoptosis by a soluble form of the Fas molecule. Science 263: 1759-1762.
    OpenUrlAbstract/FREE Full Text
  40. ↵
    Cascino, I., G. Fiucci, G. Papoff, G. Ruberti. 1995. Three functional soluble forms of the human apoptosis-inducing Fas molecule are produced by alternative splicing. J. Immunol. 154: 2706-2713.
    OpenUrlAbstract
  41. ↵
    Knipping, E., K. M. Debatin, K. Stricker, B. Heilig, A. Eder, P. H. Krammer. 1995. Identification of soluble APO-1 in supernatants of human B- and T-cell lines and increased serum levels in B- and T-cell leukemias. Blood 85: 1562-1569.
    OpenUrlAbstract/FREE Full Text
  42. ↵
    Papoff, G., I. Cascino, A. Eramo, G. Starace, D. H. Lynch, G. Ruberti. 1996. An N-terminal domain shared by Fas/Apo-1 (CD95) soluble variants prevents cell death in vitro. J. Immunol. 156: 4622-4630.
    OpenUrlAbstract
  43. ↵
    Suda, T., H. Hashimoto, M. Tanaka, T. Ochi, S. Nagata. 1997. Membrane Fas ligand kills human peripheral blood T lymphocytes, and soluble Fas ligand blocks the killing. J. Exp. Med. 186: 2045-2050.
    OpenUrlAbstract/FREE Full Text
  44. ↵
    Bijl, M., T. van Lopik, P. C. Limburg, P. E. Spronk, S. M. Jaegers, L. A. Aarden, R. J. Smeenk, G. G. Kallenberg. 1998. Do elevated levels of serum-soluble fas contribute to the persistence of activated lymphocytes in systemic lupus erythematosus?. J. Autoimmun. 11: 457-463.
    OpenUrlCrossRefPubMed
  45. ↵
    Bond, M., G. Murphy, M. R. Bennett, A. C. Newby, A. H. Baker. 2002. Tissue inhibitor of metalloproteinase-3 induces a Fas-associated death domain-dependent type II apoptotic pathway. J. Biol. Chem. 277: 13787-13795.
    OpenUrlAbstract/FREE Full Text
  46. ↵
    Scaffidi, C., I. Schmitz, J. Zha, S. J. Korsmeyer, P. H. Krammer, M. E. Peter. 1999. Differential modulation of apoptosis sensitivity in CD95 type I and type II cells. J. Biol. Chem. 274: 22532-22538.
    OpenUrlAbstract/FREE Full Text
  47. ↵
    Perlman, H., C. Georganas, L. J. Pagliari, A. E. Koch, K. Haines, III, R. M. Pope. 2000. Bcl-2 expression in synovial fibroblasts is essential for maintaining mitochondrial homeostasis and cell viability. J. Immunol. 164: 5227-5235.
    OpenUrlAbstract/FREE Full Text
  48. ↵
    Ichinose, Y., K. Eguchi, K. Migita, Y. Kawabe, T. Tsukada, T. Koji, K. Abe, T. Aoyagi, H. Nakamura, S. Nagataki. 1998. Apoptosis induction in synovial fibroblasts by ceramide: in vitro and in vivo effects. J. Lab. Clin. Med. 131: 410-416.
    OpenUrlCrossRefPubMed
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section