ADAM15 Adds to Apoptosis Resistance of Synovial Fibroblasts by Modulating Focal Adhesion Kinase Signaling

Authors


Abstract

Objective

To study the contribution of ADAM15, a disintegrin metalloproteinase that is up-regulated in the rheumatoid arthritis (RA) synovial membrane, to the characteristic resistance of RA synovial fibroblasts (RASFs) to apoptosis induction by genotoxic stress or stimulation with proapoptotic FasL, which is present at high concentrations in RA synovial fluid.

Methods

Caspase 3/7 activity and the total apoptosis rate in RASFs upon exposure to the DNA-damaging agent camptothecin or FasL were determined using enzyme assays and annexin V staining. Phosphorylated signaling proteins were analyzed by immunoblotting. RNA interference was used to silence ADAM15 expression. NF-κB activity was determined by enzyme-linked immunosorbent assay.

Results

RASFs displayed significantly higher caspase 3/7 activity upon camptothecin and FasL exposure when ADAM15 had been down-regulated by specific small interfering RNAs. Upon FasL stimulation, RASFs phosphorylated focal adhesion kinase (FAK) and c-Src (Src), and activated phosphatidylinositol 3-kinase as well as the transcription factor NF-κB. This ADAM15-dependent, FasL-induced activation of antiapoptotic kinases and NF-κB was demonstrated by a marked reduction of apoptosis upon knockdown of ADAM15 protein expression. Inhibitors specifically interfering with FAK and Src signaling, such as FAK inhibitor 14 and dasatinib, potently induce apoptosis in RASFs, with significant enhancement by the silencing of ADAM15.

Conclusion

ADAM15 contributes to apoptosis resistance in RASFs by activating the Src/FAK pathway upon FasL exposure, rendering the FAK/Src signaling pathway an interesting target for potential therapeutic intervention in RA.

ADAM15 belongs to the family of disintegrins and metalloproteinases and is a cell membrane–anchored multidomain protein composed mainly of a metalloproteinase (which is kept inactive by a pro domain), a disintegrin, and a cytoplasmic domain ([1]). ADAM15 plays a role in pathologic neovascularization ([2]). Its overexpression is correlated positively with tumor staging of esophagogastric adenocarcinoma ([3]), and it is strongly associated with aggressive forms of prostate and breast cancer ([4, 5]). Besides its role in tumorigenesis, ADAM15 is implicated in non-neoplastic conditions of extracellular matrix remodeling in degenerative joint disease ([6-8]), as well as in inflammatory diseases such as rheumatoid arthritis (RA) ([9]). ADAM15 is markedly up-regulated in the synovial membranes of RA patients, with a strong expression in the hyperplastic synovial lining layer ([10, 11]).

RA is an immune-mediated systemic disease that is predominantly manifested in the joints as a chronic inflammatory process initiated and maintained by cells of the innate and adaptive immune system interacting with activated synovial fibroblasts. These cells are predominantly implicated in the destruction of articular cartilage and bone by secreting matrix-degrading proteinases in addition to a variety of proinflammatory mediators ([12]). During the inflammatory process, the stroma and the lining of the synovial tissue forming the inner layer of the joint capsule become hyperplastic by infiltration with lymphocytes and macrophages ([13]). However, stromal cells also contribute to synovial hyperplasia. Thus, the infrequent presence of cellular apoptosis in RA synovium has been ascribed to disturbances in apoptotic pathways, resulting in the local accumulation of synovial fibroblasts ([12-15]).

Despite a low rate of occurrence, apoptotic cell death in RA synovial fibroblasts (RASFs) is mainly induced by members of the tumor necrosis factor (TNF) superfamily. In this regard, the binding of FasL to Fas, a member of the TNF death receptor family, represents a dominant trigger that initiates an apoptosis cascade via formation of the so-called death-inducing signaling complex ([16, 17]). However, FasL-induced pathways seem to be rather complex, not invariably leading to apoptotic cell death, but also being capable of conveying antiapoptotic signals under certain circumstances ([18]). Among a variety of downstream signaling cascades elicited by FasL is a pathway that leads to the activation of NF-κB and the subsequent transcriptional up-regulation of genes encoding Bcl-2 (an inhibitor of apoptosis proteins) and FLIP (the caspase 8 inhibitor), which confer an increased resistance to FasL-mediated apoptosis ([19]). Another FasL-triggered survival signaling pathway involves the activation of phosphatidylinositol 3-kinase (PI3K)/Akt ([20]).

The present investigation aimed at elucidating whether ADAM15 may contribute to increased apoptosis resistance of RASFs via its interaction with focal adhesion kinase (FAK). FAK, a nonreceptor tyrosine kinase, functions as a critical scaffolding molecule that integrates signals transmitted by integrins and growth factor receptors into triggers of growth, differentiation, and survival pathways ([21]). Indeed, FAK was previously demonstrated to be actively involved in antiapoptotic pathways triggered by oxidative stress, hyperosmotic conditions ([22-24]), or the disruption of cell–matrix contacts in anchorage-dependent cells ([25]). Moreover, direct binding of ADAM15 to FAK has recently been shown to be crucial in eliciting the antiapoptotic activation of FAK and Src in osteoarthritic (OA) chondrocytes in response to genotoxic stress ([26]).

In the present investigation, we found experimental evidence that ADAM15 exerts an antiapoptotic response in FasL-stimulated RASFs, suggesting its significant contribution to their pronounced apoptosis resistance. The activation of this ADAM15-dependent survival pathway was accompanied by the up-regulation of NF-κB and PI3K, as well as Src and FAK phosphorylation. Accordingly, a dependency of apoptosis induction in RASFs on ADAM15 expression was also demonstrated by compounds that interfere with Src/FAK signaling, such as dasatinib or FAK inhibitor 14 (FAK-14). Thus, our results uncover a critical role of ADAM15 as a potent suppressor of FasL-induced apoptosis in RASF by enhancing the cell-protecting properties of FAK signaling. This newly elucidated FAK-mediated protective mechanism consists of an enhancement of antiapoptotic NF-κB and PI3K activation, which was concomitantly elicited with the proapoptotic stimulus by Fas ligation.

MATERIALS AND METHODS

Materials

Mouse and goat anti-ADAM15 antibodies (recognizing the pro domain) were obtained from R&D Systems. Rabbit antibodies against the cytoplasmic domain of ADAM15 were from Abcam. Rabbit anti–pY397 FAK was from BioSource, and anti–pY576/577 FAK, anti–pY861 FAK, and antitubulin were from Epitomics. Rabbit anti–pY416 Src, rabbit anti–pSer32 IκBα, and rabbit anti–phospho-PI3K (p85, Y458) were from Cell Signaling Technology. FasL and TNFα were from PeproTech. Camptothecin was from Calbiochem. FAK inhibitor 14 was from Tocris Bioscience, and dasatinib (Sprycel) was from Bristol-Myers Squibb.

RA synovial tissue and cell culture

Synovial tissue was obtained during joint replacement/arthroscopic synovectomy at the Clinic of Orthopedics, University Hospital Jena. All patients met the American College of Rheumatology 1987 criteria for RA ([27]) and had established RA of >3 years' duration. A total of 80% of the patients were rheumatoid factor positive. The disease activity was moderate to severe. All patients were receiving conventional disease-modifying antirheumatic drug (DMARD) therapy (methotrexate, leflunomide, sulfasalazine, or azathioprine, with 1 patient receiving adalimumab plus leflunomide). The dosage of DMARDs was tapered prior to joint replacement/arthroscopic synovectomy, according to the national recommendations. At the time of surgery, none of the patients were receiving any antirheumatic treatment besides nonsteroidal antiinflammatory drugs and/or prednisolone at a dosage of ≤10 mg.

RASFs were isolated and grown in Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal calf serum (FCS), as previously described ([28]). For all subsequent tests, cells were grown to subconfluence (4 × 106 cells per 75-cm2 tissue culture flask), and passage 3–6 cells were used for all experiments.

The study was approved by the Ethics Committee of the University Hospital Jena. Informed consent was obtained from the patients.

Silencing of ADAM15 in synovial fibroblasts by RNA interference

Synovial fibroblasts (1 × 104 or 1 × 105) were seeded in 96- or 24-well culture plates and grown for 24 hours. Cells were treated with 20 nM Silencer Select predesigned small interfering RNAs (siRNAs; Ambion) for ADAM15 as described previously ([26]). Nonsilencing siRNA Control #1 (Ambion) was used as the negative control.

Determination of caspase 3/7 activity

Synovial fibroblasts (1 × 104) were seeded in 96-well white tissue culture plates (Greiner) and grown for 24 hours in DMEM containing 10% FCS. After silencing of ADAM15 for 40 hours, cells were treated with either camptothecin (20 μM), FasL (100 ng/ml), TNFα (100 ng/ml), or the signaling inhibitors PP2, FAK inhibitor 14, and dasatinib at the indicated concentrations. Caspase 3/7 activity was measured using the Caspase-Glo 3/7 assay (Promega) on a Mithras LB 940 luminometer plate reader (Berthold) as described previously ([26]).

Determination of apoptosis

Apoptosis was analyzed using a Cell Meter annexin V binding apoptosis assay kit from Biomol. RASFs (1 × 104) were grown in chamber slides (BD Falcon) for 24 hours, and ADAM15 was silenced for 40 hours. Apoptosis was induced with camptothecin (20 μM) or FasL (100 ng/ml) at the indicated time points. After fixation in 4% paraformaldehyde in phosphate buffered saline (PBS) and blocking with PBS–1% bovine serum albumin, apoptotic cells were stained using annexin V–iFluor 488 conjugate that specifically binds to phosphatidylserine, which is transferred to the outer leaflet of the plasma membrane during apoptosis. Nuclei were counterstained with DAPI. At least 200 cell nuclei were counted, and the percentage of cells positively stained for annexin V was determined.

Preparation of cell lysates and Western blotting

Cell lysates were prepared and immunoblotted as described previously ([26]). Signals were exposed to x-ray film, scanned, and signal densities were measured using the ImageJ program (National Institutes of Health). Results of the densitometric analysis were used to calculate the changes in protein levels.

NF-κB assay

NF-κB activation was measured using a TransAM NF-κB p65 assay (Active Motif) according to the supplier's instructions. Briefly, FasL-stimulated cell lysates were added to a 96-well plate onto which oligonucleotide containing an NF-κB consensus–binding sequence had been immobilized. The activated NF-κB p65 subunit bound to the oligonucleotide was detected with an anti–NF-κB antibody and visualized with secondary anti–horseradish peroxidase–conjugated antibody. Chemiluminescence was quantified with a luminometer plate reader.

Statistical analysis

Data presented are the mean ± SD of quadruplicate samples of at least 5 independently performed assays. Statistical significance was determined using Student's unpaired t-test. P values less than 0.05 were considered significant.

RESULTS

ADAM15-induced resistance of RASFs to different modes of apoptosis induction

We have previously shown that ADAM15 acts as an antiapoptotic factor upon induction of genotoxic stress in OA chondrocytes with camptothecin ([29]). In the present study, we analyzed whether this property of ADAM15 was restricted to OA chondrocytes and whether apoptotic stimuli other than DNA damage could also induce an increased rate of apoptosis.

ADAM15 was knocked down using 2 specific siRNAs, and protein expression was examined by immunoblotting. ADAM15 protein expression in RASFs was silenced by ∼80–90% after 48 hours of treatment with specific siRNAs, as compared to the expression in RASFs treated with scrambled negative siRNA. After 72 hours of silencing, ADAM15 was reduced by ∼75–80% as compared to the expression after treatment with the negative siRNA control (Figure 1A). After silencing of ADAM15 for 40 hours with siRNA I or siRNA II, RASFs were treated for up to 48 hours with camptothecin, FasL, or TNFα, and caspase 3/7 activity was then measured (Figure 1B). Independently of the apoptotic stimulus applied, a highly significant, increased caspase 3 activity of ∼2.0–2.5-fold was observed in RASFs when ADAM15 expression was silenced as compared to the expression in RASFs treated with scrambled control siRNA. Moreover, of all of the RASF samples analyzed (n = 10 or more), the compound that consistently induced the highest caspase 3 activity was camptothecin, followed by FasL. TNFα displayed the lowest caspase 3–inducing capacity.

Figure 1.

ADAM15 protection of rheumatoid arthritis synovial fibroblasts (RASFs) against apoptosis. A, ADAM15 expression after silencing for 48 or 72 hours with ADAM15-specific small interfering RNA (siRNA) I and siRNA II was compared to treatment with a nonsilencing siRNA control (N) or transfection agent alone (0), as determined by immunoblotting with anti ADAM15 antibodies. B, Down-regulation of ADAM15 by ADAM15 siRNAs I and II (solid bars) led to significantly increased caspase 3/7 activity as compared to treatment with a nonsilencing siRNA control (open bars) or transfection agent alone (shaded bars) upon apoptosis induction by camptothecin (left), FasL (middle), or tumor necrosis factor α (TNFα) (right) at the indicated times. Values are the mean ± SD. ∗ = P < 0.05; ∗∗ = P < 0.005; ∗∗∗ = P < 0.0005 versus nonsilencing controls at the same time point. C, Apoptosis induction in RASFs from 2 different donors following treatment with FasL for 24 or 48 hours resulted in a higher total apoptosis rate in cells silenced with ADAM15-specific siRNAs I and II as compared to treatment with a nonsilencing siRNA control. Results are representative of 5 assays. The total apoptosis rate (%) was determined by correlating the cells that stained positive for annexin V to the whole cell population visualized with the use of DAPI.

To study the total apoptosis rate, we applied fluorescence-labeled annexin V to adherent, grown RASFs that had been silenced for 40 hours with the specific siRNAs or treated with a negative siRNA control and then subsequently exposed to FasL for 24 or 48 hours. Cell nuclei were visualized with DAPI. The cells positive for annexin V were counted, and the percentage of apoptotic cells was determined. The overall apoptosis rate induced by FasL varied considerably in the different RASF samples, ranging from ∼20% to ∼40% after 48 hours (Figure 1C). Analogous to the caspase assays shown, the silencing of ADAM15 resulted in a doubling of the apoptosis rate as compared to the nonsilencing control siRNA. Thus, depending on the donor, we detected total apoptosis rates of up to 80% (Figure 2).

Figure 2.

Phosphorylation of focal adhesion kinase (FAK) and Src in rheumatoid arthritis synovial fibroblasts (RASFs) after stimulation with camptothecin or FasL, as demonstrated by immunoblotting. A, RASFs exhibited phosphorylation of Src at Y416 and of FAK at Y576, Y861, and the autophosphorylation site Y397 upon stimulation with camptothecin as well as with FasL at the indicated times. B, Silencing of ADAM15 by ADAM15-specific small interfering RNA (siRNA) I and siRNA II resulted in markedly reduced phosphorylation of FAK at Y576, Y861, and Y397, as well as of Src at Y416, upon FasL stimulation for the indicated times. Tubulin served as a loading control.

Synovial fibroblast activation of FAK and Src upon apoptosis induction with camptothecin or FasL

We previously showed that OA chondrocytes respond to genotoxic stress conferred by camptothecin with an increase in phosphorylation of FAK and Src. In the presence of ADAM15, FAK and Src signaling is considerably amplified ([26]). To study whether RASFs also activate Src and FAK in response to either genotoxic stress or receptor-triggered apoptosis, the cells were incubated for 0–60 minutes with camptothecin or FasL, and cell lysates were subjected to immunoblotting using antibodies specific for FAK Y576, Y861, or the autophosphorylation site Y397, as well as antibodies specific for Y416 of Src. Stimulation of RASFs with either camptothecin or FasL led to phosphorylation of Src at its activation site Y416 and of FAK at Y861, Y576, and the autophosphorylation site Y397 (Figure 2A).

To analyze the influence of ADAM15 on FasL-induced FAK and Src phosphorylation, ADAM15 in RASFs was silenced by 2 specific siRNAs, siRNA I and siRNA II, for 48 hours, and the phosphorylation of FAK and Src was determined by immunoblotting. All 3 tyrosines of FAK were phosphorylated to a significantly lower degree (∼2.0-fold) in the ADAM15-silenced RASFs upon FasL stimulation for 30 and 60 minutes, as compared to cells silenced with a nonsilencing siRNA or cells treated with transfection agent alone (Figure 2B). Accordingly, FasL stimulation for 15 and 30 minutes induced a stronger phosphorylation signal of Src at Y416 in RASFs that had been pretreated with either nonsilencing siRNA or transfection reagent as compared to ADAM15-silenced RASFs (Figure 2B). This clearly demonstrates that FasL signaling via Src and FAK is dependent on ADAM15 protein expression.

Induction of apoptosis in RASFs by signal transduction inhibitors of FAK and Src

To analyze whether Src and FAK signaling induced by FasL stimulation had an impact on cellular apoptotic events in RASFs, the Src inhibitor PP2 and dasatinib, which inhibits, among other kinases, Src and FAK ([30]), were applied for 24 hours, and caspase 3/7 activity was subsequently determined in the treated cells. When compared to apoptosis induction by FasL, both PP2 and dasatinib induced caspase activity to a markedly higher degree, as shown in Figures 3A and B. The impact on RASF apoptosis of another inhibitor, FAK-14, which specifically docks to the autophosphorylation site Y397 of FAK, thereby blocking its activation ([31]), was also determined. FAK-14 is capable of inducing caspase 3 activity in RASFs in the low micromolar range, but the apoptosis-inducing efficiency was highly heterogeneous and rather fine-tuned in different donors, as shown in Figure 3B for 2 representative donors. Whereas FAK-14 resulted in maximum induction of apoptosis in RASFs from donor 1 after 21 hours of incubation at a concentration of 2 μM, it was lower than that obtained with PP2 incubation but was in the same concentration range as that of incubation with FasL. The maximum apoptosis-inducing capacity for FAK-14 in RASFs from donor 2 was obtained at a concentration of 5 μM (Figure 3B). Moreover, a slightly higher FAK-14 concentration than that required for the maximum donor-adjusted apoptosis induction did not result in caspase 3 activity (e.g., 6 μM in donor 2) (Figure 3B), but led to an early detachment of the cells and necrosis (data not shown), which was not observed for FasL (100 ng/ml) or PP2 (1 and 10 μM).

Figure 3.

Induction of apoptosis by focal adhesion kinase (FAK) and Src signal transduction inhibitors in rheumatoid arthritis synovial fibroblasts (RASFs). A, Apoptosis induction by treatment with PP2 or dasatinib for 24 hours resulted in significantly higher caspase 3/7 activity as compared to treatment with FasL (100 ng/ml). Camptothecin (campto) treatment was applied at 20 μM. B, Treatment with FAK inhibitor 14 (FAK-14) induced caspase 3 activity in the low micromolar range (2–6 μM), as compared with the effects of PP2 (1 and 10 μM) and FasL (100 ng/ml) treatment, as shown in 2 representative donors (of 8 donors tested). The maximum apoptosis-inducing capacity of FAK-14 was finely tuned for each donor, since individual higher concentrations resulted in cell necrosis. Values in A and B are the mean ± SD. C, Dasatinib and PP2 showed complete inhibition of FAK phosphorylation at Y861, Y576, and Y397, as well as of Src at Y416, in RASFs at the indicated concentrations, as demonstrated by immunoblotting of cells from a representative donor. FAK-14 showed complete inhibition of FAK phosphorylation at the autophosphorylation site Y397 and ∼70% inhibition at Y576 and Y861 at concentrations of 2 μM and 4 μM. Treatment with Dulbecco's modified Eagle's medium (DMEM) or DMSO was used as the control. Tubulin served as loading control.

In addition, the influence of dasatinib, FAK-14, and PP2 on the phosphorylation of FAK and Src was analyzed by immunoblotting using phosphospecific antibodies. RASFs incubated for 6 hours with DMEM alone or with the DMSO control phosphorylated FAK at Y567, Y861, and Y397, and Src at Y416. Incubation with dasatinib (30–1,000 nM) resulted in the complete suppression of phosphorylation of Y576 and Y861 of FAK at a concentration as low as 30 nM and ∼90% for Y397 of FAK (Figure 3C). Accordingly, the activation of Src at Y416 was suppressed with 30 nM dasatinib. FAK-14 potently inhibited the phosphorylation of FAK at its autophosphorylated Y397 at a concentration of 1 μM, but higher concentrations (2 and 4 μM) were required for ∼50–70% inhibition of the phosphorylation of FAK at Y576 and Y861. Also, Src phosphorylation at Y416 was not inhibited with FAK-14 at 1 μM and was reduced to ∼50–70% at 2 and 4 μM FAK-14. Complete suppression of the phosphorylation of Src was detected with the Src inhibitor PP2.

Induction of higher levels of caspase 3 activity by signal transduction inhibitors of FAK and Src in the absence of ADAM15.

To analyze the impact of ADAM15 on apoptosis induction by the inhibition of FAK signaling, the inhibitors dasatinib, FAK-14, and PP2 were applied to RASFs that had been silenced for 48 or 72 hours with ADAM15-specific siRNAs I and II, and caspase 3/7 activity was measured. Highly significant increases in caspase 3/7 activity were observed in the ADAM15-deficient RASFs as compared to cells silenced with a negative siRNA control or with transfection reagent alone upon apoptosis induction with dasatinib (100 nM), FAK-14 (2 μM), or PP2 (10 μM) (Figure 4). Thus, the clear involvement of ADAM15 in FAK/Src signaling-mediated survival pathways was revealed. Apoptosis induction by serum withdrawal using DMEM served as a control, and this also resulted in higher levels of caspase 3 activity in ADAM15-silenced RASFs as compared to RASFs that had been silenced with a negative siRNA.

Figure 4.

Higher caspase 3 activity upon apoptosis induction with focal adhesion kinase (FAK) and Src signaling inhibitors following ADAM15 silencing in rheumatoid arthritis synovial fibroblasts (RASFs). After ADAM15 was silenced by 40 hours of treatment with specific small interfering RNA (siRNA) I and siRNA II, RASFs were incubated for 48 or 72 hours with FasL (100 ng/ml), dasatinib (100 nM), FAK-14 (2 μM), or PP2 (10 μM), and caspase 3/7 activity was determined. Apoptosis induction by the FAK/Src inhibitors dasatinib, FAK-14, and PP2 resulted in significantly increased caspase 3 activity in ADAM15-silenced RASFs as compared to treatment with a nonsilencing siRNA control (N) or transfection agent alone (0). Results are from a representative donor (of 5 donors tested). Values are the mean ± SD. ∗ = P < 0.05; ∗∗ = P < 0.005; ∗∗∗ = P < 0.0005 versus nonsilencing controls at the same time point.

Influence of ADAM15 on FasL signaling via NF-κB and PI3K

Since FasL can trigger NF-κB and the PI3K/Akt pathway, the potential modulatory role of ADAM15 was analyzed by immunoblotting using antibodies specific for phosphorylated PI3K and phospho-IκBα, as a marker of activated NF-κB. In addition, complementary investigations measuring NF-κB activity by enzyme-linked immunosorbent assay were performed. RASFs stimulated with increasing concentrations of FasL (50–250 ng/ml) displayed increasing phosphorylation of IκBα (Figure 5A), as well as increased NF-κB activity after FasL stimulation for 18 hours (Figure 5B). Silencing of ADAM15 using siRNAs I and II and subsequent stimulation with FasL revealed a ∼2–3-fold lower level of phosphorylation of IκBα, as compared to RASFs that had been pretreated with either nonsilencing siRNA or transfection agent alone (Figure 5C). Accordingly, after FasL stimulation for 12 or 18 hours, NF-κB activity was significantly lower in RASFs with down-regulated ADAM15 than in RASFs that had been silenced with a negative control or transfection agent alone (Figure 5D). Also, FasL potently phosphorylated PI3K in RASFs in a concentration-dependent manner (Figure 5A), which was considerably reduced when ADAM15 expression was silenced by specific siRNAs I and II (Figure 5C), demonstrating clear involvement of ADAM15 in the NF-κB pathway as well as the PI3K/Akt pathway.

Figure 5.

ADAM15 expression–dependent phosphatidylinositol 3-kinase (PI3K) and NF-κB activation following FasL stimulation of rheumatoid arthritis synovial fibroblasts (RASFs). A, Cell lysates that had been treated with 50–250 ng/ml of FasL showed phosphorylation of IκBα and PI3K, as demonstrated by immunoblotting. B, FasL stimulation of RASFs for 18 hours increased NF-κB activity in a concentration-dependent manner (0–500 ng/ml). C and D, After ADAM15 was silenced by 40 hours of treatment with specific small interfering RNA (siRNA) I and siRNA II, RASFs were incubated with FasL (100 ng/ml), and phosphorylation of IκBα and PI3K was visualized by immunoblotting using phosphospecific antibodies (C) or NF-κB activity was determined (D). There was stronger activation of IκBα and NF-κB in RASFs that had been silenced with a nonsilencing siRNA control (N) or transfection agent alone (0) as compared to ADAM15-silenced RASFs. Results are from a representative donor. Values in B and D are the mean ± SD. ∗ = P < 0.05; ∗∗ = P < 0.005; ∗∗∗ = P < 0.0005 versus nonsilencing controls at the same time point.

DISCUSSION

ADAM15 is highly expressed in the RA synovial membrane as well as in OA chondrocytes, whereas its expression levels are very low in normal, nondiseased cartilage and synovial tissue ([6, 10, 11]). ADAM15 has been reported to reinforce antiapoptotic pathways in OA chondrocytes upon serum and matrix withdrawal or under genotoxic stress induced by camptothecin exposure ([26, 29]). The present investigation clearly showed that the expression of ADAM15 made a considerable contribution to the capacity of RASFs to resist apoptosis induction by camptothecin-conferred DNA damage as well as by FasL- or TNFα-triggered death receptor pathways. Our study further demonstrated that ADAM15 was capable of conferring survival properties to RASFs, allowing them to withstand exposure to rather diverse apoptotic stimuli.

Likewise, several other proteins that interfere with different death receptor–triggered signaling pathways have been shown to contribute to apoptosis resistance in RASFs. Thus, it has been shown that down-regulation of FLIP, an important component of the death-inducing signaling complex, sensitizes RASFs to Fas-induced apoptosis ([32]). FLIP also conferred protective effects in TNFα-induced apoptosis. Interestingly, in this situation, TNFα as the stimulus of death receptor-mediated pathways has also been demonstrated to concomitantly trigger an NF-κB activation loop, resulting in increased FLIP expression and enhanced apoptosis resistance in the RASFs ([33]). Sentrin, which is localized to sites of invasion of the synovium, has been demonstrated to bind to the cytoplasmic tails of the Fas receptor and TNF receptor type I, thereby displaying cell-protective properties in RASFs ([34]). POSH, an SH3 domain– containing protein that serves as a scaffold in a multiprotein complex that mediates JNK activation in apoptosis, has also been found to promote synovial cell survival ([35]).

It has been shown that ADAM15 can modulate apoptosis-inducing signals elicited by genotoxic stress in OA chondrocytes via activation of FAK at the tyrosine residues Y576 and Y861, representing common targets of Src kinase–mediated phosphorylation ([21, 26]). Remarkably, our studies provide the first experimental evidence that stimulation of RASFs with FasL also results in the phosphorylation of Src and FAK. Moreover, this FasL-induced FAK/Src activation in RASFs was found to be strongly dependent on concomitant ADAM15 expression. Accordingly, ADAM15 silencing by specific siRNAs caused a pronounced reduction of the FAK/Src phosphorylation signal in response to FasL challenge. The crucial role of FAK/Src interactions with ADAM15 in providing survival signals was further corroborated by the application of the Src/FAK kinase inhibitors, dasatinib and FAK-14. Whereas both kinase inhibitors were able to induce caspase 3 activity in RASFs, the concomitant down-regulation of ADAM15 further enhanced their proapoptotic effect.

Dasatinib, a pan–Src kinase family and Bcr-Abl inhibitor, displays strong antitumor and antiproliferative activity against a variety of hematologic and solid tumor cell lines, and it has been approved for treatment of chronic myeloid leukemia ([36]). Besides inhibiting Src and other receptor kinases in the low nanomolar range, it also affects Src downstream kinases, such as FAK ([30]). In our experiments, dasatinib at low nanomolar concentrations efficiently blocked the phosphorylation of Src and of FAK at Y576 and Y861, as well as at its autophosphorylation site Y397, in RASFs. Accordingly, dasatinib was demonstrated to act as a potent inducer of apoptosis in all RASF samples analyzed and, in this respect, was considerably more efficient than FasL.

We selected FAK-14 as a rather specific kinase inhibitor that had originally been detected by structure-based molecular modeling using the National Cancer Institute database of small-molecule compounds in order to develop a potent FAK inhibitor ([31]). FAK-14 docks to the autophosphorylation site Y397 of FAK, thereby effectively inhibiting activation of the kinase at low micromolar concentrations ([31]). Our studies demonstrated the capacity of FAK-14 to induce apoptosis in RASFs at a dosage of 1–5 μM, however, with a variable-dose optimum for maximal apoptosis induction in individual donors. Moreover, FAK-14 dosing >5–10 μM easily led to fast cell detachment and necrosis, rather than apoptosis induction.

These data provide clear evidence that a functional FAK/Src signaling pathway critically contributes to cell survival in RASFs. Moreover, our data provide conclusive evidence that ADAM15 plays a pivotal role in Src/FAK signaling in RASFs, since silencing of ADAM15 acts synergistically with pharmacologic inhibition of the Src/FAK signaling pathway to significantly increase the apoptosis rate.

The precise molecular mechanism involved in the interaction of FAK/Src and ADAM15 to enhance resistance to FasL-induced apoptosis in RASFs remains to be elucidated. However, there are some hints in the literature that might at least contribute some enlightening pieces of information. Although ADAM15 is membrane-anchored ([1]) and despite our recent elucidation of the capability of its cytoplasmic tail to transmit an extracellular stimulus to intracellular signals in a transfected chondrocytic cell line ([26]), there is no specific naturally occurring extracellular ligand representing a direct (patho)physiologic trigger of ADAM15 signaling. However, we demonstrated in the model system of primary chondrocytes as well as in a human chondrocytic cell line that the cytoplasmic domain of ADAM15 can act as a scaffold for the recruitment of survival signal–transducing kinases upon challenge with serum starvation or camptothecin-induced genotoxic stress ([26, 29]). This counterregulatory survival pathway was shown to involve the direct binding of the cytoplasmic tail of ADAM15 to the C-terminal region of FAK, which in turn, bound to Src, resulting in an enhanced phosphorylation of the FAK/Src complex ([26]).

FAK is a nonreceptor protein tyrosine kinase that plays a key role in integrin-dependent as well as integrin-independent cell survival. It is well documented that the antiapoptotic FAK signaling can evoke several pathways in tumor cells, rendering it an interesting target for pharmacologic inhibition ([37]). These pathways include triggering of PI3K/Akt and NF-κB activation ([22-25, 38]). In the first step, the activation of PI3K involves binding of its p85 subunit to FAK. It is crucial that this interaction occurs and, hence, also crucial for subsequent Akt signaling that FAK becomes activated at its autophosphorylation site Y397. Intriguingly, the activation of Y397 is considerably enhanced by ADAM15 binding to the C-terminal end of FAK ([39, 40]) and may thereby contribute to reinforcement of the FAK-dependent PI3K/Akt signaling. As the PI3K/Akt pathway is of central importance for survival signals ([41]), the role of the proline-rich cytoplasmic tail of ADAM15 is likely to serve as a scaffolding domain, facilitating the formation of the multicomponent signaling complexes. Further studies are clearly needed to delineate the precise nature of the complex sequence of events and to unravel new targets for future treatment options aiming at the reduction of apoptosis resistance of the aggressive proinflammatory RASFs. In this respect, preliminary results from the present study contribute to the identification of the FAK/Src signaling pathway as a potential target of pharmacologic interference with specific kinase inhibitors.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Burkhardt had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Böhm, Burkhardt.

Acquisition of data. Böhm, Freund, Krause.

Analysis and interpretation of data. Böhm, Kinne, Burkhardt.

Ancillary