Activation of factor VII-activating protease in human inflammation: a sensor for cell death

The study was approved by the institutional medical ethics committees of the centers involved, and from all study participants or legal representatives written informed consent was obtained.

Mouse monoclonal antibodies to FSAP (anti-FSAP-4 and anti-FSAP-9), to complexed C1-inhibitor (KOK-12), to α₂-antiplasmin (AAP-20), and a control antibody (anti-IL6) were prepared at our department (all IgG1κ) [10, 24]. PE-labeled rabbit-anti-mouse F(ab')2 antibody was obtained from Dako (Glostrup, Denmark). Iscove's modified Dulbecco's medium was obtained from BioWhittaker Europe (Verviers, Belgium). Fetal calf serum was obtained from Bodinco BV (Alkmaar, The Netherlands). Penicillin and streptomycin were obtained from Gibco/Invitrogen (Groningen, The Netherlands). Etoposide, ß-mercaptoethanol, RNase, and nitro blue tetrazolium and 5'-bromo-4'-chloro-3'-indolyl phosphate (NBT/BCIP) were obtained from Sigma (Zwijndrecht, The Netherlands). NuPage 4 to 12% polyacrylamide gels, sample buffer, dithiothreitol (DTT) and nitrocellulose membranes were obtained from Invitrogen (Groningen, The Netherlands). Western blocking reagent was obtained from Roche Diagnostics (Mannheim, Germany). Streptavidin-alkaline phosphatase was obtained from Mabtech (Nacka Strand, Sweden). High performance ELISA buffer (HPE) and Poly-HRP-labeled streptavidin were obtained from Sanquin (Amsterdam, The Netherlands). (3,5,3',5')-tetramethylbenzidine (TMB) was obtained from Merck (Darmstadt, Germany). CNBr-activated sepharose and Protein G Sepharose was obtained from Pharmacia Biochem (Uppsala, Sweden).

Jurkat cells were cultured in culture medium (Iscove's Modified Dulbecco's Medium (IMDM) containing 5% (v/v) fetal calf serum (FCS), penicillin (100 IU/ml), streptomycin (100 μg/ml) and 50 μM ß-mercaptoethanol. Before apoptosis and necrosis induction, cells were washed three times with culture medium without FCS by centrifugation at 360 × g for 10 minutes and resuspended in culture medium without FCS. Cells (1 × 10⁶ cells/ml) were incubated for 48 h with etoposide in a final concentration of 200 μM to induce apoptosis. For necrosis induction, cells were incubated with 0.3% H₂O₂ at 60°C for 60 minutes.

In previous experiments recalcified citrated plasma was used instead of serum clotted in the presence of cells since microparticles in serum obscure the FACS analysis [9]. It turned out that FSAP was not activated upon recalcification and that recalcified citrated plasma removed nucleosomes from apoptotic cells as efficiently as serum [9, 10]. In the text, recalcified citrated plasma is denoted as r-plasma. Blood was obtained from healthy donors in vials containing a final concentration of 10 mM sodium citrate and were centrifuged two times at 1,300 × g. Citrated plasma was recalcified with 20 mM CaCl₂ in a glass vial and incubated at 37°C for 30 minutes. Subsequently the plasma was incubated at 4°C for 30 minutes and the formed clot was removed. The r-plasma was stored at -20°C until use. All donors were homozygous for the wild-type form of FSAP [25].

FSAP was activated as recently described [10]. Apoptotic, necrotic or living Jurkat cells were washed in HN-buffer (10 mM Hepes, 140 mM NaCl, pH 7.2) and resuspended in HN-buffer at a concentration of 2 × 10⁶ cells/ml. RNase was added to the cells to a final concentration of 10 U/ml and incubated for 30 minutes at 37°C. R-plasma was incubated for 30 minutes at 37°C with living, necrotic or apoptotic Jurkat cells (0.5 × 10⁶ cells/ml) in HN-buffer.

Apoptotic or living Jurkat cells were washed in HN-buffer (10 mM Hepes, 140 mM NaCl, pH 7.2) and resuspended in HN-buffer at a concentration of 2 × 10⁶ cells/ml. RNase was added to the cells to a final concentration of 10 U/ml and incubated for 30 minutes at 37°C. R-plasma was incubated for 30 minutes at 37°C with living, necrotic or apoptotic Jurkat cells (0.5 × 10⁶ cells/ml) in HN-buffer. After three washes with buffer B-0.5% BSA (10 mM Hepes, 150 mM NaCl, 5 mM KCl, 2 mM CaCl2 and 2 mM MgCl2) cells were incubated with an anti-FSAP antibody or an irrelevant anti-IL6 antibody for 15 minutes. Cells were washed and stained with a PE-labeled rabbit-anti-mouse F(ab')2 antibody. Cells were washed, resuspended in 150 μl buffer B- 0.5% BSA and the samples were analyzed by flow cytometry using BD LSRII flow cytometer (Becton Dickinson, Mountain View, CA, USA). Cells being double positive for propidium iodide and Annexin V were considered to be apoptotic.

R-plasma was incubated with anti-FSAP-4 antibody coupled to CNBr-activated sepharose (10 mg mAb to 500 mg of sepharose) for two hours at room temperature (RT) with gentle shaking [10]. The sepharose was washed three times with PBS, 0.02% Tween 20, 500 mM NaCl. Proteins attached to the beads were eluted by heating to 90°C for 10 minutes with SDS-PAGE sample buffer containing 50 mM dithiothreitol (DTT) and samples were applied to NuPage 4 to 12% polyacrylamide gels. After electrophoresis, Western blotting was performed on the gel according to the Western blotting protocol Novex^®. In brief, samples were blotted onto a nitrocellulose membrane and blocked with 1% Western blocking reagent in TBS-T (10 mM Tris pH 8.0, 150 mM NaCl, 0.1% Tween 20). Subsequently they were incubated with biotinylated anti-FSAP-9 recognizing the light chain of FSAP (1 μg/ml). The membranes were washed and Streptavidin-alkaline phosphatase in TBS-T 0.5% Western blocking reagent (dilution 1:500 v/v) was added. The blotting membranes were developed using NBT/BCIP and the reaction was stopped with distilled water.

Monoclonal anti-FSAP-4 or the irrelevant anti-IL6 was coupled to CNBr-activated sepharose (10 mg mAb to 500 mg of sepharose). Columns were washed five times with start buffer (10 mM Hepes, 1M NaCl, 0.02% Tween 20, pH 7.2). R-plasma incubated with apoptotic Jurkat cells or r-plasma incubated with buffer was applied to the column at a flow rate of 1 ml/minute followed by washing with five column volumes of start buffer. Elution was performed with 0.1M glycine, 140 mM NaCl, 0.02% Tween 20, pH 2.5. The eluate was adjusted to pH 7 with 1M tris-HCl, pH 8. Subsequently eluates were IgG-depleted. The Protein G Sepharose was washed three times with PBS and 1 ml of eluate was added to 0.5 ml of Protein G Sepharose and incubated head-over-head overnight at 4°C. The supernatants were diluted in SDS-PAGE sample buffer and applied to NuPage 4 to 12% polyacrylamide gels. Gels were stained with Coomassie Blue or silver staining by using a silverstaining kit.

MALDI-TOF peptide mass fingerprinting and MALDI-TOF/TOF peptide sequencing analysis was performed on the excised Coomassie Blue stained protein band at Eurosequence B.V. (Meditech Center, Groningen, The Netherlands). Identification was performed by database search using MASCOT software.

For detection of complexes of FSAP with C1-inh and AP, 96-well microtiter plates (maxisorp, Nunc) were coated for four hours at RT with a monoclonal antibody specific for C1-inh in complex [24] or with a monoclonal antibody against AP both at 2 μg/ml diluted in PBS. All further incubations were performed at RT under shaking conditions. After each step, the wells were washed five times with washing fluid (PBS, 0.02% Tween 20) with a Microplate Autowasher (Bio-tek Instruments, Inc. Winooski, VT, USA). All incubation steps were performed in high performance ELISA buffer (HPE). Plasma samples were added to the plate and incubated for 60 minutes at RT. Wells were washed five times and biotinylated anti-FSAP-4 (1 μg/ml) was added and incubated for 60 minutes. After five washes streptavidin-polymerized horseradish peroxidase (poly-HRP) conjugate (dilution 1:10,000 v/v) diluted in HPE was added to each well and incubated for 20 minutes. Plates were developed by an addition of 100 μg/ml 3,3',5,5'-tetramethylbenzidine and 0.003% hydrogen peroxide in 0.11 M sodium acetate buffer pH 5.5. Coloring reaction was stopped by adding 100 μl of 2 M H₂SO₄. The absorbance was measured at 450 nm with an ELISA plate reader (Multiskan, Thermo Labsystems, Breda, The Netherlands). As standard curve we used citrated r-plasma incubated with apoptotic cells in presence of 20 mM EDTA as described in detail above. Citrated plasma of 20 healthy donors was incubated with apoptotic cells and FSAP-serpin complexes were measured. The median of the level of complexes in these controls was arbitrarily set as 0.5 AU/ml since plasma was diluted 1:1 (v/v) with apoptotic cells. The intra- and inter assay coefficient of variation of the FSAP-AP complex ELISA was 9% and 8%, respectively. Both, the intra- and inter-assay coefficient of variation of the FSAP-C1-inh complex ELISA was 5%.

Nucleosome levels were determined by an ELISA as recently described [9]. Briefly, monoclonal antibody CLB-ANA/60, which recognizes histone 3 was used as a catching antibody. Biotinylated F(ab')2 fragments of CLB-ANA/58, which recognizes an epitope exposed on complexes of histone 2A, histone 2B and dsDNA, in combination with poly-HRP were used for detection.

Results are expressed as mean ± SEM/median with range. The Mann-Whitney rank-sum test was used to assess differences between groups at a given time. Correlations between variables were assessed by using Spearman's rank correlation corrected for multiple testing. A P-value < 0.05 was considered to be statistically significant. The cut-off values for FSAP-inhibitor complexes in healthy volunteers were calculated by a method described by Rümke et al.[26].

Since incubation of plasma with apoptotic cells induces nucleosome release via the serine protease activity of FSAP we argued that FSAP must become activated by interaction with those apoptotic cells [10]. We, therefore, analyzed whether FSAP binds to apoptotic cells. Indeed, plasma FSAP binds strongly to apoptotic cells whereas binding to living cells is negligible (Figure 1A). In addition, plasma FSAP also strongly binds to necrotic cells (data not shown). Furthermore, we could show that interaction of r-plasma with dead cells leads to conversion of single chain FSAP into active two-chain FSAP as indicated by the appearance of the light chain at 28 kDa (Figure 1B). No activation could be found upon incubation with living cells.

Since serpins form covalent complexes with their target proteases, activated FSAP will presumably form covalent complexes with its target inhibitors in plasma [27]. Hence, measurement of these complexes would offer an easy and sensitive way to monitor FSAP activation in r-plasma or plasma. To identify which serpins form complexes with activated FSAP, r-plasma was incubated with apoptotic cells and applied to an anti-FSAP affinity column. Analysis of the eluate of the r-plasma activated with apoptotic cells by SDS-PAGE under non-reducing conditions demonstrated an additional band at 97 kDa as compared to the r-plasma that was not incubated with apoptotic cells (data not shown). MALDI-TOF peptide mass fingerprinting and MALDI-TOF/TOF peptide sequencing analysis of the 97 kDa band resulted in a positive identification for FSAP and α₂-antiplasmin (AP) which is described to be an inhibitor of FSAP in purified systems [17]. Although C1-inh was reported to inhibit activated FSAP in plasma we were not able to show C1-inh to form complexes with FSAP on SDS-PAGE [16].

Our results demonstrate that upon activation with apoptotic cells FSAP forms covalent complexes with AP. To quantify these complexes we set up an ELISA to measure complexes between activated FSAP and AP. A monoclonal antibody recognizing AP was used as a catching and biotinylated anti-FSAP as detection antibody. Upon activation of FSAP in r-plasma, FSAP-AP complexes could be measured by ELISA, whereas no FSAP-AP complexes could be detected in plasma incubated with buffer (Figure 2A). No signal was detected when using an irrelevant antibody (anti-IL6) as detecting antibody demonstrating the ELISA to specifically detect FSAP-AP complexes (data not shown).

Since inhibition of FSAP with C1-inh was described by others [16], we also tested whether complexes between activated FSAP and C1-inh could be detected by ELISA. As a catching antibody a monoclonal antibody was used which exclusively reacts with C1-inh when complexed with its target proteases [24] and biotinylated anti-FSAP was used for detection. FSAP-C1-inh complexes could be measured in r-plasma after incubation with apoptotic cells whereas no FSAP-C1-inh complexes could be measured in r-plasma incubated with buffer (Figure 2B). Again no signal could be detected when an irrelevant antibody was used as detecting antibody.

After incubation with apoptotic and necrotic cells complexes with C1-inh and with AP were readily detectable (Figure 3). Only low levels of FSAP-inhibitor complexes could be detected after incubation with living cells. No complexes could be detected after incubation of plasma with buffer. Together these results suggest that measurement of complexes of FSAP with its serpins, AP and C1-inh, by ELISA is a sensitive tool to assess FSAP activation in plasma.

Because FSAP is known to undergo autoactivation we extensively analyzed how plasma levels of FSAP complexes are influenced by sample preparation. Whereas purified plasma-derived FSAP is susceptible for autoactivation, FSAP in plasma turned out to be very resistant to autoactivation. Even blood clotting does not activate FSAP (data not shown). Moreover, plasma of sepsis patients stored for three hours at room temperature contains equal amounts of FSAP-serpin complexes as plasma immediately after collection (data not shown). When plasma samples of healthy donors are incubated for three hours at 37°C also no complexes are found. Only when plasma samples of patients were incubated at 37°C we found an increase in complex levels. Together these results indicate that there is no additional complex formation after collection at room temperature, nor via auto-activation, nor via circulating cell fragments in plasma of septic patients. We then measured FSAP activation in citrated plasma of patients with increasing severity of inflammation.