Interplay between C1-inhibitor and group IIA secreted phospholipase A2 impairs their respective function

High levels of human group IIA secreted phospholipase A2 (hGIIA) have been associated with various inflammatory disease conditions. We have recently shown that hGIIA activity and concentration are increased in the plasma of patients with hereditary angioedema due to C1-inhibitor deficiency (C1-INH-HAE) and negatively correlate with C1-INH plasma activity. In this study, we analyzed whether the presence of both hGIIA and C1-INH impairs their respective function on immune cells. hGIIA, but not recombinant and plasma-derived C1-INH, stimulates the production of IL-6, CXCL8, and TNF-α from peripheral blood mononuclear cells (PBMCs). PBMC activation mediated by hGIIA is blocked by RO032107A, a specific hGIIA inhibitor. Interestingly, C1-INH inhibits the hGIIA-induced production of IL-6, TNF-α, and CXCL8, while it does not affect hGIIA enzymatic activity. On the other hand, hGIIA reduces the capacity of C1-INH at inhibiting C1-esterase activity. Spectroscopic and molecular docking studies suggest a possible interaction between hGIIA and C1-INH but further experiments are needed to confirm this hypothesis. Together, these results provide evidence for a new interplay between hGIIA and C1-INH, which may be important in the pathophysiology of hereditary angioedema.

On the basis of the combined functions of C1-INH, the congenital deficiency of C1-INH (incidence of 1:50,000) results in a kallikrein-kinin (contact) system-related disorder but with no clear signs of hemostatic problem. This disease is called C1-INH-HAE (OMIM #106,100) and is characterized by unpredictable recurrent spontaneous histamine-independent episodes involving the deeper layers of the skin and/or submucosal tissue that can take place at various tissue sites throughout the body [25].
Besides the above finding associating hGIIA and C1-INH, hGIIA sPLA 2 and its mammalian orthologs have been shown to bind and modulate activities of different molecules involved in the complement or coagulation systems. For instance, the major sPLA 2 (likely rat GIIA) purified from rat peritoneal inflammatory sites was inhibited by two large fragments of C3 complement factor [26]. hGIIA was also reported to exert anticoagulant effects by inhibiting prothrombinase activity via binding to FXa [27]. The effect was independent of phospholipid hydrolysis and due to direct interaction with FXa, as measured under both in vitro and ex vivo conditions [27]. On the other hand, regulation of C1-INH activity by other types of enzymes has already been shown. C1-INH can be degraded by serine proteases such as elastase and plasmin [28,29] or can interact with them, for example, with MBL-associated serine protease 1 (MASP-1) and MASP-2, forming protein complexes [30].
Based on the above findings, we sought to determine whether hGIIA can interact and interfere with the function of C1-INH, and vice versa, either directly or indirectly, thereby leading to impaired activation of immune cells by either of the two types of molecules.
Recombinant human group IIA sPLA 2 (hGIIA) was produced in E. coli as the N1A catalytically active mutant (the N1A mutation facilitates the removal of the initiator methionine without impacting enzymatic activity) as reported [31]. RO032107A, a specific hGIIA inhibitor, was a kind gift from Pr. Michael Gelb (University of Washington, Seattle, USA) [32].

Isolation and purification of peripheral blood mononuclear cells (PBMCs)
The study protocol involving the use of human blood cells was approved by the Ethics Committee of the University of Naples Federico II, and written informed consent was obtained from blood donors according to the principles expressed in the Declaration of Helsinki (Protocol Number 301/12). PBMCs were isolated from buffy coats of healthy donors (HBsAg − , HCV − , and HIV − ) obtained from a leukapheresis unit. Plasma was separated from cellular components by centrifugation (400 × g for 20 min at 22 °C), collected, and stored at − 80 °C. Leukocytes were separated from erythrocytes by dextran sedimentation. PBMCs were purified by Histopaque-1077 (Sigma-Aldrich, Milan, Italy) density gradient centrifugation (400 × g for 20 min at 22 °C). The cells were resuspended (10 6 cells/250 µL) in RPMI 1640 with 2 mM l-glutamine and 1% antibiotic-antimycotic solution, and incubated (37 °C, 5% CO 2 ) in 48-well plates. After 2 h, the cell medium was removed and the plates were gently washed with fresh medium. The adherent cells were resuspended in RPMI 1640 with 5% FCS, 2 mM l-glutamine, and 1% antibiotic-antimycotic solution (complete medium) and used for experiments.

ELISA assays
Concentrations of cytokines and chemokines in cell supernatants were measured using commercially available ELISA kits for IL-6 (range of detection 9.4-600 pg/mL), TNF-α (15.6-1000 pg/mL), and CXCL8 (31.3-2000 pg/ mL) (R&D Systems, MN, USA). The results obtained were normalized for the total protein content in each well, determined in cell lysates (cells lysed with 0.1% Triton X-100) by a Bradford assay, with the standard curve performed with bovine serum albumin. Cytokine release was expressed as pg or ng of cytokine/mg of total proteins.

Effect of C1-INH on hGIIA enzymatic activity
hGIIA enzymatic activity was measured as previously described, using [ 3 H]-oleate-radiolabeled E. coli membranes as a sensitive substrate for sPLA 2 s [34]. To test the inhibitory effect of C1-INH on hGIIA sPLA 2 , the recombinant enzyme (hGIIA N1A, 10 pM) was preincubated with various concentrations of recombinant or plasma-derived C1-INH in 100 µL of sPLA 2 activity buffer (100 mM Tris pH 8.0, 10 mM CaCl 2 , and 0.1% bovine serum albumin (BSA)) for 15 min at room temperature. The enzymatic activity was measured by addition of 30,000 dpm of [ 3 H]-oleate-radiolabeled E. coli membranes in 100 µL of sPLA 2 activity buffer and further incubation at 37 °C for 60 min. Reactions (200 µL) were stopped by addition of 300 µL of stop buffer (100 mM EDTA pH 8.0 and 0.1% fatty acid-free BSA). Mixtures were centrifuged at 10,000 × g for 5 min, and supernatants containing released free [ 3 H]-oleate were counted (dpm/assay). The percentage of inhibition by C1-INH is calculated relative to the enzymatic activity measured in the presence of hGIIA but absence of C1-INH, after subtraction of the background value measured in the absence of sPLA 2 . Addition of C1-INH alone has no effect on the background value measured in the absence of hGIIA.

C1-INH functional assays
C1-INH function was determined as the capacity of plasma C1-INH from healthy donors to inhibit the esterase activity of exogenous C1s with a chromogenic substrate (commercially available kit from Technoclone GmbH, Vienna, Austria). Reference ranges were as follows: 0.70 to 1.30 unit of C1-INH/mL (1 C1-INH unit corresponds to the average C1-INH activity present in 1 mL of fresh citrated normal plasma). The functional activity of plasma C1-INH was also expressed as a percentage of activity of C1-INH present in samples. In selected experiments, plasma of healthy donors was incubated (2 h, 37 °C) with and without hGIIA (3 μg/mL) or LPS (100 ng/mL). After treatment, the enzymatic activity of C1-INH was determined as above. In other experiments, hGIIA was preincubated for 2 h at 37 °C with rhC1-INH and pdC1-INH (or their absence) followed by determination of C1-INH activity. In a last group of experiments, hGIIA was preincubated with RO032107A and then incubated for 2 h at 37 °C with rhC1-INH and pdC1-INH (or their absence) followed by determination of C1-INH activity.

Surface plasmon resonance (SPR)
Real-time binding assays were performed on a Biacore 3000 Surface Plasmon Resonance (SPR) instrument (GE Healthcare, Milan, Italy). The N1A mutant of hGIIA was immobilized at 800 RU on a CM5 Biacore sensor chip, at ~ 20 µg/mL in 10 mM sodium acetate, pH 5.0, by using the EDC/NHS chemistry, with a flow rate of 2 μL/min and an injection time of 7 min. BSA was immobilized similarly as a reference channel. Binding assays were carried out by injecting 90 µL of analyte, at 30 µL/min, with various concentrations of C1-INH 10, 20, 30, 40, and 50 µM in HBS (10 mM Hepes, pH 7.4, 150 mM NaCl, 3 mM EDTA). The association rate (k on ) was monitored for 180 s, and the dissociation rate (k off ) was monitored for 300 s. The BIAevaluation analysis package (version 4.1, GE Healthcare) was used to subtract the signal from the reference channel and kinetic values were estimated by applying a 1:1 Langmuir model, as reported [35].

Fluorescence spectroscopy
A Jasco FP 8300 spectrofluorometer equipped with a 10-mm path-length quartz cuvette was employed. Data were acquired at 25 °C, using an excitation wavelength of 298 nm and a fluorescence emission wavelength ranging from 300 to 400 nm, at a 2 μM concentration of C1-INH, in 10 mM phosphate buffer, pH 7.4, and incubated in the presence of increasing concentrations of hGIIA (0-20 µM) ranging from 0 to 1.0 equivalents. Experiments were carried out in duplicates.

Molecular dynamics
The C1-INH-hGIIA model with the lowest HADDOCK score was placed in a cubic box with a water layer of 1.0 nm, neutralized with Na + and/or Cl − ions, and minimized. The steepest descent minimization stopped either when the maximum force was lower than 1000 kJ/mol/nm or when 50,000 minimization steps were performed with 0.005 kJ/ mol energy step size, Verlet cut-off scheme, short-range electrostatic cut-off, and van der Waals cut-off of 1.0 nm. AMBER99SB-ILDN force field [45], tip3p water, and periodic boundary conditions were employed. NVT and NPT equilibrations were performed for 100 ps by restraining the protein backbone, followed by 500-ns-long NPT production runs at 330 K. The iteration time step was set to 2 fs with the Verlet integrator and LINCS [46] constraint. All the simulations and their analyses were run as implemented in the Gromacs package 2020.3 [47]. Root mean squared deviations (RMSDs) were calculated from configurations sampled every 0.5 ns. Simulations were run on M100 (CINECA, Italy).

Statistical analysis
The data are expressed as mean values ± SEM (standard error mean) of the indicated number of experiments. Statistical analysis was performed with Prism 6 (GraphPad Software). Statistical analysis was performed by Student's t-test or one-way analysis of variance followed by Dunnett's test (when comparison was made against a control) or Bonferroni's test (when comparison was made between each pair of groups). Statistically significant differences were accepted when the p-value was at least ≤ 0.05.

Effects of hGIIA on the release of cytokines and chemokines from PBMCs
Upon activation, PBMCs release cytokines and chemokines such as IL-6, TNF-α, and CXCL8 [48]. In a first series of experiments, we evaluated the effects of hGIIA on the secretion of cytokines and chemokines from PBMCs. hGIIA induced the release of IL-6 ( Fig. 1a), TNF-α (Fig. 1b), and CXCL8 (Fig. 1c). For comparison, LPS, the most abundant component within the cell wall of Gram-negative bacteria and immune cell activator, was used as positive control and potently released the same cytokines [48,49]. The effect of hGIIA on the release of cytokines and chemokines was sensitive to the active site sPLA 2 inhibitor RO032107A [50] (Fig. 1d-f). Moreover, although we used highly purified recombinant hGIIA in these experiments, we excluded the possibility that the effect of hGIIA was due to small amount of LPS contamination by stimulating PBMCs with hGIIA in the presence of polymyxin B (50 µg/mL), a potent binder of LPS [51]. Polymyxin B did not influence the capacity of hGIIA to induce the release of IL-6 ( Fig. 1g), TNF-α (Fig. 1h), and CXCL8 (Fig. 1i), whereas it almost completely suppressed the production of cytokines and chemokines induced by LPS (Fig. 1g-i).

Effects of recombinant human C1-INH and plasma-derived C1-INH on PBMCs activated by hGIIA
We tested the effects of physiological concentrations of C1-INH on the cytokine/chemokine release from PBMCs. Recombinant human C1-INH (rhC1-INH) and plasmaderived C1-INH (pdC1-INH) had no effect by themselves on the release of IL-6, TNF-α, and CXCL8 (Fig. 2). The presence or absence of 5% FCS in complete medium did not change the effect of C1-INH on cytokine production (data not shown).
However, the same physiological concentrations of rhC1-INH and pdC1-INH dose-dependently inhibited the effect of hGIIA on the release of IL-6, TNF-α, and CXCL8 (Fig. 3a-c). Conversely, rhC1-INH and pdC1-INH had no inhibitory effect on LPS at inducing the secretion of cytokines/chemokines in PBMCs (Fig. 3d-f). The percentage of viable PBMCs at 16 h after treatment with the different stimuli did not differ from that of untreated cells (data not shown).

hGIIA partially impairs the activity of C1-INH to inhibit C1-esterase
Preincubation of plasma from healthy donors (containing C1-INH with normal activity) with hGIIA partially induced an inhibition of C1-INH activity whereas LPS had no effect (Fig. 4a). In another series of experiments, we preincubated rhC1-INH and pdC1-INH with or without hGIIA and measured C1-INH activity. Figure 4b shows that hGIIA alone had no effect on C1-esterase activity but partially reduced the ability of C1-INH molecules to inhibit C1-esterase activity. Moreover, the hGIIA effect was not affected by preincubation of hGIIA with RO032107A (Fig. 4c).

Effects of C1-INH on enzymatic activity of sPLA 2
To identify a possible direct interaction between hGIIA and C1-INH, we tested whether C1-INH modulates the enzymatic activity of hGIIA. When using the highly sensitive radiolabeled E. coli membranes sPLA 2 assay that requires very low concentrations of hGIIA to measure enzymatic activity, rhC1-INH and pdC1-INH had no significant effect on hGIIA enzymatic activity, even at high concentrations of complement inhibitors (Fig. 5).

C1-INH-hGIIA interaction
To further test the possibility of a direct interaction between C1-INH and hGIIA, we first used SPR as in vitro binding assay where hGIIA was immobilized on the sensorchip and pdC1-INH employed as the analyte. The overlay of sensorgrams, reported in Fig. 6a, exhibited a dose-response increase of signal. The kinetic parameters (k on = 1.74 × 10 3 1/ms and k off = 1.25 × 10 −2 1/s) allowed to estimate a K D value of 6.70 µM.
Based on the intrinsic emission of C1-INH and assuming that some aromatic residues could be involved in the formation of the complex C1-INH/hGIIA, we analyzed fluorescence emission spectra of C1-INH at increasing concentrations of hGIIA and the overlay of emission spectra is reported in Fig. 6b. Upon excitation at 298 nm, the emission intensity of C1-INH showed a dose-response quenching and a shift of λ max following the addition of hGIIA. This behavior suggested the possible involvement of aromatic solventexposed residues in the recognition site between the two proteins, even if specific further studies will be needed to ascertain this finding.
Docking results revealed that 73% of the generated structures (corresponding to the 10 lowest scoring clusters; Fig. 7A) present hGIIA bound to the larger C1-INH binding site comprising Phe-369. Indeed, Phe-369 on C1-INH participates to the binding, as well as Tyr-11 and Phe-23 on hGIIA. Furthermore, in the lowest scoring conformation, the two proteins are kept together by 13 hydrogen bonds (Fig. 7B).
The C1-INH:hGIIA complex stability was then investigated by means of atomistic molecular dynamics simulations in full water solvent. In the simulations, the temperature was kept above room temperature (330 K) to favor molecular rearrangements. The protein complex remained associated along the simulation time and the proteins maintained their conformation as evidenced by their backbone root mean squared deviation (inset in Fig. 7C). However, their reciprocal orientation changed along the simulated time (Fig. 7C). hGIIA moved on C1-INH surface leading to a final observed configuration in which two aromatics are involved on C1-INH: Phe-369 and Trp-460 (Fig. 7D). These form an aromatic interaction with Phe-63 of hGIIA. The involvement of additional hydrogen bonds, such as that between Gln-463 on C1-INH and Tyr-66 on hGIIA, further strengthens the interaction between the two

Discussion
We previously demonstrated the presence of high levels of circulating hGIIA in patients with C1-INH-HAE and a negative correlation between plasma activities of hGIIA and C1-INH [20]. In the present study, we asked  [20] that would contribute to inhibition of C1-INH activity [52]. Moreover, we dare the hypothesis that a rapid increase of circulating plasma hGIIA could induce a transient drop of C1-INH activity and contribute to development of angioedema.
In this manuscript, we also show that both rhC1-INH and pdC1-INH inhibit the effect of hGIIA on the production of cytokines from PBMCs at concentrations found in healthy donors. It will be interesting to test whether plasma from C1-INH-HAE patients with various levels of hGIIA versus C1-INH and complement produce various levels of cytokine production by PBMCs, thereby showing a complex interplay between these three factors. Such a scenario would be in line with protein complexes consisting of several proteins and playing important role in regulatory processes, cellular activation, and signaling cascades [52]. It is also in line with the fact that hGIIA can act on cells through either enzymatic activity [10] or its ability to interact with different targets including heparan sulfate proteoglycans (HSPGs) and integrins [5,9,14,16,17,19,[53][54][55][56]. This study only analyzed the effect of hGIIA on PBMCs while hGIIA and other sPLA 2 s can activate several blood and resident immune cells such as neutrophils, macrophages, eosinophils, and platelets [5,6,[57][58][59]. Thus, in the future, it will be interesting to study the effect of C1-INH on the biological roles of hGIIA in different pathophysiological conditions, and to expand our findings to test whether C1-INH impacts on the effect of various sPLA 2 s in the activation of other immune cells. In particular, it is well demonstrated that high levels of circulating sPLA 2 are found in several pathological situations and positively correlate with disease severity [9,16,17,19]. Interestingly, PLA 2 serum activity is increased in B-cell lymphoma and has been proposed as a new biomarker for B-cell lymphoproliferation [60]. Moreover, we know that acquired angioedema due to C1-inhibitor deficiency (AAE-C1-INH) is often associated with malignant B-cell lymphoma and other disorders [61]. Therefore, it is conceivable that the increase of circulating PLA2 in lymphoma could be the cause of decrease of C1-INH and so development of AAE-C1-INH. Further studies are needed to demonstrate this hypothesis. In conclusion, since C1-INH inhibits the proinflammatory effect of sPLA 2 , the data collected in this paper suggest that patients with angioedema with C1-INH deficiency could have a greater and uncontrolled inflammatory response to endogenous (human) and exogenous (bee and snake venom, etc.) sPLA 2 s compared to healthy subjects. Further study will be necessary to demonstrate this hypothesis. Author contribution Substantial contributions to the conception or design of the work; or the acquisition, analysis, or interpretation of data for the work: SL, ALF, MB, AP, DM, SF, GL, GM, RC, FP.
Drafting the work or revising it critically for important intellectual content: SL, ALF, MB, AP, DM, SF, GL.
Final approval of the version to be published: SL, ALF, MB, AP, DM, SF, GL, RC, FP, CP, GM.
Agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved: SL, ALF, MB, AP, DM, SF, GL, FP, RC, CP, GM.
Funding Open access funding provided by Università degli Studi di Napoli Federico II within the CRUI-CARE Agreement. This work was partially supported by an investigator-initiated research grant (grant IIT-ITA-002138) from Shire International GmbH, a Takeda Company, and was also supported by CINECA Awards N. HP10B3JT25, 2020, FISM 2018R4 for the availability of high-performance computing resources and support. GL acknowledges support from the Fondation Jean Valade/Fondation de France (Award FJV_FDF-00112090), the National Research Agency (AirMN (ANR-20-CE14-0024-01)), and "Investments for the Future" Laboratory of Excellence SIGNALIFE, a network for innovation on signal transduction pathways in life sciences (ANR-11-LABX-0028-01 and ANR-15-IDEX-01), and the Fondation de la Recherche Médicale (DEQ20180339193L).
Data availability Data and material supporting the reported results are available upon request.

Declarations
Institutional review board statement The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Ethics Committee of University of Naples Federico II (Protocol Number 301/12).

Consent to participate
Informed consent was obtained from all donors involved in the study and written informed consent for publication was obtained from participants.

Conflict of interest
The authors declare that they have no conflict of interest.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.