Secreted Phospholipases A2 in Hereditary Angioedema With C1-Inhibitor Deficiency

Background Hereditary angioedema (HAE) caused by deficiency (type I) or dysfunction (type II) of the C1 inhibitor protein (C1-INH-HAE) is a disabling, potentially fatal condition characterized by recurrent episodes of swelling. We have recently found that patients with C1-INH-HAE have increased plasma levels of vascular endothelial growth factors and angiopoietins (Angs), which have been associated with vascular permeability in several diseases. Among these and other factors, blood endothelial cells and vascular permeability can be modulated by extracellular or secreted phospholipases A2 (sPLA2s). Objective We sought to investigate the enzymatic activity and biological functions of sPLA2 in patients with C1-INH-HAE. Methods sPLA2s enzymatic activity was evaluated in the plasma from 109 adult patients with C1-INH-HAE and 68 healthy donors in symptom-free period and attacks. Plasma level of group IIA sPLA2 (hGIIA) protein was measured in selected samples. The effect of C1-INH-HAE plasma on endothelial permeability was examined in vitro using a vascular permeability assay. The role of hGIIA was determined using highly specific sPLA2 indole inhibitors. The effect of recombinant hGIIA on C1-INH activity was examined in vitro by functional assay. Results Plasma sPLA2 activity and hGIIA levels are increased in symptom-free C1-INH-HAE patients compared with controls. sPLA2 activity negatively correlates with C1-INH protein level and function. C1-INH-HAE plasma increases endothelial permeability in vitro, and this effect is partially reverted by a specific hGIIA enzymatic inhibitor. Finally, recombinant hGIIA inhibits C1-INH activity in vitro. Conclusion sPLA2 enzymatic activity (likely attributable to hGIIA), which is increased in C1-INH-HAE patients, can promote vascular permeability and impairs C1-INH activity. Our results may pave the way for investigating the functions of sPLA2s (in particular, hGIIA) in the pathophysiology of C1-INH-HAE and may inform the development of new therapeutic targets.

inTrODUcTiOn Hereditary angioedema due to C1-inhibitor deficiency (C1-INH-HAE) is a disabling, potentially fatal condition characterized by recurrent episodes of swelling caused by reduced levels (type I) or dysfunction (type II) of the C1-INH protein (1,2). These patients display insufficient C1-INH function to prevent bradykinin (BK) formation (3), which increases endothelial permeability and leads to recurrent episodes of swelling (i.e., angioedema attacks) involving the deeper layers of the skin and/or submucosal tissue (4)(5)(6). High concentrations of circulating BK and cleaved high-molecular weight kininogen (HK) are elevated in patients with C1-INH-HAE and further increased during angioedema attacks, with the latter correlating with attack frequency (7,8). In addition, specific BK antagonism reverts angioedema symptoms (9). Even if, BK stands as main mediator of C1-INH-HAE (10,11), the need to explain symptom variability among patients genetically deficient in C1-INH prompt on identifying additional factors that modulate endothelial cell biology and vascular permeability (12). Vascular endothelial growth factors (VEGFs) and angiopoietins (Angs) have well-established role in endothelial cells conditioning and modulation of permeability (13)(14)(15)(16) and we recently showed their increase in plasma of patients with C1-INH-HAE in symptom-free period (17). Interestingly, variants in angiopoietins 1 gene (ANGPT1) are associated with a recently described form of hereditary angioedema (HAE) (18). Nevertheless, other factors controlling endothelial cell biology and vascular permeability may intervene in C1-INH-HAE.
Owing to the ability of sPLA2s to modulate vascular permeability (either by directly activating endothelial cells or by catalyzing the production/degradation of vasoactive molecules) (36), we have analyzed the enzymatic activity and biological function of sPLA2s present in plasma from C1-INH-HAE patients in symptom-free period and during attacks.

study Population
We studied 109 C1-INH-HAE patients followed at the University of Milan, University of Naples Federico II, and University of Budapest and 68 normal healthy controls. Diagnosis of C1-INH-HAE was based on the presence of at least one clinical and labo ratory criteria as described (40). Table 1 summarizes the  clinical characteristics of patients with C1-INH-HAE and healthy  controls. Patients (104 type I and 5 type II C1-INH-HAE) belong to 76 unrelated families and their median age at symptoms onset was 6 years (interquartile range [3][4][5][6][7][8][9][10][11][12][13][14]. There were no differences between the study groups in terms of age, sex, and ethnicity. The frequency of angioedema attacks was used as an index of disease severity; therefore, patients were grouped according to the number of attacks registered during the last 12 months: 69 of them had less than 12 attacks (low frequency), while 40 complained ≥12 attacks (high frequency). Regarding symptom-free samples, blood sampling was performed at least 8 days apart from an angioedema attack in all patients. In 22 patients, blood samples were obtained also during angioedema attack. Nineteen patients were taking prophylactic therapy at the time of blood collection (4 were on tranexamic acid and 15 on attenuated androgens).

Blood sampling
The Ethical Committee of the University of Naples Federico II, University of Milan, and University of Budapest approved that plasma obtained during routine diagnostics could be used for research investigating the physiopathology of HAE and written informed consent was obtained from patients according to the principles expressed in the Declaration of Helsinki. Blood was collected during routine diagnostic procedures and the remaining plasma sample was labeled with a code which was documented into a datasheet. The controls had been referred for routine medical check-up and volunteered for the study by giving informed consent. Technicians who performed the assays were blinded to the patients' history. Blood was drawn by a clean venepuncture and minimal stasis using two types of anticoagulants: sodium citrate 3.2% and, for the measurement of cleaved HK, an inhibitor cocktail containing 100 mM benzamidine, 400 µg/ml hexadimethrine bromide, 2 mg/ml soybean trypsin inhibitor, 263 µM leupeptin, and 20 mM aminoethylbenzenesulphonyl fluoride dissolved in acid/citrate/dextrose (100 mM trisodium citrate, 67 mM citric acid, and 2% dextrose, pH 4.5). After centrifugation (2,000 g for 20 min at 22°), the plasma was divided into aliquots and stored at −80°C until used.

complement system analysis
Tube with an anti-coagulant sodium citrate 3.2% is used for separating plasma from whole blood. Plasma C1-INH was measured by radial immunodiffusion (NOR-Partigen, Siemens Healthcare Diagnostics, Munich, Germany). C4 antigen levels in Italia was measured by radial immunodiffusion (NOR-Partigen) (the method is not specific for C4 fragments) whereas in Hungary C4 levels was measured by turbidimetry (Roche Cobas Integra 800, Beckman Coulter Complement C4). The antibody employed in the Beckman Coulter C4 assay is directed against the common portion of the C4 molecule and it exhibit the same reactivity with C4 fragments as well as with the native molecule.
C1-INH function was assessed as the capacity of plasma to inhibit the esterase activity of exogenous C1s as measured on a specific chromogenic substrate by means of a commercially available kit (Technoclone GmbH, Vienna, Austria) (17). Reference ranges were 0.70-1.30 U C1-INH/ml (1 U C1-INH corresponds to the average C1-INH activity present in 1 ml of fresh citrated normal plasma). The functional activity of C1-INH was also expressed as a percentage of activity of C1-INH present in samples. Normal values of activity of C1-INH are greater than 0.7 U C1 INH/ml (>70%). According to diagnostic criteria, all patients enrolled in this study had C1-INH functional activity lower than 50% of normal (41). In selected experiments, plasma of healthy controls was incubated (2 h, 37°C) with and without hGIIA (0.5-5 µg/ml). After treatment, enzymatic activity of C1-inhibitor was assessed using commercially available MicroVue C1-Inhibitor EIA kit (Quidel, San Diego, CA, USA).

contact system analysis
The cleavage of HK was assessed by means of sodium dodecyl sulfatepolyacrylamide gel electrophoresis (SDS-PAGE) and immunoblotting analysis (a modification of the method descri bed by Berrettini et al.) (42). Samples were loaded on a 9% SDS-PAGE. After electrophoretic separation, proteins were transferred from the gel to a polyvinylidene difluoride membrane using Bio-Rad Trans-Blot ® Turbo™ Transfer System (Bio-Rad Laboratories, Hercules, CA, USA). HK was identified using goat polyclonal anti-HK light chain (Nordic, Tilburg, The Netherlands) and visualized using a biotinylated rabbit anti-goat antibody (Sigma Aldrich Co., St. Louis, MO, USA). The density of the bands obtained was measured using a Bio-Rad GS-800 densitometer. The amount of cleaved HK was expressed as a percentage of total HK.

elisa for hgiia
Human sPLA2 group IIA levels in plasma samples were determined by ELISA kit (Catalog No. MBS9303777, MyBioSource, San Diego, CA, USA). The concentration of hGIIA in plasma was tested in duplicate and determined against a standard curve for each ELISA assay. To evaluate the reliability of the assay we loaded 500 and 1,000 pg/ml of recombinant hGIIA in the ELISA wells and the spectrophotometer measured 397 and 886 pg/ml, respectively.

In Vitro Vascular Permeability assay
Endothelial cell permeability was assessed by in vitro vascular permeability assay kit (Life Technologies, Carlsbad, CA, USA). BEAC were seeded onto collagen-coated Transwell filters (1 µm pore size) at the density of 7.5 × 10 4 cells/well in a 96-well receiver plate and incubated at 37°C and 5% CO2 for 72 h. After this time, cells starvation step was performed by adding DMEM 0.5% FBS and incubation for 18 h at 37°C, 5% CO2. Plasma from patients with C1-INH-HAE was pre-incubated (20

statistical analysis
Data were analyzed with the GraphPad Prism 5 software package. Data were tested for normality using the D' Agostino-Pearson normality test. If normality was not rejected at 0.05 significance level, we used parametric tests. Otherwise, for not-normally distributed data we used nonparametric tests. Statistical analysis was performed by unpaired two-tailed t-test or two-tailed Mann-Whitney test as indicated in figure legends. Correlations between two variables were assessed by Spearman's correlation analysis and reported as coefficient of correlation (r). Plasma activity of sPLA2 and hGIIA is shown as the median (horizontal black line), the 25th and 75th percentiles (boxes) and the 5th and 95th percentiles (whiskers) of 68 controls and 109 patients. In selected experiments, the data are expressed as mean values ± SD of the indicated number of experiments. Statistical analysis was performed with Prism 5 (GraphPad Software) by one-way analysis of variance followed by Dunnett's test (when comparison was made against a control). Statistically significant differences were accepted when the p value was ≤0.05.

resUlTs
increased Plasma levels of sPla 2 enzymatic activity and hgiia Protein in Patients with c1-inh-hae compared with healthy controls We assessed the PLA2 enzymatic activity (likely attributable to sPLA2) in the plasma from 109 C1-INH-HAE patients in the symptom-free period vs 68 healthy controls matched for age and gender ( Table 1). Figure 1A shows that sPLA2 activity was increased by appreciatively twofold in C1-INH-HAE patients in symptom-free period compared with controls [sPLA2: 2.4 (1.3-3.0) vs 1.3 (0.6-1.8) U/ml median values (interquartile ranges)]. No gender or age differences in sPLA2 activity were found in both controls and patients and (see Table S1 and Figure  S1 in Supplementary Material).

The sPla 2 activity in Plasma inversely correlates with c1-inh Functional activity and Protein levels
In C1-INH-HAE patients, functional C1-INH levels are, by definition, below 50% of the normal value (40) and C4 concentrations are usually reduced and can be used as a screening test (48). We investigated whether differences in the complement component levels (C1-INH and C4) were associated with differences in sPLA2 activity. C1-INH activity negatively correlated with sPLA2 activity (r = −0.29; p < 0.01) (Figure 2A) Figure 2B). When we stratified the patients according to the concentration of C4 (less than 25 or 25-50% of normal values) no difference in sPLA2 plasma activity between these groups was found ( Figure 2C).

lack of association between sPla 2 activity in Plasma and angioedema attacks
To investigate a possible role for sPLA2s in angioedema attacks, we measured its activity in patients with less (low frequency) or more than 12 attacks (high frequency) in the last 12 months. sPLA2 activity was comparable in these two groups (Figure 4A). In addition, we compared sPLA2 activity in 22 C1-INH-HAE patients in symptom-free period and during angioedema attacks. We found that the sPLA2 activity [attack 1.2 (0.4-2) vs symptom-free period 3.0 (1.4-4.4) U/ml] was reduced in patients examined during attack compared with basal conditions ( Figure 4B).

hgiia activity in c1-inh-hae Plasma increases endothelial Permeability
Secreted phospholipases A2 can modulate endothelial cell mobility and vascular permeability (36). To gain mechanistic insight into the role of sPLA2s in C1-INH-HAE, we performed

hgiia impairs c1-inh Functional activity
Our results demonstrate that hGIIA activity in C1-INH-HAE plasma increases endothelial leakage in vitro, supporting a mecha nistic role for sPLA2s in modulating vascular permeability in vivo (i.e., C1-INH-HAE patients). This effect can be mediated by a direct modulation of endothelial cells through enzymatic activity and/or receptor-mediated mechanisms (e.g., binding to HSPGs). Nevertheless, we hypothesized that hGIIA could also directly affect C1-INH activity. To this aim, we incubated in vitro plasma from healthy controls (containing normal C1-INH) with recombinant hGIIA and assessed C1-INH functional activity. Interestingly, hGIIA concentration-dependently reduced C1-INH activity (Figure 6).

DiscUssiOn
In this study, we found that plasma sPLA2 enzymatic activity and hGIIA are increased in symptom-free C1-INH-HAE patients compared with healthy controls. sPLA2 activity is positively correlated with hGIIA plasma concentrations and inversely correlated with C1-INH activity and protein level. hGIIA in C1-INH-HAE plasma increases endothelial permeability and hGIIA impairs C1-INH functional activity in vitro. No correlation was found between plasma sPLA2 activity and the severity of angioedema and intriguingly sPLA2 activity was decreased during attacks. The local generation of BK causes angioedema in C1-INH deficiency (10). It is well established that this mediator comes from HK and is cleaved by plasma kallikrein lacking its main physiological inhibitor (11). It is still unclear whether BK generation is sufficient to cause an angioedema attack or if other mechanisms are involved. We have demonstrated that plasma levels of vascular permeability factors (i.e., VEGFs and Angs) are increased in symptom-free C1-INH-HAE patients (17). In this study, we demonstrate that sPLA2 activity is increased in plasma of C1-INH-HAE patients and negatively correlates with C1-INH plasma activity. Although multiple sPLA2s have been described in mammals, several lines of evidence support a specific role for hGIIA in C1-INH-HAE. First, hGIIA is the most represented sPLA2 in human serum and plasma (43)(44)(45)(46)(47). Second, hGIIA plasma levels are increased in C1-INH-HAE patients compared with healthy controls. Moreover, hGIIA plasma levels strongly correlates with C1-INH activity. Finally, sPLA2 activity in the plasma from both C1-INH-HAE patients and healthy controls is markedly reduced by the hGIIA-specific enzymatic inhibitor RO032107A (39).
Our results suggest a possible involvement of hGIIA in the pathogenesis of C1-INH-HAE. We show that plasma from C1-INH-HAE patients increases endothelial permeability in vitro compared with healthy donor plasma. Their effect is partially reverted by the addition of the hGIIA-specific enzymatic inhibitor RO032107A or endothelial cell treatment with heparinase, an enzyme that degrades HSPGs. It is conceivable that hGIIA in C1-INH-HAE plasma binds to HSPGs on endothelial surface and increases endothelial cell permeability in a process that requires its enzymatic activity. These results are in keeping with the evidence that sPLA2s can modulate endothelial cell permeability directly or through the release of vasoactive mediators (e.g., PGE2, VEGFs, and Angs) (17,18,31). Further studies are required to gain more insights into this model and define its relevance to C1-INH-HAE pathogenesis.
We also uncovered another possible mechanism of hGIIA involvement in C1-INH-HAE pathophysiology that is independent of its interaction with endothelial cells. We found that hGIIA impairs C1-INH activity of healthy donor plasma in a concentration-dependent manner. Modulation of C1-INH by other classes of enzymes has already been demonstrated. For example, the proteases elastase and plasmin degrade C1-INH (53,54). C1-INH also interacts with MBL-associated serine protease 1 (MASP-1), and C1-INH/MASP-1 complexes are reduced in C1-INH-HAE patients (55). Interestingly, both wild-type and catalytically inactive hGIIA bind to factor Xa of the coagulation cascade and inhibit prothrombinase activity (56). Whether hGIIA interacts with C1-INH in vitro and in vivo, and whether hGIIA enzymatic activity is required for impairing C1-INH activity require further investigation. Our plane is to study a potential interaction between C1-INH and hGIIA, thermodynamically and kinetically, employing singlemolecule in vitro assays (e.g., surface plasmon resonance and isothermal titration calorimetry). Furthermore, at a later time, it could be possible to detect this interaction on plasma samples.
It has been suggested that a systemic activation process can occur in patients with HAE (4). Accordingly, we have hypothe sized that C1-INH deficiency and inflammatory stimuli contribute to generate a variable, ongoing increase in vascular permeability that defines the threshold where localized triggers act for the development of angioedema attack (17). This hypothesis is also supported by previous findings indicating that several proinflammatory mediators such as C reactive protein (57,58) and pentraxin 3 (57) are elevated in asymptomatic C1-INH-HAE patients and during attacks. Moreover, VEGFs (17) and sPLA2 produced by activated immune cells (29,31,(59)(60)(61) are elevated in C1-INH-HAE patients. Together, these findings suggest that low-grade systemic inflammation can occur in these patients.
Comprehensive studies have identified BK as the principal mediator of vascular leakage in C1-INH-HAE-related swelling attacks (10, 62) Circulating levels of BK, markers of endothelial activation, prothrombin fragments, D-dimer (63), cytokines (e.g., TNF-α and IL-8), as well as neutrophil count and neutrophil-derived factors (e.g., elastase, myeloperoxidase, pentraxin 3) (64) are increased during attacks compared with symptom-free periods in C1-INH-HAE patients. By contrast, we found that sPLA2 activity is consistently decreased during attacks and does not correlate with canonical biomarkers of angioedema severity (i.e., cleaved HK) (65). Whatever the mechanism(s), sPLA2 is the first mediators so far identified which shows opposite beha vior during clinical remission (i.e., increase) and angioedema attacks (i.e., reduction). The reasons for this intriguing observation are unclear and command additional in vitro and in vivo investigations. Furthermore, our preliminary data show that sPLA2 activity is not modified by prophylactic treatments (either tranexamic acid or attenuated androgens) compared with untreated patients (unpublished results).
In conclusion, our study provides evidence for a possible role of hGIIA in the pathophysiology of C1-INH-HAE and also gives mechanistic insights into how hGIIA may predispose to the development of angioedema attacks.

eThics sTaTeMenT
The Ethical Committee of the University of Naples Federico II approved that plasma obtained during routine diagnostics could be used for research investigating the physiopathology of hereditary angioedema and written informed consent was obtained from patients in according to the principles expressed in the Declaration of Helsinki. Protocol number 216/16.