Antibodies to age‐β₂glycoprotein I in patients with anti‐phospholipid antibody syndrome

Summary

Anti‐phospholipid antibody syndrome (APS) is a systemic autoimmune disease characterized clinically by arterial and/or venous thromboses, recurrent abortions or fetal loss and serologically by the presence of ‘anti‐phospholipid antibodies’ (aPL). The main target antigen of the antibodies is β₂glycoprotein I (β₂GPI). Post‐translational oxidative modifications of the protein have been widely described. In this study we aimed to analyse sera reactivity to glucose‐modified β₂GPI (G‐β₂GPI). Sera collected from 43 patients with APS [15 primary APS (PAPS) and 28 APS associated with systemic lupus erythematosus (SLE) (SAPS)], 30 with SLE, 30 with rheumatoid arthritis (RA) and 40 healthy subjects were analysed by an enzyme‐linked immunosorbent assay (ELISA) using a G‐β₂GPI. Nine of 15 consecutive PAPS out‐patients (60%) and 16 of 28 SAPS (57.1%) showed serum antibodies [immunoglobulin (Ig)G class] against G‐β₂GPI (anti‐G‐β₂GPI) by ELISA. The occurrence of anti‐G‐β₂GPI was significantly higher in APS patients compared to patients suffering from SLE. No RA patients or control healthy subjects resulted positive for anti‐G‐β₂GPI. Of note, aG‐β₂GPI prompted to identify some APS patients (four PAPS and seven SAPS), who were negative in the classical anti‐β₂GPI test. Moreover, in APS patients, anti‐G‐β₂GPI titre was associated significantly with venous thrombosis and seizure in APS patients. This study demonstrates that G‐β₂GPI is a target antigen of humoral immune response in patients with APS, suggesting that β₂GPI glycation products may contain additional epitopes for anti‐β₂GPI reactivity. Searching for these antibodies may be useful for evaluating the risk of clinical manifestations.

Introduction

Anti‐phospholipid antibody syndrome (APS) is a systemic autoimmune disease characterized clinically by arterial and/or venous thromboses, recurrent abortions or fetal loss and serologically by the presence of anti‐phospholipid antibodies (aPL) 1, 2. Diagnosis of APS requires the combination of at least one clinical and one laboratory criterion 3. Anti‐cardiolipin antibodies (aCL) and anti‐β_2‐glycoprotein I antibodies (aβ₂GPI) detected by enzyme linked immunosorbent assay (ELISA), and lupus anti‐coagulant (LA) detected by clotting assays, are the recommended tests for detection of aPL 4, 5. However, aPL represent a heterogeneous family of antibodies reacting with phospholipid‐binding proteins, including not only β₂GPI 6, 7, but also different anionic phospholipids, proteins or phospholipid–protein complexes, such as prothrombin 8, protein S 9, 10, protein C 11, annexin V 12, annexin II 13, vimentin 14, oxidized low‐density lipoproteins, lyso‐bis‐phosphatidic acid (LBPA) and sulphatides, etc. 15, 16, 17. β₂GPI, an abundant plasma glycoprotein involved in clotting mechanisms and lipid pathways 18, is the most common target for aPL associated frequently with vascular cell dysfunction 19, thrombotic events and proatherogenic mechanisms 20, 21, 22. The aPL action mechanism forms the basis of APS pathophysiology. Moreover, only autoimmune‐type aCL are usually dependent upon the presence of β₂GPI and anti‐β₂GPI have been involved in the expression of LA activity, an in‐vitro phenomenon associated with an increased risk of arterial and/or venous thromboembolic events. Understanding the mechanisms of how this abundant self‐plasma protein becomes a target of pathogenic autoantibodies will improve the knowledge of APS pathophysiology.

The molecular structure and location of the major epitopic region(s) of the β₂GPI molecule are controversial, although they were recently characterized fully 23. Several studies have investigated whether the immune response is directed to native β₂GPI 22, 23, 24, 25 or to cryptic or neoepitopes 26, 27. Decisive events generating cryptic or neoepitopes include β₂GPI binding to anionic surfaces, such as phospholipids, and oxidative modifications that alter phospholipid binding 27, 28, 29, 30. Many mechanisms may be responsible for generating neoepitopes, and multiple mechanisms receive support from the heterogeneous antigenic specificities in β₂GPI‐specific antibodies. One candidate mechanism is non‐enzymatic glycosylation (glycation), a process that leads to the formation of early, intermediate and advanced glycation end products (AGEs), able to modify self‐molecule structures and functions. Even though glycation is present physiologically and is modulated by several factors, diet, ageing as well as disorders of glucose metabolism and systemic autoimmune diseases associated with inflammation and oxidative stress may favour the formation and accumulation of these products 31, 32, 33. AGEs represent new epitopes and contain new antigenic structures, thereby possibly contributing to the generation of autoimmune responses. Recently, we showed that several potential glycation sites are present within the β₂GPI primary structure and that, after in‐vitro exposure to glucose, β₂GPI was sugar‐modified and this modification probably consisted of an AGE formation. Our results on the ability of AGE‐β₂GPI to activate human monocyte‐derived immature dendritic cells suggest a possible role for glycation in the increase of β₂GPI immunogenicity 34. Although there is accumulating evidence that AGEs are involved in the progression of inflammatory and immune‐mediated diseases 35, further investigations are needed to clarify the role of glycation in generating new antigenic epitopes in β₂GPI possibly contributing to the heterogeneous specificity of aPL.

This study was designed to investigate whether β₂GPI in vitro treated with glucose (G‐β₂GPI) is a target of humoral response in APS patients and whether anti‐G‐β₂GPI may be associated with clinical features.

Materials and methods

Patients

This study included 73 consecutive out‐patients attending the Lupus Clinic of the Sapienza University of Rome. Forty‐three patients had APS, diagnosed according to the Sydney Classification Criteria 3, primary (PAPS, n = 15) or APS associated with systemic lupus erythematosus (SLE) (SAPS, n = 28); 30 patients had SLE fulfilling the ACR revised criteria for the classification of SLE 36. Sera were collected at several times and stored at −20°C until use. Finally, 30 patients with RA and 40 healthy subjects (normal blood donors) matched for age and sex were also studied as controls. This study was approved by the local ethic committees and participants gave written informed consent in accordance with the Declaration of Helsinki.

Bioinformatic analysis of β₂GPI

The primary structure of β₂GPI (Accession no. P02749) was analysed by means of tools available online at http://sysbio.unl.edu/SVMTriP/index.php, to predict the presence of flexible and functional sites, including antigenic epitopes 37, related closely or structurally to potential glycation sites 34. The most recommended epitope(s) within β₂GPI sequence were calculated using 20 amino acids as the default epitope length search for putative epitopes on the web server, as reported 38.

Preparation of glucose‐modified β₂GPI

Human native β₂GPI was purchased from Calbiochem (La Jolla, CA, USA). Purity of β₂GPI > 98% was checked by mass spectrometry, as reported by the manufacturer.

To prepare glucose‐modified β₂GPI (G‐β₂GPI), human β₂GPI was dissolved in glycation buffer solution (GB) (0·144 g/l KH₂PO₄, 0·426 g/l Na₂HPO₄) pH 7·4, at 10 μg/ml final concentration and frozen immediately at −80°C under sterile conditions. Then, β₂GPI aliquots were incubated in the presence of 250 mM glucose or mannitol (Sigma‐Aldrich, Milan, Italy) in the dark, at 37°C for increasing intervals of time (0, 12 h, 10 and 22 days) (in sealed vials), as described previously 34, 39. As glycation control, a highly purified preparation of human albumin (Sigma‐Aldrich) was treated with D‐glucose under the same conditions used for β₂GPI.

Characterization of glucose‐modified β₂GPI

Size exclusion chromatography was performed by fast protein liquid chromatography (FPLC; Pharmacia, Uppsala, Sweden) interfaced to an ultraviolet (UV) monochromator detector (ProteomeLab PF2D Protein Fractionation System; Beckman Coulter, Brea, CA). Twenty micrograms of native or G‐β₂GPI resuspended into 50 μl of phosphate‐buffered saline (PBS) without calcium and magnesium (PBS^−/−) were injected onto a Superose S12 Pharmacia column equilibrated in PBS^−/−, pH 7·4. Elution was carried out with a 0·4 ml/min flow rate at 22°C, PBS^−/− as elution buffer, and protein peaks were detected under UV recording (optical density at 214 nm). Purity of β₂GPI and its molecular weight under denatured conditions was checked by electrophoretic analysis 34.

UV spectrophotometry

The UV absorption characteristics of G‐β₂GPI and native‐β₂GPI were recorded on a UV/visible spectrophotometer (dual beam Uvikon 860 Instrument; Kontron, Zurich, CH) between 200 and 400 nm, using 1 cm optical length UV quartz cuvettes, as described previously 40, with some modifications: G‐β₂GPI was analysed onto the ‘sample’ quartz cuvette, while native‐β₂GPI was analysed onto the ‘reference’ quartz cuvette. In this manner, the differential absorption spectrum indicates the spectral regions of β₂GPI affected by glucose‐induced modifications.

Non‐tryptophan fluorescence studies

Fluorescence studies were carried out as reported previously 34, 39. Steady‐state fluorescence emission spectra of β₂GPI preparations were collected with a FluoroMax‐2 spectrofluorometer (Jobin Yvon‐Spex, Edison, NJ, USA) using an excitation wavelength of 370 nm, equal bandwidths for excitation and emission (5/5). Emission fluorescence at 450 nm was measured by subtracting contributions of sugar solutions from the fluorescence of the sugar‐β₂GPI mixture.

ELISA for aCL and anti‐β₂GPI antibodies

aCL and anti‐β₂GPI ELISA kits were obtained from Inova Diagnostics Inc. (San Diego, CA, USA). ELISA was performed for all the patients’ and healthy subjects’ sera according to the manufacturer's instructions. aβ₂GPI were also tested by ELISA, as reported below.

LA assay

LA for all the patients’ and healthy subjects’ sera was studied in two coagulation systems, a dilute sensitized activated partial thromboplastin time (aPTT) and a dilute Russell's viper venom time (dRVVT), followed by confirmatory test, using reagents and instrumentation by Hemoliance Instrumentation Laboratory (Lexington, MA, USA).

ELISA for glucose‐modified β₂GPI

Ninety‐six‐well polystyrene plates were coated and incubated overnight at 4°C with 1 µg/well of native‐β₂GPI or G‐β₂GPI in 0·05 μM NaHCO₃ buffer, pH 9·5. Coated plates were incubated overnight at 4°C and then washed three times with PBS containing 0·05% Tween 20 (PBS‐T). Plates were blocked for 2 h at room temperature with 100 µl of 1% bovine serum albumin (BSA) in PBS. After washing three times with PBS‐T, the wells were incubated for 1 h at room temperature with 100 µl of patient sera diluted 1 : 100 in the blocking buffer. Each serum was analysed in triplicate. Goat polyclonal anti‐β₂GPI (Affinity Biologicals Inc., Ancaster, ON, Canada) was used as positive control. After three washes with PBS‐T, the plates were incubated for 1 h at room temperature with horseradish peroxidase (HRP)‐conjugated antibodies, anti‐human immunoglobulin (Ig)G (Sigma‐Aldrich) or anti‐goat IgG (Dako Italia S.p.A., Cernusco sul Naviglio, Italy) diluted in the blocking buffer. The plates were washed three times with PBS‐T; the bound peroxidase was then revealed with 100 µl of O‐phenylenediamine dihydrochloride buffer and colour development was stopped with H₂SO₄ 0·2 M for 5 min. Absorbance was measured at 492 nm in a microplate reader. Data were presented as the mean optical density (OD) corrected for background (wells without coated antigen). Forty sera from healthy subjects were also tested and a cut‐off value was established at the mean of OD ± 3 standard deviations (s.d.) of normal human sera. Parallel experiments were performed with an identical procedure but without coated G‐β₂GPI.

Absorption tests

Sera of patients with APS, positive for both aβ₂GPI and anti‐G‐β₂GPI, and of healthy subjects, were diluted 1 : 100 in the blocking buffer and incubated with native β₂GPI (100 μg/ml) for 1 h at 37°C and then overnight at 4°C. The mixture was centrifuged for 15 min at 27 000 g at 4°C, according to Alessandri et al. 16; the supernatant fraction was kept as absorbed serum and tested for anti‐G‐β₂GPI and anti‐native‐β₂GPI antibodies by ELISA, as reported above.

Statistical analysis

All the statistical procedures were performed by GraphPad Prism software Inc. (San Diego, CA, USA). Normally distributed variables were summarized using the mean ± s.d. and non‐normally distributed variables by the median and range. Differences between numerical variables were tested with the Wilcoxon test. Correlation was tested with Spearman's rank order or Pearson's correlation coefficient. For comparison of categorical variables or percentages, Fisher's exact and χ² tests were used when appropriate. P‐values less than 0·05 were considered significant.

Results

Clinical and serological characteristics of APS and SLE patients

All patients enrolled into this study were Caucasian. APS patients were 39 females and four males with a median age of 45 years (range = 17–75), and a median disease duration of 7 years (range = 0·5–24). SLE patients were 28 females and two males, with a median age of 42 years (range = 25–64) and a median disease duration of 6 years (range = 1–22). The clinical characteristics of APS and SLE patients are reported in Table 1. All patients were screened for aCL, anti‐β₂GPI (tested by both ELISA methods with overlapping results) and LA; the prevalence of the antibodies is reported in Table 2.

Table 1.

Clinical characteristics of patients.

Characteristics n (%)	PAPS n = 15	SAPS n = 28	SLE n = 30
Arterial thrombosis	9 (60)	11 (39·3)	0
Venous thrombosis	4 (26·7)	16 (57·1)	0
Pregnancy morbidity	4 (26·7)	5 (17·9)	3 (10)
Livedo reticularis	6 (40)	12 (42·9)	5 (16·6)
Thrombocytopenia	4 (26·7)	8 (28·6)	1 (3·3)
Migraine	3 (20)	2 (7·1)	2 (6·6)
Seizures	2 (13·3)	2 (7·1)	1 (3·3)

PAPS = primary anti‐phospholipid antibody syndrome; SLE = systemic lupus erythematosus; SAPS = APS associated with SLE. 

Table 2.

Prevalence of antibodies of patients.

Autoantibodies n (%)	PAPS n = 15	SAPS n = 28	SLE n = 30
aCL	12 (80)	22 (78·6)	3 (10)
aβ2GPI	5 (33·3)	9 (32·1)	1 (3·3)
LA	8 (53·3)	16 (57·1)	1 (3·3)
aG‐β2GPI	9 (60)	16 (57·1)	8 (26·6)

PAPS = primary anti‐phospholipid antibody syndrome; SLE = systemic lupus erythematosus; SAPS = APS associated with SLE; aCL = anti‐cardiolipin antibodies; LA = lupus anti‐coagulant.

Bioinformatic analysis of β2GPI

Analysis of functional and immunological sites within β₂GPI sequence showed that the most recommended epitope corresponds to the residues 51–70. Of note, this region contains the residue with the highest potentiality to be glycated (#63: score = 0.967, as calculated by Netglycate‐1·0 prediction), as published previously 34 (Supporting information, Table S1).

Characterization of glucose‐modified β₂GPI

Purified β₂GPI preparation was incubated for increasing times in the presence of D‐ glucose and characterized as described in the Methods section.

In order to verify whether G‐β₂GPI might undergo a significant molecular size modification, a size exclusion chromatographic separation was carried out. UV recording at 214 nm of eluate showed the formation of a molecular size increase due probably to the formation of glucose‐β₂GPI adducts (G‐β₂GPI) (see arrow in Fig. 1a). Analysis of the chromatographic profile showed a molecular size increase in the 22‐day glycated β₂GPI corresponding to approximately 18% of total protein.

Figure 1.

(a) Chromatographic elution of native and glucose‐modified β₂glycoprotein I (G‐β₂GPI). Twenty micrograms of highly purified β₂GPI incubated previously with sugar were injected onto Superose S12 fast protein liquid chromatography (FPLC) column as described in the Methods section and protein peaks were detected under ultraviolet (UV) recording (optical density at 214 nm). The dashed line indicates that G‐β₂GPI elutes as a complex protein population involving higher molecular weight complexes (highlighted by the arrow). A semiquantitative evaluation of such polymeric G‐β₂GPI, corresponding to the left‐side shoulder of the dashed line, indicates that under these non‐denaturating chromatographic conditions, such complexes may represent the 15–20% of total G‐β₂GPI proteins; (b) differential absorption spectrum between native β₂GPI and G‐β₂GPI. Using a dual‐beam UV spectrophotometer, absorbance of G‐β₂GPI and native‐β₂GPI were analysed. The absorption spectrum shows the different spectral regions of β₂GPI affected by glucose‐induced modification. The analysis indicates that three main regions are affected: the change of absorbance observed at region corresponding to 310–340 nm suggests the formation of non‐enzymatic glycosylation adducts 34, whereas the changes of absorbance observed at 260–280 and 215–240 suggest structural modifications occurring on the β₂GPI protein folding.

UV absorption spectrum indicated that three main regions of β₂GPI were affected by glucose treatment. The change of absorbance observed at region corresponding to 310–340 nm suggests the formation of non‐enzymatic glycosylation adducts, as reported by published studies 40, whereas the changes of absorbance observed at 260–280 and 215–240 nm suggest structural modifications occurring on β₂GPI protein folding (Fig. 1b).

Fluorescence at 450 nm, specific for AGE formation, was measured and confirmed the creation of time‐dependent AGE products (Supporting information, Table S2).

Detection of antibodies to glucose modified‐β₂GPI by ELISA

Nine of 15 consecutive PAPS out‐patients (60%) and 16 of 28 SAPS (57·1%) showed serum antibodies (IgG class) against G‐β₂GPI (anti‐G‐β₂GPI). No significant difference was found between PAPS and SAPS patients (Table 2). Of note, four sera from patients with PAPS and seven with SAPS were positive for anti‐G‐β₂GPI but negative for anti‐native‐β₂GPI (Fig. 2, arrows).

Figure 2.

Relationship between anti‐native β₂glycoprotein I (β₂GPI) and anti‐glucose‐modified GPI (G‐β₂GPI) antibodies in primary anti‐phospholipid antibody syndrome (PAPS) and APS associated with systemic lupus erythematosus (SLE) (SAPS) patients. Sera from patients with PAPS and SAPS were analysed by enzyme‐linked immunosorbent assay (ELISA) for the detection of anti‐native‐β₂GPI and anti‐G‐β₂GPI immunoglobulin (Ig)G. The arrows indicate the sera from patients (PAPS or SAPS) positive for anti‐G‐β₂GPI but negative for anti‐native‐β₂GPI. The horizontal line shows the cut‐off level calculated as mean optical density (OD) value in control healthy subjects’ sera ± 3 standard deviations (s.d.).

The occurrence of anti‐G‐β₂GPI was significantly higher in APS patients (25 of 43, 58·1%) compared to patients suffering from SLE (26·6%) (P = 0·037). No RA patients or control healthy subjects resulted positive for aG‐β₂GPI (Fig. 3).

Figure 3.

Anti‐glucose‐modified β₂glycoprotein I (G‐β₂GPI) antibodies in patients and healthy subjects. Sera from patients with anti‐phospholipid antibody syndrome (PAPS) and APS associated with systemic lupus erythematosus (SLE) (SAPS), SLE, rheumatoid arthritis (RA) and of healthy subjects were analysed by enzyme‐linked immunosorbent assay (ELISA) for the detection of anti‐G‐β₂GPI immunoglobulin (Ig)G. All the sera from RA patients and healthy subjects were negative for anti‐G‐β₂GPI IgG antibodies. The horizontal line shows the cut‐off level calculated as mean optical density (OD) value in control healthy subjects’ sera ± 3 standard deviations (s.d.).

Sera of patients with APS, positive for both anti‐β₂GPI and anti‐G‐β₂GPI, were absorbed with native β₂GPI and then tested for anti‐G‐β₂GPI antibodies, or alternatively for anti‐native‐β₂GPI, by ELISA. Reactivity with G‐β₂GPI showed significant inhibition (63% inhibition, P < 0·001) by first absorbing sera with native β₂GPI (Fig. 4a). As expected, reactivity with native‐β₂GPI showed almost complete inhibition (83·8% inhibition, P < 0·001) (Fig. 4b). These findings suggest that anti‐G‐β₂GPI recognize specific glycation‐related epitopes.

Figure 4.

Absorption tests for detection of anti‐glucose‐modified β₂glycoprotein I (G‐β₂GPI) specificity. Sera of patients with anti‐phospholipid antibody syndrome (APS), positive for both anti‐native β₂GPI and anti‐G‐β₂GPI, and of healthy subjects (controls), unabsorbed or absorbed with native β₂GPI, were analysed by enzyme‐linked immunosorbent assay (ELISA) for the detection of: (a) anti‐G‐β₂GPI IgG, mean absorbance of unabsorbed sera [optical density (OD) 2·264] was set to 100%. Absorbed patients’ sera versus unabsorbed patients’ sera: P < 0·001; absorbed controls’ sera versus unabsorbed controls’ sera: P > 0·05. (b) anti‐β₂GPI IgG, mean absorbance of unabsorbed sera (OD 1·116) was set to 100%. Absorbed patients’ sera versus unabsorbed patients’ sera: P < 0·001; absorbed controls’ sera versus unabsorbed controls’ sera: P > 0·05.

Associations of antibodies to glucose‐modified β₂GPI with classical aPL and clinical features

Analysis of the correlations between anti‐G‐β₂GPI OD and classical aPL in APS patients revealed a significant association with aCL (P = 0·013) and anti‐native β₂GPI (P = 0·009). Furthermore, a significant correlation was found between anti‐G‐β₂GPI with venous thrombosis (P = 0·017) and seizure (P = 0·027) in these patients.

Discussion

This study provides new findings showing that G‐β₂GPI is a target antigen of humoral immune response in patients with APS and suggests a possible usefulness of anti‐G‐β₂GPI to improve the diagnosis of this disease.

Several studies have suggested that post‐translational oxidative modifications of β₂GPI may affect antigenic properties of the molecule. Among these, glycosylation processes may play a relevant role 41. In this study we analysed the presence of antibodies to glucose‐modified β₂GPI by glycation (non‐enzymatic glycosylation). Glycation is the non‐enzymatic addition or insertion of saccharide derivatives to proteins, lipids or nucleic acids, leading to the formation of intermediary Schiff bases and Amadori products and, finally, to irreversible AGEs 42. Glycation reaction, similarly to other decisive events, including β₂GPI binding to anionic surfaces, such as phospholipids, and oxidative modifications 29, 30, 43, 44, may induce a significant misfolding effect on the β₂GPI structure, contributing to the expression of cryptic or neoepitopes recognized by the immune system. In a previous investigation we demonstrated, by bioinformatic analyses of the β₂GPI primary structure, that several potential glycation sites are present within the molecule 34. The intriguing co‐localization of high glycation sites with potential epitopes found in the present study suggests that sugar‐induced modifications occurring on β₂GPI may be able to affect its antigenic behaviour.

In our previous investigation we also showed that a 10‐day β₂GPI treatment with glucose induces a protein modification probably consisting of AGE‐β₂GPI formation, and that this glucose‐modified protein is able to activate human monocyte‐derived immature dendritic cells 34. In the present study, we investigated the effects of a longer incubation (22 days) with glucose on β₂GPI structure. G‐β₂GPI was studied under non‐denaturing conditions, in order to evaluate the formation of glycation end products and/or adducts with sugar characterized by non‐covalent bonds. The chromatographic analysis of G‐β₂GPI indicated the presence of a protein population with shorter elution time, i.e. with larger molecular size. Peak integration of the chromatogram suggested that this molecular size shift could involve a significant portion of β₂GPI protein.

Therefore, we verified that glucose treatment may induce a significant misfolding effect on the β₂GPI structure, probably leading to expression of cryptic or neoepitopes recognized by the immune system. Upon modification with glucose, increase in UV absorbance was recorded. This finding provided a useful insight into the structural perturbation of β₂GPI protein probably inducing the formation of glycation adducts. The fluorescence spectrophotometric analysis of G‐β₂GPI confirmed the creation of time‐dependent advanced glycation end products (AGEs).

In this study we observed that anti‐G‐β₂GPI were present in a significant percentage of patients with APS. Interestingly, ELISA for anti‐G‐β₂GPI prompted to identify some APS patients (four PAPS and seven SAPS), who were negative in the ‘classical’ anti‐β₂GPI test.

The mechanisms by which G‐β₂GPI accumulates and may trigger a humoral immune response in APS patients are not completely known. A possible explanation is that G‐β₂GPI accumulation occurs in patients with APS driven by oxidative stress and inflammation, thereby initiating a local autoimmune process and becoming the target of autoimmune responses. The role of oxidative post‐translational modification of β₂GPI has been described widely 45, 46. In particular, it can produce new antigens that are not represented in the thymus, so that autoreactive T cells can escape negative selection and move to the periphery. Alternatively, the way by which antigen‐presenting cells process oxidized β₂GPI may be different from that processing the reduced form. In both cases an autoimmune response can be triggered. Thus, as epitope dominance is influenced by protein structure, glycation events may change the molecular context of β₂GPI epitopes (by altering secondary or tertiary structure), thus permitting the efficient presentation of cryptic and neodeterminants 45. According to our previous results 34, 44, we can hypothesize that a pro‐oxidant and proinflammatory microenvironment predisposes local β₂GPI to glycation and/or oxidation, thereby initiating a local autoimmune process.

Another interesting finding of this study is the observation that anti‐G‐β₂GPI were correlated significantly with several clinical manifestations of APS, including venous thrombosis and seizure. To date, the most commonly investigated antigenic target in APS patients is β₂GPI, and anti‐β₂GPI antibodies represent a highly specific test for diagnosis of the syndrome. In particular, IgG anti‐β₂GPI antibodies comprise a family of antibodies which recognize different epitopes of the protein. In recent years, several studies showed that antibodies to domain I (DI), that specifically recognize the glycine40‐arginine43 epitope, showed a good correlation with thrombosis and pregnancy morbidity 47, 48. Moreover, anti‐β₂GPI antibodies with DI specificity were found in the majority of APS patients and were associated significantly with LA and venous thrombosis 49. Our findings extend these data, strongly suggesting a relationship between the presence of anti‐G‐β₂GPI and thrombosis. This observation is not surprising. Indeed, since 1985, Vlassara et al. showed that AGE‐modified proteins can be bound to a special receptor receptor for advanced glycation end‐product (RAGE) 50, and the same binding has been shown for AGE‐β₂GPI 34. Engagement of RAGE results in intracellular signalling, which leads to activation of the proinflammatory transcription factor nuclear factor kappa B (NF‐κB), with consequent production of cytokines, adhesion molecules, prothrombotic and vasoconstrictive gene products 34, 51. For example, the endothelial surface is changed in a way that coagulation events are favoured; indeed, AGE‐bound RAGE on the endothelium may result in alteration of the cell surface structure, inducing a procoagulant endothelium, via reduced thrombomodulin activity 52 concomitant with increased tissue factor expression 53.

Taken together, our data demonstrate the existence of anti‐G‐β₂GPI in APS patients, suggesting that β₂GPI glycation products may represent additional epitopes for anti‐β₂GPI reactivity, possibly contributing to the heterogeneous specificity of aPL 54. Searching for these antibodies may be useful for evaluating the risk of clinical manifestations in APS patients.

Disclosure

The authors declare that they have no disclosures.

Author contributions

M.S. designed the study and drafted the manuscript, B.B. designed the study and carried out the experiments, A.C. carried out the experiments, E.P. designed the study and carried out the experiments, F.F. carried out the experiments (preparation of G‐β₂GPI and chromatographic analyses), C.A. and S.T. collected the sera samples, characterized the patients clinically and performed statistical analysis, R.M. drafted the manuscript and revised the manuscript critically, F.C. and G.V. revised the manuscript critically, R.R. designed the study and drafted the manuscript.

Supporting information

Additional Supporting information may be found in the online version of this article at the publisher's web‐site:

Table S1. Bioinformatic analysis of functional and immunological sites within β2GPI sequence.

Table S2. Time‐dependent non‐tryptophan fluorescence of sugar‐treated β2GPI.