CUTTING EDGE IN AUTOIMMUNITY

Immunoreactivity and avidity of IgG anti-b2-glycoprotein I antibodies from patientswith autoimmune diseases to different peptideclusters of b2-glycoprotein I

A. Artenjak • I. Locatelli • H. Brelih • D. M. Simonic • Z. Ulcova-Gallova •

J. Swadzba • J. Musial • T. Iwaniec • L. Stojanovich • F. Conti • G. Valesini •

T. Avcin • J. W. Cohen Tervaert • Y. Shoenfeld • M. Blank • A. Ambrozic •

S. Sodin-Semrl • B. Bozic • S. Cucnik

� Springer Science+Business Media New York 2014

Abstract The pathogenicity of antibodies against b2-glycoprotein I (anti-b2GPI) depends on multiple factors such as

subclass type, epitope binding and avidity. Due to their large heterogeneity, their impact on antiphospholipid syndrome

(APS) onset is still not fully clarified. We studied the binding characteristics of IgG anti-b2GPI with known avidity from

sera of 201 autoimmune patients (87 with APS, 67 with APS associated with systemic lupus erythematosus (SLE), 47 with

only SLE) to six b2GPI peptides corresponding to amino acid clusters on domains I–II, II, III and III–IV by indirect ELISA

and evaluated their association with clinical features of APS. Peptides A (LKTPRV; domain I–II), B (KDKATF; domain

IV) and C (TLRVYK; domain III) were derived from a hexapeptide phage display library previously shown to react with

pathogenic monoclonal anti-b2GPI. Peptides D (NGPANSK; domain III), E (YNPLWFV; domain II) and F (KMDGNHP;

domain III–IV) represent surface amino acid clusters on b2GPI. The percentage of patients positive for peptides were

observed as follows: 30.3 % for peptide D, 28.90 % for B, 25.9 % for C, 24.9 % for E, 24.4 % for F and 10.0 % for A. The

anti-peptide antibodies in studied serum samples were predominantly of heterogeneous avidity, followed by law avidity

anti-peptide antibodies, whereas only a few were of high avidity. Positive and negative correlations were found between

several anti-peptide antibodies and the rate of thrombosis. Our results indicated diverse reactivity of IgG anti-b2GPI to

different epitopes on b2GPI. Classification of IgG anti-b2GPI into subgroups regarding epitope specificity and avidity

could represent an additional tool in understanding their pathogenicity in APS.

Keywords Antiphospholipid syndrome � Anti-b2GPI antibodies � b2GPI peptides � Anti-b2GPI-related peptides

antibodies � Avidity

A. Artenjak

Lek Pharmaceuticals d.d., Sandoz Biopharmaceuticals Menges,

Kolodvorska 27, 1234 Menges, Slovenia

I. Locatelli � B. Bozic

Faculty of Pharmacy, University of Ljubljana, 1000 Ljubljana,

Slovenia

H. Brelih � D. M. Simonic � A. Ambrozic � S. Sodin-Semrl �B. Bozic � S. Cucnik (&)

Laboratory for Immunology, Department of Rheumatology,

University Medical Centre Ljubljana, Vodnikova 62,

1000 Ljubljana, Slovenia

e-mail: sasa.cucnik@guest.arnes.si

Z. Ulcova-Gallova

Department of Gynaecology and Obstetrics, Genetics and

Medical School of Charles University, 32600 Plzen-Cernice,

Czech Republic

J. Swadzba � J. Musial � T. Iwaniec

Department of Medicine, Jagiellonian University Medical

College, 31-008 Krakow, Poland

L. Stojanovich

Bezhanijska Kosa, University Medical Centre, Belgrade

University, 11080 Belgrade, Serbia

F. Conti � G. Valesini

Dipartimento Medicina Interna e Specialita Mediche,

Rheumatology, Lupus Clinic, La Sapienza, University of Rome,

00161 Rome, Serbia

T. Avcin

Department of Allergology, Rheumatology and Clinical

Immunology, University Medical Centre, University Children’s

Hospital Ljubljana, 1000 Ljubljana, Slovenia

S. Cucnik

123

Immunol Res

DOI 10.1007/s12026-014-8578-0

Introduction

Antiphospholipid syndrome (APS) is an autoimmune dis-

ease defined by vascular (thrombotic) and/or obstetric

complications, in the presence of antiphospholipid anti-

bodies (aPL), i.e., anticardiolipin antibodies, antibodies

against b2-glycoprotein I (anti-b2GPI) and/or lupus anti-

coagulants [1]. The formulation of the original classifica-

tion criteria for APS was performed in Sapporo, Japan

(1998) [2], with a revision in Sydney, Australia, in 2004

[1]. In order for a patient to be diagnosed with APS, one

clinical and one laboratory criterion must be present [1, 3].

APS may be associated with other diseases, mostly sys-

temic lupus erythematosus (SLE) [4]. SLE has been clas-

sified according to the criteria established by the American

Rheumatism Association [5] and revised by the American

College of Rheumatology [6].

aPL are not only a laboratory marker for APS, but can

also be actively involved in disease pathogenesis [7–9].

Furthermore, aPL are considered as a non-traditional risk

factor for atherosclerosis-based cardiovascular diseases,

such as ischemic stroke, coronary artery disease and

peripheral artery disease in patients without overt autoim-

munity [10, 11].

Among the most clinically significant autoantibodies in

APS is anti-b2GPI, directed against b2GPI [8], previously

described as apolipoprotein H. b2GPI is a multifunctional

glycosylated protein present in plasma in concentrations

*180 mg/l (range 20–300 mg/l) [8, 12, 13]. The *50 kDa

protein consists of 326 amino acids, which are folded into

five sushi domains. Domains I–IV are composed of

approximately 60 amino acids, while the 5th domain con-

sists of 82 amino acids. Domain V also has specific char-

acteristics, such as a positively charged, lysine-rich region,

a hydrophobic loop and C-terminal extension, allowing

b2GPI to bind to negatively charged phospholipids [14–

17]. Due to specific interactions between 1st and 5th

domain of b2GPI, the protein exists in two conformations,

i.e., circular plasma and open fishhook conformations [18].

After binding to phospholipids, b2GPI changes the con-

formation from circular to open (‘‘J’’), exposing a cryptic

epitope which is recognized by anti-b2GPI directed against

domain I [19].

Although routine laboratories detect and interpret anti-

b2GPI as one set of autoantibodies, they actually represent

a very heterogeneous group directed against the same

antigen. Due to their polyclonal nature, subpopulations of

anti-b2GPI vary in epitope recognition [8, 20–23], avidity

[24–27] and mechanisms of action [7, 8, 28], which results

in their different pathogenicity and clinical relevance [8,

22, 23, 25, 29, 30]. A subset of IgG anti-b2GPI directed

against domain I on b2GPI seems to have a prominent role

in the pathology of thrombotic complications in patients

with APS [19]. In contrast, a subpopulation of IgA anti-

b2GPI directed against domain IV has been linked to

atherosclerosis in a study using different b2GPI domain-

deleted mutants and native b2GPI [30]. An insight into the

diversity of these autoantibodies was provided by a study

with monoclonal anti-b2GPI. Peptides corresponding to

amino acid sequences on domains I–II, III and IV of b2GPI

were identified using hexamer phage display libraries [28].

These peptides inhibited in vitro/in vivo development of

experimental APS triggered by monoclonal anti-b2GPI,

indicating a possible therapeutic use of synthetic peptides

[28]. Furthermore, a clinical study on 295 patients with

APS and the above-mentioned peptides stressed their

clinical relevance in recognition of different domains of

b2GPI. Antibody positivity for the peptide located on

domain III significantly predicted recurrent spontaneous

abortions, while positivity for the peptide found between

domains I and II was negatively associated with overall

thrombosis [31].

Avidity must also be considered when describing the

diversity of anti-b2GPI subpopulations and their clinical

relevance. Several research groups/clinical studies indi-

cated that difference in strength of binding to the same

antigen/same epitope could be an important marker of the

pathogenicity of the specific anti-b2GPI subgroups [25, 29,

32]. An association between high avidity (HAv) IgG anti-

b2GPI and thrombotic or obstetric complications in

patients with APS was reported. This is in contrast to low

avidity (LAv) IgG anti-b2GPI, which were more prevalent

in autoimmune patients without APS [29, 33]. The

involvement of HAv IgG anti-b2GPI in the pathogenesis of

APS was also confirmed in an in vitro study on human

coronary artery endothelial cells [34]. In this study, HAv

IgG anti-b2GPI were shown to increase the expression of

inflammatory and chemotactic cytokines leading to higher

migration of monocytes. These effects were further inten-

sified in the presence of the major acute-phase protein

serum amyloid A [34], denoting that along with

J. W. Cohen Tervaert

Clinical and experimental Immunology, Maastricht University,

Universiteitssingel 40, 6229 ER Maastricht, The Netherlands

Y. Shoenfeld � M. Blank

Sheba Medical Center, The Zabludowicz Center for

Autoimmune Diseases, 52621 Tel-Hashomer, Israel

M. Blank

Department of Human Microbiology, Sackler Faculty of

Medicine, Tel Aviv University, 69978 Ramat Aviv, Israel

S. Sodin-Semrl

Faculty of Mathematics, Natural Sciences and Information

Technologies, University of Primorska, 6000 Koper, Slovenia

Cutting Edge in Autoimmunity

123

lipopolysaccharide [7], serum amyloid A as a marker of

inflammation also represents a trigger or a second hit

uncovering the full pathological potential of anti-b2GPI

[34]. Using heptamer phage display libraries, two binding

epitopes of HAv IgG anti-b2GPI have been characterized:

FNPYWYV and QGOAHSK [27]. These sequences mim-

icked specific amino acid clusters on domains II and III of

b2GPI and are accessible to HAv IgG anti-b2GPI in the

fishhook and circular conformations. In contrast, the

sequence KMDGNHP has been characterized as a surface-

binding epitope for LAv IgG anti-b2GPI and is located

between domains III and IV of b2GPI. It is accessible for

binding of LAv IgG anti-b2GPI only in open fishhook

conformation [27].

The major aim of this study was to identify and define

the binding characteristics of IgG anti-b2GPI from sera of

autoimmune patients with defined diagnoses (APS in the

absence and the presence of SLE and SLE alone) to sep-

arate b2GPI peptides corresponding to amino acid clusters

on domains I–II, II, III and III–IV. The binding and avidity

of IgG anti-b2GPI to the selected six peptide sequences of

b2GPI were addressed and evaluated for their clinical

value and association with clinical/serological features of

APS.

Materials and methods

Selection of patients’ sera

The study which was approved by the National Ethical

Committee (#163/02/09) and is a part of the National

Research Programme #P3-0314 was primarily conducted

on 479 serum samples from patients with different auto-

immune diseases (predominantly APS, APS associated

with SLE, SLE, etc.). The patients enrolled in this study

were collected from seven European centers within the

framework of the European Forum on aPL [29]. Due to the

lack of inter-laboratory standardization for detection of

aPL [35], all samples were retested for positivity of IgG

anti-b2GPI by our in-house indirect enzyme-linked

immunosorbent assay (ELISA) [36, 37]. Positivity for IgG

anti-b2GPI was confirmed in 226 samples, and clinical

relevance of IgG anti-b2GPI avidity was previously tested

and reported by Cucnik et al. [29]. In the current cross-

sectional study, we used 201 of 226 serum samples for

determination of immunoreactivity and avidity of IgG anti-

b2GPI to six b2GPI-related peptides that correspond to

sequence or surface amino acids on different domains of

b2GPI (see section ‘‘Selection of peptides’’). From the

starting 226 samples which were positive for IgG anti-

b2GPI and had determined their avidity, nine were omitted

due to insufficient volume of the sera for all experiments/

studies and 16 due to diagnostic overlaps with other

autoimmune diseases (Sjogren‘s syndrome, systemic scle-

rosis, urticarial vasculitis and autoimmune thrombocyto-

penic purpura). From 201 serum samples appropriate for

the b2GPI-related IgG anti-peptide antibodies (anti-pep-

tide) immunoreactivity and avidity study, 87 patients were

diagnosed with APS, 67 with APS associated with SLE and

47 with only SLE (Fig. 1). The following clinical param-

eters were evaluated in correlation to anti-peptide positivity

and avidity: (a) thrombotic events together or separate as

arterial, venous or microvascular thrombosis and (b) in

female patients also obstetric manifestations.

Selection of peptides

Immunoreactivity and avidity against six b2GPI-related

peptide sequences corresponding to different domains of

b2GPI were studied with a set of in-house indirect IgG anti-

peptide ELISA (Fig. 2). Peptides A (58LKTPRV63; domains

I–II), B (208KDKATF213; domain IV) and C (133TLRVYK138;

domain III) were derived from a hexapeptide phage display

library as epitope targets for pathogenic monoclonal anti-

Fig. 1 Patient enrollment. *Patients positive for IgG anti-b2GPI with

Sjogren‘s syndrome, systemic sclerosis, urticarial vasculitis and

autoimmune thrombocytopenic purpura were excluded due to small

frequencies/group sizes and for sake of clarity of statistical analysis.

Anti-b2GPI antibodies against b2-glycoprotein I, APS antiphospho-

lipid syndrome, SLE systemic lupus erythematosus

Cutting Edge in Autoimmunity

123

b2GPI [28] and were previously shown to have clinical rele-

vance [31]. Peptides D (N143G142P139A141N144S140K138;

domain III), E (Y96N98P116L115W111F81V64; domain II) and F

(K231M161D165G163N164H159P157; domains III–IV) represent

surface amino acid clusters on b2GPI and were found, using a

heptapeptide phage display library by Zager et al. [27], to have

binding epitopes for HAv and LAv IgG anti-b2GPI.

Anti-peptide elisa for immunoreactivity and avidity

detection

Synthetic b2GPI-related peptides (GL Biochem (Shanghai)

Ltd., Shanghai, PR China) were diluted with phosphate-

buffered saline, pH 7.4, 0.15 M (PBS) to final concentra-

tions 10 lg/ml/well for peptides B, D, E, F and 20 lg/ml/

well for A, C and used in the in-house indirect IgG anti-

peptide ELISA. Briefly, 100 ll of diluted peptides were

added into each well of 96-well plates (high binding Costar

EIA-RIA plates, clear, flat bottom, non-sterile, polystyrene;

Corning Incorporated, Corning, NY, USA) and incubated

overnight (*16 h) at 4 �C. The plates were then washed

one time with 250 ll of PBS and blocked with 200 ll of

3 % bovine serum albumin (BSA)-PBS for 1 h at room

temperature (RT). After washing the plates four times with

250 ll PBS, 100 ll of serum samples diluted with 0.5 %

BSA-PBS 1:100 were added onto the plates and incubated

for 2 h at RT. For determination of relative strength of

binding (avidity), serum samples were diluted 1:100 in

0.5 % BSA-PBS with increasing concentrations of NaCl:

0.15, 0.25, 0.5, 1 and 2 M. In the next step, the plates were

washed four times with 250 ll of PBS and incubated with

100 ll of 1:1,000 diluted goat anti-human IgG (Fc) con-

jugate (Accurate Chemical and Scientific corp., Westbury,

NY, USA) for 1 h at RT. In the final step, the plates were

again washed four times with 250 ll of PBS and 100 ll of

the substrate p-nitrophenylphosphate in diethanolamine

solution (pH 9.8) was added onto the plates. The reaction

was stopped when the positive control reached the value of

1000 mili optical density (mOD) at wavelength 405 nm

with reference wavelength being 680 nm. A serum sample

from a patient with APS highly positive for IgG anti-b2GPI

served as a positive control and in our preliminary anti-

peptide ELISA studies was shown to be highly immuno-

reactive to all six peptides. Cutoff values for determination

of positivity to each peptide were set for each peptide

separately as 99th percentile in mOD based on results/

responses of 100 tested sera from health blood donors. The

relative binding strength (avidity) was determined with

chaotropic variant of anti-peptide ELISA and presented as

% of binding at higher ionic strength (500 mM NaCl)

compared with binding at 150 mM NaCl. Avidity was

arbitrarily defined as LAv, heterogeneous (HetAv) and

HAv, if the binding at higher ionic strength (500 mM

NaCl) was under 25 %, between 25 and 65 %, or [65 %,

respectively, as compared to binding at 150 mM NaCl.

Statistical analysis

Statistical analysis was conducted using IBM SPSS Sta-

tistics program version 22 (IBM Corporation, Armong,

NY, USA) for applying logistic regression, multinomial

logistic regression, Mann–Whitney U test and Kruskal–

Wallis test. Logistic regression model was developed for

assessment of influence of anti-peptide positivity on the

occurrence of clinical manifestations (e.g., venous, arterial,

microvascular, overall thrombosis and obstetric manifes-

tations) or underlying disease-type subgroups (APS vs.

APS ? SLE, APS vs. SLE and APS ? SLE vs. SLE). In

this model, the presence of anti-peptide positivity was used

as an independent dichotomous variable. In total, 18 dif-

ferent dichotomous variables were produced (six separate

peptides and some selected combinations of different

peptides), namely positivity of each peptide (A, B, C, D, E

and F), mutual positivity of selected pairs of peptides

(A&E, B&C, B&D, B&F, C&D, C&F and D&F), mutual

positivity of selected triplets of peptides (B&C&D,

B&C&F, B&D&F and C&D&F) and the mutual positivity

of a quadruplet of peptides (B&C&D&F). Selection of

pairs, triplets and quadruplets was based on the peptide

position of domains of b2GPI, as estimated by previous

research (clinical and pathological relevance of anti-

b2GPI) [8, 22, 23, 26, 27, 29, 30]. Consequently, we

selected combinations of peptides A and E that are located

Fig. 2 Schematic presentation of peptide positions on b2GPI

domains. Only the peptide position on the domains is exact

Cutting Edge in Autoimmunity

123

on domains I and II and all combinations of peptides B, C,

D and F that are located on domains III and IV (Fig. 2).

Due to the large number of variables tested in the regres-

sion model, the forward conditional method was used for

choosing the significant variables.

Multinomial logistic regression was used for studying

the relationship between anti-peptide positivity and IgG

anti-b2GPI avidity presented as three arbitrarily defined

categories (HAv, HetAv or LAv). As before the multi-

nomial logistic regression model, anti-peptide positivity

was tested as 18 independent dichotomous variables and

forward conditional method was used for choosing the

significant variables.

For studying the influence of binding strength of anti-

peptide antibodies on the occurrence of clinical manifes-

tations or underlying disease type, Mann–Whitney U test or

Kruskal–Wallis test were used depending on whether there

were two or three groups compared. The results were

considered statistically significant when p \ 0.05 (i.e.,

significance level). In case of multiple pairwise compari-

sons, the Bonferroni correction was applied resulting in

significance level to be set at 0.0167, if three groups were

compared.

Results

From 201 enrolled patients, 43.3 % had APS, 33.3 % had

APS associated with SLE and 23.4 % had SLE. The mean

age of patients was 40.3 years and ranged between 12 and

80 years. The majority of patients were female (77.1 %).

76.6 % experienced thrombotic complications (arterial,

venous or microvascular) and 34.8 % of female patients

experienced obstetric manifestations. All included patients

were positive for IgG anti-b2GPI and had IgG anti-b2GPI

avidity determined. From 67 patients with HAv IgG anti-

b2GPI, 88.1 % were diagnosed with APS (independently

or as associated with SLE) and 11.9 % that had only SLE.

73.1 % of these patients had thrombotic complications

(venous, arterial and/or microvascular) and 53.3 % of

female patients had obstetric manifestations. In compari-

son, from 45 patients with LAv IgG anti-b2GPI, 64.4 %

were diagnosed with APS (independently or associated

with to SLE), whereas 35.6 % had SLE. Patients positive

for LAv IgG anti-b2GPI experienced 57.8 % thrombotic

disorders (venous, arterial and/or microvascular) and

27.0 % of female patients had obstetric episodes.

Observed numbers of positive sera for anti-peptide

antibodies were as follows: 20 for A (LKTPRV; domains

I–II), 58 for B (KDKATF; domain IV), 52 for C

(TLRVYK; domain III), 61 for peptide D (NGPANSK;

domain III), 50 for E (YNPLWFV; domain II) and 49 for F

(KMDGNHP; domains III–IV) (Table 1). Due to

insufficient amount of sera, some samples positive for

peptides were excluded in the coming step (5 positive

samples for peptide A, 1 for B, 2 for C, 9 for D, 7 for E and

4 for F), when binding strength for anti-peptide antibodies

was determined. The median binding strength was 28.7 %

for peptide A, 37.1 % for B, 32.7 % for C, 33.8 % for D,

27.5 % for E and 28.2 % for F. The anti-peptide antibodies

in studied serum samples were predominantly of HetAv (8

for peptide A, 43 for B, 40 for C, 37 for D, 29 for E and 29

for F) followed by LAv anti-peptide antibodies (4 for

peptide A, 11 for B, 10 for C, 13 for D, 13 for E and 14 for

F), whereas only a few were of HAv (3 for peptides A and

B, 2 for D and F and 1 for E) (Table 2). Further, statistical

analyses were conducted on determined anti-peptide avid-

ity presented as binding strength and not categorically as

arbitrarily defined groups (e.g., high, heterogeneous and

low anti-peptide avidity).

In order to evaluate the influence of anti-peptide posi-

tivity on the occurrence of clinical manifestations or

underlying disease-type subgroups or anti-b2GPI avidity, a

logistic regression modeling was applied. Here, mutual

positivity of selected combinations of peptides was also

tested. Among 201 patients, the number of serum samples

with mutual positivity of pairs of peptides A&E, B&C,

B&D, B&F, C&D, C&F and D&F were 17, 38, 47, 36, 39,

41 and 39, respectively. Number of serum samples with

mutual positivity of triplets of peptides B&C&D, B&C&F,

B&D&F and C&D&F were 32, 31, 33 and 34, while 28

serum samples were mutually positive for peptides

B&C&D&F. Using logistic regression, we observed a

statistically significant negative prediction of the positivity

for peptide D on the occurrence of venous thrombosis

(odds ratio (OR) 0.465, 95 % confidence interval (CI)

(95 % CI) 0.247–0.879, p = 0.018). We also observed a

positive prediction of positivity for peptide E on arterial

thrombosis occurrence (OR 3.441, 95 % CI 1.331–8.901,

p = 0.011), while mutual positivity for peptides B and F

had a negative prediction for arterial thrombosis (OR

0.153, 95 % CI 0.045–0.517, p = 0.003). Patients mutually

positive for peptides B, C and F had a *2.5 times smaller

odds to experience overall thrombosis (either venous,

arterial and/or microvascular) as compared to patients

negative for any of these peptides (OR 0.407, 95 % CI

0.185–0.874, p = 0.021). Patients positive for peptide B

had a *4.1 times higher odds for having only APS without

SLE as compared to patients diagnosed with APS associ-

ated with SLE (OR 0.242, 95 % CI 0.087–0.673,

p = 0.007), while on the other hand, positivity for peptides

C and F was a significant predictor for diagnosis of APS

associated with SLE as compared to only the APS diag-

nosis (OR 7.105, 95 % CI 2.292–22.028, p = 0.001). A

significant negative prediction was also observed for pos-

itivity to peptide A and the presence of HAv IgG anti-

Cutting Edge in Autoimmunity

123

b2GPI in comparison with LAv IgG anti-b2GPI (OR

0.167, 95 % CI 0.033–0.845, p = 0.031). A negative pre-

diction was observed for overall thrombosis in the HAv

anti-b2GPI patients subgroup when samples were positive

for peptide C and F (OR 0.170, 95 % CI 0.036–0.805,

p = 0.026) (Table 3).

In order to evaluate the impact of anti-peptide avidity,

the relationships between binding strength of anti-peptide

positive samples and clinical/serological parameters were

analyzed. A significant difference was observed between

binding strengths to peptides B and F in groups with

present HAv, HetAv and LAv IgG anti-b2GPI. For peptide

B, we observed the highest median binding strength in

HAv IgG anti-b2GPI group, followed by HetAv and LAv

IgG anti-b2GPI groups (44.3 vs. 36.6 vs. 31.7 %,

p = 0.021). Similarly, for peptide F, the highest median

binding strength was noted for HAv IgG anti-b2GPI

positive group, followed by LAv and HetAv anti-b2GPI

Table 1 Patients’ clinical and laboratory characteristics and anti-peptide positivity to b2-glycoprotein I-related peptides

All patients HAv anti-b2GPI HetAv anti-b2GPI LAv anti-b2GPI

Number of patients 201 67 89 45

Age in years; mean (SD; range) 40.3 (12.3; 12–80) 38.6 (10.0; 16–67) 40.0 (12.3; 20–76) 43.3 (14.9; 12–80)

Number of female patients [(%)] 155 (77.1) 45 (67.2) 73 (82.9) 37 (82.2)

Entire APS [n (%)] 154 (76.6) 59 (88.1) 66 (74.2) 29 (64.4)

Only APS [n (%)] 87 (43.3) 40 (59.7) 35 (39.3) 12 (26.7)

APS ? SLE [n (%)] 67 (33.3) 19 (28.4) 31 (34.8) 17 (37.8)

SLE [n (%)] 47 (23.4) 8 (11.9) 23 (25.8) 16 (35.6)

VT [n (%)] 88 (43.8) 37 (55.2) 39 (43.8) 12 (26.7)

AT [n (%)] 61 (30.3) 17 (5.4) 29 (32.6) 15 (33.3)

mT [n (%)] 12 (6.0) 1 (1.5) 4 (4.5) 7 (15.6)

T [n (%)] 134 (66.7) 49 (73.1) 59 (66.3) 26 (57.8)

OM [n/female (%)] 54/155 (34.8) 24/45 (53.3) 20/73 (27.4) 10/37 (27.0)

Positivity of peptide A [n (%)] 20 (10.0) 2 (3.0) 11 (12.4) 7 (15.6)

B [n (%)] 58 (28.9) 18 (26.9) 27 (30.3) 13 (28.9)

C [n (%)] 52 (25.9) 12 (17.9) 27 (30.3) 13 (28.9)

D [n (%)] 61 (30.3) 17 (25.4) 30 (33.7) 14 (31.1)

E [n (%)] 50 (24.9) 11 (16.4) 25 (28.1) 14 (31.1)

F [n (%)] 49 (24.4) 11 (16.4) 25 (28.1) 13 (28.9)

anti-b2GPI antibodies against b2-glycoprotein I, APS antiphospholipid syndrome, AT arterial thrombosis, HAv high avidity, HetAv heteroge-

neous avidity, LAv low avidity, mT microvascular thrombosis, OM obstetric manifestations, SD standard deviation, SLE systemic lupus

erythematosus, VT venous thrombosis, T overall thrombosis

Table 2 Anti-peptide avidity and binding strength to b2-glycoprotein I-related peptides

Peptide Positivity (n) Samples enrolled in anti-peptide

avidity study (n)aMedian binding

strength (range) (%)bAnti-peptide avidityc

HAv (n) HetAv (n) LAv (n)

A 20 15 28.7 (16.9–80.7) 3 8 4

B 58 57 37.1 (10.6–89.8) 3 43 11

C 52 50 32.7 (13.0–62.1) 0 40 10

D 61 52 33.8 (14.6–65.4) 2 37 13

E 50 43 27.5 (9.3–73.0) 1 29 13

F 49 45 28.2 (13.7–111.0) 2 29 14

HAv high avidity, HetAv heterogeneous avidity, LAv low avidity, n number of samples/patientsa Some peptides positive samples were excluded from further avidity analysis/determination due to insufficient sera (nine samples for peptide A,

seven for B, four for C, five for D, one for E and two for F)b Binding strength was determined with chaotropic variant of anti-peptide ELISA and calculated as A500 mM NaCl/A150 mM NaCl

c Anti-peptide avidity was arbitrary defined as LAv, HetAv or HAv, if the binding at higher ionic strength (500 mM NaCl) was under 25 %,

between 25 and 65 % or [65 % as compared to binding at 150 mM NaCl, respectively

Cutting Edge in Autoimmunity

123

groups (41.8 vs. 28.4 vs. 26.5 %, p = 0.044). The binding

strength for peptide B differed in patients with/without

arterial and microvascular thrombosis with the higher

median binding strength for the presence of arterial

thrombosis compared with those without arterial throm-

bosis (41.9 vs. 34.4 %, p = 0.033). Interestingly, the

median binding strength was higher in patients positive for

peptide B without microvascular thrombosis compared

with patients positive for peptide B with microvascular

thrombosis (38.8 vs. 24.3 %, p = 0.049). Additionally, a

difference between binding strength to peptide B was

observed also among patients diagnosed either with APS,

APS associated with SLE or SLE alone with the highest

median binding strength for APS group followed by APS

associated with SLE and SLE groups (41.2 vs. 34.3 vs.

32.2 %, p = 0.022). A difference in strength of binding for

peptide C was also observed in female patients with pre-

vious obstetric manifestation compared with female

patients without such complications (36.8 vs. 32.0 %,

p = 0.039) (Table 4).

Table 3 Significant influences of anti-peptide positivity on clinical/serological outcomes

Outcome Anti-peptide positivity Odds ratio

(95 % CI)

p value

Venous thrombosis D 0.465 (0.247–0.879) 0.018

Arterial thrombosis E 3.44 (1.33–8.90) 0.011

B&F 0.153 (0.045–0.517) 0.003

Overall thrombosis B&C&F 0.407 (0.185–0.874) 0.021

APS ? SLE versus APSa B 0.242 (0.087–0.673) 0.007

C&F 7.11 (2.29–22.0) 0.001

HAv versus LAv anti-b2GPIb A 0.167 (0.033–0.845) 0.031

Overall thrombosis (HAv anti-b2GPI subgroup) C&F 0.170 (0.036–0.805) 0.026

anti-b2GPI antibodies against b2-glycoprotein I, APS antiphospholipid syndrome, HAv high avidity, HetAv heterogeneous avidity, LAv low

avidity, SLE systemic lupus erythematosusa Multiple comparisons of dependent variable were performed (APS vs. APS ? SLE, APS vs. SLE and APS ? SLE vs. SLE) and Bonferroni

correction used—significant result when p \ 0.0167b Multinominal logistic regression model was used to asses this correlation (HAv vs. HetAv, HAv vs. LAv and HetAv vs. LAv anti-b2GPI

groups). Significant results are shown for relation between HAv vs. LAv anti-b2GPI groups

Table 4 Significant relationships between anti-peptide binding strength and serological/clinical outcomes

Peptide Testing outcome Median (range) binding strength (%) n p value Statistical test

B AT- 34.4 (10.6–89.8) 44 0.033 M–W (U = 174.0)

AT? 41.9 (32.4–62.8) 13

B mT- 38.8 (10.6–89.8) 53 0.049 M–W (U = 43.0)

mT? 24.3 (15.7–37.1) 4

B APS 41.2 (21.9–89.8) 27 0.022 K–W (v22 = 7.661)

APS ? SLE 34.3 (15.7–49.1) 15

SLE 32.2 (10.6–51.7) 15

B HAv 44.3 (26.5–85.5) 17 0.021 K–W (v22 = 7.769)

HetAv 36.6 (16.0–89.8) 27

LAv 31.7 (10.6–50.1) 13

C OM-a 32.0 (13.0–56.6) 30 0.039 M–W (U = 117.0)

OM?a 36.8 (24.2–62.1) 13

F HAv 41.8 (14.5–50.6) 10 0.044 K–W (v22 = 6.229)

HetAv 26.5 (14.8–111.0) 22

LAv 28.4 (13.7–74.8) 13

Anti-b2GPI antibodies against b2-glycoprotein I, APS antiphospholipid syndrome, AT arterial thrombosis, HAv high avidity, HetAv heteroge-

neous avidity, K–W Kruskal–Wallis test, LAv low avidity, mT microvascular thrombosis, M–W Mann–Whitney U test, n number of patients, OM

obstetric manifestations, SLE systemic lupus erythematosusa Only female patients positive for peptide F were enrolled in this model

Cutting Edge in Autoimmunity

123

Discussion

In this study, we evaluated the immunoreactivity and avidity

(presented as binding strength) against six peptide clusters

corresponding to amino acid clusters on different domains on

b2GPI, discovered with hexamer [28] and heptamer [27]

phage display libraries. Peptides A, B and C were identified

as in vitro/in vivo inhibitors for monoclonal anti-b2PGI

antibodies [28], while peptides D and E were characterized

as surface-binding epitopes for HAv IgG anti-b2GPI and

peptide F as a binding epitope for LAv IgG anti-b2GPI [27].

Diverse reactivity of human patient sera to peptides A–F and

their clinical and/or serological relevance confirmed the

dispersity of polyclonal IgG anti-b2GPI in antigen recogni-

tion. We report that not only the positivity to specific pep-

tides but also anti-peptide avidity is of relevance when

studying clinical/serological correlations.

Peptides allocated to domains III and IV (peptides B, C,

D and F) mainly carried negative predictions for positivity

and clinical correlations of APS (Table 3). The positivity

of same sera for peptides B and F was negatively associ-

ated with arterial thrombosis in contrast to peptide E

(domain II) for which positivity was a predictor for arterial

thrombosis. The mutual positivity for peptides B, C and F

was also a negative predictor for overall thrombosis. The

positivity for peptides C and F was also negatively asso-

ciated with overall thrombosis in the subgroup of patients

with present HAv IgG anti-b2GPI. Further, positivity for

peptide D was also a negative predictor for venous

thrombosis. Interestingly, positivity to peptide B more

likely occurred in patients with only APS as compared to

patients with APS associated with SLE, while the opposite

association was detected for positivity of same sera for

peptides C and F. On the other hand, the positivity for

peptide A was a stronger predictor for the presence of LAv

anti-b2GPI as compared to HAv IgG anti-b2GPI. The

clinical study by Shoenfeld et al. on 295 patients with APS

showed that the positivity for peptide A (58LKTPRV63;

domains I–II) was negatively associated with overall

thrombosis [34]. The study also revealed that anti-peptide

C antibodies were a significant predictor of recurrent

spontaneous abortions [34], which was not confirmed in

our study. However, peptide C was significantly involved

in the prediction of overall thrombosis (negatively) and

disease type (APS associated with SLE). In a recent study,

Banzato et al. [38] tested 22 triple positive for aPL (lupus

anticoagulant, IgG aCL, IgG anti-b2GPI), 15 double

positive (IgG aCL, IgG anti-b2GPI), 9 single positive (IgG

anti-b2GPI) patient sera and 20 control serum samples for

reactivity to synthetic domain I (full-length; correctly fol-

ded; amino acid sequence 1–62) and two peptide sequences

on domains I (amino acid sequence 40–52) and I–II (amino

acid sequence 57–70; a larger copy of peptide C) with a

competitive inhibition anti-b2GPI ELISA. They found no

differences in inhibition between patient/control samples

for both peptides in contrast to inhibition with domain I and

concluded that triple aPL-positive patients have a signifi-

cantly higher titer of anti-domain I antibodies as compared

to double or single aPL-positive patients [38]. Another

study with domains I–V of b2GPI-specific peptides

(including elongated peptides A–C) was conducted by

Muller et al. [39] on a surface plasmon resonance (SPR)

biosensor system. No differences between 21 serum sam-

ples from APS patients and age-/sex-matched controls were

detected in SPR signals for all peptides [39], in contrast to

higher SPR responses for patient serum samples compared

with controls when immobilized b2GPI was used [40],

indicating a problem in detection of antibody binding to

short amino acid sequences. Zager et al. [26] used SPR

chips with immobilized b2GPI and noted that antibody-

antigen interactions between polyclonal HAv IgG anti-

b2GPI and b2GPI are presumably more stable, monovalent

and less dependent on antigen density as compared to

polyclonal LAv IgG anti-b2GPI, which form less stable

complexes, are more dependent on antigen density and

require bivalent interactions. Pierangeli et al. worked with

b2GPI-derived peptides and observed a significantly

decreased thrombus size in mice injected with IgG isolated

from an APS patient and treated with an elongated version

of peptide A (NTLKTPRVGGC) as compared to a control

peptide (scrambled elongated peptide A; GTKGCP

NVRLT). These results indicated a possible therapeutic

application of short synthetic peptide that mimics amino

acid sequence between domains I and II of b2GPI as it

inhibited in vivo thrombogenic properties of aPL in mice

[41].

The second step was the evaluation of anti-peptide

avidity of sera positive for peptides and its clinical/sero-

logical correlations (Table 4). The highest median binding

strength to peptides B and F was observed for the groups of

patients with HAv IgG anti-b2GPI as compared to patients

with HetAv and LAv IgG anti-b2GPI. The highest median

binding strength to peptide B was also noted for the groups

of patients with APS followed by the APS associated with

SLE group, with the lowest binding strength for SLE

group. The binding strength of peptide B was of impor-

tance for arterial with the higher median binding strength in

patients who previously experienced arterial thrombosis as

compared to those who did not. Interestingly, in contrast to

arterial thrombosis, a lower median binding strength was

observed for patients positive for peptide B with previous

microvascular thrombotic events as compared to those

without events. Although we did not observe a significant

association between positivity to peptide C

(133TLRVYK138; domain III) and obstetric manifestations

as was the case reported by Shoenfeld et al. [31], we

Cutting Edge in Autoimmunity

123

observed a significantly higher anti-peptide C avidity

between female patients who experienced obstetric mani-

festations and those who did not. Avidity is an important

aspect in describing the binding characteristics of anti-

bodies; however, it seems that in our tests (indirect anti-

peptide ELISA with BSA blocking) avidity was affected by

shorter amino acid sequences, because the measured anti-

peptide avidity was lower (Table 2) compared with anti-

b2GPI avidity from our previous studies [24, 29, 32, 33].

The results on immunoreactivity and avidity in the current

study are in part controversial. For instance, the immuno-

reactivity to peptide D (domain III) was negatively asso-

ciated with venous thrombosis (Table 3); however, in

previous studies, peptide D was found as a possible-bind-

ing epitope for HAv anti-b2GPI, associated with throm-

botic as well as obstetric complications in APS patients

[24, 29, 32, 33]. Positivity for peptide E (domain II) rec-

ognized by HAv anti- b2GPI [27], was a predictor for

arterial thrombosis (Table 3). Expectedly, positivity for

peptide F (domains III–IV) found as a binding epitope for

LAv anti-b2GPI [27] was involved in negative prediction

of arterial and overall thrombosis and as a predictor for

APS ? SLE as compared to diagnosis of only APS

(Table 3). However, interestingly, although peptide F was

found to be a binding epitope for LAv anti-b2GPI [27], we

observed a significant correlation between anti-peptide F

binding strength and anti-b2GPI avidity. The highest

median anti-peptide F avidity was detected for HAv anti-

b2GPI (Table 4), which once again elucidates the diversity

of human polyclonal IgG anti-b2GPI subpopulations.

In conclusion, in the last few years, a theory has been

emerging that points to IgG anti-b2GPI directed against

domain I as having the most clinical relevance [7, 8, 19].

However, current as well as previous clinical and transla-

tional studies [25, 27–29] on anti-b2GPI and peptide pos-

itivity also clearly indicate that other anti-b2GPI

subpopulations have to be taken into account to delineate

APS pathology. Furthermore, anti-b2GPI and anti-peptide

avidity should be considered, especially in context of APS

clinical manifestations, such as thromboses and obstetric

complications. Our results indicate a wide dispersity in

antibody avidity and recognition of peptides corresponding

to amino acid sequences on different domains of b2GPI

and their clinical relevance. Direct comparison between

anti-b2GPI and anti-peptide positivity/avidity cannot be

equally evaluated due to different interactions between

antibody/whole antigen and antibody/peptides. This delin-

eation depends on the secondary structure of b2GPI, which

can strengthen the epitope/paratope recognition and bind-

ing strength, versus the lack of these interactions in pep-

tides. Thus, classification of anti-b2GPI into subgroups

regarding epitope specificity and avidity could represent an

additional tool in understanding the pathogenicity of IgG

anti-b2GPI in APS. In the future, studying IgG anti-b2GPI

subpopulations as evaluated by positivity/avidity of anti-

bodies against all domains and specific domains of b2GPI

should be addressed, especially in the context of their

clinical impact.

Acknowledgments The study is in accordance with the Helsinki

Declaration of 1975, revised in 1983 and was approved by the

National Ethical Committee (No. 163/02/09). This work was funded

by the National Research Program Grant No. P3-0314 from the

Ministry of Higher Education, Science and Technology, Slovenia.

Conflict of interest The authors declare no conflict of interest.

References

1. Miyakis S, Lockshin MD, Atsumi T, Branch DW, Brey RL,

Cervera R, et al. International consensus statement on an update

of the classification criteria for definite antiphospholipid syn-

drome (APS). J Thromb Haemost. 2006;4(2):295–306. doi:10.

1111/j.1538-7836.2006.01753.x.

2. Wilson WA, Gharavi AE, Koike T, Lockshin MD, Branch DW,

Piette JC, et al. International consensus statement on preliminary

classification criteria for definite antiphospholipid syndrome:

report of an international workshop. Arthritis Rheum. 1999;42(7):

1309–11. doi:10.1002/1529-0131(199907)42:7\1309:AID-ANR1[3.

0.CO;2-F.

3. Cervera R, Asherson RA. Antiphospholipid syndrome. In:

Shoenfeld Y, Cervera R, Gershwin ME, editors. Diagnostic criteria

in autoimmune diseases. Totowa: Humana Press; 2008. p. 9–14.

4. Cervera R, Piette JC, Font J, Khamashta MA, Shoenfeld Y,

Camps MT, et al. Antiphospholipid syndrome: clinical and

immunologic manifestations and patterns of disease expression in

a cohort of 1,000 patients. Arthritis Rheum. 2002;46(4):1019–27.

doi:10.1002/art.10187.

5. Tan EM, Cohen AS, Fries JF, Masi AT, McShane DJ, Rothfield

NF, et al. The 1982 revised criteria for the classification of sys-

temic lupus erythematosus. Arthritis Rheum. 1982;25(11):

1271–7.

6. Hochberg MC. Updating the American College of Rheumatology

revised criteria for the classification of systemic lupus erythe-

matosus. Arthritis Rheum. 1997;40(9):1725.

7. Meroni PL, Borghi MO, Raschi E, Tedesco F. Pathogenesis of

antiphospholipid syndrome: understanding the antibodies. Nat

Rev Rheumatol. 2011;7(6):330–9. doi:10.1038/nrrheum.2011.52.

8. de Laat B, Mertens K, de Groot PG. Mechanisms of disease:

antiphospholipid antibodies-from clinical association to patho-

logic mechanism. Nat Clin Pract Rheumatol. 2008;4(4):192–9.

doi:10.1038/ncprheum0740.

9. Espinosa G, Cervera R. Antiphospholipid syndrome: frequency,

main causes and risk factors of mortality. Nat Rev Rheumatol.

2010;6(5):296–300. doi:10.1038/nrrheum.2010.47.

10. Artenjak A, Lakota K, Frank M, Cucnik S, Rozman B, Bozic B,

et al. Antiphospholipid antibodies as non-traditional risk factors

in atherosclerosis based cardiovascular diseases without overt

autoimmunity. A critical updated review. Autoimmun Rev.

2012;11(12):873–82. doi:10.1016/j.autrev.2012.03.002.

11. Shoenfeld Y, Gerli R, Doria A, Matsuura E, Cerinic MM, Ronda

N, et al. Accelerated atherosclerosis in autoimmune rheumatic

diseases. Circulation. 2005;112(21):3337–47. doi:10.1161/

CIRCULATIONAHA.104.507996.

Cutting Edge in Autoimmunity

123

12. Lin F, Murphy R, White B, Kelly J, Feighery C, Doyle R, et al.

Circulating levels of beta2-glycoprotein I in thrombotic disorders

and in inflammation. Lupus. 2006;15(2):87–93.

13. Horstman LL, Jy W, Bidot CJ, Ahn YS, Kelley RE, Zivadinov R,

et al. Antiphospholipid antibodies: paradigm in transition.

J Neuroinflammation. 2009;6:3. doi:10.1186/1742-2094-6-3.

14. Lozier J, Takahashi N, Putnam FW. Complete amino acid

sequence of human plasma beta 2-glycoprotein I. Proc Natl Acad

Sci USA. 1984;81(12):3640–4.

15. Hunt JE, Simpson RJ, Krilis SA. Identification of a region of beta

2-glycoprotein I critical for lipid binding and anti-cardiolipin antibody

cofactor activity. Proc Natl Acad Sci USA. 1993;90(6):2141–5.

16. Schwarzenbacher R, Zeth K, Diederichs K, Gries A, Kostner GM,

Laggner P, et al. Crystal structure of human beta2-glycoprotein I:

implications for phospholipid binding and the antiphospholipid

syndrome. EMBO J. 1999;18(22):6228–39. doi:10.1093/emboj/

18.22.6228.

17. Sodin-Semrl S, Frank M, Ambrozic A, Pavlic J, Sustar V, Cucnik S,

et al. Interactions of phospholipid binding proteins with negatively

charged membranes: b2-glycoprotein I as a model mechanism. In:

Leitmannova Liu A, editor. Advances in planar lipid bilayers and

liposomes, 1st ed. Amsterdam: Elsevier; 2008. p. 243–73.

18. Agar C, van Os GM, Morgelin M, Sprenger RR, Marquart JA,

Urbanus RT, et al. Beta2-glycoprotein I can exist in 2 confor-

mations: implications for our understanding of the antiphospho-

lipid syndrome. Blood. 2010;116(8):1336–43. doi:10.1182/blood-

2009-12-260976.

19. de Laat B, Derksen RH, van Lummel M, Pennings MT, de Groot PG.

Pathogenic anti-beta2-glycoprotein I antibodies recognize domain I

of beta2-glycoprotein I only after a conformational change. Blood.

2006;107(5):1916–24. doi:10.1182/blood-2005-05-1943.

20. Wang MX, Kandiah DA, Ichikawa K, Khamashta M, Hughes G,

Koike T, et al. Epitope specificity of monoclonal anti-beta

2-glycoprotein I antibodies derived from patients with the anti-

phospholipid syndrome. J Immunol. 1995;155(3):1629–36.

21. Kasahara H, Matsuura E, Kaihara K, Yamamoto D, Kobayashi K,

Inagaki J, et al. Antigenic structures recognized by anti-beta2-gly-

coprotein I auto-antibodies. Int Immunol. 2005;17(12):1533–42.

doi:10.1093/intimm/dxh330.

22. Mahler M, Norman GL, Meroni PL, Khamashta M. Autoanti-

bodies to domain 1 of beta 2 glycoprotein 1: a promising can-

didate biomarker for risk management in antiphospholipid

syndrome. Autoimmun Rev. 2012;12(2):313–7. doi:10.1016/j.

autrev.2012.05.006.

23. Ambrozic A, Avicin T, Ichikawa K, Kveder T, Matsuura E,

Hojnik M, et al. Anti-beta(2)-glycoprotein I antibodies in children

with atopic dermatitis. Int Immunol. 2002;14(7):823–30.

24. Cucnik S, Kveder T, Krizaj I, Rozman B, Bozic B. High avidity

anti-beta 2-glycoprotein I antibodies in patients with anti-

phospholipid syndrome. Ann Rheum Dis. 2004;63(11):1478–82.

doi:10.1136/ard.2003.017939.

25. de Laat B, Derksen RH, de Groot PG. High-avidity anti-beta2

glycoprotein I antibodies highly correlate with thrombosis in

contrast to low-avidity anti-beta2 glycoprotein I antibodies.

J Thromb Haemost. 2006;4(7):1619–21. doi:10.1111/j.1538-

7836.2006.02002.x.

26. Zager U, Irman S, Lunder M, Skarabot M, Musevic I, Hodnik V,

et al. Immunochemical properties and pathological relevance of

anti-beta2-glycoprotein I antibodies of different avidity. Int

Immunol. 2011;23(8):511–8. doi:10.1093/intimm/dxr043.

27. Zager U, Lunder M, Cucnik S, Kveder T, Rozman B, Bozic B.

Immunodominant epitopes of beta2-glycoprotein I for high

avidity antibodies. In: 8th International congress on autoimmu-

nity. Granada, Spain: Geneva: Kenes International; 2012.

28. Blank M, Shoenfeld Y, Cabilly S, Heldman Y, Fridkin M, Kat-

chalski-Katzir E. Prevention of experimental antiphospholipid

syndrome and endothelial cell activation by synthetic peptides.

Proc Natl Acad Sci USA. 1999;96(9):5164–8.

29. Cucnik S, Kveder T, Artenjak A, Ulcova Gallova Z, Swadzba J,

Musial J, et al. Avidity of anti-beta2-glycoprotein I antibodies inpatients with antiphospholipid syndrome. Lupus. 2012;21(7):764–5.

doi:10.1177/0961203312440057.

30. Iverson GM, von Muhlen CA, Staub HL, Lassen AJ, Binder W,

Norman GL. Patients with atherosclerotic syndrome, negative in

anti-cardiolipin assays, make IgA autoantibodies that preferentially

target domain 4 of beta2-GPI. J Autoimmun. 2006;27(4):266–71.

doi:10.1016/j.jaut.2006.09.007.

31. Shoenfeld Y, Krause I, Kvapil F, Sulkes J, Lev S, von Landen-

berg P, et al. Prevalence and clinical correlations of antibodies

against six beta2-glycoprotein-I-related peptides in the anti-

phospholipid syndrome. J Clin Immunol. 2003;23(5):377–83.

32. Cucnik S, Bozic B, Kveder T, Tomsic M, Rozman B. Avidity of

anti-beta2-glycoprotein I and thrombosis or pregnancy loss in

patients with antiphospholipid syndrome. Ann N Y Acad Sci.

2005;1051:141–7. doi:10.1196/annals.1361.055.

33. Cucnik S, Kveder T, Ulcova-Gallova Z, Swadzba J, Musial J,

Valesini G, et al. The avidity of anti-beta2-glycoprotein I anti-

bodies in patients with or without antiphospholipid syndrome: a

collaborative study in the frame of the European forum on anti-

phospholipid antibodies. Lupus. 2011;20(11):1166–71. doi:10.

1177/0961203311406308.

34. Artenjak A, Kozelj M, Lakota K, Cucnik S, Bozic B, Sodin-Semrl S.

High avidity anti-b2-glycoprotein I antibodies activate human cor-

onary artery endothelial cells and trigger peripheral blood mono-

nuclear cell migration. Eur J Inflamm. 2013;11(2):385–96.

35. Favaloro EJ, Wong RC. Laboratory testing for the antiphospho-

lipid syndrome: making sense of antiphospholipid antibody

assays. Clin Chem Lab Med. 2011;49(3):447–61. doi:10.1515/

CCLM.2011.064.

36. Cucnik S, Ambrozic A, Bozic B, Skitek M, Kveder T. Anti-

beta2-glycoprotein I ELISA: methodology, determination of cut-

off values in 434 healthy Caucasians and evaluation of mono-

clonal antibodies as possible international standards. Clin Chem

Lab Med. 2000;38(8):777–83. doi:10.1515/CCLM.2000.111.

37. Avcin T, Ambrozic A, Bozic B, Accetto M, Kveder T, Rozman

B. Estimation of anticardiolipin antibodies, anti-beta2 glycopro-

tein I antibodies and lupus anticoagulant in a prospective longi-

tudinal study of children with juvenile idiopathic arthritis. Clin

Exp Rheumatol. 2002;20(1):101–8.

38. Banzato A, Pozzi N, Frasson R, De Filippis V, Ruffatti A, Bison

E, et al. Antibodies to domain I of beta(2)glycoprotein I are in

close relation to patients risk categories in antiphospholipid

syndrome (APS). Thromb Res. 2011;128(6):583–6. doi:10.1016/

j.thromres.2011.04.021.

39. Muller C, Thaler M, Schlichtiger A, Schreiegg A, Balling G,

Steigerwald U, et al. beta2-glycoprotein I-derived peptides as

antigenic structures for the detection of antiphospholipid anti-

bodies. J Thromb Haemost. 2010;8(9):2073–5. doi:10.1111/j.

1538-7836.2010.03987.x.

40. Metzger J, von Landenberg P, Kehrel M, Buhl A, Lackner KJ,

Luppa PB. Biosensor analysis of beta2-glycoprotein I-reactive

autoantibodies: evidence for isotype-specific binding and differ-

entiation of pathogenic from infection-induced antibodies. Clin

Chem. 2007;53(6):1137–43. doi:10.1373/clinchem.2006.079632.

41. Pierangeli SS, Blank M, Liu X, Espinola R, Fridkin M, Ostertag

MV, et al. A peptide that shares similarity with bacterial antigens

reverses thrombogenic properties of antiphospholipid antibodies

in vivo. J Autoimmun. 2004;22(3):217–25. doi:10.1016/j.jaut.

2004.01.002.

Cutting Edge in Autoimmunity

123