The epitope study of α-fodrin autoantibody in primary Sjögren's syndrome

Abstract

Alpha-fodrin, an intracellular organ-specific cytoskeleton protein, was identified recently as an autoantigen associated with Sicca- and Sjögren's syndrome (SS). Identification of the antigenic determinants of α-fodrin is a prerequisite to developing highly sensitive and specific anti-α-fodrin antibodies, which provides potential means for the diagnosis of primary Sjögren's syndrome (pSS) in patients. Based on the structure and predicted antigenic sites of α-fodrin protein with 560 amino acids (α-fodrin 560), we prepared a set of overlapping recombinant protein fragments covering antigenic epitopes and synthesized a set of peptides derived from the α-fodrin protein. These recombinant proteins and synthesized peptides were subjected to screening with pSS patients sera, respectively. The peptide with the strongest immunoreactivity was used as antigenic peptide to define further the role of anti-α-fodrin-peptide antibodies in the sera of 135 patients with pSS, 48 patients with systemic lupus erythematosus (SLE), 88 patients with rheumatoid arthritis (RA) and 83 normal controls. Our data showed that the N-terminal peptide of amino acids 46–59 (N46) of α-fodrin 560 was the epitope with strongest antigenicity. The prevalences of anti-N46 peptide antibodies (α-N46PA) in patients with pSS, SLE, RA and normal controls were 78.5%, 10.4%, 21.6% and 6.0%, respectively. The sensitivity and specificity of the autoantibodies in pSS were 78.5% and 86.8%, respectively. These results suggest the α-N46PA which shows highest sensitivity and specificity is of significance to develop an effective diagnostic approach for pSS.

Introduction

Primary Sjögren's syndrome (pSS) is a systemic rheumatic disease characterized by progressive lymphocytic and plasma cell infiltration of the salivary and lachrymal glands, and the presence of several autoantibodies in the blood [1]. Data have demonstrated that immunological factors, certain autoantigens and autoanibodies play important roles in the pathogenesis of organ-specific autoimmunity in the lacrimal and salivary glands in pSS. Haneji et al. identified a 120-kDa fragment of the ubiquitous cytoskeletal protein α-fodrin as an autoantigen in an NFS/sld mouse model of human pSS [2]. Furthermore, recombinant α-fodrin induced an antigen-specific proliferative response in spleen cells from immunized NFS/sld-3Tx mice and peripheral blood monocytes from SS patients, suggesting that this protein is a candidate organ-specific autoantigen that may be responsible for the development of autoimmune lesions in human SS [3]. More recently, several studies suggested that anti-α-fodrin antibody might be of value in the diagnosis of SS and may serve as a marker of disease activity [4–9].

The identification and characterization of pathogenetically disease-relevant autoantigens is a key issue in autoimmunity. To evaluate the potential of anti-α-fodrin antibodies in the diagnosis of SS, a conserved α-fodrin peptide was used as the antigen in our previous study [10]. The results suggest that anti-α-fodrin peptide antibodies (α-FPA) are serological markers in pSS, and occur in the early stages of SS. Recently, Shiari et al. investigated the epitope specificities of IgG antibodies using overlapping fusion proteins of the N-terminal part of α-fodrin as antigens and suggested that major and initial B cell epitopes reside specifically in N-terminal amino acids 36–132 and could be used for the diagnosis of pSS [11].

In this study, we analysed the antigenic determinants of α-fodrin with 560 amino acids (designated herein as α-fodrin 560) using computer analysis and immunological analysis about the overlapping recombinant α-fodrin protein fragments covering antigenic epitopes and the synthesized peptides derived from the α-fodrin protein. We then examined the sensitivity and specificity of anti-α-fodrin-antigenic-peptide antibodies for pSS. We evaluated further whether the presence of anti-α-fodrin-peptide antibodies was correlated with anti-α-fodrin antibodies and the most frequent pSS antibodies. To our knowledge, this is the first report so far to evaluate the prevalence of anti-α-fodrin epitope peptide antibodies in a cohort of Chinese patients with pSS.

Materials and methods

Patients

All serum samples were provided by the hospitals in Beijing. Serum samples from 135 patients with pSS (115 females, 20 males) and 25 secondary SS associated with RA or SLE were examined. Their mean [± standard deviation (s.d.)] age was 54 (± 6.5) years, and the mean (s.d.) disease duration was 8.0 (± 3.0) years. Serum samples from 48 patients with systemic lupus erythematosus (SLE), 88 with rheumatoid arthritis (RA) and 83 healthy subjects were analysed as controls. All the patients with pSS were evaluated carefully based on the European–American classification criteria [12]. Patients with RA and SLE fulfilled the American College of Rheumatology (ACR) criteria for classification.

Structure analysis of α-fodrin 560

The biophysical characteristics of α-fodrin 560 encoded by the 1680 base pairs (bp) of α-fodrin ORF sequence (GenBank Accession number: BC053521) was analysed using the ProtScale program [Swiss Institute of Bioinformatics (SIB), Geneva, Switzerland]. Antigenic peptides were determined according to the method by Jameson and Wolf (1990) [13]. Predictions are based on a table that reflects the occurrence of amino acid residues in experimentally known segmental epitopes.

Overlapping protein expression and immunoblotting detection

Using α-fodrin cDNA plasmid (a gift from Dr Hayashi) as template, a set of cDNA constructs expressing α-fodrin N-terminal amino acids 1–59, 1–128, 1–167, 1–238, 1–284, 1–360, 1–423 and 1–485 were created by polymerase chain reaction (PCR) cloning into expression plasmid pGEX-4T-2 (GE Healthcare Amersham, Uppsala, Sweden). Recombinant α-fodrin overlapping proteins were prepared and detected as described previously [14]. Briefly, the purified recombinant fusion proteins were separated by sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS-PAGE) and electroblotted to polyvinylidene difluoride (PVDF) membrane. Blocked membranes were incubated with the representative serum samples [diluted 1:50 in phosphate-buffered saline (PBS)] from five patients with pSS (IgG anti-α-fodrin antibody-positive) and subsequently with biotinylated goat anti-human IgG secondary antibodies. Immunoreactive signals were captured with diaminobenzidine tetrahydrochloride (DAB) detection reagents. Moreover, the parts of α-fodrin 560, with the exception of the minimal immunoreactivity-positive recombined protein segments, were expressed and detected with the same serum mixture samples above, then the antigenic determinant regions of α-fodrin were recognized from these approaches.

Design and synthesis of peptides

The peptides were designed on the basis of the combined analyses of the antigenic determinant regions. The synthesized peptides of 13–20 amino acids covered the antigenic determinant regions and contained the corresponding predicted antigenic sites to the limit. All the peptides were synthesized commercially and purified by reversed-phase high pressure liquid chromatography (HPLC) with purity > 95%.

Enzyme linked immunosorbent assay (ELISA)

Peptides (1 µg/ml, in 0.05 M carbonate buffer, pH 9.6) were dispensed into a 96-well microplate (100 µl/well) and then incubated at 4°C overnight. After washing with PBS containing 0.5% Tween-20 (PBS-T), normal rabbit sera were added and incubated at 37°C for 1 h to block the non-specific binding. The same representative serum samples (each 10 µl) from 30 patients with pSS (anti-α-fodrin antibody-positive) diluted with 900 µl of sample buffer and were added into each well, followed thereafter by standard ELISA procedures.

Isolation of peptide-specific antibodies

To affinity purify the anti-peptide antibodies, we coupled the selected antigenic peptide (5 mg peptide/g of dried Sepharose powder) to Sepharose 4B (Pharmacia, Uppsala, Sweden), according to the manufacturer's instructions. Serum samples diluted in PBS were applied to the columns. The columns were washed with PBS and bound Ig were eluted with 0.1 M glycine (pH 2.5) and dialysed against PBS.

From the results with ELISA and detection of the antibodies affinity purified against the peptide, we noted that the peptides showed comparable immunoreactivities with the α-fodrin 560 as the strongest antigenic peptide.

Detection of antibodies against the strongest antigenic peptide of α-fodrin in patients

To establish a standard for the ELISA, 135 sera obtained from patients with pSS were measured. Flat-bottomed microtitre plates were coated with the strongest antigenic peptide of α-fodrin (1 µg/ml) at 4°C overnight. Antibodies against the strongest antigenic peptide of α-fodrin in patients were detected by ELISA, as stated above. The results of detection for antibodies were expressed as optical density (OD) units ± s.d. An OD value greater than 2 s.d. of the normal control sera was considered positive.

In addition, IgG anti-α-fodrin, anti-SSA, anti-SSB and anti-nuclear antibodies (ANA) were determined using commercially available ELISA kits according to the manufacturer's instructions (ORGenTec GmbH, Mainz, Germany). P-values less than 0.05 were considered to be significant.

Results

Prediction of antigenic peptide sites of the α-fodrin protein

Based on a semi-empirical method through making a statistics of appearance frequency of each amino acid in known segmental epitopes, we predicted 23 antigenic sites of the α-fodrin 560 protein and found that they were clustering in the N-terminus and the middle (Fig. 1).

Fig. 1.

The possible antigenic sites of α-fodrin 560. We have predicted 23 antigenic sites of the α-fodrin protein, and found they cluster in the N-terminus and the middle from their antigenic propensity. Predictions are based on a table that reflects the occurrence of amino acid residues in experimentally known segmental epitopes. Segments are reported only if they have a minimum size of eight residues. The reported accuracy of method is aproximately 75%.

Expression and purification of recombinant overlapping proteins

The DNA gel electrophoresis showing the expected PCR amplifications of α-fodrin fragments is presented in Fig. 2a. The recombinant proteins were expressed in Escherichia coli BL21 and purified by affinity chromatography, thrombin digestion and gel filtration by Sephacryl S100 column. As shown by SDS-PAGE, glutathione S-transferase (GST)-fused α-fodrin N-terminal fragments amino acids 1–59, 1–128, 1–167, 1–238, 1–284, 1–360, 1–423 and 1–485 were prepared with high purity (> 95%) (Fig. 2b).

Fig. 2.

The fragments of a series of α-fodrin cDNA (a) constructs amplified by polymerase chain reaction (PCR). M. 100 base pairs (bp) DNA ladder marker. 1–8. The eight overlapped cDNA segments of α-fodrin. These fragments were prepared from the 1785 cDNA including 1680 cDNA of α-fodrin using specific primer pairs, including 1–1560 (1), 1–1374 (2), 1–1185 (3), 1–957 (4), 1–819 (5), 1–606 (6), 1–489 (7) and 1–282 bp (8). Sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS-PAGE) (b) analysis of expression and purification of the eight overlapping protein segments of α-fodrin 560. 1–8. The purified recombinant α-fodrin overlapping proteins, which encompassed aa 1–485 (1), 1–423 (2), 1–360 (3), 1–284 (4), 1–238 (5), 1–167 (6), 1–128 (7) and 1–59 (8), respectively, of α-fodrin 560. Immunoassays (c) for the recombinant eight recombined overlapping fusion proteins of α-fodrin 560 by Western blot. 1–8. The result of Western blot of eight recombined overlapping fusion proteins of α-fodrin 560 on a mixture of five anti-α-fodrin antibody-positive pSS sera. G. The result of Western blot of glutathione S-transferase (GST) protein on the same mixture of five anti-α-fodrin antibody-positive pSS sera. N. The result of Western blot of the recombined overlapped fusion protein segments mixture on a mixture of five anti-α-fodrin antibody-negative primary Sjögren's syndrome (pSS) sera.

Immunoresponses of the recombinant overlapping proteins to pSS sera and the determination of the epitope region

Figure 2c showed that all recombinant overlapping proteins had strong immunoresponses to the mixture of five anti-α-fodrin antibody-positive pSS sera. The major immunoreactive bands appeared close to the respective molecular weight of the overlapping proteins of the theoretical estimation. Because similar immunoreactivities were observed in all the eight overlapping protein recombinants, they might share a similar epitope site, possibly within the N-terminal 59 amino acids (N59) of α-fodrin 560, which are included in all other recombinant fragments.

Moreover, the protein segment of α-fodrin, except the possibly shared epitope amino acids (1–59 aa), was expressed and detected with Western blot above (Fig. 3a–c). The results showed that the recombinant protein segment had no immunoresponses to the same serum mixture samples as above; the antigenic determinant region of α-fodrin 560 was then recognized from codons 1–59 aa.

Fig. 3.

Polymerase chain reaction (PCR) amplification (a) of the gene of α-fodrin except the possibly shared epitope gene [1–282 base pairs (bp)]. 1. 100 bp DNA ladder marker. 2. The PCR DNA of part of the α-fodrin except the possibly shared epitope gene (1–282 bp). Sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS-PAGE) analysis (b) of expression and purification of the protein segment of α-fodrin except the possibly shared epitope amino acids (1–59 aa). 1. Protein ladder. 2. The purified recombinant fusion protein segment of α-fodrin except the possibly shared epitope amino acids (1–59 aa). Immunoassays (c) for the recombinant fusion protein segment of α-fodrin except the possibly shared epitope amino acids (1–59 aa) by Western blot using the same mixture of five anti-α-fodrin antibody-positive primary Sjögren's syndrome (pSS) sera above as primary antibody. 1. The recombinant fusion protein segment of α-fodrin except the possibly shared epitope amino acids (1–59 aa) on the same mixture of five anti-α-fodrin antibody-positive pSS sera. 2. The eighth recombined overlapping protein segment (1–59 aa) on the same mixture of five anti-α-fodrin antibody-positive pSS sera.

Screening the antigenic peptide at epitope region by ELISA

On the basis of hydropathy plot (data not shown), five peptides located in this region (1–59 aa) were selected and synthesized. Table 1 summarizes the ELISA data from the screening of the five peptides; two peptides, N25 and N46, have comparable immunoreactivities with the α-fodrin 560, with positive detection rates of 83% and 100%, respectively.

Table 1.

Statistics of enzyme-linked immunosorbent assay (ELISA) reactivity of the five synthesized peptides in primary Sjögren's syndrome (pSS) sera.

Peptides	Sequence	Position	OD values mean (s.d.)	P-values*
α-fodrin		1–560	0.73 (0.25)
1–59	MDPSGVKVLETAEDIQERRQQVLDRYHRFK ELSTLRRQKLEDSYRFQFFQRDAEELEKW	1–59	0.77 (0.37)	n.s.
N1	MDPSGVKVLETAE	1–13	0.05 (0.02)	< 0.01
N16	QERRQQVLDRYHR	16–28	0.10 (0.04)	< 0.01
N25	RYHRFKELSTLRR	25–37	0.71 (0.21)	n.s.
N38	QKLEDSYRFQFFQ	38–50	0.15 (0.05)	< 0.01
N46	FQFFQRDAEELEKW	46–59	0.78 (0.16)	n.s.

^*The significance of the differences (P) in ELISA reactivities to α-fodrin and the different synthesized peptides was analysed by Student's t-test.; n.s.: not significant.

Moreover, the antibodies affinity purified against the N46 peptide recognized the α-fodrin 560 in Western blot analysis. Such reactivity was not observed in other peptides and the recombinant protein segment of α-fodrin with the exception of the possibly shared epitope amino acids (1–59 aa) (data not shown). These findings suggest that the N46 peptide was the epitope with strongest antigenicity.

Prevalence of antibodies to α-fodrin antigenic peptide (N46) in pSS

As shown in Table 2, the IgG anti-α-fodrin-N46 peptide antibodies (α-N46PA) were present in 106 of 135 (78.5%)patients with pSS, 13 of 25 (52%) with secondary SS, five of 48 (10.4%) with SLE, 19 of 88 (21.6%) with RA and five of 83 (6.0%) with normal blood donors. The prevalence of α-N46PA in pSS was much higher than that in SLE, RA and normal controls (P < 0.01 in each case).

Table 2.

Prevalence of α-N46PA in patients with Sjögren's syndrome (SS), rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE).

	Patients (n)	α-N46PA Positive no (%)	OD values mean (SD)
Primary SS	135	106 (78.5)	0.41 (0.08)
Secondary SS	25	13 (52)	0.43 (0.09)
SLE	48	5 (10.4)	0.37 (0.09)
RA	88	19 (21.6)	0.39 (0.09)
Healthy controls	83	5 (6.0)	0.11 (0.05)

In comparison to the concentrations of α-N46PA in SS, SLE and RA, no significant difference was found in terms of OD values (Table 2). However, the levels of α-N46PA in the patient groups was much higher than in healthy controls (P < 0.01 in each case), suggesting a role of α-N46PA in these autoimmune disorders.

Comparison of α-N46PA and other autoantibodies in pSS

The sensitivity of α-N46PA in pSS was much higher than anti-SSA, anti-SSB, ANA and anti-α-fodrin antibodies (Table 3). The specificity of α-N46PA (86.8%) was similar to that of anti-SSB (89.0%) and anti-α-fodrin (84.5%), and was higher than anti-SSA (67.3%) and ANA (45.0%) in the diagnosis of pSS. Taken together, α-N46PA appeared to be a more valuable serological parameter than antibodies to SSA, SSB and ANA in SS in terms of sensitivity and specificity.

Table 3.

Sensitivity and specificity of α-N46PA in 135 patients with primary Sjögren's syndrome (SS).

Antibodies	Positive (no.)	Sensitivity (%)	Specificity (%)
α-N46PA	106	78.5	86.8
Anti-α-fodrin	82	60.7	84.5
Anti-SSA	71	52.6	67.3
Anti-SSB	32	23.7	89.0
ANA	52	38.5	45.0

ANA: anti-nuclear antibodies.

Discussion

The fodrin subunit is cleaved in association with apoptosis, and the 120-kDa fragment (α-fodrin) is a breakdown product of the α subunit. Several studies have suggested that the anti-α-fodrin antibodies to 120-kDa cleaved α-fodrin as a disease marker of SS, because it exhibited high sensitivity and specificity [2,4–9]. Moreover, that the antigenic α-fodrin fragments were expressed in abundance in glandular epithelial cells may perpetuate the autoimmune response by revealing previously ‘cryptic’ epitopes, resulting in the production of autoantibodies to α-fodrin and T cell reactivities to epithelial cells leading to glandular destruction. Therefore, structural analysis is required for expression of the antigenic epitope recognized by α-fodrin antibodies in pSS sera and may provide a more sensitive and specific diagnostic serological marker and new strategies for specific therapies, such as vaccination with analogue peptides.

In our study, the overlapping recombinant α-fodrin protein fragments covering antigenic epitopes and the synthesized peptides derived from the α-fodrin protein were screened in sera from pSS patients. To avoid the inadequacy of the computer to predict all antigenic epitope positions, especially the antigenic epitope positions consistent with complex things (eg. protein conformation; protein flexibility), the overlapped recombinant proteins were used to apply the computer analysis to the epitope of α-fodrin. Synthesized peptides containing the corresponding predicted antigenic sites two peptides, N25 (codons 25–37) and N46 (codons 46–59), showed strong immunoreactivities.

In order to analyse further the significance of the strongest antigenic peptide antibody in pSS, we used the N46 peptide as the antigen and developed an ELISA for the detection of antibodies against the N46 peptide. Our results suggest that anti-α-fodrin-N46 peptide antibody (α-N46PA) is a serological marker in pSS. It is more sensitive and specific than ANA, anti-SSA and anti-SSB. Our study confirms that the α-N46PA test is potentially an additional diagnostic tool for pSS where anti-SSA/SSB antibodies are detected less frequently (data not shown).

The mechanism of α-fodrin production has been well studied recently. It has been shown that the majority of antigens targeted by autoantibodies in the sera of patients with SS are susceptible to granzyme B. Generation of unique fragments by granzyme B is a universal feature of autoantigens [15]. Alpha-fodrin is cleaved by granzyme B. It is possible that the antigenic peptide used in this study is a critical antigenic region involved in the immune responses in pSS. We speculate that cleavage and altered distribution of α-fodrin in glandular epithelial cells may induce secretary function impairment and perpetuate an autoimmune response to α-fodrin, leading to autoantibody production and glandular destruction in pSS. Further studies exploring mechanisms for the structural modification of α-fodrin in lachrymal and salivary gland epithelial cells may provide a clue to the pathogenesis of pSS, and further investigation into these issues is needed.

It is not yet known whether α-N46PA can be considered as a marker of pSS pathogenesis. It will be of interest to perform a sequential follow-up study of the level of α-N46PA in order to evaluate whether it is correlated with the clinical course or the therapeutic response. Although the pathophysiological role of the N46 peptide remains to be determined, this new antigenic target appears to be implicated in pSS. Further studies are needed to determine whether this new antibody plays a specific role in pSS and, if so, whether it is a marker of pSS pathogenesis.