Identification of new autoantibody specificities directed at proteins involved in the transforming growth factor β pathway in patients with systemic sclerosis

Systemic sclerosis (SSc) is a connective tissue disorder characterised by excessive collagen deposition in the dermis and internal organs, vascular hyperreactivity and obliteration phenomena [1]. A large number of autoantibodies have been identified in the sera of SSc patients. Antinuclear antibodies (ANAs), usually detected by indirect immunofluorescence on HEp-2 cells, are identified in 90% of patients [2]. Some of them are disease-specific and mutually exclusive: anticentromere antibodies (ACAs), associated with limited cutaneous SSc (lcSSc) and possibly pulmonary arterial hypertension (PAH); anti-topoisomerase I antibodies (ATAs), associated with diffuse cutaneous SSc (dcSSc) and interstitial lung disease (ILD); and anti-RNA polymerase III antibodies, associated with dcSSc and scleroderma renal crisis (SRC) [3]. In addition, other autoantibodies have been found in the sera of SSc patients and include antifibrillarin, antifibrillin 1, anti-Th/To, anti-PM/Scl [3], antifibroblast [4‐6] and anti-endothelial cell antibodies [7‐9]. Overall, the only specific autoantibodies routinely tested for in SSc patients are ACAs, ATAs and, more recently, anti-RNA polymerase III antibodies.

Thus, approximately 10% of SSc patients have no routinely detectable autoantibodies, and for 20% to 40% of those with detectable ANAs, the nuclear target antigens of these ANAs have not been identified [2]. Therefore, further work is warranted to better determine the disease subset and prognosis for these patients. The specification of new autoantibodies could help in understanding the pathophysiology of SSc and reveal new diagnostic and/or prognostic markers.

Using a proteomic approach combining two-dimensional electrophoresis (2-DE) and immunoblotting, we recently identified target antigens of antifibroblast antibodies in patients with PAH [10]. In this work, using a similar proteomic approach with total and enriched nuclear protein extracts of HEp-2 cells as sources of autoantigens, we systematically analysed autoantibodies in SSc patients and identified a number of new target antigens for these autoantibodies.

In the present work, we have identified a number of new target antigens for autoantibodies in SSc patients that are either shared among patients or specific to a given phenotype. For some antigens, including TPI, SOD2, hnRNP L and lamin A/C, IgG reactivity was higher in sera pools from patients than in pools from HCs. TPI, a glycolytic enzyme localised in the cytoplasm, is one of the nine proteins specifically identified in whole saliva from patients with dcSSc as compared with HCs [17]. Interestingly, we recently identified another glycolytic enzyme, α-enolase, as a target of antifibroblast antibodies in SSc patients, particularly those with ILD and/or ATAs [18, 19]. SOD2 is a mitochondrial metalloenzyme that catalyses the dismutation of the superoxide anion to hydrogen peroxide and oxygen and protects against reactive oxygen species (ROS). Thus, autoantibodies directed against SOD2 might impair the enzyme function and favour ROS accumulation. This finding could be relevant to the pathogenesis of SSc, because a major increase in ROS level is a hallmark of SSc [20]. Interestingly, Dalpke et al. [21] reported that a hyperimmune serum against SOD2 inhibited the protective effects of SOD2 on endothelial cells exposed to oxidative stress. In addition, downregulation of SOD2 expression was described in osteoarthritis [22], and anti-TPI antibodies have been identified in several autoimmune conditions, including neuropsychiatric systemic lupus erythematosus (SLE) [23], and in osteoarthritis [24].

Lamins A and C are both encoded by the LMNA gene and represent major constituents of the inner nuclear membrane. Mutations of this gene have been identified in a number of conditions, including Hutchinson-Gilford progeria syndrome [25], which represents a major differential diagnosis of juvenile SSc. The most frequent mutation responsible for progeria creates a truncated progeria mutant lamin A (progerin), which accumulates within the nuclei of human vascular cells and may be directly responsible for vascular involvement in progeria [26]. The identification of lamin as a major target of autoantibodies in SSc patients precludes the potential role of modified and/or dysfunctional lamin and/or antilamin autoantibodies in the pathogenesis of SSc. Antilamin antibodies were found in sera from patients with SLE [27] and antiphospholipid syndrome [28] as well as in a patient with linear morphea [29].

HnRNP L is a nuclear protein associated with hnRNP complexes and takes part in the processing of pre-mRNA. Anti-hnRNP L antibodies were identified in a small cohort of SSc patients in association with anti-hnRNP A/B antibodies [30]. HnRNP L was also identified as a target of autoantibodies in New Zealand White × BXSB mice with SLE and antiphospholipid syndrome [31].

Our analysis revealed that PRDX2, cofilin 1 and calreticulin were specifically recognised by IgG from phenotypic subsets of patients with unidentified ANAs. Other target antigens listed in Table 4 might also be relevant and should be tested in further work. PRDX2 is a peroxidase that eliminates endogenous ROS produced in response to growth factors such as platelet-derived growth factor (PDGF). PRDX2 influences oxidative and heat stress resistance [32] and inhibits PDGF signalling and vascular remodelling [33]. Interestingly, PRDX2 has recently been identified as a target of anti-endothelial cell antibodies in systemic vasculitis [34].

Cofilin 1 is a regulator of actin depolymerisation. Cofilin is a major effector of nicotinamide adenine dinucleotide phosphate (NADPH) oxidase 1-mediated migration, and NADPH oxidase 1 plays a critical role in neointima formation by mediating vascular smooth muscle cell migration, proliferation and extracellular matrix production [35]. Moreover, regulation of the phosphorylation state of cofilin controls PDGF-induced migration of human aortic smooth muscle cells [36]. Anti-cofilin 1 antibodies have been detected in a few patients with rheumatoid arthritis, SLE or polymyositis and/or dermatomyositis [37].

Calreticulin is an endoplasmic reticulum chaperone and an intracellular calcium-binding protein and thus is involved in signal transduction pathways. In apoptotic cells, calreticulin is translocated to the cell surface, conferring immunogenicity of cell death [38]. Calreticulin has been described as a potential cell surface receptor involved in cell penetration of anti-DNA antibodies in patients with SLE [39]. Anticalreticulin antibodies have been reported in patients with celiac disease and SLE [40, 41].

Interestingly, we determined that several autoantigens recognised by IgG from SSc patients were involved in the TGF-β pathway. In the pathophysiology of SSc, fibroblast proliferation and accumulation of extracellular matrix result from uncontrolled activation of the TGF-β pathway and from excess synthesis of connective tissue growth factor, PDGF, proinflammatory cytokines and ROS [3]. Thus, increased expression and/or modified structure or fragmentation in the presence of ROS of a number of proteins involved in the TGF-β pathway could trigger specific immune responses in these patients. Casciola-Rosen et al. [42] reported on the sensitivity of scleroderma antigens to ROS-induced fragmentation in this setting, possibly through ischemia-reperfusion injury as the potential initiator of the autoimmune process in SSc.

The combined use of 2-DE and immunoblotting offers an interesting approach to identifying target antigens of autoantibodies [10, 13]. We used HEp-2 cells as sources of autoantigens because these cells are routinely used to detect ANAs. Although not directly relevant to the pathogenesis of SSc, we thought it more appropriate to use these cells as sources of autoantigens because we were looking for additional targets to ANAs. Additional validation studies with sera from patients with other connective tissue diseases are necessary. In addition, 2-DE and immunoblotting were not adapted to test a large number of sera, and thus further experiments using ELISA with recombinant proteins are necessary, which will allow for validation of the target antigens and screening of a large number of patients.

However, our work has several additional limitations. Less than 1,000 protein spots were stained in the reference gel of the total protein extract. Therefore, a number of proteins were probably lost at each step of the technique, depending on their charge, molecular weight, subcellular localisation and/or abundance in the cell. Topoisomerase II is not detected by traditional methods of 2-DE [43], and we failed to identify topoisomerase I or centromeric protein B as target antigens of IgG autoantibodies, whereas these antigens are easily detected in 1-D gels [6, 44, 45]. Anti-topoisomerase I and anti-RNA polymerase III antibodies preferentially recognise a discontinuous or conformational epitope that may not be detected in 2-D gels [46, 47]. As expected, none of the identified antigens was located at the cell surface, since protein extraction for 2-DE does not allow the identification of membrane proteins.

We have identified new target autoantigens in SSc patients, a number of which are involved in the TGF-β pathway. Although these data must be confirmed by other groups and in large cohorts of patients with SSc or other connective tissue diseases, these new autoantibody specificities could represent major advances in the diagnosis and prognosis of patients with SSc.