Introduction
Systemic lupus erythematosus (SLE) is a chronic, usually life-long, potentially fatal autoimmune disease characterised by an increased production of autoantibodies, impairment of B- and T-cell functions, cytokine production, and immune complex deposition. SLE is manifested, for example, in neurological, dermal, haematological, musculoskeletal, and renal symptoms [
1]. Central nervous system (CNS) involvement has been reported to occur in 14% to 75% of patients with SLE and is a major factor contributing to morbidity and mortality in patients [
2]. The aetiology of neuropsychiatric SLE (NPSLE) includes autoantibody production specific for brain structures, immune complex depositions, microangiopathy, and intrathecal production of proinflammatory cytokines. Seizures, stroke, depression, psychoses, and disordered mentions are manifestations of this disease [
3]. Beneficial treatment in the form of cytotoxic drugs is available [
4] but requires early recognition of CNS involvement. Due to the multiple pathogenic mechanisms causing NPSLE, there is no single confirmatory diagnostic test for NPSLE. Several clinical, laboratory, and radiographic test findings are reported to be abnormal in some but not all patients. Magnetic resonance imaging (MRI) of the brain has been shown to be valuable in detecting even minor NPSLE-induced lesions [
5]. Pleocytosis and elevated protein levels are found in some but not all NPSLE patients. Elevated concentrations of IgG in cerebrospinal fluid (CSF) IgG-albumin ratio, IgG index, and the presence of oligoclonal bands have all been described with varying frequencies. Increased levels of interleukin (IL)-1 [
6], IL-6 [
6,
7], IL-8 [
8], and interferon-gamma (IFN-γ) [
9] have been found in CSF of NPSLE patients. We have reported earlier that patients with NPSLE displayed elevated CSF levels of matrix metalloprotease-9 well as intrathecal neurofilament (NFL) and glial fibrillary acidic protein (GFP) [
10], which are markers for neuronal and astroglial brain damage.
The tumour necrosis factor (TNF) family ligands BAFF (B-cell activating factor of TNF family) and APRIL (a proliferation-inducing ligand) are implicated in several immunological phenomena such as peripheral B-cell survival, CD40L-independent antibody production and isotype switching, autoimmunity as well as tumour cell growth [
11,
12]. BAFF is expressed on the cell surface or cleaved and secreted [
12], while APRIL is cleaved from the Golgi and solely exists as a secreted soluble ligand [
13]. BAFF and APRIL share two receptors: B-cell maturation antigen (BCMA) and transmembrane activator and calcium-modulating cyclophilin ligand interactor (TACI), which are found mainly on B cells and plasma cells [
14]. In addition, BAFF binds to BAFF receptor found mainly on B cells, plasma cells, and some subsets of T cells [
15,
16], while APRIL interacts with heparin sulfate proteoglycans, which likely constitutes a third receptor for APRIL [
17].
In the context of autoimmunity, both BAFF and APRIL are implicated in the establishment and/or maintenance of autoimmune disease. Abnormal serum levels of BAFF and APRIL have been observed in patients with rheumatoid arthritis [
18], Sjögren syndrome [
19], and SLE [
20]. In SLE patients, increased serum levels of BAFF, APRIL, and BAFF/APRIL heterotrimers correlate with anti-double-stranded DNA autoantibodies and disease activity [
21]. Gene polymorphism of APRIL has been reported to be associated with SLE [
22]. The association between enhanced levels of BAFF and autoimmune disease in humans has been substantiated in mice rendered transgenic or deficient for this cytokine. Mice overexpressing BAFF develop a lupus-like phenotype characterised by high titres of anti-DNA antibodies, hypergammaglobulinaemia, and glomerulonephritis [
23], while mice lacking BAFF are deficient in mature B cells and marginal zone B cells [
24]. The association between elevated circulating levels of BAFF and polyclonal hypergammaglobulinaemia extends to humans as well. Increased serum and/or plasma levels of BAFF have been documented in SLE, rheumatoid arthritis, and Sjögren syndrome [
25‐
27], all conditions associated with polyclonal hypergammaglobulinaemia. Overexpression of APRIL, in contrast, has not been associated with autoimmunity in mice but leads to enhanced IgM production, T-cell-independent type 2 humoral responses, and T-cell survival [
28], while lack of APRIL is associated with enhanced numbers of effector/memory T cells and impaired IgA responses [
29,
30]. However, APRIL and BAFF/APRIL heterotrimers have been found to be elevated in sera and target organs of autoimmune disease patients, including SLE, Sjögren syndrome, multiple sclerosis (MS), and myasthenia gravis [
31‐
34]. The purpose of this study was to examine BAFF and APRIL levels in the CSF of SLE patients with or without NPSLE in order to investigate whether BAFF and/or APRIL could have a role in the pathology of NPSLE.
Discussion
To our knowledge this study is the first to show that SLE patients have elevated levels of APRIL and BAFF in CSF. SLE patients displayed levels of APRIL in CSF that were 24-fold higher than those of healthy controls, and levels of BAFF that were 200-fold higher, thus suggesting high intrathecal levels of APRIL and BAFF to be a feature of SLE. Although increased levels of both BAFF and APRIL have been reported earlier [
27,
41], these cytokines were measured in serum, where the differences between SLE patients and healthy controls were more modest. Due to the greater strictness of the NPSLE criteria in comparison with the 1987 American College of Rheumatology Case definitions for SLE [
35] that we have previously used [
10] and that we also employed in this study, patients with mild NPSLE could potentially be grouped with SLE patients without CNS inflammation. Nevertheless, we were able to find significant differences in APRIL levels but not in BAFF levels between SLE patients and NPSLE patients.
Mechanisms associated with the pathogenesis of NPSLE include anti-neuronal antibodies, anti-phospholipid antibody-associated thrombosis, and (rarely) vasculitis by immune complex depositions. Thus, pathogenic antibody formation is strongly associated with the manifestations of NPSLE. We found elevated levels of the antibody-inducing cytokine APRIL in the CNS of NPSLE patients compared with SLE patients without CNS involvement. Elevated levels of IL-6 in CSF have been reported in NPSLE patients [
40] and could be confirmed in the present patient material. Since BAFF, APRIL, and IL-6 are important players in the survival, differentiation, and isotype switching of B cells, they may have an important role in the aetiology of NPSLE.
Analyses of covariance revealed no correlation between serum and CSF levels of BAFF or APRIL, suggesting that APRIL and BAFF in CSF are produced locally instead of systemically and subsequently passed to the CSF. A potential source of APRIL and BAFF in the CNS are the astrocytes of the brain. In MS patients, astrocytes have been established as producers of both BAFF and APRIL. APRIL in the CNS was expressed only by reactive astrocytes and increased in MS lesions [
42]. Similarly, the transcript level of BAFF was 10-fold higher in MS lesions compared with normal CNS [
43]. Collectively, this suggests that astrocytes do not constitutively express high levels of APRIL and BAFF, but rather upregulate the expression of these cytokines upon inflammation. BAFF and APRIL are expressed by a number of other cell types, including monocytes, macrophages, dendritic cells (DCs), neutrophils, and T cells [
44]. Although the CNS was initially considered an immunologically privileged site, one of the primary events that occur during CNS affecting autoimmune disease such as MS is the recruitment of immunocompetent lymphocytes to the brain [
45]. In lupus-prone MRL-Ipr mice, T cells pass the blood-brain barrier to infiltrate the brain [
46,
47]. Even though T-cell infiltration in the CNS of SLE patients has not been firmly established, T cells cannot be ruled out as a source of intrathecal BAFF and APRIL in SLE patients.
Analyses of covariance between APRIL and BAFF levels in the CSF of SLE patients suggested that BAFF and APRIL covary and thus are regulated together. IFN-α, IFN-γ, IL-10, and CD40L are important in the upregulation of both BAFF and APRIL [
48,
49]. SLE patients have large increases in the levels of IFN-α, IL-10, and soluble CD40L. Furthermore, T and B cells from SLE patients express highly increased levels of CD40L [
50‐
52]. IFN-α, IL-10, and CD40L therefore could account for the upregulation of BAFF and APRIL seen in SLE patients. How BAFF and APRIL production in the CNS is regulated is currently not known but warrants further investigation.
SLE has been described as a disease of B-cell hyperactivity [
53]. A number of studies concur in showing that DCs play a major role in B-cell development, mostly through the production of cytokines such as BAFF, IL-12, IL-6, and IFN-α [
54]. CD40L treatment of bone-marrow-derived DCs from lupus-prone B6.TC mice induced higher production of IL-6, IL-10, and TNF-α than in B6 mice [
55]. In an attempt to investigate whether IL-6, BAFF, and APRIL are regulated together, perhaps through the involvement of DCs, we analysed covariance between IL-6 and APRIL or BAFF. Although IL-6 levels were increased in the CSF of NPSLE patients, we could not correlate IL-6 to APRIL or BAFF levels. IL-6 is not known to induce BAFF and APRIL. In the development of humoral immunity, BAFF particularly supports the survival of immature transitional and mature B cells [
24]. IL-6, on the other hand, promotes the transition of plasmablasts to plasma cells [
56]. BAFF, APRIL, and IL-6 therefore likely have complementary roles in the autoantibody formation seen in SLE.
BAFF and APRIL are known to form homotrimers [
57,
58]. However, it has been shown that BAFF and APRIL can associate with each other to form a heterotrimer capable of stimulating B-cell proliferation. The levels of heterotrimers of BAFF/APRIL are nondetectable in the serum of healthy controls, whereas they are increased in patients with autoimmune diseases such as SLE [
59]. We do not know whether the antibodies used in our BAFF- and APRIL-specific ELISAs recognise only monomeric and homotrimeric molecules of BAFF and APRIL or whether they are also able to recognise heterotrimers of BAFF/APRIL. If CSF levels of BAFF/APRIL heterotrimers are increased in the CSF similarly to the levels seen in serum, and the BAFF- and APRIL-specific antibodies used in our study are unable to recognise heterotrimers, we may have underestimated the levels of BAFF and APRIL in CSF. Regardless of this possibility, we did find elevated levels of BAFF and APRIL in the CSF of SLE patients.
Evidence suggests a major contribution of B cells to the pathogenesis of NPSLE, an effect supported by the salutary effect of the B-cell-depleting monoclonal antibody rituximab in this condition [
60]. BAFF receptors and CD20 overlap in many B-cell subsets, but they differ in certain populations such as plasma cells, which express BCMA but not CD20. Blockage of BAFF therefore might provide distinct advantages over B-cell depletion therapy that targets CD20 (that is, rituximab). Another concern is that rituximab treatment as a side effect causes elevated BAFF levels in serum [
61]. Therefore, upon cessation of treatment, newly generated immature cells could be exposed to high levels of BAFF, causing a resurgence of autoimmunity. This, in theory, could be averted by targeting BAFF. The potentially beneficial effects of BAFF and/or APRIL blockage are underscored in experimental autoimmune encephalomyelitis (a mouse model of MS). Besides demyelating disease, these mice have elevated immunoglobulin levels and spontaneously secrete autoantibodies to DNA, which eventually leads to autoimmune manifestations similar to SLE. Treatment with a BAFF/APRIL antagonist (soluble BCMA-Fc) is able to inhibit these autoimmune manifestations [
62].
Since BAFF and APRIL have a major impact on B-cell survival, T-cell function, and antibody production [
11,
12], overexpression of these cytokines could contribute to increased lymphocyte survival and proliferation, potentially leading to increased leucocyte infiltration into the CNS. Moreover, upregulation of BAFF expression
in vivo can result in the rescue of self-reactive B cells from elimination [
63]. Collectively, these findings suggest that treatment with BAFF and/or APRIL antagonists could be beneficial in SLE treatment. A phase II study in rheumatoid arthritis and SLE patients using LymphoStat-B, a fully humanised antibody specific for human BAFF (also known as Belimumab), showed modest efficacy in rheumatoid arthritis [
64], while a phase II study in SLE patients did not meet the primary efficacy endpoints [
65]. The decoy receptor TACI-Ig (Atacicept) preventing the binding of BAFF and APRIL to the receptor TACI on B cells led to improvements in animal models of lupus [
66] and arthritis [
67]. Its clinical value is currently under investigation in rheumatoid arthritis and SLE.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
KE designed the study, was involved in the management of the study, and helped to interpret the results. AG-C was involved in the management of the study, performed the laboratory investigations, analysed the data statistically, helped to interpret the results, and prepared the manuscript. ET was involved in the management of the study, collected the clinical data, and helped to interpret the results. AG-C and ET contributed equally to this work. All authors read and approved the final manuscript.