CpG oligodeoxynucleotide stimulates production of anti-neutrophil cytoplasmic antibodies in ANCA associated vasculitis

Abstract

Background: Wegener's Granulomatosis and Microscopic Polyangiitis are life-threatening systemic necrotizing vasculitides of unknown aetiology. The appearance of circulating antibodies to neutrophil cytoplasmic antigens (ANCA) is strongly associated with the development of the disease. A link between infection and disease has long been suspected, and the appearance of ANCA antibodies has been reported following bacterial and viral infections. The depletion of circulating B cells with monoclonal antibody therapy can induce remission, and this observation suggests a pathogenic role for B cells in this disease. As bacterial DNA is known to induce B cell proliferation and antibody production via TLR-9 stimulation, we have explored the possibility that unmethylated CpG oligodeoxynucleotide, as found in bacterial and viral DNA, may play a role in stimulating circulating autoreactive B cells to produce ANCA in patients with vasculitis.

Results: We have confirmed that unmethylated CpG oligonucleotide is a potent stimulator of antibody production by PBMC in vitro. The stimulation of PBMC with CpG oligonucleutides resulted in the production of similar amounts of IgG in both ANCA+ patients and normal controls. In spite of this, PR3 ANCA+ patients synthesised significantly higher amount of IgG ANCA than normal controls. In MPO ANCA+ patients, there was a tendency for patients to produce higher amount of ANCA than controls, however, the difference did not reach significance. Furthermore, we were able to detect circulating MPO-reactive B cells by ELISpot assay from the peripheral blood of 2 MPO+ ANCA vasculitis patients. Together, this indicates that circulating anti-neutrophil autoreactive B cells are present in ANCA+ vasculitis patients, and they are capable of producing antibodies in response to CpG stimulation. Of note, CpG also induced the production of the relevant autoantibodies in patients with other types of autoimmune diseases.

Conclusion: Circulating ANCA autoreactive B cells are present in patients with ANCA+ vasculitis. The production of ANCA from these cells in response to unmethylated CpG stimulation lead us to propose that stimulation of these cells by immunostimulatory DNA sequences such as CpG oligodeoxynucleotide during infection may provide a link between infection and ANCA associated vasculitis. This phenomenon may also apply to other antibody mediated autoimmune diseases.

Figure 1

CpG-B induces ANCA production from vasculitis patients in vitro. PBMCs isolated from patients with active ANCA associated vasculitis, 5 PR3⁺ and 5 MPO⁺ ANCA patients, were cultured with CpG-B and IL-2. Each patient assay was paired with a healthy control. After 12 days of culture, supernatants were harvested. IgG concentration and supernatant reactivity to either PR3 or MPO was measured by ELISA. The amount of IgG detected in the supernatants was of 5.5 ± 2.2 μg mL^-1 in the patients compared to 4.1 ± 1.2 μg mL^-1 in the control group (A). Figure B shows the reactivity of the supernatants from PR3⁺ ANCA patients towards PR3 antigen. The difference against control individuals was highly significant (P = 0.0082). Figure C shows the reactivity of the supernatants from MPO⁺ ANCA patients towards MPO antigen. The difference was not significant (P = 0.072) although their supernatants showed a clear tendency towards higher reactivity compared to controls. There was no correlation between patients' serum ANCA titre at the time of the assay and their in vitro production of ANCA in response to CpG-B as shown in D (r² = 0.172).

Figure 2

In vitro ANCA production is induced by CpG-B but not by other B cell stimulants. In addition to CpG-B + IL2, PBMCs from 5 different pairs were also stimulated with either IL2 alone, LPS + IL2, pokeweed mitogen (PWM) + IL2, or inactivated staphylococcus aureus + IL2. A representative set of results is shown from a PR3⁺ ANCA patient (A) and a MPO⁺ ANCA patient (B).

Figure 3

Detection of circulating B cells capable of producing ANCA in response to CpG-B. PBMCs from 2 MPO⁺ ANCA vasculitis patients were cultured with CpG-B and IL-2. These PBMCs had not undergone enrichment for B cells prior to culture. After 5 days culture, cells were transferred into ELISpot wells which had been coated with either myeloperoxidase (MPO) in duplicates or foetal calf serum (FCS) as a control antigen. After overnight culture, IgG antibody producing cells against these antigens were detected by anti-human IgG conjugate. The total number of IgG producing B cells was measured by coating the wells with polyclonal anti-human IgG. The results from a patient as shown in this figure are representative of results from 2 patients. Fig A shows ELIspot plate with total IgG producing cells in the first column followed by the detection of anti-MPO B cells in duplicates in the middle columns and finally cells against the control antigen. The numbers of spots counted are depicted in Figure B. In spite of both patient and control having similar number of IgG producing cells, the number of anti-MPO B cells is higher in the MPO⁺ patient. This result coincide with those from a parallel experiment where PBMCs from this pair of individuals were cultured in the presence of CpG-B to measure their in vitro production of anti-MPO by ELISA as shown in (C).

Figure 4

CpG-B also induced the production of relevant IgG autoantibodies in patients with other autoimmune diseases. PBMCs from 4 patients with autoimmune thyroiditis and 3 patients with anti-phospholipid antibody syndrome were cultured with CpG-B and IL-2. Each patient assay was paired with a healthy control. After 12 days of culture, supernatants from autoimmune thyroiditis patients were tested by ELISA for IgG anti-thyroperoxidase (TPO) as shown in A. Supernatants from anti-phospholipid antibody syndrome patients were tested for IgG anti-cardiolipin antibodies (ACLA) as shown in B.