Streptococcal IdeS: therapeutic potential for Guillain–Barré syndrome

Plasma exchange and intravenous immunoglobulin are effective in treating Guillain–Barré syndrome (GBS) probably because the former removes IgG autoantibodies and complement and the latter inhibits complement activation subsequent to the autoantibody binding to peripheral nerve antigens. IgG degrading enzyme of Streptococcus pyogenes (IdeS) can cleave the pathogenic autoantibodies into F(ab’)2 and Fc. The purpose of this study is to show whether IdeS has novel therapeutic potential for GBS. Sera with anti-ganglioside IgG antibodies from 15 patients with GBS or Miller Fisher syndrome were used. We tested whether IdeS cleaved the anti-ganglioside IgG antibodies and inhibited deposition of activated complement component on ELISA plates. IdeS efficiently cleaved IgG and blocked complement activation mediated by anti-GM1, anti-GD1a and anti-GQ1b IgG antibodies. IdeS has therapeutic potential for GBS and related conditions.

Scientific RepoRts | 5:10809 | DOi: 10.1038/srep10809 Results IdeS efficiently cleaved IgG and blocked complement activation mediated by anti-GM1, anti-GD1a and anti-GQ1b IgG antibodies. The established assays demonstrated the binding of autoantibodies to each ganglioside and the deposition of active complement component. The detection of Fc domain was obscured due to cleavage by IdeS and subsequent rinsing out. Nevertheless F(ab') 2 remained to bind stably to the ganglioside coating on the microtiter plates ( Fig. 1A and 2A). The clearance of Fc depended on the concentration of IdeS (Fig. 1B) as well as the time after the addition of IdeS to the serum (Fig. 1C). The cleaving effect emerged in a few minutes and reached the maximum in one hour. IdeS cleaved all the anti-GM1, anti-GD1a and anti-GQ1b IgG antibodies (Fig. 1D). Fc deposition was degraded by IdeS (10 μ g/ml), whereas F(ab') 2 deposition remained unaltered. Thus, the subsequent complement deposition mediated by anti-ganglioside IgG autoantibodies was inhibited by IdeS, resulting in the blocking of the C3 deposition (Fig. 2B). The blocking effect depended on the concentration of IdeS as well as the clearance of Fc (Fig. 2C). In contrast, IdeS did not affect the binding of anti-GM1 IgM antibodies (Fig. 1E).

Discussion
Anti-ganglioside antibodies binding followed by complement activation causes nerve injury in the axonal form of GBS [2][3][4] . Removal or scavenging of pathological autoantibodies is thought to be one of the most important mechanisms of plasma exchange and IVIG 6,11 . Complement inhibitors such as eculizumab and nafamostat mesilate prevented nerve injury in animal models of GBS 12,13 . Therefore, therapeutic approaches to decreasing both pathological antibodies and the subsequent deposition of complement by immunotherapies are considered to effectively ameliorate the severity and outcome of GBS in patients.
In the current study, we showed that IdeS successfully cleaved the pathological antibodies and blocked the activation of complement. It is considered that F(ab') 2 domain of autoantibodies binding to ganglioside cannot cause pathogenesis independently in GBS because it lacks Fc domain which binds and activates immune effectors. This suggests that IdeS could suppress the activation of complement on the axolemma of motor fibers at the nodes of Ranvier where pathological anti-ganglioside autoantibodies deposit and thereby prevent nerve injury. We confirmed that IdeS cleaved the Fc domains similarly both before and after anti-ganglioside antibodies bind to ganglioside (data not shown). This suggests that IdeS could cleave circulating anti-ganglioside IgG antibodies before their binding to the antigen and remove the Fc domains of the immune complexes binding to the nerve antigens.
Animal experiments have shown that IdeS treatment is strong enough to prevent or cure the diseases induced by autoantibodies. IdeS treatment of mice with collagen-antibody-induced arthritis reduced the severity of the arthritis when administered within 24 hours of the onset of clinical arthritis and prevented an antibody-induced relapse 14 . In a mouse model of immune thrombocytopenic purpura with IgG antibodies against platelet surface antigens, profoundly thrombocytopenic animals were treated and cured by a single injection of IdeS 15 . Severe albuminuria in mice induced by anti-glomerular basement membrane antibodies was completely prevented by IdeS, accompanied by a significant reduction of the deposition of complement components 16 . In a mouse model of neuromyelitis optica induced by anti-aquaporin-4 antibodies, IdeS treatment greatly reduced pathological lesions 17 . These reports showed that IdeS reduced the pathogenesis caused by antibody-mediated complement-dependent cytotoxicity and antibody-dependent cell-mediated cytotoxicity. Other autoimmune diseases with inflammation induced by pathological autoantibodies could be possible candidates for IdeS treatment.
Anti-IdeS antibodies may be present in humans and may increase in number after therapeutic use of IdeS because IdeS is a heterologous protein secreted by S. pyogenes. Streptokinase, another enzyme secreted by several streptococci has been used as an effective and safe thrombolysis medication by activating plasminogen 18 . Repeated use of IdeS could increase the number of anti-IdeS IgG antibodies, which might neutralize and suppress the effect of IdeS and produce anti-IdeS antibodies, which may cause severe adverse reactions such as hypersensitivity. Therefore, single rather than repeated use of IdeS is advised. From this viewpoint, GBS is one of the most promising candidates for IdeS treatment because GBS is an acute and monophasic disease. Furthermore, IdeS may increase risk of infection in GBS patients because of its characteristic as a virulence factor for microbes. There is another concern that naturally occurring, anti-hinge antibodies may generate complement C3b 2 -containing immune complexes which could stimulate complement amplification 19 . However, a phase I clinical trial in healthy volunteers was safely executed 20 . Based on those results, subjects have been recruited for a phase II trial designed to evaluate the safety and efficacy of IdeS in patients with chronic kidney diseases who are scheduled for kidney transplantation with donor-specific HLA antibodies because they require plasma exchange before the transplantation 21 .
Clinical trials performed in Europe and North America showed plasma exchange is effective for GBS 1 , suggesting that IgG autoantibodies play a crucial role for the development of AIDP, a predominant form of GBS in western countries. IdeS has been developed for patients who require plasma exchange. We believe that IdeS is effective for demyelinating GBS as well, and we are also planning a phase II trial in patients with GBS.  In conclusion, we found that IdeS cleaved the pathological anti-ganglioside antibodies and blocked the subsequent complement activity mediated by autoantibodies. These results suggest that IdeS is an effective treatment for GBS.

Methods
Serum samples. The study included patients with GBS, Miller Fisher syndrome and multifocal motor neuropathy, admitted to Dokkyo Medical University (Tochigi, Japan) who fulfilled the published diagnostic criteria 22,23 . Serum samples were chosen from the patients with GBS associated with anti-GM1 (n = 5) or anti-GD1a IgG antibodies (n = 5); Miller Fisher syndrome associated with anti-GQ1b IgG antibodies (n = 5); or multifocal motor neuropathy associated with anti-GM1 IgM antibodies (n = 5), which had been kept at -80 °C before use. All experiments were performed in accordance with NMEC Ethical Guidelines on Research Involving Human Subjects 24 . Written informed consent was obtained from every patient. Normal human sera were obtained from five healthy subjects as a source of complement. The study was approved by the Dokkyo Medical University Ethics Committee and National University Singapore Medical Research Ethics Committee.
ELISA. Phosphate-buffered saline (PBS) containing 0.5% casein sodium salt was used for each dilution. The serum dilutions were titrated so that the optical density values showed from 2.0 to 3.0 with anti-ganglioside antibody-positive sera. Sera from patients with anti-GM1, anti-GD1a and anti-GQ1b IgG as well as anti-GM1 IgM antibodies were diluted at 1:500 and added to respective ganglioside-coated plates. The microtiter plates were incubated for one hour at 37 °C. To detect IgG binding, peroxidase-conjugated goat anti-human IgG antibodies recognizing the Fc (1:2,000; Sigma-Aldrich, Dorset, UK) or F(ab') 2 domain (1:2,000; Thermo Scientific, Rockford, IL) were used. To detect IgM binding, peroxidase-conjugated rabbit anti-human IgM antibodies (1:2,000; Sigma-Aldrich) were used after washing with PBS containing 0.05% Tween 20 (PBS-T). A complement deposition assay was performed in quintuplicate as described previously with minor modification 25 . Normal human sera diluted at 1:100 were added as a source of complement. After incubation for one hour at 37 °C and subsequent washing with PBS-T, peroxidase-conjugated goat anti-human C3 antibodies (1:2,000; MP Biomedicals LLC, Solon, OH, USA) were added. After further incubation and washing with PBS-T, each plate was developed.

IdeS treatment. IdeS was kindly provided by Hansa Medical AB (Lund, Sweden). Diluted IdeS
in 50 μ l of PBS was used in different concentrations from 1 ng/ml to 100 μ g/ml in individual wells of the ELISA plate to generate a dose-response curve. To investigate whether IdeS cleaves Fc domain of anti-ganglioside antibodies and blocks the deposition of complement mediated by anti-ganglioside antibodies, patients' sera mixed with IdeS were incubated on microtiter plates for 1 hour at 37 °C. The process is represented schematically in Fig. 1A and 2A. The optimal blocking dose of IdeS was determined to be 10 μ g/ml and this dose was used in the sequential assays. PBS, which did not contain IdeS, was used as a control and the relative IgG binding rates and the C3 deposition rates were evaluated respectively.
Statistics. Data were analyzed using JMP statistical discovery software version 10.0.2 (SAS Institute Inc., Cary, NC). Statistical significance was evaluated using the Mann-Whitney U-test. Differences where P < 0.05 were considered significant.