All MG patients were recruited at Huashan Hospital, Fudan University. Ten immunotherapy-naïve generalized MG patients who fulfilled the diagnostic standard (17) with confirmed relatively high titers of AChR-Ab ( Supplementary Table 1 ) and ten sex- and age-matched healthy controls were included in this study. The study protocol was approved by the ethics committee of Huashan Hospital, Fudan University. Peripheral blood samples were obtained after informed consent.

For PTMG and active EAMG studies, female Lewis rats 6–8 weeks of age (weighting 140–160 g) were purchased from the Vital River Laboratory Animal Co. Ltd. (Beijing, China) and housed in ventilated grommet cages (4 animals per cage) in the animal facility of the Institute in a controlled environment (12-h light/dark cycles; 23°C; specific pathogen free) with standard rat chow and water ad libitum. Soft food and water gels were placed on the bottom of the cages when an animal showed signs of weakness to ensure the animals were adequately fed and hydrated (18, 19). The studies were conducted following a protocol approved by the Administrative Panel on Laboratory Animal Care of Fudan University.

Total IgG was purified from sera of MG patients and healthy control using a Melon gel IgG purification kit (Thermo Fisher Scientific, Waltham, MA) and was concentrated using Amicon Ultra centrifugal filter units (Millipore, Billerica, MA) according to the manufacturer’s instructions.

AChR-expressing TE671 cells (American Type Culture Collection, ATCC) were plated onto 96-well microplates at 50,000 cells/well and cultured in DMEM (Gibco BRL, Grand Island, NY) supplemented with 10% heat-inactivated fetal bovine serum (FBS, Gibco, Grand Island, NY), 100 units/ml penicillin, and 100 μg/ml streptomycin (Gibco) at 37°C/5% CO₂ for 24 h. Cells were washed with complement-dependent cytotoxicity (CDC) buffer-Hanks’ buffered saline solution 10 mM HEPES, 1 mM sodium pyruvate, and 10 mg/ml Gentamicin (Invitrogen, Carlsbad, CA), and incubated at 37°C for 60 min with various concentrations of MG-IgG in CDC buffer in the presence or absence of CRIg/FH (60 μg/ml). Pooled normal human complement serum (10%, Innovative Research, Novi, MI) was used as a source of complement and incubated with TE671 cells in a final volume of 50 μl for 30 min at 37°C. Following MG-IgG treatment, TE671 cells were washed in CDC buffer, harvested, and then stained with 1 μl/ml fixable viability dye eFlour 780 (Invitrogen, Carlsbad, CA) for live/dead cell determination for 30 min at 4°C. The percentage of cell death was determined by acquiring flow cytometry data using Attune NxT Flow Cytometer (Thermo Fisher, Ramsey, MN). The percentage of fixable viability dye eFlour 780-positive cells (compared to total cells) was used to evaluate the degree of cell death.

TE671 cells were plated onto coverslips in 48-well plates and cultured at 37°C/5% CO₂ for 24 h, and afterwards were incubated at 37°C for 60 min with MG-IgG (200 μg/ml) in the presence or absence of CRIg/FH (60 μg/ml). Pooled normal human complement serum (10%) was incubated with TE671 cells in blocking buffer (PBS containing 1% bovine serum albumin) for 30 min. Cells were then rinsed in PBS, fixed in 4% paraformaldehyde for 15 min. After fixation, cells were incubated for 2 h with rabbit anti-C5b-9n antibody (Abcam, Cambridge, MA), followed by mouse anti-rabbit IgG-conjugated Alexa Flour 488 and α-Bungarotoxin (BTX)-conjugated Alexa Flour 555 (Invitrogen, Carlsbad, CA) in 1% bovine serum albumin (BSA) for 30 min at room temperature. Following DAPI staining (Invitrogen, Carlsbad, CA) at a 10 μg/ml concentration, fluorescence was imaged on an Olympus fluorescence microscope. The quantitation of immunofluorescence images was performed with ImageJ.

PTMG was induced with anti-AChR mAb35 in female Lewis rats (18). The mAb35 (20 pmol/100 g) was administered i.p. 2 h after initial treatment with CRIg/FH (1 mg/100 g i.p). Rats were monitored for signs of clinical disease and weighed at 24 and 48 h post disease induction. Rats were sacrificed when weight loss exceeded 15% of starting weight; all remaining rats were sacrificed at 48 h. Clinical scores were as follows: 0, no weakness; 1, fatigable or weakness is only observed after exercise; 2, clinical signs of weakness present before exercise, hunched posture, or head down; 3, severe clinical signs of weakness: no ability to grip, hindlimb paralysis, respiratory distress/apnea, weight loss > 15%, immobility or moribund; 4, death (18). Diaphragms were harvested for histological analysis.

For active EAMG, rats were randomly divided into three groups (n = 8/group): the complete Freund’s adjuvant (CFA) control group, the active EAMG group, and the CRIg/FH treatment group. Rat AChR amino acid 97-116 peptide (R97-116, DGDFAIVKFTKVLLDYTGHI) was synthesized by Glbiochem (Shanghai, China) as described previously (20). For active EAMG induction, rats were immunized subcutaneously at the tail base with 50 μg R97-116 peptide emulsified in CFA mixed with 1 mg of Mycobacterium tuberculosis (strain H37RA; Difco, Detroit, MI) (21). Then, the rats were boosted on day 30 with the same dose of R97-116 in incomplete Freund’s adjuvant. The control group was injected with Freund’s adjuvant emulsified in PBS instead of the R97-116 peptide. Rats in the treatment group were given an intraperitoneal injection of CRIg/FH (1 mg/100 g) on day 29 and every 3 days thereafter. The remaining rats in CFA and active EAMG groups received intraperitoneal injection of PBS every 3 days. All animals were weighed at the beginning of the experiment. After the first immunization, body weights and clinical scores were assessed every other day in a blinded manner by two researchers at the same time until Day 62. Sera were collected for antibody and cytokine measurements on day 62 and stored at −80°C until use.

Sandwich ELISA was performed to detect the anti-R97-116 antibodies. R97-116 (5 μg/ml) was coated onto 96-well plates (Corning, Corning, NY) at 4°C overnight. The plates were washed with PBS-T (PBS with 0.05% Tween 20) on the following day and then blocked for 2 h with 10% fetal calf serum at room temperature (RT). Rat serum samples diluted at 1:500 (100 μl/well) were added and incubated at RT for 2 h. After five washes, washed plates were incubated for 1 h with HRP-conjugated rabbit anti-rat IgG (1:2,000) at RT. Finally, TMB substrate solution was added, and the reaction was allowed to develop at RT in the dark. Plates were read at a wavelength of 450 nm and the results were expressed as OD value ± SD.

The serum anti-AChR-IgG titer was measured with a live cell-based assay, using H9c2(2-1) cells lines (National Collection of Authenticated Cell Culture, Shanghai, China) expressing the rat muscle AChR (22). The H9c2(2-1) cells were seeded on the glass coverslips in 48-well plates for 24 h in DMEM (Gibco, Grand Island, NY) supplemented with 10% FBS (Gibco, Grand Island, NY) and 1% penicillin/streptomycin (Gibco, Grand Island, NY), then fixed by 4% paraformaldehyde for 15 min and washed three times with PBS. After blocking with 3% BSA for 1 h, serial dilutions of mAb35 or serum samples (in 0.5% BSA) were added and incubated at 4°C overnight. The rabbit anti-rat IgG conjugated with Alexa Flour 488 (1:1,000) and α-BTX conjugated with Alexa Flour 555 (1:500, Invitrogen, Carlsbad, CA) were incubated in 1% BSA for 60 min at room temperature. Following DAPI staining (Invitrogen, Carlsbad, CA), fluorescence was imaged on an Olympus fluorescence microscope. The quantitation of immunofluorescence images was performed with ImageJ (23). The mAb35 antibody was used at dilutions starting at 1:50 (0.174 mg/ml) and titrated in twofold serial dilutions. Fluorescence titration curve was obtained by plotting mean fluorescence intensity values against the log10 of the mAb35 concentration. Serum samples from active EAMG and CRIg/FH-treated rats were diluted twofold starting at 1:10 up to 1:1,280. Mean fluorescence intensities (serum samples diluted 1:80) were measured and used for calculation of anti-AChR antibody titers based on the titration curve.

For analysis of MAC deposition at the NMJ, 10-μm transverse sections of rat frozen diaphragm muscle were sliced in both the control and experimental groups. Slides were fixed in cold acetone. After washing with PBS, the sections were blocked for 90 min with 4% BSA at room temperature and incubated overnight at 4°C with mouse anti-rat C5b-9n antibody (1:2,000, Santa Cruz, Heidelberg, Germany). After washing in PBS, Alexa Fluor488-conjugated anti-mouse IgG (diluted at 1:1,000, Abcam, San Francisco) and Alexa Fluor555-conjugated AChR-binding α-BTX (diluted at 1:2,000, Invitrogen, Carlsbad, CA) were incubated for 60 min. The sections were washed and then viewed with an Olympus fluorescence microscope.

Mononuclear cells (MNCs) were isolated from spleens of immunized rats on day 62 and prepared by grinding tissues through a cell strainer (Becton Dickenson, NJ) in lymphocyte separation fluid (Dakewe, Shanghai, China). MNCs were incubated with the following fluorochrome-conjugated monoclonal antibodies for 30 min at 4°C: anti-CD4 FITC and anti-CD25 PE (all from BD Bioscience, San Jose, CA). After the surface staining, cells were fixed and permeabilized with Intracellular Fixation and Permeabilization Buffer or Foxp3/Transcription Factor Buffer according to the manufacturer’s recommendations. After wash, cells were stained with anti-IFN-γ Alexa Flour 647A, anti-IL-17 PE, and anti-Foxp3 APC (eBioscience, San Diego, CA) to detect the intracellular or intranuclear markers. The expression of surface, intracellular, and intranuclear markers was analyzed by Attune NxT Flow Cytometer (Thermo Fisher, Ramsey, MN), and the results were analyzed using FlowJo software.

According to the previous studies in MG (24, 25), associated serum cytokines including IFN-γ, IL-2, IL-6, IL-10, and IL-17 were analyzed according to the manufacturer’s instructions, using a custom-made rat 5-plex kit. Briefly, following pre-wetting the filter plates, 25 μl of each sample/standard (in duplicate) was added to the bottom of the microplate wells. Then, 25 μl of bead suspension was added, and the plates were sealed with agitation overnight at 4°C. After the plates were washed twice, 25 μl of detection antibody was added; the plate was incubated for 60 min at room temperature. After incubation, 25 μl of streptavidin-PE was added to each well and incubated for 30 min at room temperature. The plate was again washed and 120 μl of reading buffer was added. The plate was read on a Bio-plex 200 analyzer.

Active EAMG rats were induced with R97-116 peptide as described above and euthanized on day 45. Conventional CD4⁺CD25^- T cells were sorted from spleens of active EAMG rats by flow cytometry using BD FACS Aria II Cell sorter (BD Biosciences, San Jose, CA). The conventional T cells were cultured for 5 days in the presence of CRIg/FH with plate-bound anti-CD3 (2.5 μg/ml, Peprotech, Rocky Hill, NJ), anti-CD28(2.5 μg/ml, Peprotech), IL-2 (100U/ml, Peprotech), and TGF-β (1 ng/ml Sino Biological, Beijing, China) to induce iTreg differentiation. Foxp3⁺ cells were detected with flow cytometry.

Group assignments, data recording, and data analysis were blinded to the operator. Data distribution was tested via Kolmogorov test. For normal distributions, the difference between the groups was calculated by one-way or two-way ANOVA comparisons followed by multiple comparison test. Post-hoc tests were run only if F achieved p < 0.05 and there was no significant variance inhomogeneity. For non-normal distributions, the difference was calculated by Kruskal–Wallis test. When comparing two different groups, unpaired t-tests were used to determine statistical significance. Statistical analysis was performed using Prism 7 software (Graph Pad, La Jolla, CA) and a p-value < 0.05 was considered statistically significant.

We first investigated the ability of CRIg/FH to suppress complement activation induced by the serum antibodies of MG patients. We collected the IgG fraction from the sera of 10 immunotherapy-naïve and AChR-Ab seropositive MG patients ( Supplementary Table 1 ) to activate the complement classical pathway. The CDC effect was measured in the AChR-expressing TE671 cells by incubating with the purified serum IgG and using normal human sera as a complement source. We observed that the IgG from MG patients’ sera induced marked cell death in a dose-dependent manner compared to IgG from healthy control sera, while administration of CRIg/FH could potently suppress the CDC effect on TE671 cells ( Figures 1A, B ). We further explored the protective effect of CRIg/FH on TE671 cells by detecting MAC deposition on the cell membrane using Immunocytochemistry assay stained by C5b-9n antibody after incubation with the gMG patient-derived IgG and normal human sera. The result showed that CRIg/FH also remarkably reduced MAC assembly on TE671 cells ( Figures 1C, D ). Therefore, we demonstrated that CRIg/FH could effectively protect AChR-expressing TE671 cells from complement-mediated injury induced by gMG patient-derived IgG against AChR.

The specific antibody against AChR may result in the NMJ injury promptly via complement activation (18); next, we evaluate the ability of CRIg/FH to prevent the induction of PTMG induced by passively transferring mAb35 into healthy rats. The results showed that mAb35 injection promptly induced the progressive and significant body weight loss ( Figure 2A ) and weakness represented by the clinical score ( Figure 2B ) at 24 and 48 h after injection, which may result from the complement attack determined by the extensive MAC deposition at the NMJ ( Figures 2C, D ). However, the pretreatment with CRIg/FH displayed a potent protective role in PTMG development. We observed no weight loss and little increase in clinical score after CRIg/FH pretreatment in PTMG animals ( Figures 2A, B ). The results of muscle pathological detection further confirmed that MAC deposition at the NMJ was near completely suppressed by the CRIg/FH pretreatment ( Figures 2C, D ). Thus, these findings demonstrated that CRIg/FH could prevent the development of PTMG by inhibiting complement activation.

We next evaluated the therapeutic potential of CRIg/FH in the active EAMG model rats immunized with AChR peptide (R97-116). Compared to the adjuvant control, immunization with AChR peptide caused significant body weight loss starting at day 46 through the end of the experiment ( Figure 3A ), and progressive exacerbation in clinical symptoms starting from day 40 until the end of the experiment ( Figure 3B ). In the CRIg/FH-treated group, we injected drug (1 mg/100 g, i.p.) every 3 days starting from day 29, and found a strong therapeutic effect determined by the measurement of body weight loss and clinical score. As shown in Figure 3A , the body weight of CRIg/FH-treated rats started to decrease on day 56, which was delayed, compared to the rats in the active EAMG group; moreover, the body weight loss was also significantly lower than the control active EAMG rats. In addition, the observation in clinical scores also showed similar results. The clinical scores started to significantly increase until day 52 in the CRIg/FH-treated group, in which the increase occurred 12 days later, and the score was significantly lower than the active EAMG group without CRIg/FH treatment ( Figure 3B ).

Considering the role of complement activation in B-cell development and autoantibody production via its components such as CR2 (26) and FH (27), we further detected the production of specific antibodies against R97-116 and AChR in these rats with sera collected on day 62. The levels of anti-R97-116 IgG were measured by ELISA and the titers of anti-AChR IgG were determined by immunofluorescence using the cell-based assay described above. We observed that the combined immunization with R97-116 peptide and CFA effectively induced the production of specific antibody against R97-116 peptide and AChR in active EAMG model rats, while CRIg/FH treatment could remarkably reduce the serum levels of both anti-R97-116 and anti-AChR IgG ( Figures 3C–F ). These results suggested a potential mechanism for CRIg/FH preventing active EAMG development via inhibition of specific antibody production. Together, we further demonstrate that CRIg/FH holds the therapeutic potential in the active EAMG model, by both regulating humoral response and directly inhibiting complement activation.

In inflammatory diseases, complement activation may induce the production of cytokines and chemokines through the regulation of the downstream cellular immunity (1, 28, 29). A bunch of pro-inflammatory cytokines has been reported to be upregulated in active EAMG and MG, including IFN-γ, IL-2, IL-6, and IL-17 (30–33). Thus, we measured the serum concentration of IFN-γ, IL-2, IL-6, IL-17, and IL-10 to further understand the protective mechanisms of CRIg/FH in active EAMG rats. After immunizing with AChR peptide R97-116 in CFA, the active EAMG rats exhibited a significant elevation of IFN-γ, IL-2, IL-6, and IL-17 (pro-inflammatory), and decrease of IL-10 (anti-inflammatory) compared to the CFA alone control treatment ( Figure 4 ). However, CRIg/FH treatment remarkably reduced the serum levels of IFN-γ, IL-2, IL-6, and IL-17 in treated rats when compared to that in untreated active EAMG rats. More importantly, CRIg/FH treatment strongly reduced the serum levels of these cytokines to that of the adjuvant control ( Figure 4 ). In addition, CRIg/FH treatment failed to alter the serum level of IL-10 in active EAMG rats ( Figure 4 ). Therefore, we determined that CRIg/FH could potently downregulate the production of multiple pro-inflammatory cytokines in active EAMG rats.

Treg cells play a critical role in the prevention of MG by suppressing the effector CD4⁺ T cell subsets and maintaining immune homeostasis and self-tolerance (34, 35). The blockage of complement C3aR and C5aR may promote TGF-β1-induced Treg differentiation (36), and both of CRIg and FH have been reported to suppress T-cell activation (37–39). Therefore, we examined the effect of CRIg/FH treatment on the development of T helper and Treg cells in active EAMG rats. Significant increase of Th1 (CD4⁺IFN-γ⁺) and Th17 (CD4⁺Th17⁺) and decrease of Treg (CD4⁺CD25⁺ Foxp3⁺) cells in the spleen of active EAMG rats were observed compared to the CFA alone treatment rats ( Figure 5 ), indicating the involvement of various CD4⁺ T subset cells in active EAMG development. However, CRIg/FH only upregulated Treg subsets while displaying no significant effect on the Th1 and Th17 cells ( Figure 5 ).

We further assessed whether CRIg/FH could promote Treg differentiation induced by TGF-β in vitro. The isolated CD4⁺CD25^- conventional T cells from the spleens of Lewis rats were cultured in the presence of anti-CD3/CD28, IL-2 and TGF-β to induce iTreg differentiation. The additional CRIg/FH treatment may significantly promote iTreg differentiation at 60 μg/ml ( Figure 6 ). Therefore, these findings revealed another mechanism by which CRIg/FH may protect from MG via the promotion of Treg differentiation.