A Functional Human-on-a-Chip Autoimmune Disease Model of Myasthenia Gravis for Development of Therapeutics

Abstract

Myasthenia gravis (MG) is a chronic and progressive neuromuscular disease where autoantibodies target essential proteins such as the nicotinic acetylcholine receptor (nAChR) at the neuromuscular junction (NMJ) causing muscle fatigue and weakness. Autoantibodies directed against nAChRs are proposed to work by three main pathological mechanisms of receptor disruption: blocking, receptor internalization, and downregulation. Current in vivo models using experimental autoimmune animal models fail to recapitulate the disease pathology and are limited in clinical translatability due to disproportionate disease severity and high animal death rates. The development of a highly sensitive antibody assay that mimics human disease pathology is desirable for clinical advancement and therapeutic development. To address this lack of relevant models, an NMJ platform derived from human iPSC differentiated motoneurons and primary skeletal muscle was used to investigate the ability of an anti-nAChR antibody to induce clinically relevant MG pathology in the serum-free, spatially organized, functionally mature NMJ platform. Treatment of the NMJ model with the anti-nAChR antibody revealed decreasing NMJ stability as measured by the number of NMJs before and after the synchrony stimulation protocol. This decrease in NMJ stability was dose-dependent over a concentration range of 0.01-20 μg/mL. Immunocytochemical (ICC) analysis was used to distinguish between pathological mechanisms of antibody-mediated receptor disruption including blocking, receptor internalization and downregulation. Antibody treatment also activated the complement cascade as indicated by complement protein 3 deposition near the nAChRs. Additionally, complement cascade activation significantly altered other readouts of NMJ function including the NMJ fidelity parameter as measured by the number of muscle contractions missed in response to increasing motoneuron stimulation frequencies. This synchrony readout mimics the clinical phenotype of neurological blocking that results in failure of muscle contractions despite motoneuron stimulations. Taken together, these data indicate the establishment of a relevant disease model of MG that mimics reduction of functional nAChRs at the NMJ, decreased NMJ stability, complement activation and blocking of neuromuscular transmission. This system is the first functional human in vitro model of MG to be used to simulate three potential disease mechanisms as well as to establish a preclinical platform for evaluation of disease modifying treatments (etiology).

Keywords: acetylcholine receptor; autoantibodies; disease pathology; human-on-a-chip; induced pluripotent stem cells (iPSCs); microphysiological systems; myasthenia gravis; neuromuscular junction (NMJ).

FIGURE 1

Pathogenic mechanisms of anti-nAChR antibody (Ab). (I) Blocking abs interfere with ACh binding site interrupting neuromuscular transmission. (II) Modulating Abs that bind to two nAChRs causing receptor endocytosis leading to AChR degradation and reduction of expression. (III) Binding Abs that complex with a distinct location on the nAChR, triggering the activation of the complement cascade and subsequent formation of the membrane attack complex (MAC). Scheme adopted from Conti-Fine (Conti-Fine et al., 2006). The alpha (α) refers to the alpha subunit(s) on the nAChR.

FIGURE 2

Schematic of NMJ chamber with primary SKM innervated by WT iPSC-MNs and electrically stimulated by electrodes for either direct (SKM-side) or indirect (MN-side) stimulation. Phase images indicate the morphology of co-culture on SKM innervated by axons progressing through tunnels (right) and WT iPSC derived MNs (left). Scale bars = 100 microns.

FIGURE 3

Anti-nAChR IgG effects on the NMJ. (A) A dose response curve established from NMJ stability of SKM dosed with the anti-nAChR Ab (diamond) from 0 to 20 μg/mL resulting in an IC50 of 3.4 μg/mL (N = 2, 3–10 replicates). A non-specific IgG antibody (triangle) did not significantly alter NMJ stability at 2 and 10 μg/mL (N = 2, 7–8 replicates). (B) ICC images of SKM revealing the antibody-antigen complex with the binding of anti-nAChR Ab (red channel) to the cell surface near the endplate marked by BunTX (green channel), overlay of red and green channels (upper left) and phase image with DAPI (lower right). Scale bars = 50 microns. (C) The number of contracting SKM under direct stimulation at dosing concentrations from 0 to 20 μg/mL (N = 2, 3–10 replicates). Statistics: One-way ANOVA followed by Dunnett’s test (alpha = 0.05) ^∗∗p < 0.020, ^∗∗∗p < 0.002. (D) ICC images of SKM with anti-nAChR Ab staining (red channel), sodium channel (Pan, green channel), overlay of red and green channels (upper left) and phase image on lower right. Scale bars = 50 microns.

FIGURE 4

Anti-nAChR Abs induce receptor internalization. (A) ICC images of SKM stained for anti-nAChR Ab with and without triton (top panel), incubated with 2 μg/mL or 10 μg/mL anti-nAChR Ab, middle and bottom panels, respectively, and later fixed and blocked with and without triton. Scale bars = 50 microns. (B) The corrected total cell fluorescence between the surface receptors (black bars) and internalized receptors (gray bars) for each condition. The percent change in fluorescence is represented as a bar graph for nAChR internalization.

FIGURE 5

Direct effects of hCS and anti-nAChR on SKM. (A) Demonstrates the number of contracting myotubes under direct stimulation after dosing with 0.05% hCS (N = 2, 8 replicates), non-specific IgG (2 μg/mL, N = 3, 8 replicates), anti-nAChR-Ab (2 μg/mL, N = 2, 10 replicates) and a combination of 0.05% hCS and 2 μg/mL anti-nAChR-Ab (N = 2, 9 replicates). (B) ICC images of SKM incubated with 2 μg/mL of anti-nAChR with 0.05% hCS for 3 h and stained with anti-C3 (red channel) and BunTX (green channel). Scale bars = 50 microns.

FIGURE 6

Complement activation reduces SKM and NMJ fidelity. SKM fidelity at 2 Hz under direct and indirect stimulation after incubation with 0.05% hCS (N = 2, 6 replicates), non-specific IgG1 (N = 3, 7 replicates), anti-nAChR Ab and a combination of 0.05% hCS (N = 2, 4 replicates) and 2 μg/mL of anti-nAChR Ab (N = 2, 4 replicates). (A) Representative stimulation traces generated at 2 Hz from both direct (SKM stimulation) and indirect (MN stimulation) of a myotube under the five conditions. Mean fidelity of peaks under direct (B) and indirect (C) stimulation. Significant reduction of fidelity with antibody and complement under direct and indirect stimulation and antibody only under indirect stimulation. Statistics: Fisher LSD Method ^∗p < 0.050; ^∗∗p < 0.020; ^∗∗∗p < 0.010.