Potential therapeutic benefit of C1-esterase inhibitor in neuromyelitis optica evaluated in vitro and in an experimental rat model

Abstract

Neuromyelitis optica (NMO) is an autoimmune demyelinating disease of the central nervous system in which binding of anti-aquaporin-4 (AQP4) autoantibodies (NMO-IgG) to astrocytes causes complement-dependent cytotoxicity (CDC) and inflammation resulting in oligodendrocyte and neuronal injury. There is compelling evidence for a central role of complement in NMO pathogenesis. Here, we evaluated the potential of C1-esterase inhibitor (C1-inh) for complement-targeted therapy of NMO. C1-inh is an anti-inflammatory plasma protein with serine protease inhibition activity that has a broad range of biological activities on the contact (kallikrein), coagulation, fibrinolytic and complement systems. C1-inh is approved for therapy of hereditary angioedema (HAE) and has been studied in a small safety trial in acute NMO relapses (NCT 01759602). In vitro assays of NMO-IgG-dependent CDC showed C1-inh inhibition of human and rat complement, but with predicted minimal complement inhibition activity at a dose of 2000 units in humans. Inhibition of complement by C1-inh was potentiated by ∼10-fold by polysulfated macromolecules including heparin and dextran sulfate. In rats, intravenous C1-inh at a dose 30-fold greater than that approved to treat HAE inhibited serum complement activity by <5%, even when supplemented with heparin. Also, high-dose C1-inh did not reduce pathology in a rat model of NMO produced by intracerebral injection of NMO-IgG. Therefore, although C1r and C1s are targets of C1-inh, our in vitro data with human serum and in vivo data in rats suggest that the complement inhibition activity of C1-inh in serum is too low to confer clinical benefit in NMO.

Trial registration: ClinicalTrials.gov NCT01759602.

Conflict of interest statement

Competing Interests: This work described in this manuscript was funded, in part, by a sponsored research agreement from a commercial source (CSL Behring). This does not alter the authors' adherence to PLOS ONE policies on sharing data and materials.

Figures

Figure 1. Action of C1-inh on NMO-IgG-dependent cytotoxicity involve the classical complement pathway.

NMO-IgG binds to cell surface AQP4 on astrocytes, which activates the classical complement pathway by C1q binding to the Fc region of NMO-IgG leading to formation of the membrane attack complex (MAC) (left). C1-inh binds to and inactivates C1r and C1s, preventing C1q action on C4 and downstream complement activation.

Figure 2. C1-inh inhibits complement-dependent cytotoxicity produced by NMO-IgG in AQP4-expressing cell cultures.

A. CDC assay showing 1 h incubation of NMO-IgG, human complement (HC) and C1-INH to CHO cells expressing human M23-AQP4, followed by assay of cytotoxicity with AlamarBlue. B. Percentage cytotoxicity as a function of C1-inh concentration in cells incubated with human complement (2%) and recombinant NMO-IgG rAb-53 (1 µg/ml) (left) or individual or pooled human NMO sera (100 µg/ml) (right). C. Study as in B in cells incubated with rAb-53 (0.5 µg/ml) and rat serum (2%). Data are mean ± S.E. for 4 measurements per condition.

Figure 3. C1-inh inhibition of NMO-IgG-dependent cytotoxicity depends on complement and NMO-IgG concentration.

A. Percentage cytotoxicity as a function of C1-inh concentration in cells incubated with 1 µg/ml rAb-53 and indicated concentration of human complement. Data are mean ± S.E. for 4 measurements per condition. Inset shows percentage inhibition of cytotoxicity as a function of time after pre-incubation of 2 µM C1-inh and 2% human complement prior to addition AQP4-expressing CHO-cell that were pre-incubated with 1 µg/ml rAb-53. B. Study as in A but with 0.5 µg/ml rAb-53.

Figure 4. Polysulfated macromolecules potentiate C1-inh inhibition of NMO-IgG-dependent CDC.

A. Percentage cytotoxicity as a function of C1-inh concentration in cells incubated with human complement (2%) and rAb-53 (1 µg/ml) in the absence or presence of 2 µM heparin, 0.5 µM dextran sulfate or 0.02 µM fucoidan (mean ± S.E., n = 4). Insets show chemical structures. B. Percentage cytotoxicity as a function of concentration of polysulfated macromolecules in cells incubated with human complement (2%) and rAb-53 (1 µg/ml) in the absence or presence of 1 µM C1-inh.

Figure 5. Intravenous administration of high-dose C1-inh in rats does not inhibit complement activity.

A. C1-inh was administered to rats by tail vein injection, and serum samples obtained at specified times were bioassayed for NMO-IgG-dependent CDC by incubation of AQP4-expressing CHO-cells with 2% rat serum with 1 µg/ml rAb-53 on percentage cytotoxicity produced by rat serum at different times after administration of 600 units/kg C1-inh (or vehicle control) (left), and 1.5 mg/kg heparin, without or with 600 units/kg C1-inh (right) (mean ± S.E., 4 rats per condition).

Figure 6. High-dose C1-inh does not reduce brain pathology in a rat model of NMO produced by intracerebral injection NMO-IgG.

AQP4, GFAP and MBP immunofluorescence at 2 days after injection of 10 µg rAb-53. C1-inh (300 units/kg) (or saline control) was administered intravenously just before and 1 day after intracerebral injection of NMO-IgG. The needle tract is shown in yellow line, the white line demarcates lesion areas with loss of AQP4 and GFAP immunofluorescence, and the white dashed line demarcates the penumbra areas with reduced AQP4 but normal GFAP immunofluorescence (left). Summary of lesion areas showing data for individual rats (mean ± S.E., n = 4). Differences not significant.