Abstract

Neuromyelitis optica (NMO) is a serious inflammatory, demyelinating disease affecting the spinal cord and optic nerves. IgG antibodies to aquaporin-4 (AQP4) in patients' sera are now taken as evidence for an NMO spectrum disease (Lennon et al, 2004, 2005). Since their discovery, a variety of assays have been used to detect the AQP-4 antibody. Most are specific but detection of antibodies binding to cells expressing AQP4 (cell-based assay, CBA) are the most sensitive (reviewed in Waters and Vincent 2008). These assays use either the M1 isoform of AQP-4 (Takahashi et al, 2007) or either M1 or M23 isoform tagged N-terminally with EGFP (Waters et al, 2008). However, the M23 isoform forms large arrays in astrocytes (Furman et al, 2003) and previously we have found that clustering an antigen provides a more sensitive target for the detection of antibodies by CBA (Leite et al, 2008). We, therefore, compared the CBA using AQP4 M23-EGFP with AQP4 M23 without EGFP. AQP4 M23 was expressed very strongly on the cell surface in dense clusters and, in assays of sera sent for diagnosis, was 30% more sensitive than M1 or M23-EGFP (32 vs 23 positive AQP4 antibodies in 1 month).