Altered aquaporin-4 (AQP4) expression has been reported in brain edema, tumors, muscular dystrophy, and neuromyelitis optica. However, the plasma membrane organization of AQP4 and its interaction with proteins such as the dystrophin-associated protein complex are not well understood. In this study, we used sucrose density gradient ultracentrifugation and 2D blue native/sodium dodecyl sulfate-polyacrylamide gel electrophoresis and showed the expression of several AQP4 multi-subunit complexes (pools) of different sizes, ranging from >> 1 MDa to similar to 500 kDa and containing different ratios of the 30/32 kDa AQP4 isoforms, indicative of orthogonal arrays of particles of various sizes. A high molecular weight pool co-purified with dystrophin and beta-dystroglycan and was drastically reduced in the skeletal muscle of mdx3cv mice, which have no dystrophin. The number and size of the AQP4 pools were the same in the kidney where dystrophin is not expressed, suggesting the presence of dystrophin-like proteins for their expression. We found that AQP2 is expressed only in one major pool of similar to 500 kDa, indicating that the presence of different pools is a peculiarity of AQP4 rather than a widespread feature in the AQP family. Finally, in skeletal muscle caveolin-3 did not co-purify with any AQP4 pool, indicating the absence of interaction of the two proteins and confirming that caveolae and orthogonal arrays of particles are two independent plasma membrane microdomains. These results contribute to a better understanding of AQP4 membrane organization and raise the possibility that abnormal expression of specific AQP4 pools may be found in pathological states

No abstract

Grapevine leafroll-associated virus 3 (GLRaV-3) is a major pathogen of grapevine. A previously described single-chain fragment variable (scFv) antibody (scFvLR3), directed against the coat protein (CP) of GLRaV-3, was expressed in Escherichia coli and used to develop a diagnostic ELISA kit. The antibody was fused to the light chain constant domain of human immunoglobulin to create the bivalent reagent C(L)-LR3, which was purified from the periplasmic fraction, giving a yield of ~5 mg/l. The sensitivity of the reagent against recombinant GLRaV-3 CP was 0.1 ng. The sensitivity, specificity and durability of the reagent was similar to a commercial kit. The C(L)-LR3 showed a weak cross-reaction in immune electron microscopy assays to GLRaV-1 and -7, but not with the phylogenetically more distant GLRaV-2. A fully recombinant kit was developed with the inclusion of a recombinant GLRaV-3 CP expressed in bacteria, thus avoiding problems associated with virus propagation and purification. This system represents a rapid, simple, sensitive and standardized diagnostic protocol for GLRaV-3 detection.