Aquaporin-4 (red) and EAAT2 (green) colocalize in normal spine.

Neuromyelitis optica (NMO), also known as Devic's disease, results in MS-like lesions in the optic nerves and along the spine. Yet unlike MS, NMO is associated with production of a specific autoantibody known as NMO-IgG. Clinicians use NMO-IgG to diagnose the disease, but they don't know how the antibody contributes to its symptoms—sight impairment, paraplegia, and loss of limb, bladder, and bowel sensation.

In 2005, the same group identified NMO-IgG's target as aquaporin-4, a water channel protein concentrated in astrocyte membranes along the blood-brain barrier. However, the finding was somewhat perplexing because myelin damage occurs on nerve cells, not on astrocytes. Now the team confirms that astrocytes are indeed the relevant target, and show how autoantibody binding can lead to demyelination. Clue: when astrocytes hurt, their duties lapse, too.

When NMO-IgG binds to aquaporin-4, they show, the levels of the astrocyte glutamate transporter EAAT2 drop. And no transporter, no glutamate regulation. Astrocytes themselves aren't sensitive to changes in glutamate levels, but neurons and oligodendrocytes are. These cells rely on EAAT2 to mop up excess glutamate from extracellular space. Accumulation of the neurotransmitter can be toxic to myelin-making oligodendrocytes.

After spelling out the pathway in astrocyte in vitro assays, the team examined human tissue. Sure enough, NMO lesions along cadaver spines lacked both aquaporin-4 and EAAT2. Lesions from MS patients show no such deficiencies, highlighting another way in which the demyelinating disorders differ.

If the groups' results are confirmed in vivo, drug development could be straightforward. Therapeutic trials for glutamate antagonists, created to treat other neurodegenerative diseases like Lou Gehrig's disease (or ALS), are already underway.