Evidence Grows for Pathogenic Antibody
American Academy of Neurology meeting report
SAN DIEGO—For years, MS researchers have searched in vain for a holy grail: a protein target of autoimmune reactivity that leads to disease. Neurologists at the Technical University Munich in Germany recently identified an antibody target that may be the elusive target: the potassium channel KIR4.1. On March 20, the group presented new findings about this target channel at the annual meeting of the American Academy of Neurology in San Diego, California, showing that it is missing from brain lesions in MS patients.
Previously, the researchers purified immunoglobulin G (IgG)—the most common type of antibody—from blood samples from MS patients and tested whether any bound to protein targets isolated from brain tissue. Indeed, IgG antibodies from some patients bound to a membrane protein the scientists identified as KIR4.1. The channel was not targeted by antibodies from healthy people or from patients with other neurological diseases. Nearly half of MS patients carried the KIR4.1-specific antibody, according to work published last July in the New England Journal of Medicine (Srivastava et al., 2012; see “Channeling MS”).
In the new work, the researchers wanted to learn more about the ion channel in human brain tissue, where little was known about the protein. They isolated the KIR4.1 antibody from MS patients’ blood and used it to stain postmortem brain tissue sections from healthy people. They found the channel not on neurons but on the surface of glial cells—non-neuronal cells that include astrocytes and myelin-producing oligodendrocytes. “The channel was strongly expressed on oligodendrocytes and a subset of astrocytes,” which were clustered around blood vessels, said Lucas Schirmer, who analyzed the brain samples. Bernhard Hemmer, the lead investigator, said the expression pattern of KIR4.1 “fits nicely with what we know about MS pathology: It’s expressed on oligodendrocytes, which we know are a target of MS.”
KIR4.1 function in astrocytes, which communicate with blood vessel epithelial cells, oligodendrocytes, and neurons, might also be relevant to MS. Astrocytes help maintain a stable brain environment, in part by sustaining the delicately balanced ion concentrations required by neurons for their ultrafast signaling. For example, astrocytes take up potassium through the KIR4.1 channel, which allows potassium to flow into—but not out of—cells, particularly when extracellular levels become too high. These cells are then thought to pass the excess potassium through gap junctions directly into blood vessels, a role that could explain the presence of such KIR4.1-expressing astrocytes in perivascular regions.
In addition to the channel’s location, the researchers also determined the specific structure of the channel that binds the MS patients’ antibody. KIR channels exist in two forms: homotetramers, which are made of four subunits of the same type, or heterotetramers, which are made of two different subunit types. “The subunit makeup somehow alters how they’re regulated,” Schirmer said. The antibodies isolated from MS patients bound only to the homotetramers, which were known to be the only form to appear on oligodendrocytes.
In collaboration with the UK Multiple Sclerosis Tissue Bank, the scientists then looked at expression of KIR4.1 in nearly 100 postmortem brain samples from patients with MS. First they sorted the samples according to lesion type. Early inactive (also known as chronic active) lesions were characterized by a high density of infiltrating peripheral macrophages and microglia (resident immune cells of the central nervous system) in the area surrounding the lesions, which were not incorporating myelin fragments, and few inflammatory cells in the demyelinated core of the lesion. Active lesions in white matter (regions with myelinated axons) were characterized by ongoing demyelination, with active immune cells phagocytosing myelin fragments. Chronic late inactive demyelinated lesions had little or no macrophage or microglial activity, but bore damage called a glial scar.
The researchers found varying patterns of antibody staining in these different lesion types. KIR4.1 reactivity was lost in early inactive lesions, but the tissue stained positive for glial fibrillary acidic protein (GFAP), an astrocyte marker. That pattern indicated that the astrocytes were still present but the channel was not, Hemmer said. In active lesions, channel staining was missing from perivascular astrocytes and from oligodendrocytes, many of which had apparently met their demise following immune attack. In line with this idea, phagocytes containing KIR4.1-associated cellular debris were observed.
In late inactive lesions that exhibit lots of demyelination, but not inflammation, the researchers were surprised to see KIR4.1 restored on perivascular astrocytes. “We saw an extensive glial scar that was positive for both KIR4.1 and GFAP,” Hemmer said, “so we see that the channel is upregulated once the inflammation is gone.” Finally, when they looked at remyelination that appeared in some lesions, they saw a subset of cells that were positive for KIR4.1 and for Olig2, a marker expressed by both oligodendrocyte precursor cells and by mature oligodendrocytes. “We can say those probably are new oligodendrocytes, and at a certain stage of development they express KIR4.1 again,” Hemmer added.
So might an antibody that attacks KIR4.1 lead to MS? According to Hemmer, their findings “could support the idea that KIR4.1 is a target” of autoantibodies that trigger this disease. For example, the fact that KIR4.1 was lost from astrocytes in early lesions—even as other molecules are being upregulated—suggests that the protein has selectively gone missing, perhaps as a result of antibody binding. Even before beckoning an all-out immune attack, antibody binding would likely disrupt KIR-channel activity and compromise astrocytes’ ability to buffer potassium. “At least in rodents, we know that KIR4.1 is the major regulator of potassium in the brain. It’s responsible for siphoning and buffering potassium,” Hemmer said. At this point, he says, “it’s all indirect evidence” that attack of KIR4.1 by the antibody found in MS patients initiates or contributes to the disease process, “but our findings would be compatible with that idea.”
The hunt for an MS-related autoantibody was recently reinvigorated by the identification of an antibody to the water channel aquaporin 4 (AQP4) in patients with neuromyelitis optica (NMO), historically categorized as a form of MS but now recognized as a separate disease (Roemer et al., 2007). That antibody turned out to be pathogenic, revealing new details about the disease process and how to diagnose and treat it. Brian Weinshenker, a neurologist at the Mayo Clinic in Rochester, Minnesota, contributed to that work on NMO but was not involved in the current study. He wondered whether the patients whose brains were used to see the pattern of KIR4.1 reactivity in the different lesion types were also seropositive for the antibody themselves, and whether that status conferred a particular pattern of degeneration. “In the case of NMO, AQP4-IgG was associated in blinded analysis with a unique phenotype of demyelinating disease,” found only in patients seropositive for the AQP4 antibody, Weinshenker said. In the current study, blood samples were not available from the donors whose brains were used, but Hemmer is currently working to determine whether MS patients who carry the antibody experience a different disease progression as seen by MRI or disability than those who don’t. Whether or not the KIR4.1 antibody leads to MS, it might serve as a valuable tool to differentiate potential subsets within the MS patient population.
Even if such an autoantibody to KIR4.1 does contribute to MS, it’s unlikely to be the sole cause of the disease, Hemmer says. “We know for sure that genes and environment play a big role,” Hemmer said. Certain genetic variants probably set the stage for the immune system to become autoreactive, but an antibody such as this one found in MS patients could tip the scales toward disease. Although many proteins have been considered candidates for the triggering antigen, none has panned out yet. “In our hands, KIR4.1 shows promise [for this role],” Hemmer said.
The initial finding of the antibody last year has since been corroborated by Kevin O’Connor, a neuroimmunologist at Yale School of Medicine in New Haven, Connecticut. “We reproduced [Hemmer’s] assay, and we were able to detect antibodies to KIR4.1 in patients with MS,” O’Connor said. He has not yet assayed enough patients to say how many in his sample carry the protein, or whether it will approach the 47% reported by Hemmer. “Different labs working with different populations, even in different countries—you won’t get the same rate,” O’Connor said.
If KIR4.1 does turn out to be a trigger for MS, it probably won’t be the only one. “It’s very likely there will be additional antibodies found that play a role in this disease,” Hemmer said. In hunting for them, he said, researchers need to look for targets that live at the cell surface, where antibodies could access them. Some proposed targets might serve as good biomarkers but may not significantly contribute to disease. For example, he said, “it’s hard to believe that an antibody to a nuclear protein would contribute to pathology, because it wouldn’t get to its target.”
To find new antibodies, MS researchers face a major challenge in developing better screening tools. “Many proteins that are modified by antibodies—including KIR4.1—form three-dimensional structures that are not easy to build in a dish,” Hemmer said. O’Connor agreed. “Many people try to express a protein in a dish and use it [to screen for antibodies] and expect it to work. It’s far, far more complicated than that,” he said. “This assay is very difficult to do.” Through extensive communications with Hemmer’s group, “we figured out all the technical details, and it does work.”
Despite the challenges, it seems clear that identification of autoantibodies in both NMO and MS will lead to improved understanding, diagnosis, and hopefully treatment of the diseases.
Key open questions
- Does the KIR4.1 antibody affect disease phenotype or progression?
- Does the presence of the antibody predict conversion from clinically isolated syndrome to MS?
- Is the antibody present before disease begins in people who later get MS?
- Is there a genetic component to KIR4.1 antibody expression?
- Could blocking antibodies to KIR4.1 be a treatment option for MS?
Thumbnail image on landing page. “San Diego palm trees sidewalk piers” by Jon Sullivan, via Wikimedia Commons and Public Domain Images, public domain.
See our coverage of the 2012 AAN meeting
- MS Gene Tests Inch Toward the Clinic
- Positive Drug Data for MS
- No Go for MS Combination Therapy
- Eyeing Neurodegeneration
- AAN Final Roundup
Suggested by May Han
Multiple sclerosis (MS) and neuromyelitis optica (NMO) are inflammatory demyelinating disorders that affect the central nervous system, and their clinical presentations are often difficult to differentiate. Identification of an autoantibody, NMO-IgG (anti-aquaporin 4, AQP4), has made a major step in diagnosing NMO. However, the issue still remains for NMO-IgG negative spectrum of disorders and MS cases that primarily affect the optic nerve and spinal cord. Work by Dr. Lucchinetti's group at the Mayo Clinic distinguishes the pathological features that clearly distinguish MS from NMO. Moreover, pathogenesis of NMO was elaborated by careful analysis of the postmortem NMO tissue. This article teaches us how much we can learn about complex disorders from analysis of the diseased tissue.