26 November 2008. Gumming up the brain years before a person may notice memory loss or other hints of Alzheimer disease, amyloid-β (Aβ) has fueled much of AD research, and efforts to develop Aβ-targeted therapeutics have consumed billions of dollars. With more than 200 talks and posters devoted to some aspect of the pesky peptide, this year’s Society for Neuroscience (SfN) meeting, held 15-19 November in Washington, DC, left our small crew of Alzforum reporters with considerable ground to cover. In the coming days, we will highlight a small handful of recent Aβ developments, focusing primarily on unpublished data and work not previously covered on the Alzforum. This article describes two new studies on Aβ clearance—one proposing myelin basic protein (MBP) as a novel Aβ-degrading enzyme, another presenting serum response factor (SRF) and myocardin (MYOCD) as transcriptional co-activators that regulate clearance of brain amyloid. Aβ metabolism is under increased scrutiny in part because accumulating evidence suggests that Aβ immunotherapy works well to decrease plaque load, but may need help from other clearance mechanisms to completely flush Aβ out of the brain vasculature (ARF related news story). As always, ARF welcomes your comments on these and/or other SfN presentations on related topics.
MBP may be best known for its involvement in multiple sclerosis. Indeed, antibodies to the protein, a major component of the myelin sheaths that insulate nerve fibers, play a role in MS pathogenesis. Still, the new SfN data isn’t the first to link MBP to AD. Several studies have suggested that neurodegeneration in AD tracks with myelination patterns during growth and maturation (Bartzokis et al., 2006; Bartzokis et al., 2004). Another investigation identified myelination of peripheral neurons as a major physiological function of BACE1, which catalyzes the first of two proteolytic steps needed to release Aβ from its precursor (see Willem et al., 2006 and ARF related news story). Last year, William Van Nostrand and colleagues at Stony Brook University in New York reported that MBP could bind mutant Aβ that causes cerebral amyloid angiopathy (CAA) and inhibit its assembly into fibrils in vitro (Hoos et al., 2007). Curiously, they noticed in those assays that if MBP was incubated with Aβ for longer periods, the Aβ seemed to disappear. Consistent with this observation, another group reported autocatalytic activity in MBP purified from human brain, and suggested that this cleavage mechanism could play a role in generating autoreactive peptides in MS (D’Souza et al., 2005). Meichen Liao, a graduate student in Van Nostrand’s lab, reasoned that if MBP has enzymatic activity, this behavior could explain the vanishing Aβ seen in the earlier in vitro studies. Based on her emerging data presented publicly for the first time at SfN, the answer appears to be “yes.”
Assessing degradation of monomeric Aβ by purified human MBP in vitro, Liao found that after 24 hours, Aβ40 levels dropped by 50 percent and Aβ42 dipped to about 20 percent of starting levels. In a cell culture—Cos-1 cells transfected with MBP—levels of exogenous Aβ40 and Aβ42 fell by 50 and 60 percent, respectively, after 48 hours. To address whether MBP can degrade fibrillar Aβ, Liao and colleagues performed two in vitro assays. In the first, they coaxed synthetic Aβ42 peptide to form fibrils (by incubating at 37 degrees C for six days), then used thioflavin T fluorescence to determine the change in Aβ fibril content over two days in the presence or absence of MBP. For the second assay, the team mixed unlabeled and FITC-labeled Aβ42 in a 9:1 ratio, aged them for six days at 37 degrees C to promote fibril formation, then two days later, assessed fibrillar Aβ content as the fluorescence of pelleted material in samples treated with or without MBP. By both assays, the amount of fibrillar Aβ dropped by about 50 percent after MBP treatment compared to controls.
Having established that MBP degrades monomeric and fibrillar Aβ in vitro, Liao tested whether MBP could degrade Aβ deposits formed in vivo in Tg2576 mice, a widely studied AD model that overexpresses mutant human amyloid precursor protein (APP) and develops both plaques and vascular Aβ deposits. When fresh cortical slices from the brains of aged Tg2576 mice were incubated with purified human MBP, both plaque load and vascular amyloid, detected as thioflavin S staining, dropped about 60 percent relative to pre-MBP levels, Liao reported. This effect was blocked by addition of the serine proteinase inhibitor, phenylmethane sulfonylfluoride (PMSF).
“These are still early days for this particular study, but I think they've got some nice data showing there's something interesting going on,” said Cindy Lemere of Brigham and Women's Hospital and Harvard Medical School, Boston, in a phone conversation after the SfN meeting.
During Q&A and in conversations after the talk, several scientists expressed concern that the experiments could have picked up enzymatic activity of a contaminant in the purified human MBP preps. Liao said that silver stain gels and mass spectrometry experiments have consistently identified a single protein in their MBP samples. Furthermore, MBP purified from a variety of cell types—including human white matter, mouse brain, bacterial recombinant systems, and adenovirus-infected mammalian cells—seems to exhibit the same activity, she said, reducing the likelihood of a contaminating enzyme in the preps. To Lemere, the experiments showing Aβ degradation in transfected Cos-1 cells, which do not normally express MBP, provided the most compelling evidence against the contamination possibility. “That would be pretty remarkable if the same factor that's in the brain extracts is in the Cos-1 cells,” she told ARF.
Speaking with this reporter at the SfN meeting, Van Nostrand acknowledged limitations with interpreting the data thus far. “Right now we’re just looking at purified MBP,” he said. “We don’t know if it has this effect when wrapped around myelin sheaths.” To address whether native MBP can influence Aβ accumulation in vivo, his group is crossing MBP knockout mice with two APP-overexpressing transgenic models—Tg2576 and another strain, SwDI (Davis et al., 2004), which expresses a human APP gene with Swedish, Dutch, and Iowa mutations, leading to a massive buildup of vascular amyloid and parenchymal plaques in the brain.
During the same SfN session, Robert Bell, a graduate student in the lab of Berislav Zlokovic at the University of Rochester, New York, described a transcriptional mechanism that may regulate clearance of vascular Aβ in the brain. This work, which will appear in the February issue of Nature Cell Biology (DOI: 10.1038/ncb1819), extends an earlier study by Zlokovic's group showing that AD patients have increased levels of the transcriptional co-activators myocardin (MYOCD) and serum response factor (SRF) (see ARF related news story). MYOCD is restricted to cardiac muscle and smooth muscle cells, whereas SRF is found in all cell types. Bell and colleagues found that MYOCD/SRF pathway regulates Aβ clearance via low-density lipoprotein receptor-related protein 1 (LRP1), a key Aβ clearance receptor (see ARF related news story).—Esther Landhuis.