New Tack on Aβ Oligomer Role in Disease and Treatment
First published at alzforum.org
As the case builds that soluble oligomers are the most toxic forms of Aβ in Alzheimer’s disease, researchers are working out how best to characterize and target these species. As evidenced by presentations at the 14th International Athens/Springfield Symposium on Advances in Alzheimer Therapy, held March 9 to 12 in Athens, Greece, progress has been slow and breakthroughs remain elusive. Scientists presented snippets of data that suggest Aβ oligomers are physically separate from neurofibrillary tangles, hinting that each of the two spreads through the brain on its own, not together. Others pinpointed a signaling pathway by which Aβ oligomers can lead to the hyperphosphorylation and missorting of tau, and antibodies that might target them therapeutically. “Certainly oligomers are a treatment target,” wrote John Morris, Washington University School of Medicine, St. Louis, to Alzforum. “However, the precise relationship of oligomers to amyloid plaques and neurofibrillary tangles has yet to be determined,” he emphasized. Morris was not involved in the research.
Scientists have puzzled for years about how amyloid plaques and neurofibrillary tangles might intersect, since they form in the brain at different stages of disease and in different places (see Thal et al., 2014). Could there be a tighter relation between soluble oligomers and tangles? So far, much of the data on Aβ oligomers and their location in the brain has come from animal studies. However, Ezio Giacobini, Geneva University Hospitals, who co-organized the conference, wanted to find out where oligomers appear in people’s brains and how they relate to other protein pathologies. He examined postmortem brain tissue from 30 older people with pathologically confirmed mild to severe Alzheimer’s disease and from 43 cognitively healthy controls. Using immunohistochemistry, he detected tau tangles with AT8, amyloid with 4G8, and Aβ oligomers with NU1, an antibody that specifically recognizes synthetic Aβ oligomers. He found that while oligomers and amyloid plaques turned up together in the entorhinal cortices of AD patients, tangles appeared in the hippocampi.
The results suggest that certain oligomers are unlikely to cause tau pathology, said Karen Ashe, University of Minnesota, Minneapolis. However, some forms of oligomer might, she noted. Ashe recently reported that oligomers fall into two main classes. Type 1 are formed from monomers and distributed throughout the brain, whereas type 2 appear stuck close to amyloid plaques that catalyze their formation (June 2015 news). Based on the pattern of staining around plaques, NU1 appears to label type 2 oligomers, Ashe said. Giacobini said that is not certain. Ashe said the results leave open the possibility that type 1 oligomers could co-localize with tangles. A type 1-specific antibody would be required to find out, but none is available, she said. Sylvain Lesné, University of Minnesota, Minneapolis, agreed that multiple antibodies will be required to label different oligomer species and get a fuller picture of their distribution throughout the brain.
Giacobini went on to examine human brain tissue by western blot. In extracts from controls with and without plaques, and from AD patients, he found an NU1-positive band running at 55kDa. AD patients had 40 percent more of this protein in the temporal cortex than did controls. Giacobini told Alzforum that since this oligomer was found in healthy people, he believes it is a normal occurrence in the brain and has some function. Though he detected it in extracts from the cerebellum, which normally are devoid of plaques, he speculated that it reaches toxic levels in disease. While based on its size, this 55kDa protein could be a 14-mer, Giacobini said it may be a dodecamer, like Aβ*56 that was isolated by Ashe’s group. Aβ*56 tracks with cognitive decline in the Tg2576 mouse model of AD, and correlates with neurodegenerative markers in the cerebrospinal fluid of humans (Lesné et al., 2006; Lesné et al., 2008; Zahs and Ashe, 2013).
Lesné, who was not involved in the work, cautioned that the 55kDa band that Giacobini saw may not be Aβ*56. It could be another oligomer with a similar molecular weight to Aβ*56, or it could be a cleavage fragment of the amyloid precursor protein (APP), he suggested. When Lesné measured Aβ*56 in postmortem brain tissue, he found markedly less in those with AD compared to controls, unlike the 55kDa oligomer (Lesné et al., 2013).
For his part, Lesné is looking into the effects of Aβ oligomers on neurons. Scientists have previously reported that oligomers of synthetic Aβ cause hyperphosphorylation and missorting of tau into dendrites (Sep 2010 news on Zempel et al., 2010; De Felice et al., 2008). In Athens, Lesné presented data suggesting Aβ*56 plays a role in this process as well. In the synapses of cultured primary cortical neurons, he found that the oligomer bound NMDA receptors.
This opened the receptors’ ion channels, allowing calcium ions to flood in and activate the Ca2+-dependent calmodulin kinase (CaMKIIα). CaMKIIα then doubled phosphorylation of tau at serine 416. Tau also missorted into the dendritic compartment of these neurons. Blocking CaMKIIα with its endogenous inhibitor, CN21, prevented tau from being hyperphosphorylated and missorted. Lesné saw a similar uptick in activated CaMKIIα in Tg2576 and J20 mice. He also detected elevated CaMKIIα in the inferior temporal gyrus of postmortem brains of cognitively normal elderly.
“We think these results could explain the tau pathology detected in the inferior temporal gyrus of aged subjects at risk for AD,” Lesné wrote to Alzforum. This is one plausible link between Aβ and tau pathology, said Giacobini, though he pointed out that the bulk of these results come from transgenic mice.
Lesné noted that there are likely multiple oligomeric species of Aβ that interact with other receptors and alter tau in different ways. He previously reported that Aβ dimers bind the cellular form of the prion protein, which binds to and activates the Src kinase Fyn, which then hyperphosphorylates tau (Larson et al., 2012). Different oligomer-specific effects on neurons could predominate at various stages of disease, he added, since he had previously found that the relative abundance of Aβ dimers, trimers, and Aβ*56 oligomers changed with age and disease (Lesné et al., 2013).
Some researchers are exploring antibodies that could target various Aβ oligomeric species therapeutically.William Klein, Northwestern University, Evanston, Illinois, who co-founded Acumen Pharmaceuticals, has worked with scientists there to develop an oligomer-specific antibody they called ACU-193 (formerly 19.3). This is the humanized version of a mouse antibody raised to ADDLs, soluble oligomers of synthetic Aβ generated by a procedure developed in Klein’s lab. Their data suggest that ACU-193 binds both type 1 and 2 oligomers, but not Aβ monomers or fibrils, wrote William Goure of Acumen to Alzforum.
In Athens, Klein presented preclinical data on this antibody and introduced a set of single-chain variable fragment antibodies that selectively target oligomer subspecies. In cultured hippocampal neurons, ACU-193 bound no thioflavin-positive plaques. It did bind oligomers, preventing their interaction with the membrane, and this rescued long-term potentiation as well as the oligomer-induced calcium overload. In Tg2576 mice at any age, ACU-193, given intravenously, also bound oligomers. It caused no microhemorrhages, which have been a side effect of some therapeutic antibodies for Aβ. In behavioral assays, the murine precursor of ACU-193, ACU-3B3, reduced hyperactivity in the open field and Y-arm tests in J20 mice, as well as learning deficits seen in the visible, but not hidden version of the Morris water maze. Enough antibody crossed into the brain in mice, rats, dogs, and rhesus monkeys, to be detected by immunofluorescence, and it bound Aβ oligomers with nanomolar affinity, Klein said.
The company has not directly examined binding in postmortem human AD tissues, Goure wrote, but Brian Bacskai and colleagues at Massachusetts General Hospital, Charlestown, found that ACU-3B3 bound soluble oligomers from postmortem AD brain tissue, said Klein. Klein and colleagues previously reported that ACU-193 and ACU-3B3 bind oligomers in human cerebrospinal fluid, and detect more oligomers in people with AD (Savage et al., 2014; Yang et al., 2015). Klein said that given these positive preclinical results and good safety data, he envisions clinical trials for ACU-193 and is trying to raise funds.
Given lingering uncertainty about which forms of oligomer are relevant to Alzheimer’s, Giacobini said it was hard to know whether this antibody would work in humans, and at which stage of disease it would be most effective. Ashe agreed. "Not all oligomers are alike, and people are beginning to accept the idea that there may be more than one type,” she told Alzforum. “It’s important to understand more about the different types of oligomer, and characterize antibodies in terms of whether they target type 1 and/or type 2." David Brody, Washington University, St. Louis, Missouri, agreed. “There is a lot of controversy about which oligomers exist in the human brain; many of the synthetic, cell culture, and transgenic animal-derived forms seem quite different from what has been observed in humans,” he said.
—Gwyneth Dickey Zakaib
6 April 2016