9 April 2010. Taking a page from research on previously identified mutations causing early-onset Alzheimer disease, Japanese scientists have now incorporated an unusual variant of amyloid precursor protein (APP) into a new mouse model. Aβ peptides produced from this APP mutant readily oligomerize in vitro, and mice that overexpress it show synaptic and cognitive impairments, tau hyperphosphorylation, glial activation, and neuron loss—all in the absence of amyloid plaques—according to a study in the April 7 Journal of Neuroscience. Despite some researchers’ concerns about the evidence for the presence of Aβ oligomers and the absence of Aβ fibrils in these mice, their striking phenotype makes them a powerful tool for understanding this APP mutation and how it causes disease.
Researchers led by Hiroshi Mori of Osaka City University, Japan, first identified the rare APP mutation (E693Δ) in patients who presented with clinically typical Alzheimer disease, yet had low brain amyloid on in vivo imaging by positron emission tomography (Tomiyama et al., 2008 and ARF related news story). The E693Δ mutation produces a form of Aβ that lacks glutamate-22 (E22Δ), and shows a recessive pattern of inheritance in affected families. Mutant Aβ arising from this APP variant forms oligomers but no detectable fibrils in vitro, and inhibits long-term potentiation more strongly than wild-type Aβ peptides when injected into rat cerebral ventricles.
To see whether this oligomerization-prone Aβ mutant could bring on other pathological features of AD in vivo, first author Takami Tomiyama and colleagues generated transgenic mice expressing the E693Δ APP under the mouse prion promoter. They put these animals through various immunohistochemical, behavioral, and electrophysiology tests, comparing them with transgenic mice expressing wild-type human APP under the same promoter. APP expression levels in the two strains were low (i.e., comparable to endogenous mouse APP and about 1/10 that of mutant APP in Tg2576 mice) and differed at most twofold (higher in wild-type) among the lines used for immunohistochemistry and behavioral analyses.
In contrast to the Tg2576 controls, which had abundant plaques in the cerebral cortex and hippocampus at 24 months, age-matched E693Δ APP mice had no discernible plaques in these brain areas. The mutant APP transgenics did, however, show extensive intraneuronal Aβ staining in these regions, whereas the wild-type APP mice did not.
Since E22Δ Aβ in their hands oligomerizes quickly in transfected cells (Nishitsuji et al., 2009), the researchers presumed the intraneuronal Aβ signal in the E693Δ APP mice corresponded to oligomers. This appears to be the case, as shown in a separate set of cortical and hippocampal stainings using an antibody (NU-1) that selectively detects Aβ oligomers (Lambert et al., 2007). E693Δ APP mice racked up oligomeric Aβ in the cortex and hippocampus in an age-dependent fashion, starting at eight months, whereas no NU-1 signal came up in 24-month-old wild-type APP transgenic or non-transgenic controls.
Strangely, Western blots of Aβ-immunoprecipitated brain extracts revealed oligomers, which are presumed to be soluble, in the detergent-insoluble fractions—more in E693Δ than wild-type APP mice, and predominantly in the most stringent extractions done with formic acid. Scientists contacted for comment on the study raised this at best as a curiosity, at worst a key flaw. Karen Hsiao, University of Minnesota, Minneapolis, and colleagues wondered whether it could be an artifact of the extraction protocol. They noted that the current study could be solidified with biochemical data showing that Aβ oligomers are compartmentalized intracellularly, to confirm the NU-1 immunohistochemical stainings. The authors acknowledge in their paper that the detection of Aβ oligomers largely in insoluble fractions flies in the face of conventional belief, and noted in an e-mail to ARF that they are performing further studies to characterize both extracellular and intracellular Aβ oligomers.
The researchers also found no evidence that the intraneuronal Aβ in E693Δ APP mice forms fibrils. Whereas brain sections from 24-month-old Tg2576 mice showed strong staining with the amyloid-binding dye thioflavin S, tissue from age-matched E693Δ APP mice showed no signal besides a weak, diffuse staining within neurons in the cortex and hippocampus. Other scientists raised the possibility that the inability of these mice to form plaques could arise from their low expression of APP. Some pointed out that the Western blot showing Aβ oligomers in the detergent-insoluble fractions was cut such that it only showed monomers, dimers, and trimers, but not Aβ*56 and other higher-weight oligomeric forms.
What’s less disputable are the striking phenotypes seen in the E693Δ APP mice. Compared with wild-type APP transgenic and non-transgenic controls, these animals show an age-dependent drop in synaptic density, measured in brain sections stained with an antibody to the presynaptic marker synaptophysin. In addition, E693Δ APP mice had defects in short- and long-term synaptic plasticity in electrophysiology studies, and did worse in the Morris water maze (which measures spatial memory). The researchers documented all these defects starting around eight months, the age at which oligomeric Aβ was first detected in the E693Δ APP mice.
Beyond the synaptic and behavioral woes, brain sections of E693Δ APP transgenic mice showed evidence of abnormal tau phosphorylation (judged by staining with antibodies to pathological tau, PHF-1, and MC1) and glial activation (measured by antibodies to astrocyte and microglia markers GFAP and Iba-1, respectively). These abnormalities showed up starting at 12 months, and worsened with age. By 24 months, E693Δ APP transgenics also showed appreciable neuronal loss (judged by staining with an antibody to the mature neuron marker NeuN) in the hippocampal CA3 region, though not in cerebral cortex.
Amid concerns about the biochemical data, the E693Δ APP mouse appears to be the first demonstration “that neuronal loss was induced by Aβ oligomers alone in vivo in the absence of amyloid plaques,” the authors write. Despite uncertainty as to whether the intraneuronal Aβ is exclusively oligomeric, “the paper provides evidence for a provocative, new role for APP and intracellular fragments and will be of great interest to the AD community,” wrote Cindy Lemere, Brigham and Women’s Hospital, Boston, in an e-mail to ARF.—Esther Landhuis.
Tomiyama T, Matsuyama S, Iso H, Umeda T, Takuma H, Ohnishi K, Ishibashi K, Teraoka R, Sakama N, Yamashita T, Nishitsuji K, Ito K, Shimada H, Lambert MP, Klein WL, Mori H. A Mouse Mode of Amyloid Beta Oligomers: Their Contribution to Synaptic Alteration, Abnormal Tau Phosphorylation, Glial Activation, and Neuronal Loss In Vivo. J. Neurosci. 7 April 2010;30(14):4845-4856. Abstract