Read a PDF of the entire series.

Call it nanonova. Okay, the analogy is far-fetched, for unlike a supernova, an exploding neuron hardly spews out tons of debris at lightning speed. But if Charlie Glabe is right, an exploding neuron does expel material that then triggers the formation of new, stable matter in its surroundings, and this is how amyloid plaques are born. This, at least, was one of the more stellar narratives to emerge from the 11th International Conference on Alzheimer’s and Parkinson’s Diseases, held 6-10 March 2013 in Florence, Italy. In a provocative talk that reverberated in the hallways, Glabe, at the University of California, Irvine, showed images of dying neurons, their nuclei—yes, nuclei–stuffed with amyloid. The work came out of a project to create tools to distinguish between the different flavors of the Aβ peptide, in particular to detect the small oligomeric forms that many scientists believe to be the most deadly. The data could help resurrect an old hypothesis, which holds that ruptured neurons form the core of neuritic plaques. On the broader question of how to visualize and track oligomeric forms of pathogenic proteins, numerous talks and posters at AD/PD 2013 showcased conformation-specific antibodies.

image

Neuritic plaque contains fibrillar amyloid (red), APP-CTF (green), and DNA (blue), suggesting its nuclear origin. Image courtesy of Ricardo Albay III

“This is a bold idea,” Gunnar Gouras at Lund University, Sweden, told Alzforum. “Glabe has been a trailblazer in this area.” Gouras pointed out that traditional ways of looking at amyloid often involve grinding up brain tissue, but antibodies such as the one Glabe made can see native protein forms in unfixed tissue, providing a glimpse into the natural pathophysiology. Other scientists expressed surprise at seeing evidence of Aβ in the neuron’s nucleus, a place the peptide is not thought to frequent.

Glabe previously generated antibodies specific for oligomeric Aβ by using a synthetic peptide, and he is developing an active vaccination strategy with this peptide as a potential therapy (see ARF related news story; ARF news story). In Florence, Glabe asked a broader question: Is the wide variation in Aβ species truly significant for the pathogenesis and heterogeneity of Alzheimer’s disease? To address this, Glabe immunized rabbits with Aβ42 and analyzed the resulting antibodies on an array dotted with 132 different forms of amyloid, including Aβ monomers, oligomers, fibrils, and islet amyloid polypeptide. He selected for further analysis 24 antibody clones that each gave different patterns of reactivity on the array. This process generated several new research tools, he noted, such as antibody M31, which stains only vascular amyloid deposits.

Do Dying Neurons Seed Plaques?
Most of his AD/PD talk, however, focused on antibody M78, which recognizes fibrillar protein forms made up of parallel β-sheets. M78 is not specific for Aβ; it also reacts with fibrillar APP and α-synuclein deposits, Glabe said. In aged human brains with some AD pathology, M78 stains nuclear and perinuclear regions of neurons, oligodendrocytes, and astrocytes in most brain regions. It seems to mark an early stage of disease, as M78-positive brains came from people who had either normal cognition or mild cognitive impairment (MCI). M78 staining goes down as more plaques deposit.

Glabe was particularly intrigued by the abnormal morphology of M78-positive neurons. Their nuclear envelope was swollen, DNA fragmented, and cell bodies filled with multivesicular bodies. Glabe used 3xTg mice to examine these changes with age. At 10 months, these animals showed the earliest M78 staining. It was perinuclear and co-localized with β-CTF immunoreactivity, implying that APP may have been cleaved by BACE1 nearby. By 12 months, M78 staining showed up in the nucleus, while β-CTF remained on the outside. By 14 months, the antibody marked the remains of a nucleus containing diffuse DNA, in the center of a neuritic plaque surrounded by ruptured membranes that stain for β-CTF. The plaque center also stained with the neuronal marker NeuN, confirming its neuronal origin. Glabe pointed out that all of these molecules maintained the same spatial relationships with each other observed two months earlier. The findings imply that amyloid accumulation lyses nuclei, Glabe said.

“Each senile plaque arises from the demise of a single neuron,” Glabe concluded. In his model, what scientists call “dystrophic neurites” are not neurites at all, but rather the remains of the exploded nuclear envelope and endoplasmic reticulum.

Glabe noted that this is not a new idea. Michael D’Andrea at The RW Johnson Pharmaceutical Research Institute, Spring House, Pennsylvania, suggested in the year 2000 that lysed neurons form amyloid plaques (see D’Andrea et al., 2001). Numerous other researchers, such as Gerd Multhaup at the Free University of Berlin, Germany, and Frank LaFerla at the University of California, Irvine, have seen intraneuronal amyloid, although controversy has erupted over whether this is actually Aβ or APP (see ARF Webinar). Reports of intranuclear Aβ, on the other hand, are quite rare, and turned heads in Florence. Glabe is careful to note that M78 is not necessarily detecting Aβ in the nucleus; it could be aggregated APP or C-terminal fragments of APP.

Numerous researchers at the conference praised the quality of the work and debated the questions it raises. “There is definitely a cell soma at the heart of plaques,” said Virgil Muresan at the University of Medicine and Dentistry of New Jersey, Newark. He pointed out that the perinuclear compartment, where Glabe first sees Aβ and β-CTF accumulation, plays an important role in APP processing, making it a logical place for deposits to start. Christian Haass at Ludwig-Maximilians University, Munich, Germany, observed that, because M78 recognizes several types of β-sheet fibrils, further work is needed to determine exactly what protein the antibody is detecting in the nucleus. He questioned whether Aβ itself could get in, given that the peptide is typically exported. Maybe it moves in after the nuclear envelope breaks down, he suggested. Gouras suggested that live imaging of cells would help nail down the progression of events, although he acknowledged that this would be technically challenging to do. Another unanswered question is how Aβ gets inside cells in the first place. It may simply accumulate in endocytic or secretory compartments where it is processed or, alternatively, cells may take up oligomeric Aβ from the outside, Glabe said.

Specific α-Synuclein Antibodies for Parkinson’s Disease Research
Meanwhile, Parkinson’s disease researchers are also developing antibodies to various toxic forms of α-synuclein, the main pathogenic protein in Parkinson’s disease. They did not deploy the antibodies to trace the provenance of Lewy bodies, but to instead characterize their potential for biomarker or therapeutic use. Therese Fagerqvist, working with Lars Lannfelt, Martin Ingelsson, and Joakim Bergström at Uppsala University, Sweden, immunized mice with α-synuclein oligomers made using an in-house process (see Näsström et al., 2011). In collaboration with BioArctic Neuroscience AB, Stockholm, Sweden, she selected for two monoclonal antibodies, mAb38F and mAb38E2, that have 150 times more affinity for α-synuclein oligomers than monomers but show no cross-reactivity with aggregates of tau, Aβ, or other synucleins. These antibodies detect pathology in A30P α-synuclein transgenic mice earlier than do commercial antibodies such as Syn-1, Fagerqvist reported. The new antibodies revealed that α-synuclein oligomer levels increase 10-fold with age in the brain and spinal cord of A30P mice. Higher oligomer levels in the endoplasmic reticulum coincide with behavioral symptoms, she noted (see Fagerqvist et al., 2013). The antibodies prevent α-synuclein oligomerization in cellular assays. When the researchers injected a related antibody, mAb47, into 14-month-old mice weekly for three and a half months, it lowered α-synuclein oligomer levels in brain extracts. The researchers are currently testing behavior.

In contrast, Gabor Kovacs at the Medical University of Vienna, Austria, described an antibody, 5G4, that he said reacts only with disease-associated forms of α-synuclein, both fibrillar and oligomeric (see Kovacs et al., 2012). 5G4 was developed in collaboration with biotech company AJ RoboScreen GmbH, Leipzig, Germany. An ongoing clinical study with 200 participants is testing its use as a biomarker of PD (see press release). In his talk, Kovacs also claimed that 5G4 detects pathology with more sensitivity than commercial antibodies. It stains deposits in the neuronal cytoplasm and processes, as well as in astrocytes. In ultrastructural studies done in collaboration with Lajos László at Eötvös University of Science, Budapest, Hungary, the researchers also see α-synuclein in endosome-like structures in neurons, supporting findings that the protein may enter cells through this route (see ARF related news story). In neurons, small α-synuclein aggregates were surrounded by clusters of mitochondria, while astrocytes contained fibrillar structures. PD patients who rapidly developed dementia had higher levels of antibody staining in limbic regions compared to those with stable cognition or slowly progressing dementia, implying that spread of α-synuclein pathology to this brain area might explain some cases of dementia associated with Lewy pathology. The antibody gives a signal even from old brain sections that have been preserved for years in formalin, suggesting it could be useful in re-evaluating archival material, Kovacs added.

Several other groups presented posters in this area. Omar El-Agnaf at United Arab Emirates University, Al Ain, has made several monoclonal antibodies specific for either oligomeric (Syn-O1, Syn-O2, Syn-O3, Syn-O4) or fibrillar (Syn-F1, Syn-F2) forms of α-synuclein, which he reported to have low affinity for monomers and do not cross-react with aggregates of tau, Aβ, other forms of amyloid, or synuclein. In previous work using a different antibody, El-Agnaf showed higher levels of α-synuclein oligomers in the cerebrospinal fluid of PD patients compared to controls, suggesting this could make a biomarker (see ARF related news story). Likewise, Harald Weninger and Markus Mandler at AFFiRiS AG, Vienna, Austria, described in a poster the development of their α-synuclein antibody, PD01. It curtails α-synuclein pathology and neurodegeneration in PD/DLB mice while improving their performance in the Morris water maze, Weninger said (see ARF related news story and Schneeberger et al., 2012). He stayed mum on the details of how the antibody was made or what form of α-synuclein it recognizes, however. The antibody crosses the blood-brain barrier and is in Phase 1 trials for PD.

The preceding studies represent only a smattering of the work in this area, all of which may provide researchers with new tools for investigating the role of specific conformations of pathogenic proteins in disease. We invite readers to add other examples through Alzforum commentary.—Madolyn Bowman Rogers.

Comments

  1. I'd like to clarify and extend my quoted comments about the interesting talk by Charlie Glabe at the AD/PD meeting. I was mainly referring to his new images of nuclear aggregates. Charlie did also show very nice new images of plaques appearing to arise from neuron cell bodies, although as he indicated, this part is not new. We had shown β amyloid aggregates in neurons in human AD brains back in 2000, and Michael D’Andrea very nicely showed this as described in the Alzforum news story. More recently, Bob Vassar’s group convincingly showed this in their 5xFAD mice (Eimer and Vassar, 2013). It seems quite clear that some plaques, particularly dense core plaques with dystrophic neurites, represent degenerated neurons. A challenge with interpreting such neuron cell body-associated plaques is to understand how they arise. What led our group away from cell soma was our subsequent immuno-EM for Aβ42, which pointed to dystrophic neurites and synapses as preferential sites of Aβ42 accumulation.

    At least three pieces of information highlight synapses as critical sites for Aβ:

    1. The physical association of Aβ42 accumulation with neuritic/synaptic pathology (Takahashi et al., 2002).

    2. That adding Aβ1-42 targets synapses in cultured neurons (Lacor et al., 2004).

    3. Conditional knockout of PS1 preferentially led to accumulation of APP CTFs within synaptic terminals (Saura et al., 2005; while this paper emphasized presynaptic terminals, abundant, albeit anatomically selective, postsynaptic accumulation was also pronounced; see Gouras et al., 2010).

    Moreover, APP is transported to synapses and preferentially processed to Aβ as a result of synaptic activity. We therefore favor that even when neuron cell bodies appear to be sites of plaque formation, synaptic sites within neuritic trees (also when associated with neuron soma) are initial sites of APP/Aβ aggregation, and thus the initial nidus of plaque formation. This scenario would also fit better with the presence of plaques in anatomic areas with abundant synaptic connections but few, if any, neuron soma, such as in the outer molecular layer of the dentate and the CA1 lacunosum-moleculare of hippocampus, as well as the superficial layer of cortex. We further envision that hyperexcitable neurons secreting excess extracellular Aβ contribute to Aβ aggregation within the intraneuronal Aβ accumulating, hypoactive neurons.

    The presence of nuclear aggregates was the most novel and surprising part of Charlie Glabe’s interesting talk. Moreover, I am aware that Gerd Multhaup’s group also has some intriguing data on β amyloid within the nucleus.

    References:

    . Neuron loss in the 5XFAD mouse model of Alzheimer's disease correlates with intraneuronal Aβ42 accumulation and Caspase-3 activation. Mol Neurodegener. 2013;8:2. PubMed.

    . Intraneuronal Alzheimer abeta42 accumulates in multivesicular bodies and is associated with synaptic pathology. Am J Pathol. 2002 Nov;161(5):1869-79. PubMed.

    . Synaptic targeting by Alzheimer's-related amyloid beta oligomers. J Neurosci. 2004 Nov 10;24(45):10191-200. PubMed.

    . Intraneuronal beta-amyloid accumulation and synapse pathology in Alzheimer's disease. Acta Neuropathol. 2010 May;119(5):523-41. PubMed.

  2. We see strikingly similar immunoreactive signals in brain sections from 15-month-old immune-challenged transgenic mice (3xTgAD and ArcAβ lines) and in postmortem AD samples using anti-APP (22C11) and anti-Aβ1-40/42 antibodies, which label dystrophic neurites and plaque cores (Fig. 7, Krstic et al., 2012). Both DAPI and FluoroJade B signals were also detected in the cores of plaques. While FluoroJ was quite intense, the DAPI signal was more diffuse, very much in line with Charlie Glabe's findings.

    However, based on the high affinity of these dyes for acidic residues (see Mazzini et al., 1997), we recently revised our interpretation of degenerating neurons as the source these DAPI/FluoroJ signals. We rather hypothesize that these dyes strongly label the acidic core of the plaque that is associated with large neuritic varicosities and axonal leakages, putative hotspots of amyloid-β peptide generation (Krstic and Knuesel, 2013).

    This view is in line with experimental data obtained from a large collection of postmortem human brain slices showing extensive axonal leakages that are associated with amyloid plaques (Xiao et al., 2011), and many other findings that we have recently summarized in our review (Krstic and Knuesel, 2013).

    References:

    . Systemic immune challenges trigger and drive Alzheimer-like neuropathology in mice. J Neuroinflammation. 2012;9:151. PubMed.

    . Interaction of DAPI with pepsin as a function of pH and ionic strength. Biophys Chem. 1997 Sep 1;67(1-3):65-74. PubMed.

    . The origin and development of plaques and phosphorylated tau are associated with axonopathy in Alzheimer's disease. Neurosci Bull. 2011 Oct;27(5):287-99. PubMed.

    . Deciphering the mechanism underlying late-onset Alzheimer disease. Nat Rev Neurol. 2013 Jan;9(1):25-34. PubMed.

Make a Comment

To make a comment you must login or register.

References

News Citations

  1. San Diego: The Future of AD—Can We Vaccinate?
  2. DC: Anti-Aβ Treatments Old and New
  3. Tau, α-Synuclein Spread: Crazy Stuff—How Might It Work?
  4. An Oligomer Test for PD—Could One for AD Be Close Behind?
  5. Stockholm: New Strategies for Immunotherapy

Webinar Citations

  1. Intraneuronal Aβ: Was It APP All Along?

Paper Citations

  1. . Evidence that neurones accumulating amyloid can undergo lysis to form amyloid plaques in Alzheimer's disease. Histopathology. 2001 Feb;38(2):120-34. PubMed.
  2. . The lipid peroxidation products 4-oxo-2-nonenal and 4-hydroxy-2-nonenal promote the formation of α-synuclein oligomers with distinct biochemical, morphological, and functional properties. Free Radic Biol Med. 2011 Feb 1;50(3):428-37. PubMed.
  3. . Monoclonal antibodies selective for α-synuclein oligomers/protofibrils recognize brain pathology in Lewy body disorders and α-synuclein transgenic mice with the disease-causing A30P mutation. J Neurochem. 2013 Jan 31; PubMed.
  4. . An antibody with high reactivity for disease-associated α-synuclein reveals extensive brain pathology. Acta Neuropathol. 2012 Feb 28; PubMed.
  5. . Vaccination for Parkinson's disease. Parkinsonism Relat Disord. 2012 Jan;18 Suppl 1:S11-3. PubMed.

Other Citations

  1. Read a PDF of the entire series.

External Citations

  1. 3xTg mice
  2. A30P α-synuclein transgenic mice
  3. press release
  4. Phase 1 trials

Further Reading