To what extent is AD an acceleration of normal aging? This decades-old question receded in favor of the view that AD is a separate process from normal aging when studies showed that patterns of neuronal loss are different in aging and AD. Early AD entails selective loss of neuronal projections, such as the perforant path connecting the entorhinal cortex and the dentate gyrus of the hippocampus. Advanced AD features neuronal loss that far exceeds and differs in its regional pattern from that of normal aging. This anatomical data is undisputed and highlights the selective regional vulnerabilities in AD. But postmortem stereology is limited in that it assesses the brain after the initial pathogenesis has occurred, and therefore says less about cause than about features of disease progression. Hence, the question of aging versus AD continues to be debated. FAD mutations speak to both sides of the issue. They reduce the age at onset, diminishing the importance of age and validating overexpression approaches, but at the same time, even people whose FAD mutations flood their brains with Aβ from an early age appear healthy until their 40s. The current distinction between aging and AD is called into question by emerging comparisons of gene expression patterns between normally aging people and others with dementia.
One way to address the question is to study synaptic aging. Is Alzheimer disease a problem of synaptic maintenance? When synapses degenerate, is it a problem of "synaptosis" or "synecrosis"? In other words, does an active program inflict death by complement, aberrant reactivation of MCH class 1 proteins, or other outside signals, or does the synapse fall apart as synaptic organizing molecules disappear? GFP- and YFP-expressing transgenic mice allow repeated live imaging of peripheral synapses over time to monitor how synapses age normally and the building of a knowledge base for understanding how they age in AD models. Application of this method to neuromuscular junctions show that as these synapses age, they become fragmented and the postsynaptic side loses transmitter receptors and nerve contact. Aging nerve terminals sprout, leading to multiple innervations of a given postsynaptic site. Axonal dystrophy and balloon-like distensions of the axon are common. Over time, there is a net loss of innervation. The aging synaptic cleft widens in parallel with changes in the molecular composition of the basal lamina. The aging synapse loses expression of synaptic organizing molecules including agrin and certain types of laminin. Genetic deletion of one laminin form accelerates the morphological changes associated with aging synapses, suggesting that the molecular loss is one cause of the aging, rather than merely a correlate. Together, these results suggest that age-related synapse loss may be quite passive, a form of dedifferentiation where a developmental program rewinds itself as proteins that organize the synapse are progressively lost.
A related question concerns the unit that changes in aging and AD: do individual synapses weaken and disappear one at a time, as happens in activity-dependent synaptic plasticity, or does a given neuronal arbor lose all its synapses just before the axon retracts, much as a winter tree sheds its leaves? The two imply different proximal insults and mechanisms of synaptic loss. The answer to this question is still not clear, but studies of the YFP transgenic mice show that aging muscles tend to have fewer but larger motor units than in young mice, suggesting that some units lose their branches and die back, whereas others gain branches, overextending themselves before they, too, die. Synaptic activity is likely to be an important factor in this process, and could determine selective vulnerability of certain circuits and brain regions.
Technical problems of accessibility and size preclude application of this technique to CNS synapses, so it is unknown whether central circuits age similarly. It is already possible to examine the aging "neural unit," and rapidly improving methods may soon allow imaging of structural details at individual synapses. Further imaging and molecular studies of age-dependent changes in synaptic organizing molecules are needed to address this knowledge gap in AD research.
Live multiphoton microscopy enables prospective brain imaging of the fate of synapses in AD models, albeit at a lower level of resolution. For example, studies monitoring the temporal sequence of events in AD pathogenesis in APP and tau transgenic mouse models are challenging the conventional wisdom that AD represents a steady process of continuous decline. This work suggests instead that AD is punctuated by a series of fast, catastrophic changes. During the 20 years that AD lasts, association cortices lose most of their neurons, and spine and dendritic changes are too numerous to count. Even so, the underlying process may be one of spurts of sudden changes at the cellular and molecular level.
Clinical AD follows a prodromal phase of MCI, during which the brain already contains abundant plaques, tangles, and gliosis, and a majority of neurons have died in certain cortical areas. To identify which changes happen first, scientists monitor over time the same brain area in APP transgenic mice. This work shows that plaques do not grow gradually in size over months, as had been predicted; rather, they appear suddenly from one day to the next in their full size and then remain stable for months. This is consistent with the idea that they might precipitate out of solution. Likewise, within days of the appearance of a plaque, neurites abutting it become crooked and dystrophic. Some neurites stay that way for months; others break. Direct application of anti-Aβ antibody clears both plaques and neuritic dystrophy within a week. Dendrites in the vicinity of a plaque rapidly form new spines, but the spines are unstable, leading to net loss of dendritic spines near plaques within days of the appearance of a new plaque. Several studies have documented spine loss and dendritic changes near plaques (e.g., Tsai et al., 2004; Spires et al., 2005). These synaptic changes are a function of the APP transgene and independent of age.
Tangles in humans correlate with cognitive decline and neuronal loss, but neuronal loss exceeds the number of tangles (Gomez-Isla et al., 1996). Tangles can be imaged in Tg4510 mutant human tau transgenic mice, which develop tangle-like pathology and dramatic neuron loss in cortex and dentate gyrus (Spires, 2006). In these mice, individual tangle-bearing neurons remain stable for months even while neurons die in large numbers. Imaging reagents that fluoresce after a specific caspase has cleaved them indicate that tangle-bearing neurons are likely to activate caspases. They cleave tau, and still these neurons survive for months. Caspases remain present in the neuron after transgene expression is turned off. Markers of apoptosis are absent, yet monitoring of a given visual field captures occasional instances of a fluorescent neurite disintegrating over the course of 2 hours. This is seen only in Tg4510 mice, not in APP-transgenic or control mice. Together, these new data suggest that tangles and caspase activation precede additional biological changes that destroy the neuron by a still-mysterious mechanism.
The rapid appearance of fully grown plaques and the sudden collapse of tangle- and caspase-bearing neurites after months of apparent stability raise new questions about which biological changes dominate the long prodromal phase of human AD. Other mouse studies reporting that spine loss and synaptic changes precede plaques (e.g., Jacobsen et al., 2006), and work from multiple labs showing spine loss in response to oligomeric Aβ, keep alive the related debate about which Aβ species most damages spines. Imaging labels for diffusing, low-molecular-weight species of Aβ and tau are needed to resolve these issues in vivo.
Multiphoton microscopy of normally aging mouse brain has generated data to suggest that a large majority of dendritic spines remain stable and could serve to store long-term memories (Grutzendler et al., 2002; Zuo et al., 2005). More recent multiphoton imaging data implicate cortical microglia in dendritic spine plasticity. They show microglia to be highly plastic, reacting to an injury within hours by becoming activated and extending processes toward the injury (Davalos et al., 2005). The time course of this microglial activation parallels changes in spine dynamics following injury. The tools are in place for monitoring the dynamics of activated microglia around amyloid plaques.
Other areas of synaptic biology are expanding rapidly and should be explored for links to AD pathogenesis. They include the study of signaling between synapse and nucleus that enables those forms of long-lasting neuronal plasticity that require gene transcription. These pathways are being established in several laboratories (e.g., Martin and Zukin, 2006) and could be tested for their robustness in aging neurons and AD models. Another area that is already informing the study of AD is that of neuronal endocytosis. Ongoing work in the field is largely focusing on defects in vesicle transport, autophagy, and APP endocytosis and processing, yet additional proteins involved in the membrane traffic that controls synaptic vesicle recycling could expand this research area. Proteins of interest include the phosphoinositide phosphatase synaptojanin, an enzyme overexpressed in Down syndrome brain whose gene lies in the Down's region of chromosome 21. A cycling synaptic vesicle needs synaptojanin to shed its clathrin coat before it can be reloaded with transmitter. Proteins of growing interest also include the GTPase dynamin, a mechanoenzyme that, together with accessory proteins, pinches invaginated vesicles off the membrane (Roux et al., 2006). Researchers are beginning to understand that some isoforms of dynamin are sufficient for basal endocytosis, whereas the brain-specific version dynamin1 handles increased synaptic demand during intense stimulation/excitation. The field is beginning to generate hypotheses about dynamin's role in AD pathogenesis, but none have yet been independently confirmed and widely accepted. For example, dynamin has been implicated in genetic AD risk (Kuwano et al., 2006), in APP endocytosis (see ARF Eibsee conference report), and downstream of Aβ action on NMDA receptors (Kelly and Ferreira, 2006).
The question of Aβ effects on postsynaptic receptors of excitatory transmission has become an area of active investigation after a report proposed a negative feedback loop in which synaptic activity would increase APP processing, and Aβ in turn would restrain activity (Kamenetz et al., 2003). The same group has since expanded this initial data. New results suggest that Aβ likely destabilizes synapses by recruiting some of the postsynaptic signaling mechanisms that downregulate AMPA receptors during normal instances of long-term depression (LTD). This would, in effect, generate a situation of Aβ-induced chronic LTD that would weaken first the synapse and then the spine carrying it. How loss of AMPA receptors causes the spines to disappear is unclear, but researchers know that the intracellular tails of these receptors serve as organizing principals for other components of the postsynaptic density. Other labs have similarly implicated Aβ in glutamate receptor loss (see Chang et al., 2006; Cirrito et al., 2005; Snyder et al., 2005; Almeida et al., 2005). Together, the studies are generating interest in the role of LTD in aging and AD research. Big questions remain. Researchers have not definitively shown with which receptors Aβ may interact directly as opposed to which ones are affected secondarily, or which forms of Aβ have these effects in vivo. They also have not yet defined a physiological function of Aβ on glutamate receptors vis-a-vis an age- or disease-related one, or reached consensus on pre- versus postsynaptic generation of Aβ. α7 nicotinic acetylcholine receptors also are likely to play a role in Aβ-induced synaptic dysfunction, but a clear pathway has not been delineated in vivo.
Another question that is being actively studied concerns whether APP cleavage products other than Aβ can impair synaptic function. Some research indicates that the cytoplasmic C-terminus generated by BACE and γ-secretase cleavage undergoes further caspase cleavage and then becomes toxic to synaptic transmission. Called C31, this fragment could account for some of the behavioral abnormalities in APP transgenic mice. Evidence for the toxicity of C31 exists in vitro and in vivo. The latter suggests that a mutation abolishing the caspase cleavage site on APP's C-terminus, when crossed with APP-transgenic mice, rescues the deficits in hippocampal synaptic transmission and in the Morris water maze that several labs have established for the APP-transgenics alone. The synaptic rescue occurs in the presence of a full complement of amyloid plaques. These data point to an intracellular pathway by which aberrant APP processing could first lead to synaptic apoptosis and later to the death of the whole neuron (Galvan et al., 2006; also independent, unpublished work). The hypothesis proposes that high levels of Aβ are necessary but insufficient to cause the synaptic damage and neuron death seen in AD. Rather, Aβ induces formation of APP complexes that trigger the toxic caspase cleavage (Shaked et al., 2006). This work needs more study, particularly on the precise function of C31. This question, like many in AD mouse genetics, would benefit from the development of knock-in models, as well as cell-type-specific inducible knockouts of genes of interest in given brain regions, such as the CA1/CA3 synapse in the hippocampus.
- Tsai J, Grutzendler J, Duff K, Gan WB. Fibrillar amyloid deposition leads to local synaptic abnormalities and breakage of neuronal branches. Nat Neurosci. 2004 Nov;7(11):1181-3. PubMed.
- Spires TL, Meyer-Luehmann M, Stern EA, McLean PJ, Skoch J, Nguyen PT, Bacskai BJ, Hyman BT. Dendritic spine abnormalities in amyloid precursor protein transgenic mice demonstrated by gene transfer and intravital multiphoton microscopy. J Neurosci. 2005 Aug 3;25(31):7278-87. PubMed.
- Gómez-Isla T, Price JL, McKeel DW, Morris JC, Growdon JH, Hyman BT. Profound loss of layer II entorhinal cortex neurons occurs in very mild Alzheimer's disease. J Neurosci. 1996 Jul 15;16(14):4491-500. PubMed.
- Spires TL, Orne JD, Santacruz K, Pitstick R, Carlson GA, Ashe KH, Hyman BT. Region-specific dissociation of neuronal loss and neurofibrillary pathology in a mouse model of tauopathy. Am J Pathol. 2006 May;168(5):1598-607. PubMed.
- Jacobsen JS, Wu CC, Redwine JM, Comery TA, Arias R, Bowlby M, Martone R, Morrison JH, Pangalos MN, Reinhart PH, Bloom FE. Early-onset behavioral and synaptic deficits in a mouse model of Alzheimer's disease. Proc Natl Acad Sci U S A. 2006 Mar 28;103(13):5161-6. Epub 2006 Mar 20 PubMed.
- Grutzendler J, Kasthuri N, Gan WB. Long-term dendritic spine stability in the adult cortex. Nature. 2002 Dec 19-26;420(6917):812-6. PubMed.
- Zuo Y, Lin A, Chang P, Gan WB. Development of long-term dendritic spine stability in diverse regions of cerebral cortex. Neuron. 2005 Apr 21;46(2):181-9. PubMed.
- Davalos D, Grutzendler J, Yang G, Kim JV, Zuo Y, Jung S, Littman DR, Dustin ML, Gan WB. ATP mediates rapid microglial response to local brain injury in vivo. Nat Neurosci. 2005 Jun;8(6):752-8. PubMed.
- Martin KC, Zukin RS. RNA trafficking and local protein synthesis in dendrites: an overview. J Neurosci. 2006 Jul 5;26(27):7131-4. PubMed.
- Roux A, Uyhazi K, Frost A, De Camilli P. GTP-dependent twisting of dynamin implicates constriction and tension in membrane fission. Nature. 2006 May 25;441(7092):528-31. PubMed.
- Kuwano R, Miyashita A, Arai H, Asada T, Imagawa M, Shoji M, Higuchi S, Urakami K, Kakita A, Takahashi H, Tsukie T, Toyabe S, Akazawa K, Kanazawa I, Ihara Y, . Dynamin-binding protein gene on chromosome 10q is associated with late-onset Alzheimer's disease. Hum Mol Genet. 2006 Jul 1;15(13):2170-82. PubMed.
- Kelly BL, Ferreira A. beta-Amyloid-induced dynamin 1 degradation is mediated by N-methyl-D-aspartate receptors in hippocampal neurons. J Biol Chem. 2006 Sep 22;281(38):28079-89. PubMed.
- Kamenetz F, Tomita T, Hsieh H, Seabrook G, Borchelt D, Iwatsubo T, Sisodia S, Malinow R. APP processing and synaptic function. Neuron. 2003 Mar 27;37(6):925-37. PubMed.
- Chang EH, Savage MJ, Flood DG, Thomas JM, Levy RB, Mahadomrongkul V, Shirao T, Aoki C, Huerta PT. AMPA receptor downscaling at the onset of Alzheimer's disease pathology in double knockin mice. Proc Natl Acad Sci U S A. 2006 Feb 28;103(9):3410-5. PubMed.
- Cirrito JR, Yamada KA, Finn MB, Sloviter RS, Bales KR, May PC, Schoepp DD, Paul SM, Mennerick S, Holtzman DM. Synaptic activity regulates interstitial fluid amyloid-beta levels in vivo. Neuron. 2005 Dec 22;48(6):913-22. PubMed.
- Snyder EM, Nong Y, Almeida CG, Paul S, Moran T, Choi EY, Nairn AC, Salter MW, Lombroso PJ, Gouras GK, Greengard P. Regulation of NMDA receptor trafficking by amyloid-beta. Nat Neurosci. 2005 Aug;8(8):1051-8. PubMed.
- Almeida CG, Tampellini D, Takahashi RH, Greengard P, Lin MT, Snyder EM, Gouras GK. Beta-amyloid accumulation in APP mutant neurons reduces PSD-95 and GluR1 in synapses. Neurobiol Dis. 2005 Nov;20(2):187-98. PubMed.
- Galvan V, Gorostiza OF, Banwait S, Ataie M, Logvinova AV, Sitaraman S, Carlson E, Sagi SA, Chevallier N, Jin K, Greenberg DA, Bredesen DE. Reversal of Alzheimer's-like pathology and behavior in human APP transgenic mice by mutation of Asp664. Proc Natl Acad Sci U S A. 2006 May 2;103(18):7130-5. PubMed.
- Shaked GM, Kummer MP, Lu DC, Galvan V, Bredesen DE, Koo EH. Abeta induces cell death by direct interaction with its cognate extracellular domain on APP (APP 597-624). FASEB J. 2006 Jun;20(8):1254-6. PubMed.
No Available Further Reading