The Persistence of Memory
http://www.100md.com
《新英格兰医药杂志》
We often do not realize how much we need something until it is gone — and Alzheimer's disease shows us how debilitating it is to lose memory. Although we now know a lot about the biochemistry of the disease, there are some gaps in our knowledge about the links between small changes in molecules at the surface of neurons and the symptoms of disease. Because of this, we have generated only a few ideas about how to prevent damage to the brain. A recent study by Gong et al.,1 however, suggests a new way of thinking about memory deficits in patients with Alzheimer's disease and hints at new therapeutic opportunities.
Two molecules have enjoyed the limelight in the context of Alzheimer's disease: the small -amyloid peptide (A) and the microtubule-binding protein tau. Argument about the relative importance of each molecule has, at times, reached almost religious proportions, but it is possible to at least approximate the course of the disease in the neuron (Figure 1).2 An early event is the generation of A from the amyloid precursor protein, which is in turn controlled by presenilin 1 and presenilin 2, proteases buried in the cell membrane. (Mutations in the genes encoding amyloid precursor protein, presenilin 1, and presenilin 2 lead to early-onset Alzheimer's disease.) The liberated A peptides aggregate on the outside of the cell and are deposited as amyloid plaques. Some time later, through a process that remains obscure, the amyloid aggregates trigger abnormal phosphorylation of tau within the neuron. This abnormal phosphorylation is more closely correlated with neuronal death than is the deposition of amyloid aggregates, and it seems likely that tau plays a requisite role in Alzheimer's disease and related disorders.
Figure 1. Molecules and Memory Formation.
The processing of amyloid precursor protein by presenilin and -secretase generates fragments of the -amyloid peptide (A) in the neuron. These fragments aggregate to form relatively soluble oligomers that are thought to damage synapses, perhaps by lodging in the membrane (as shown). Directly or indirectly, toxic A species interfere with the production of cyclic AMP by adenylate cyclase. Normally, cyclic AMP binds to the regulatory subunit of protein kinase A, which allows the catalytic subunit to phosphorylate the transcription factor CREB (cyclic AMP–responsive element–binding protein), leading to synaptic strengthening and, eventually, to memory formation. The regulatory subunit of protein kinase A is degraded by the proteasome. Recognition of these proteins by the proteasome is aided by chains of ubiquitin. Ubiquitin carboxy-terminal hydrolase L1 (UCHL1) recycles the ubiquitin molecules, which are then available in a cytoplasmic pool of free ubiquitin monomers. Blocking UCHL1, as Gong et al. did, may interfere with the dissociation of the regulatory subunit from the catalytic subunit of protein kinase A, mimicking the effects of A-mediated synaptic damage. They found that blocking UCHL1 in a mouse model of Alzheimer's disease resulted in impaired memory formation. In contrast, increasing UCHL1 activity antagonized the effect of A and promoted efficient memory formation, at least in the mouse model. Dashed lines indicate inhibited steps in the pathway.
There are good models of A processing and deposition in mice that overexpress amyloid precursor protein and presenilin 1 simultaneously.3 Gong et al. used one such model (the APP/PS1 mouse) and showed that the mice had a synaptic deficit, probably caused by small, relatively soluble aggregates of A; this observation was consistent with the results of previous studies. Then the investigators showed something quite surprising and potentially important: that this memory deficit can be largely resolved by the overexpression of a second protease, ubiquitin carboxy-terminal hydrolase L1 (UCHL1). Abundant in neurons, this protease is a component of the ubiquitin–proteasome system. Gong et al. used a neat trick here — they modified UCHL1 (by adding a short peptide to it) so that it could breach the blood–brain barrier and the neuronal membrane. This allowed them to inject UCHL1 into the peritoneum of the mouse.
The proteasome mediates the controlled degradation of surplus or dysfunctional proteins; ubiquitin is the molecular tag that latches onto these proteins and routes them to the proteasome. UCHL1 has a vital role in this garbage system: it recycles ubiquitin — once ubiquitin has off-loaded its cargo into the proteasome — so that ubiquitin is again available for targeting redundant or otherwise compromised proteins.
It turns out that the process of adding ubiquitin to specific proteins at the neuronal synapse is critical to the maintenance of synaptic plasticity,4 probably owing to a role of the proteasome in flexibly modulating levels of different proteins over time. Indeed, when Gong et al. inhibited UCHL1 in the APP/PS1 mouse, they observed a suppression of synaptic plasticity similar to that observed in mice overexpressing A. The molecular details are not completely clear, but they involve protein kinase A (Figure 1), which is known to be involved in learning and memory and is regulated by the ubiquitin–proteasome system.
So do we simply inject patients with UCHL1 and look for cognitive improvement? This seems premature, given that the APP/PS1 mouse model does not reflect all aspects of Alzheimer's disease in humans. Many mouse models of the disease, including the APP/PS1 mouse, do not have the late effects of the human disease, in which the brain may lose a third of its mass. In contrast, only a few neurons are lost in the mouse models. Presumably, the model used by Gong et al. represents the very early stages of the disease, before tau gets involved and neurons are really "lost." The next hurdle is to show the benefit of UCHL1 in situations in which neuronal loss is more extensive.3 Nevertheless, the results of Gong et al. suggest a surprising strategy for countering cognitive decline in patients with Alzheimer's disease.
No potential conflict of interest relevant to this article was reported.
Source Information
From the Laboratory of Neurogenetics, National Institute on Aging, Bethesda, MD.
References
Gong B, Cao Z, Zheng P, et al. Ubiquitin hydrolase Uch-L1 rescues beta-amyloid-induced decreases in synaptic function and contextual memory. Cell 2006;126:775-788.
Hardy J, Selkoe DJ. The amyloid hypothesis of Alzheimer's disease: progress and problems on the road to therapeutics. Science 2002;297:353-356.
McGowan E, Eriksen J, Hutton M. A decade of modeling Alzheimer's disease in transgenic mice. Trends Genet 2006;22:281-289.
Yi JJ, Ehlers MD. Ubiquitin and protein turnover in synapse function. Neuron 2005;47:629-632.(Mark R. Cookson, Ph.D., a)
Two molecules have enjoyed the limelight in the context of Alzheimer's disease: the small -amyloid peptide (A) and the microtubule-binding protein tau. Argument about the relative importance of each molecule has, at times, reached almost religious proportions, but it is possible to at least approximate the course of the disease in the neuron (Figure 1).2 An early event is the generation of A from the amyloid precursor protein, which is in turn controlled by presenilin 1 and presenilin 2, proteases buried in the cell membrane. (Mutations in the genes encoding amyloid precursor protein, presenilin 1, and presenilin 2 lead to early-onset Alzheimer's disease.) The liberated A peptides aggregate on the outside of the cell and are deposited as amyloid plaques. Some time later, through a process that remains obscure, the amyloid aggregates trigger abnormal phosphorylation of tau within the neuron. This abnormal phosphorylation is more closely correlated with neuronal death than is the deposition of amyloid aggregates, and it seems likely that tau plays a requisite role in Alzheimer's disease and related disorders.
Figure 1. Molecules and Memory Formation.
The processing of amyloid precursor protein by presenilin and -secretase generates fragments of the -amyloid peptide (A) in the neuron. These fragments aggregate to form relatively soluble oligomers that are thought to damage synapses, perhaps by lodging in the membrane (as shown). Directly or indirectly, toxic A species interfere with the production of cyclic AMP by adenylate cyclase. Normally, cyclic AMP binds to the regulatory subunit of protein kinase A, which allows the catalytic subunit to phosphorylate the transcription factor CREB (cyclic AMP–responsive element–binding protein), leading to synaptic strengthening and, eventually, to memory formation. The regulatory subunit of protein kinase A is degraded by the proteasome. Recognition of these proteins by the proteasome is aided by chains of ubiquitin. Ubiquitin carboxy-terminal hydrolase L1 (UCHL1) recycles the ubiquitin molecules, which are then available in a cytoplasmic pool of free ubiquitin monomers. Blocking UCHL1, as Gong et al. did, may interfere with the dissociation of the regulatory subunit from the catalytic subunit of protein kinase A, mimicking the effects of A-mediated synaptic damage. They found that blocking UCHL1 in a mouse model of Alzheimer's disease resulted in impaired memory formation. In contrast, increasing UCHL1 activity antagonized the effect of A and promoted efficient memory formation, at least in the mouse model. Dashed lines indicate inhibited steps in the pathway.
There are good models of A processing and deposition in mice that overexpress amyloid precursor protein and presenilin 1 simultaneously.3 Gong et al. used one such model (the APP/PS1 mouse) and showed that the mice had a synaptic deficit, probably caused by small, relatively soluble aggregates of A; this observation was consistent with the results of previous studies. Then the investigators showed something quite surprising and potentially important: that this memory deficit can be largely resolved by the overexpression of a second protease, ubiquitin carboxy-terminal hydrolase L1 (UCHL1). Abundant in neurons, this protease is a component of the ubiquitin–proteasome system. Gong et al. used a neat trick here — they modified UCHL1 (by adding a short peptide to it) so that it could breach the blood–brain barrier and the neuronal membrane. This allowed them to inject UCHL1 into the peritoneum of the mouse.
The proteasome mediates the controlled degradation of surplus or dysfunctional proteins; ubiquitin is the molecular tag that latches onto these proteins and routes them to the proteasome. UCHL1 has a vital role in this garbage system: it recycles ubiquitin — once ubiquitin has off-loaded its cargo into the proteasome — so that ubiquitin is again available for targeting redundant or otherwise compromised proteins.
It turns out that the process of adding ubiquitin to specific proteins at the neuronal synapse is critical to the maintenance of synaptic plasticity,4 probably owing to a role of the proteasome in flexibly modulating levels of different proteins over time. Indeed, when Gong et al. inhibited UCHL1 in the APP/PS1 mouse, they observed a suppression of synaptic plasticity similar to that observed in mice overexpressing A. The molecular details are not completely clear, but they involve protein kinase A (Figure 1), which is known to be involved in learning and memory and is regulated by the ubiquitin–proteasome system.
So do we simply inject patients with UCHL1 and look for cognitive improvement? This seems premature, given that the APP/PS1 mouse model does not reflect all aspects of Alzheimer's disease in humans. Many mouse models of the disease, including the APP/PS1 mouse, do not have the late effects of the human disease, in which the brain may lose a third of its mass. In contrast, only a few neurons are lost in the mouse models. Presumably, the model used by Gong et al. represents the very early stages of the disease, before tau gets involved and neurons are really "lost." The next hurdle is to show the benefit of UCHL1 in situations in which neuronal loss is more extensive.3 Nevertheless, the results of Gong et al. suggest a surprising strategy for countering cognitive decline in patients with Alzheimer's disease.
No potential conflict of interest relevant to this article was reported.
Source Information
From the Laboratory of Neurogenetics, National Institute on Aging, Bethesda, MD.
References
Gong B, Cao Z, Zheng P, et al. Ubiquitin hydrolase Uch-L1 rescues beta-amyloid-induced decreases in synaptic function and contextual memory. Cell 2006;126:775-788.
Hardy J, Selkoe DJ. The amyloid hypothesis of Alzheimer's disease: progress and problems on the road to therapeutics. Science 2002;297:353-356.
McGowan E, Eriksen J, Hutton M. A decade of modeling Alzheimer's disease in transgenic mice. Trends Genet 2006;22:281-289.
Yi JJ, Ehlers MD. Ubiquitin and protein turnover in synapse function. Neuron 2005;47:629-632.(Mark R. Cookson, Ph.D., a)