Experimental evidence and plain common sense tell us that neurons lose activity in Alzheimer disease, but a new study surprisingly suggests that under certain conditions, neurons may become more active. This work comes from Marc Aurel Busche and colleagues at the Technische Universitaet, Maximilians-Universitaet, and the Center for Integrated Protein Science (all in Munich, Germany), as well as Novartis Institutes for Biomedical Research in Basel, Switzerland, and is published in today’s Science. Using in vivo calcium-imaging, the scientists examine layer 2/3 cortical neurons in transgenic mice that overproduce both mutant amyloid precursor protein and presenilin (APP23xPS45 mice). The investigators describe a decrease in activity in about one-third of layer 2/3 neurons, but an increase in the activity of 21 percent of neurons in the same area, as measured by spontaneous calcium bursts. What is most notable is that the hyperactive neurons were found directly in the vicinity of amyloid plaques.

Declines in neuronal activity in Alzheimer disease are not surprising, given the loss of synapses and dendritic spines that occurs even in early stages of the disease. In agreement with this idea, previous studies examining Alzheimer disease animal models showed that amyloid-β (Aβ) inhibits synaptic currents. For example, Hsia et al. (1999) found that declines in synaptic density, neurons, and synaptic transmission precede amyloid deposition in APP transgenic mice. Chang et al. (2006) observed that AMPA receptor-mediated currents and spontaneous neuronal currents decline in double amyloid-precursor protein and presenilin knock-in mice. What made the current study different from these, and similar, investigations is that it suggests that amyloid deposition may not simply suppress neuronal activity, but may create an imbalance, specifically causing both hyperactivity and reduced activity in brain circuits.

The investigators simultaneously visualized cortical neurons and amyloid plaques by injecting the calcium-indicator dye Oregon Green 488 BAPTA-1 AM and fluorescent thioflavin S, which labels fibrillar Aβ. Injections were made into the cortex. Calcium imaging is known to indicate action potential firing in neurons, so by monitoring cortical calcium activity the researchers were able to infer increases or decreases in cortical neuronal activity. Stable cortical neuronal activity was detected in the cortices of wild-type mice, but in the double transgenic mice, cortical activity was quite different. Only about half of the neurons in the transgenic animals showed normal activity, whereas the remaining cortical cells were either abnormally silent or hyperactive. Hyperactive neurons were found only within very close proximity to the border of plaques, whereas silent cells and normally active cells were found throughout the cortex. Proportions of silent cells increased as the distance from the plaques increased.

The scientists wondered if the same type of transgenic animals in which amyloid plaques had not yet been deposited also had similar aberrant neuronal activity. They repeated their calcium imaging studies in animals between 1.5-2 months of age. These double transgenic mice had normal cortical activity, comparable to wild-type mice. They performed as well on tests of spatial memory—the Morris water maze and Y-maze—as wild-type animals. In contrast, six- or eight-month-old double transgenic mice with amyloid plaques performed poorly on these tests of memory.

Busche and colleagues were then interested in understanding the source of this hyperactivity in the vicinity of plaques. To determine that the increases in calcium activity (transients) were caused by action potential firing and not another source of calcium release, the researchers blocked calcium transients using the sodium-channel blocker, tetrodotoxin. This was effective in all four of the animals tested. They were also interested in knowing whether the hyperactive neurons possessed more intrinsic excitability or whether increased synaptic activity was causing the hyperactivity. Because ionotropic glutamate receptor blockers abolished all calcium transients in these neurons, the researchers concluded that increased synaptic activity was occurring, not increased intrinsic neuronal excitability. They speculated that decreased inhibition via the neurotransmitter GABA could be causing the hyperactivity, since the benzodiazepine, diazepam (which increases the opening of GABAA receptor channels) reduced the hyperactivity. Antagonizing GABAA receptors with gabazine increased calcium transients in the hyperactive neurons, further supporting the possibility that the cells were being disinhibited via decreased GABA activity.

Olga Garaschuk, one of the study investigators, stated in an e-mail to Alzforum that “…this study improves our understanding of the pathophysiological mechanisms of Aβ amyloidosis. It also identifies hyperactive cells as a possible target for pharmacological treatment. Fortunate location of these cells in the vicinity of Aβ plaques makes targeting of pharmacological substances easier, because many substances are binding or can be targeted to plaques.”

The physiological consequence of hyperactive neurons near plaques still remains to be understood, and similar activity in humans with Alzheimer disease needs to be confirmed. Interestingly, epileptic seizure incidence increases in people with Alzheimer disease. Similar hyperactive neurons may exist near neuronal plaques in the brains of people with Alzheimer disease, potentially causing seizures. Another physiological consequence implied by these findings could be that imbalances in cortical neuronal activity could contribute to cognitive dysfunction. This idea is particularly supported by the observation that memory impairments in the transgenic mice corresponded with Aβ deposition and the presence of hyperactive neurons. If either or both of these possibilities turn out to be the case, targeting hyperactive neurons might inhibit seizures or even improve memory.

Another study has, in fact, indicated possible changes in the balance between hyperactivity and reduced activity in an animal model of Alzheimer disease, and further suggests an association of this activity with seizures. Researchers led by Lennart Mucke described spontaneous, non-convulsive seizures in transgenic mice that overproduce the human amyloid precursor protein (see Palop et al., 2007 and ARF related news story). Seizures (in other words, hyperactive neurons), measured in both cortical and hippocampal neurons, were accompanied by increases in sprouting of GABAergic axons and synaptic inhibition, specifically in the dentate gyrus of the hippocampus, an area critical to formation of certain types of memory. Palop and colleagues proposed that activation of those inhibitory circuits was to compensate for the hyperactivity in the cortical and hippocampal networks.

Although the scientists who conducted the current study have not yet observed the hyperactive cortical neurons over the long term, future investigations examining the ultimate fate of these cells may be in the works. When asked about the prospect of imaging activity in these neurons over a longer duration, Garaschuk told ARF via an e-mail, “already this is now possible. However, the cells would need to be re-stained with a calcium indicator dye (by inserting a micropipette into the brain) each time before imaging.” She noted that a recent study published in Nature Methods (Mank et al., 2008) described the advance of a genetically encoded calcium indicator TN-XXL, which allowed calcium increases in neurons to be measured repeatedly, even up to several weeks. This technology could allow for long-term calcium imaging of the hyperactive neurons associated with amyloid plaques, further elucidating the role that these cells could potentially play in the brains of people with Alzheimer disease.—Alisa Woods

Alisa Woods is a freelance writer in Brooklyn, New York.


  1. This is a very interesting article. From our perspective, there are several important points: first, it highlights a plaque-specific alteration of neural function, but emphasizes that this occurs in the vicinity around plaques, suggesting a halo effect of a potentially soluble mediator, which is critical to alter activity. Second, it uses an alternative technical strategy to the one utilized by my colleague Brian Bacskai (Kuchibhotla et al., 2008), which used gene transfer of a genetically encoded Ca2+ reporter to observe neurites near plaques, as opposed to loading of an AM molecular dye to monitor neuronal perikarya, to come to a similar conclusion—that neurons in the vicinity of plaques have a profound dyshomeostasis of calcium regulation. Third, that neural systems are markedly disrupted by the presence of plaques, with functional alterations leading to both increased and decreased activity affecting a large percentage of neurons. This neural system collapse provides a plausible biological underpinning for the cognitive impairment in AD.


    . Abeta plaques lead to aberrant regulation of calcium homeostasis in vivo resulting in structural and functional disruption of neuronal networks. Neuron. 2008 Jul 31;59(2):214-25. PubMed.

  2. This is a great paper that confirms our and others’ work that senile plaques are a focal source of toxicity. The implications of the hyperactivity are unknown; however, it is apparent that a disruption to the normal network activity in the brain is obvious. These data are in line with our own measurements of altered resting calcium concentration in neurons in animals with senile plaques and suggest that a more detailed examination of the calcium hypothesis of AD is warranted.

    We have not looked at spontaneous activity of neuronal calcium signaling in the brain, so we cannot confirm these results as yet, but it would be interesting to probe whether these observations share similarities to physiologically relevant evoked responses.

    In sum, this paper confirms that calcium imaging as an indicator of neuronal activity, and calcium imaging as an indicator of dyshomeostasis of calcium concentration, is an important future direction for our understanding of the disruption of the cellular and network disruption of neuronal signaling in AD.

  3. This is a very interesting paper and consistent with our results from primary hippocampal neurons derived from APPswe transgenic rats. We have recently shown that modest overexpression of APPswe results in increased frequency but unaltered amplitude of spontaneous calcium oscillations in these transgenic neurons (Kloskowska et al., 2008a). Furthermore, we found that the baseline of calcium oscillations was significantly higher than in control neurons (Kloskowska et al., 2008b). Thus, recent data implies that the relationship between Aβ and neuronal activity is more complex than has been thought.


    . APPswe mutation increases the frequency of spontaneous Ca2+-oscillations in rat hippocampal neurons. Neurosci Lett. 2008 May 9;436(2):250-4. PubMed.

    . The APP670/671 mutation alters calcium signaling and response to hyperosmotic stress in rat primary hippocampal neurons. Neurosci Lett. 2008 Oct 31;444(3):275-9. PubMed.

  4. A very timely article is presented here by Bushe et al., demonstrating a relationship between amyloid plaques and intraneuronal calcium signaling dysregulation using 2-photon imaging techniques in vivo. The authors found diverging populations of neurons in plaque-expressing APPxPS45 mice relative to controls, in that one group decreased calcium transient activity while another “hyperactive” group in close proximity to plaques increased frequency of calcium transients. Reduction in inhibitory synaptic tone resulting from impaired GABAergic transmission is provided as the mechanism.

    Overall, this technically demanding study adds to the existing literature demonstrating that neuronal calcium alterations are an integral component of AD pathology—whether through PS mutations, extracellular plaques, or other AD-linked pathways. This is a new view on how AD pathology can affect neuronal signaling, and will hopefully spin off several follow-up studies.

    What I do find lacking (perhaps will be addressed in follow-up?) is 1) a mechanism for why GABAergic neurons may be more vulnerable to plaque pathology, since they are loaded with calcium buffers and tend to be less susceptible to excitotoxicity, and 2) a means to determine how/if the “quiet” cells are GABAergic, and if the hyperactive cells are pyramidals or another non-inhibitory other subtype.

    Much of cortex contains a granular layer (L. IV) containing largely interneurons, with the exception of the prefrontal region, and it is not clear where in frontal cortex these recordings are taking place. This may change the negative feedback circuit anatomy and have some implications here. Since the pharmacology examines local transmitter effects, it is difficult to make assumptions about the source. The possible role of intracellular calcium waves, triggered by glutamatergic activity but generated by intracellular calcium sources, is also a possible mechanism but is not yet addressed in these studies.

    Recently, similar approaches were used in a study from Brian Bacskai's group (Kuchibhotla et al., 2008), and although the general idea is also that calcium levels are disrupted in neurons close to plaques, some significant differences in the details are evident. Although temporal domains of calcium signaling were not the primary focus, it does not appear that calcium oscillations were observed in their cortical neurons. Rather, they observed increases in steady-state calcium levels in neuronal processes close to plaques. Similar to our findings, neurons from control mice do regulate their calcium levels quite well over time (Stutzmann et al., 2006), as do neurons from the AD mice sufficiently removed from plaque deposits. Although mechanisms for this increased steady-state calcium load aren't investigated, it does not appear synaptic/activity-dependent in nature. Granted, there are differences in the techniques and calcium indicator approaches used in the two studies, so as much as I'd like to avoid this euphemism...more studies are needed.


    . Abeta plaques lead to aberrant regulation of calcium homeostasis in vivo resulting in structural and functional disruption of neuronal networks. Neuron. 2008 Jul 31;59(2):214-25. PubMed.

    . Enhanced ryanodine receptor recruitment contributes to Ca2+ disruptions in young, adult, and aged Alzheimer's disease mice. J Neurosci. 2006 May 10;26(19):5180-9. PubMed.

    View all comments by Grace Stutzmann
  5. The increase in activity in some of the neurons near the amyloid plaques in the layer of cortical neurons is surprising, but it makes sense when it is framed, just as the authors did, in the “synaptic failure hypothesis.” This finding could very well be related to the observation that endocytosis of glutamate receptors is observed in certain animal models of AD. Dysfunction of Ca2+ kinetics is probably causing the calcium overload the authors mention (seen in another study).

    Would it be interesting to look specifically at which neurons are affected? Could it be that GABAergic neurons are being affected the most, and their dysfunction thereby resulting in an overall increase in excitation? The main question still to be answered is why is it that some of the cortical neurons decrease activity, while others increase neuronal activity? In the conclusion, the authors mention anatomical remodeling in synaptic inputs as a possible cause. This explanation, although plausible, does not clarify within which mechanism are the chemical signals from Aβ affecting the neurons. Could it be that defective neurogenesis is also occurring, adding to the changes in neuronal function?

    Interestingly, distance from the plaques seems to be the determining factor of whether increase or decrease in activity is seen (the mechanism AB uses to affect neurons could be similar to what is seen in neural development, where chemical gradients determine the “identity” of a cell). The fact that Aβ is the main factor in the hyperactivation of neurons close to the cortical layer 2/3 was strengthened by the results of the predisposing APP23xPS45 mice study. Unfortunately, the fact that these predisposing mice present normal cortical activity and Ca2+ function rules out the use of these biological markers in pharmacological studies for prevention of AD. However, because the APP mice do present behavioral deficits in the discriminatory water maze and the Y-maze, there may be a possibility that an in between age (between two months and six months) could serve as a model for testing of preventive drugs.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Do "Silent" Seizures Cause Network Dysfunction in AD?

Paper Citations

  1. . Plaque-independent disruption of neural circuits in Alzheimer's disease mouse models. Proc Natl Acad Sci U S A. 1999 Mar 16;96(6):3228-33. PubMed.
  2. . 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.
  3. . Aberrant excitatory neuronal activity and compensatory remodeling of inhibitory hippocampal circuits in mouse models of Alzheimer's disease. Neuron. 2007 Sep 6;55(5):697-711. PubMed.
  4. . A genetically encoded calcium indicator for chronic in vivo two-photon imaging. Nat Methods. 2008 Aug 10; PubMed.

Further Reading


  1. . Abeta plaques lead to aberrant regulation of calcium homeostasis in vivo resulting in structural and functional disruption of neuronal networks. Neuron. 2008 Jul 31;59(2):214-25. PubMed.

Primary Papers

  1. . Clusters of hyperactive neurons near amyloid plaques in a mouse model of Alzheimer's disease. Science. 2008 Sep 19;321(5896):1686-9. PubMed.