News & Views | Published:

Neurodegenerative disorders

Neural synchronization in Alzheimer's disease

Nature volume 540, pages 207208 (08 December 2016) | Download Citation

Electrical oscillations generated by neural circuits are disrupted in Alzheimer's disease. Restoring these oscillations in mouse models activates immune cells to clear disease-associated amyloid-β protein from the brain. See Article p.230

An intriguing feature of the brain is its generation of electrical oscillations through the synchronized activity of networks of neurons. The frequencies of such brain rhythms can span several orders of magnitude, from slow oscillations in the delta (0.5–3 hertz) range to gamma (30–90 Hz) and ultrafast (90–200 Hz) ranges. These rhythms have been implicated in fundamental neural processes, such as attention, sensory perception, learning and memory1. Moreover, disruption of gamma oscillations has been observed in several neurological conditions, including brain trauma, schizophrenia and Alzheimer's disease2. On page 230, Iaccarino et al.3 present evidence in mice that disruption of gamma oscillations might contribute to the accumulation of amyloid-β protein in the brain — a hallmark of Alzheimer's disease.

The brains of people with Alzheimer's show desynchronized electrical activity and loss of oscillatory activity, particularly of gamma-frequency oscillations2,4,5. Mice carrying genetic mutations that cause Alzheimer's in humans also show altered electrical activity6 and, in one model, restoration of normal gamma oscillations has been shown to reduce memory deficits7. It is not known, however, whether these changes in neural activity have a causal role in the biological changes underlying disease progression or represent secondary phenomena.

Iaccarino and colleagues explored the effects of gamma activity on amyloid-β (Aβ) accumulation in a mouse model of Alzheimer's disease known as 5XFAD. The authors showed that these mice exhibit low gamma activity, similar to other mouse models of the disease7,8. To restore this activity, mice were genetically engineered to express the light-activated ion-channel protein channelrhodopsin in neurons called fast-spiking parvalbumin interneurons, which are involved in the generation of gamma oscillations9. Activation of channelrhodopsin using a fibre-optic light source implanted in the hippocampus region of the animals' brains induced synchronous neuron firing and gamma oscillations. Remarkably, this led to a major reduction in the accumulation of disease-associated Aβ deposits in the hippocampus.

How could augmentation of gamma oscillations have such a major effect? The first clue came when Iaccarino and co-workers analysed RNA transcripts in the hippocampus of 5XFAD mice, and found that gamma stimulation increased the expression of many genes involved in the function of resident immune cells called microglia. Some of the most affected genes were associated with phagocytosis — the process by which immune cells take up extracellular material for clearance and degradation. Furthermore, gamma stimulation caused microglia to adopt an 'activated' shape and induced increased intracellular Aβ levels, consistent with Aβ phagocytosis (Fig. 1). Biochemical experiments suggested that gamma stimulation also altered the processing of amyloid precursor protein, indicating reduced generation of Aβ. Thus, gamma stimulation might have a bimodal effect, both reducing Aβ generation and increasing its clearance by microglia.

Figure 1: Gamma oscillations stimulate the clearance of amyloid-β protein deposits.
Figure 1

Abnormal aggregation of amyloid-β (Aβ) protein in the brain is associated with Alzheimer's disease. Aβ aggregates might accumulate and promote neurodegeneration in part because immune cells called microglia cannot effectively clear the protein. In addition, synchronized patterns of electrical activity in the brain known as gamma oscillations are disrupted in Alzheimer's. Iaccarino et al.3 restored gamma oscillations in a mouse model of the disease. Such gamma stimulation led to recruitment of microglia to sites of Aβ deposition. The microglia adopted an activated shape, and consequently engulfed and degraded Aβ.

The authors then explored a non-invasive approach to stimulating natural production of gamma oscillations. Previous work10 showed that visual stimulation using flickering light at specific frequencies induced gamma oscillations in the brain's visual cortex. In an elegant series of experiments, the authors found that exposing 5XFAD mice to a 40-Hz flickering light for just 1 hour augmented gamma oscillations and markedly reduced Aβ levels for 12 to 24 hours. Repeating the treatment for 7 days reduced the load of harmful Aβ deposits called plaques in the visual cortex by about 60%, suggesting that there could be a longer-term effect. No effect on Aβ levels was seen after exposure to other frequencies or a random-frequency light flicker. Aβ deposition is therefore highly sensitive to specific patterns of neural activation.

Light-flicker treatment induced gamma oscillations and reduced Aβ levels in ageing wild-type mice, consistent with a physiological role for gamma stimulation in regulating Aβ metabolism. Interestingly, the treatment also reduced the build-up of tau proteins in a mouse model of frontotemporal dementia, suggesting that gamma activity might have broad effects on protein homeostasis in the brain.

In the past ten years, dysfunction in neural networks has emerged as a potential contributor to the biology and clinical symptoms of Alzheimer's disease11. First, it was observed that neural activity could regulate local concentrations of Aβ in the interstitial fluid that surrounds cells in the brain12. Subsequently, the topography of neural connections was shown to direct the spread of tau and Aβ protein aggregates in mouse models of neurodegenerative disease13,14. Moreover, circuit connectivity between the hippocampus and the cortex has been proposed to be a primary driver of the stepwise spread of intracellular tau aggregates called neurofibrillary tangles between these brain regions in people with Alzheimer's15. Iaccarino and colleagues suggest that not only connectivity, but also patterns of neural activity — in particular, gamma oscillations — play a part. In this model, dysfunctional neural-network activity promotes the accumulation and spread of misfolded proteins that, in turn, cause further circuit disruption in a harmful positive feedback loop.

The observation that microglia clear Aβ in response to specific patterns of neural activity raises several questions. Does synchronous neural firing result in the secretion of factors that attract microglia and activate their phagocytic activity? Could such factors be used as therapeutic agents? Does this mode of microglial stimulation have a role in the function and maintenance of the synaptic connections between neurons in healthy brains? Recent reports16,17 suggest that microglia mediate synaptic pruning, which occurs as a normal part of brain development and might become dysfunctional in Alzheimer's disease. The process might involve phagocytosis and turnover of synaptic structural components — a function that could be co-opted for clearance of Aβ and other protein aggregates.

Treating neurodegenerative disorders by non-invasive modulation of neural networks is an intriguing prospect. The present study used light to stimulate gamma oscillations in the visual cortex, which is relatively unaffected by Alzheimer's disease. It will be important to determine whether other approaches can stimulate gamma oscillations more globally, in brain regions that are affected in Alzheimer's and other disorders. For example, behavioural interventions such as meditation have been shown to increase gamma oscillations18. Electrical stimulation of deep brain regions — an effective approach in drug-resistant Parkinson's disease — could potentially be adapted to stimulate gamma oscillations in specific brain regions. These are just two of many potential therapeutic approaches likely to arise from a greater understanding of the role of neural networks in neurodegenerative disorders.

Notes

References

  1. 1.

    Physiol. Rev. 90, 1195–1268 (2010).

  2. 2.

    & Neuron 52, 155–168 (2006).

  3. 3.

    et al. Nature 540, 230–235 (2016).

  4. 4.

    et al. Neurobiol. Aging 26, 165–171 (2005).

  5. 5.

    et al. Clin. Neurophysiol. 115, 1332–1339 (2004).

  6. 6.

    et al. Neuron 55, 697–711 (2007).

  7. 7.

    et al. Cell 149, 708–721 (2012).

  8. 8.

    et al. Neuron 90, 740–751 (2016).

  9. 9.

    , , & Nature 459, 698–702 (2009).

  10. 10.

    , & Trends Neurosci. 30, 309–316 (2007).

  11. 11.

    & Nature Neurosci. 13, 812–818 (2010).

  12. 12.

    et al. Neuron 48, 913–922 (2005).

  13. 13.

    , , & Nature Rev. Neurosci. 16, 109–120 (2015).

  14. 14.

    et al. Alzheimers Res. Ther. 5, 49 (2013).

  15. 15.

    , & Annu. Rev. Pathol. 3, 41–66 (2008).

  16. 16.

    , & Curr. Opin. Neurobiol. 36, 128–134 (2015).

  17. 17.

    et al. Science 352, 712–716 (2016).

  18. 18.

    , , , & Proc. Natl Acad. Sci. USA 101, 16369–16373 (2004).

Download references

Author information

Affiliations

  1. Liviu Aron and Bruce A. Yankner are in the Department of Genetics, Harvard Medical School, Boston, Massachusetts 02115, USA.

    • Liviu Aron
    •  & Bruce A. Yankner

Authors

  1. Search for Liviu Aron in:

  2. Search for Bruce A. Yankner in:

Corresponding author

Correspondence to Bruce A. Yankner.

About this article

Publication history

Published

DOI

https://doi.org/10.1038/540207a

Further reading

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Newsletter Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing