Exercise fends off some of the ills of aging, maybe even Alzheimer’s disease, but just how physical activity benefits the brain is unclear. In a new study in the September 7 Science, researchers led by Rudy Tanzi at Massachusetts General Hospital, Charlestown, used a combination of drugs and gene therapy to mimic the effects of exercise on memory in a mouse model of AD. When given treatments that promoted production of new neurons in the hippocampus and hiked levels of nurturing trophic factors, mice that lounged around all day maintained their cognitive prowess as if they had run on a wheel for hours.

  • Drugs, gene therapy mimic exercise benefit in AD mice.
  • Treated mice were as smart as mice running hours per day.
  • Therapy boosted neurogenesis and boosted BDNF, but did not reduce plaques.

The gene-therapy interventions used in the study are not likely to fly in people. But the pharmacological interventions? Tara Spires-Jones and Craig Ritchie, University of Edinburgh, think they are promising. “In the best-case scenario, assuming these results are replicated in other models and are relevant to human disease, this study suggests that we could bottle the effects of exercise to prevent or treat dementia,” they wrote in an accompanying commentary in Science. Don’t cancel your gym membership just yet—such exercise mimetics are years away, at best, scientists agree.

Epidemiological data has shown that regular exercisers reduce the risk of dementia (for example, Mar 2018 news and Hamer et al., 2018). In mice, exercise reduces Aβ pathology and improves cognitive function (Mar 2005 conference news). It’s been known for some time that exercise fuels the birth of new neurons in the hippocampus in adult mice and that new neurons play a role in learning and memory (van Praag et al., 1999; Toda et al., 2018). Could stimulating neurogenesis in sedentary mice mimic the effects of exercise?

Sweat Equity. Exercise benefits the brain via multiple pathways, some of which can be mimicked with drugs and gene therapy to improve cognition in AD mice. [Image adapted from Spires-Jones and Ritchie, 2018; courtesy of Science/AAAS.]

To find out, first author Se Hoon Choi took advantage of P7C3, a small molecule discovered in a screen for drugs that enhance neurogenesis in adult mice (Jul 2010 news; Sep 2014 news). P7C3 and similar compounds promote survival of new neurons, and protect older neurons, too, in mouse models of Parkinson’s disease, ALS, and traumatic brain injury (Oct 2012 news; Yin et al, 2014). In a rat model of AD, a related compound prevented deterioration in cognition, but it did not reduce amyloid load (Vorhees et al., 2017). 

In their study, Choi and colleagues injected P7C3 once daily into sedentary, two-month-old 5XFAD mice. One month later, Choi injected lentiviral vectors expressing the neuron progenitor survival factor WNT3 into their hippocampi to further drive proliferation of progenitors. The researchers gave another group of mice access to running wheels for three hours each day. After three months, either regimen enhanced hippocampal neurogenesis by an average of 50 percent. Not all the mice ramped up neurogenesis in response to exercise, but those that did also performed better in three different maze-navigation tasks. However, P7C3/WNT –treated mice did no better in the mazes than untreated animals whether they had a boost in neurogenesis or not.

Replacement Therapy.

Hippocampal tissue from a 5XFAD mouse has few newborn neurons (top, bright-red cells), but such cells abound after treatment with P7C3. [Image adapted from Choi et al., 2018; courtesy of Science/AAAS.]

As expected, the exercising mice had less amyloid plaque burden than their untreated, sedentary counterparts. P7C3/WNT treatment did not reduce amyloid.

Did exercise help the 5xFAD mice navigate because it reduced amyloid plaques, or are other factors at play beyond neurogenesis that are not evoked by the P7C3/WNT mimetic? Exercise also induces cytokines and neurotrophins, particularly brain-derived neurotrophic factor (BDNF), and indeed, Choi found that exercise, but not P7C3/WNT3, boosted hippocampal levels of BDNF and the BDNF stimulator FNDC5. Wheel running also elevated the synaptic markers PSD95 and SYP, suggesting that exercise not only enhances the birth of new neurons in the hippocampus, but also promotes their survival and plasticity.

To similarly mobilize neurotrophic support in the P7C3/WNT 5xFAD mice, the scientists turned to AICAR, an AMP kinase activator that increases BDNF expression (Guerrieri and van Praag, 2015). In 5.5-month old P7C3/WNT-treated mice, daily injections of AICAR for two weeks resulted in better memory scores, similar to the effect induced by exercise. AICAR by itself did not change neurogenesis, plaque levels, or cognition.

Even as the combination treatment helped memory, it failed to reduce amyloid plaque in the mice. This lends further support to the idea that amyloid plaques are a poor marker of disease progression, Spires-Jones and Ritchie wrote.

“The study nicely confirms that adult-generated neurons play a critical role in hippocampal resilience, and it suggests that exercise mimetics are indeed a possible route to preventive or therapeutic strategies," wrote Gerd Kempermann, Center for Regenerative Therapies, Dresden, Germany, in an email to Alzforum (full comment below). However, given the low cost and low risk of exercise, Kempermann noted in most cases, simple physical activity would be the better choice.

Shouldn’t people just get off the couch and exercise? Mimetics may benefit elderly people who have trouble staying active at the level required for brain health, said Tanzi. “Our mice run on the wheel for three hours a day. They love it, and they have nothing else to do, but it’s a lot of work,” he said. “To get the equivalent exercise in elderly AD patients is probably impossible. If we can use a drug/gene therapy to mimic the effects of exercise on cognitive ability, that gives us hope that we can translate these benefits from mice to humans.” A scientist told Alzforum that P7C3 analogs are being developed for human use by Proneurotech, Inc. of South San Francisco, California. The company declined to comment on the status of its research program.

It is important to replicate the results in different AD models, said Fernando Gomez-Pinilla, University of California Los Angeles. He also wonders how long the cognitive effects last, and how the treatments affect tau pathology, considered a better marker for cognitive decline than amyloid. Though adult neurogenesis occurs only in the hippocampus and the olfactory, Helen Scharfman of the Nathan Kline Institute in New York wondered about other regions of the brain. “The authors focus on the hippocampus and dentate gyrus, but this is not necessarily the only area to think about, and not the site where the disease necessarily starts,” she wrote.—Pat McCaffrey.


  1. This is a very complex study, looking at how physical exercise might build protective effects against cognitive impairment in a genetic model of AD. The idea is that adult hippocampal neurogenesis plays a central role, at least as far as the hippocampus is concerned. The authors now show that combining a genetic enhancement of neurogenesis with the external application of BDNF can mimic the beneficial effects of exercise in terms of delaying or compensating the AD-like pathology and functional impairment. The study thereby nicely confirms that adult-generated neurons play a critical role in providing hippocampal resilience, and it suggests that exercise mimetics are indeed a possible route for preventive or therapeutic strategies. This speaks to the common quip, “If exercise were a pill, I would take it.”

    Given its low cost and its essentially dearth of side effects, one might wonder whether in most cases simple physical activity is not the better choice, but conceptually the finding is of course very interesting. What one would have to look at is the context. AD does not only affect the hippocampus but adult neurogenesis is only found in the hippocampus, not in other cortical regions.

  2. I think this is a carefully executed study that combines chemical and viral approaches to deliver neurogenic factors into the hippocampus for amelioration of cognitive decline and amyloid pathology in an established mouse model of β-amyloidosis.

    This is one of the well-tested paradigms in clinical and preclinical studies: Exercise and adult neurogenesis supports cognitive maintenance and amyloid clearance. We have previously shown that AAV-mediated expression of FGF2 in the hippocampus in APP/PS1 mice enhances neurogenesis, hippocampal LTP, and spatial learning, while reducing amyloid load (Kiyota et al., 2011). Similar findings were observed upon AAV-mediated gene transfer of the anti-inflammatory cytokine IL-4 in the hippocampal region of APP/PS1 mice (Kiyota et al, 2010).

    The new paper by Choi et al. elegantly shows that exercise or the combination of adult neurogenesis plus BDNF is necessary for the beneficial effect. The BDNF requirement could be due to the genetic difference between the 5xFAD they use and APP/PS1 mice, or due to the function of genes (P7C3, an NAMPT activator, and Wnt3a+BDNF vs. FGF-2). Validation of the key findings with additional AD mouse models is desirable. Since the study selected adult hippocampal neurogenesis (AHN) responder mice (ProAHN) from non-responders, one would wonder how successful this combination approach is for the induction of ProAHN, and if it could be improved to be 100 percent. It is also of interest to know how long the cognitive improvement can be maintained with this therapeutic approach.

    It may be challenging to translate this finding to humans, since a recent paper shows a lack of adult hippocampal neurogenesis in people (Sorrells  et al., 2018), although another lab reports its persistence throughout aging (Boldrini et al., 2018). This is a controversial field even now, and this study sheds new light on the potential therapeutic intervention using adult neurogenesis.


    . FGF2 gene transfer restores hippocampal functions in mouse models of Alzheimer's disease and has therapeutic implications for neurocognitive disorders. Proc Natl Acad Sci U S A. 2011 Dec 6;108(49):E1339-48. PubMed.

    . CNS expression of anti-inflammatory cytokine interleukin-4 attenuates Alzheimer's disease-like pathogenesis in APP+PS1 bigenic mice. FASEB J. 2010 Aug;24(8):3093-102. PubMed.

    . Human hippocampal neurogenesis drops sharply in children to undetectable levels in adults. Nature. 2018 Mar 15;555(7696):377-381. Epub 2018 Mar 7 PubMed.

    . Human Hippocampal Neurogenesis Persists throughout Aging. Cell Stem Cell. 2018 Apr 5;22(4):589-599.e5. PubMed.

  3. Rumors that physical exercise might be able to counteract dementia (and also major depressive disorder, MDD) in a BDNF-dependent fashion have been rife in the field for more than 10 years. Nevertheless, a causal connection between physical exercise-induced increased BDNF level in the brain and amelioration of both diseases is still lacking. The Choi et al. paper now manages to provide more insights what this connection could be in AD.

    Thus, the study provides several lines of evidence that adult hippocampal neurogenesis (AHN) and BDNF elevation, only together, can prevent the  AD-related memory dysfunction at six months of age in transgenic mice that model the disease, if both manipulations have been started by two months after birth. They further show that BDNF elevation can only exert its beneficial effect against AD progression if AHN is not (artificially) blocked—a condition that usually does not occur anyway. Really compelling is the finding that genetic and pharmacological stimulation of AHN combined with pharmacologically (AICAR)-induced BDNF increase at 5.5 months can also ameliorate AD pathology.Last but not least, the authors provide independent data sets for male and female mice, showing mostly similar effects for both sexes.

    Interesting as these findings are, the manipulations the authors employ represent a means to avoid or delay AD onset rather than a treatment strategy that could start close to or even after disease onset. Thus, future studies are of utmost importance that will address the following questions:

    1. Can we find ways to increase brain BDNF levels:
      1. … at later stages of AD progression to provide a therapy for patients who are already symptomatic?
      2. … without prescribing aerobic exercise which is often not possible at older ages?
      3. … without using rather invasive genetic manipulations required to provide artificial sources for AHN and BDNF?
    2. Can the extremely low numbers of AHN-derived neurons in aged animals account mechanistically for AD amelioration alone?
    3. Can combined action of AHN and BDNF elevation also restore reduced numbers of synaptic spines in pre-existing neurons, and counteract hyperexcitability and inefficient synaptic transmission of neurons, also known hallmarks of AD pathology?
    4. What other (to be identified) running-associated increased growth factors and decreased cytokines team up with BDNF elevation to account for the beneficial effects of physical exercise in dementia?
    5. Do we need to manipulate several drug targets simultaneously to ameliorate AD pathology, including therapies that avoid the pro-inflammatory actions and at the same time boost the protective effects of microglia?
    6. Can antibody based therapies reducing the Aß burden in the brain even be revived as just one element of an effective combination therapy that tackles AD from many different angles?

    Maybe we are currently heading into a new era in AD research that is defining and treating AD as a multifactorial systemic disease requiring a balanced multidrug and multi-treatment approach.

Make a Comment

To make a comment you must login or register.


News Citations

  1. 44-Year Study Ties Midlife Fitness to Lower Dementia Risk
  2. Sorrento: More Fun, Less Amyloid for Transgenic Mice
  3. Mouse Screen Yields Pro-neurogenesis Elixir
  4. Neuron-Protecting P7C3 Compounds Take Steps Toward the Clinic
  5. Compound to the Rescue in Parkinson’s, ALS Models

Research Models Citations

  1. 5xFAD (B6SJL)

Paper Citations

  1. . Physical activity and trajectories in cognitive function: English Longitudinal Study of Ageing. J Epidemiol Community Health. 2018 Jun;72(6):477-483. Epub 2018 Feb 6 PubMed.
  2. . Running increases cell proliferation and neurogenesis in the adult mouse dentate gyrus. Nat Neurosci. 1999 Mar;2(3):266-70. PubMed.
  3. . The role of adult hippocampal neurogenesis in brain health and disease. Mol Psychiatry. 2018 Apr 20; PubMed.
  4. . P7C3 neuroprotective chemicals block axonal degeneration and preserve function after traumatic brain injury. Cell Rep. 2014 Sep 25;8(6):1731-40. Epub 2014 Sep 15 PubMed.
  5. . (-)-P7C3-S243 Protects a Rat Model of Alzheimer's Disease From Neuropsychiatric Deficits and Neurodegeneration Without Altering Amyloid Deposition or Reactive Glia. Biol Psychiatry. 2017 Nov 6; PubMed.
  6. . Exercise-mimetic AICAR transiently benefits brain function. Oncotarget. 2015 Jul 30;6(21):18293-313. PubMed.

Further Reading

No Available Further Reading

Primary Papers

  1. . Combined adult neurogenesis and BDNF mimic exercise effects on cognition in an Alzheimer's mouse model. Science. 2018 Sep 7;361(6406) PubMed.
  2. . A brain boost to fight Alzheimer's disease. Science. 2018 Sep 7;361(6406):975-976. PubMed.