Taking certain hypertension medications has been linked to a lower risk of Alzheimer’s disease—but does that happen via lower blood pressure, or some other way? In the September 30 Science Translational Medicine, researchers led by Robert Vassar at Northwestern University in Chicago and Rudolph Tanzi at Massachusetts General Hospital, Boston, provide evidence for the latter. They identified a rare genetic variant in the blood-pressure gene angiotensin-converting enzyme (ACE) that associated with AD. Surprisingly, the variant did not affect blood pressure in a knock-in mouse model. Instead, it accumulated on neuronal cell membranes, killing off hippocampal neurons as the mice aged. Females fared worse, developing neuroinflammation and memory problems as well. Treating the mice with medications that block ACE1 or its receptor prevented the deficits.

  • A rare coding variant in ACE boosts protein levels and associates with AD.
  • Knock-in mice lose hippocampal neurons as they age, and females develop neuroinflammation.
  • The variant worsens degeneration in a mouse model of amyloidosis.

Vassar believes the data delineate another pathway involved in AD degeneration, separate from amyloid and tau, that explains some of the characteristics of the disease. “Our study gives us some insight into three of the mysteries of Alzheimer’s: why is age the primary risk factor, why do women have increased susceptibility, and why is there selective neuron loss in the hippocampus?” Vassar told Alzforum.

Giulio Pasinetti at the Icahn School of Medicine at Mount Sinai, New York, appreciated the therapeutic implications. “This is an excellent paper, showing a potential pathogenic mechanism by which ACE might increase the odds of developing AD … [it] provides a rationale for repurposing ACE inhibitors,” Pasinetti wrote to Alzforum (full comment below).

Neuronal Apocalypse. By 14 months of age, mice carrying one copy of the ACE R1284Q variant (middle) have lost more neurons (red) than controls (left), while mice homozygous for the variant (right) have lost far more. [Courtesy of Cuddy et al., Science Translational Medicine.]

Pasinetti had previously reported that an angiotensin receptor blocker (ARB), a common blood pressure medication, improved memory and cut plaque load in a mouse model of amyloidosis (Oct 2007 news). In people, an epidemiological study of 800,000 veterans linked ACE inhibitors and ARBs to a lower risk of dementia (Jan 2010 news). ARBs also associated with lower plaque load in cognitively impaired people (Sep 2012 news). 

Supporting this epidemiologic data, genome-wide association studies recently pegged ACE as an AD risk gene (Apr 2018 news; Kunkle et al., 2019; Jansen et al., 2019). In particular, an insertion in the noncoding region of ACE that boosts expression associates with a higher AD risk in women (Sleegers et al., 2005). 

Vassar and colleagues searched for additional risk variants in ACE by analyzing whole-genome sequencing from 446 families in the National Institute of Mental Health AD Genetics Initiative Study Sample. In seven families, they found a coding variant, R1279Q, that associated with AD, with nine of 11 carriers having the disease. Additional analysis of the National Institute on Aging family dataset confirmed the association.

First author Leah Cuddy generated a knock-in mouse that expressed the homologous ACE variant, R1284Q. Immunostaining revealed these knock-ins had more ACE1 protein in their brains than did wild-type. The protein co-localized with neuronal, but not microglial or astrocyte markers, suggesting it is exclusively expressed in neurons.

In both knock-ins and wild-types, ACE1 levels rose with age. This increase was most pronounced in females. ACE1 cleaves angiotensin I to angiotensin II, which hikes blood pressure. While knock-ins had more angiotensin II in their brains than did wild-types, curiously, their blood pressure was normal.

What might ACE1 be doing in the brain? At 8 months old, knock-ins and wild-types had no obvious brain differences. But by 14 months, both male and female knock-ins had smaller hippocampi, striata, and amygdalae than controls, indicative of atrophy. In vitro studies strengthened this association. Cultured forebrain neurons isolated from the knock-ins were more prone to apoptosis than were their wild-type counterparts.

Female knock-ins fared worse than male. In addition to hippocampal atrophy, they developed neuroinflammation, with larger numbers of activated microglia and astrocytes than in the hippocampi of controls. They lost hippocampal synapses, as seen by a drop in the presynaptic marker synaptophysin, and performed poorly in fear-conditioning tests and the Morris water maze. Their sleep rhythms were off, with higher alpha and theta power in the hippocampus. Disrupted theta rhythms have been linked to AD and memory problems (Aug 2010 news; Mar 2020 news). 

It is unclear how the ACE1 variant does this. In cultured knock-in neurons, more of the protein accumulated on the cell surface, and less was shed into the media than in wild-type cultures. Vassar believes the variant decreases cleavage of cell-surface ACE1, rather than affecting expression directly. Either way, the high levels of exposed ACE1 produced more angiotensin II, which bound to its receptor AT1R on neurons. That triggered phosphorylation of the kinase Erk, activating downstream pathways implicated in autophagy, aging, and AD. Vassar thinks this signaling may be the source of neurotoxicity. Hippocampal neurons express higher levels of AT1R than those in other brain regions, and R1284Q ACE1 knock-ins had about 50 percent more phospho-Erk in hippocampus than did controls.

Microglia also express AT1R, and its activation stimulates proinflammatory signaling (Labandeira-Garcia et al., 2017). Intriguingly, some prior research has found that microglia in female mice are more prone to inflammation than those in males, hinting at a possible mechanism for the gender difference in the knock-ins (Jul 2019 conference news). 

To the Rescue. Neurons (red) pack the hippocampus of an aged wild-type mouse (far left), die off in an ACE R1284Q knock-in mouse (left), but are preserved in knock-ins treated with an ACE inhibitor (right) or ARB (far right). [Courtesy of Cuddy et al., Science Translational Medicine.]

If angiotensin II causes toxicity, could inhibiting it preserve neurons? To test this, the authors treated R1284Q ACE1 knock-in and wild-type mice with either the ACE1 inhibitor captopril or the AT1R blocker losartan from 9 to 15 months of age. Both drugs prevented hippocampal atrophy, and in females, squelched neuroinflammation as well (see image above). The authors are currently testing memory in the treated mice.

How does this pathway relate to Alzheimer’s disease? The authors crossed the knock-ins with 5XFAD mice and evaluated their brains at 6 months. ACE1 has been reported to degrade Aβ in vitro, but in contrast to some previous mouse studies of angiotensin receptor blockers, the authors found no difference in plaque load in 5XFAD mice carrying the ACE1 variant (Oct 2005 news). However, the variant accelerated hippocampal atrophy in all 5XFAD mice and, in females, activated microglia more. To Vassar, this suggests a two-hit hypothesis. “Maybe ACE1 increasing with age, together with amyloid buildup, combines to trigger neuroinflammation and degeneration,” he suggested.

Edo Richard at Radboud University Medical Centre, Nijmegen, The Netherlands, noted that it remains unclear how hypertension contributes to neurodegeneration. “[These data] may provide an important lead for future research aiming to bridge the knowledge gap spanning links among blood pressure, antihypertensive medication, neurodegeneration and dementia,” he wrote to Alzforum (full comment below).

To find out if their findings might be relevant to people, the authors examined postmortem cortical and hippocampal sections from seven AD and 13 age-matched control brains. They found higher levels of ACE1 in AD cortex, as well as fewer ACE1-positive neurons in AD hippocampus. The data suggest angiotensin II signaling could play a role in AD even in the absence of an ACE1 mutation, Vassar said.

Vassar believes it would be worth testing brain-penetrant ACE inhibitors and ARBs in people with preclinical AD. In this case, the human equivalent of the ACE inhibitor and ARB doses used in the mouse studies would be twice the maximum prescribed for hypertension. The authors have not yet tested if lower doses could be effective. Because these drugs lower blood pressure, it might be necessary to titrate up dosage in people with normal blood pressure, Vassar noted. In a prevention paradigm, he believes lower doses might be sufficient. If they work, they could eventually be combined with treatments that target amyloid or tau, he suggested.

Richard agreed. “Using specific antihypertensive drugs in selected populations as one of several approaches may contribute to the prevention of cognitive decline and dementia. Future clinical studies should provide further evidence on whether such an approach is viable,” Richard wrote. Drug repurposing is often attempted in Alzheimer’s disease and many other areas of medical research, but rarely successful (e.g., Nov 2019 news; saracatinib; or, most recently Edwards 2020). In an unrelated note, the new SARS-CoV-2 virus spike protein binds ACE-2.—Madolyn Bowman Rogers

Comments

  1. This is an important study that contributes to the understanding of the complex relationship between blood pressure, antihypertensive drug use, and the risk of dementia. This study partly fills a knowledge gap between genetics and epidemiology. Previous GWAS had linked variants in the angiotensin-converting enzyme (ACE) gene to an increased risk of AD. Although not conclusive, we have linked the use of specific antihypertensive medications, including non-dihydropiridine calcium-channel blockers (CCBs) and angiotensin 1 receptor blockers, to a lower risk of dementia in observational studies (van Middelaar et al., 2017). Recent meta-analyses of such studies could not confirm these class effects, but primarily compared each antihypertensive drug to all other antihypertensive drugs together, potentially masking class effects if more than one class is associated with a lower dementia risk. (Peters et al., 2020; Ding et al., 2020). The underlying mechanisms of specific class effects are speculative, and the current study of Cuddy et al. helps elucidate these mechanisms.

    The authors used CRISPR-Cas9-mediated gene editing to create a knock-in mouse with a rare variant of the ACE gene, which in human studies was highly associated with Alzheimer’s disease. These knock-in mice had more neurodegeneration, particularly in the hippocampus, and more signs of neuroinflammation with increasing age. In addition, they had more impaired memory. All of this in the absence of a difference in blood pressure over the life course of the mice.

    Treatment of these mice, with antihypertensive medications that cross the blood-brain barrier and can inhibit ACE 1 and angiotensin II receptor signaling, prevented these changes from occurring. Interestingly enough, the effects all seemed to be independent of blood pressure, which strongly suggests a class effect of these drugs, and may suggest that this could, in theory, also be a treatment for persons without hypertension—of course limited by the potential side effects of blood pressure-lowering drugs in persons without elevated blood pressure.

    Crossing the ACE knock-in mouse with a mouse model of cerebral amyloidosis did not impact amyloid-β pathology, but did seem to accelerate the effect of ACE1 on neurodegeneration. This may provide an important lead for future research aiming to bridge the knowledge gaps spanning links among blood pressure, antihypertensive medication, neurodegeneration, and dementia. How exactly hypertension contributes to neurodegeneration is still not properly understood.

    The reasons for the sex-specific effects found in this study, with female KI mice having more severe neurodegeneration and neuroinflammation, remain speculative, I think. The authors suggest several possible underlying mechanisms. In human observational studies there is no consistent sex-specific antihypertensive class-effect, but this may not have been addressed in detail before. There is, however, accumulating data that sex differences may play a role in the effects of ACE on metabolic processes and cardiovascular disease.

    In an upcoming paper in Neurology, we suggest a potential explanation for why AT1 receptor blockers but not ACE inhibitors are associated with a decreased risk of dementia in epidemiological studies, whereas this animal model suggests both should do the job. It may be that ACE inhibitors and AT1 receptor blockers have different effects in the renin-angiotensin system, where the former decreases angiotensin-II activity, and the latter, together with dihydropyridine calcium channel blockers, increase angiotensin-II activity. Also, in human studies, confounding by indication bias is an important factor obscuring analyses of antihypertensive class effects.

    These results, together with the results from epidemiological studies, do suggest that repurposing of cheap, out-of-patent, antihypertensive drugs to slow down or prevent cognitive decline and dementia should be explored further. Most researchers will by now agree that there will not be a silver bullet for AD and that pursuing AD treatment or prevention by solely following the amyloid-path will most likely not lead to spectacular results. In a combined approach, using specific antihypertensive drugs in selected populations may thus contribute to the prevention of cognitive decline and dementia. Future clinical studies should provide further evidence on whether such an approach is viable, leads to clinical benefits, and does not lead to unacceptable risks.

    Further delineation of the target population for such interventions is warranted. After all, a knock-in mouse with a rare variant of the ACE1 gene, is not a human being.

    References:

    . Lower dementia risk with different classes of antihypertensive medication in older patients. J Hypertens. 2017 May 13; PubMed.

    . Investigation of antihypertensive class, dementia, and cognitive decline: A meta-analysis. Neurology. 2020 Jan 21;94(3):e267-e281. Epub 2019 Dec 11 PubMed.

    . Antihypertensive medications and risk for incident dementia and Alzheimer's disease: a meta-analysis of individual participant data from prospective cohort studies. Lancet Neurol. 2020 Jan;19(1):61-70. Epub 2019 Nov 6 PubMed.

  2. This is an excellent paper by Cuddy et al., showing a potentially pathogenic mechanism by which ACE might increase the odds of developing AD. This pathogenic mechanism targeted the hippocampus in the brain, but not the cortex or cerebellum, thus one wonders what the mechanism(s) of selective neuronal vulnerability might be. Nevertheless, the findings in this article provide critical insight into the role of the ACE1 in AD pathogenesis. More importantly, the evidence presented provides a rationale for repurposing ACE inhibitors in the protection against AD.

    This article corroborates previous findings showing ACE inhibition could be exploited to develop AD therapeutics. For example, we showed convincingly that ARBs, including Valsartan, lowered Aβ accumulation, and also attenuated the development of Aβ-mediated cognitive deterioration, even when delivered at a dose much lower than that used for hypertension treatment in humans, suggesting a mechanism independent of vascular effects. Taken together, it’s reasonable to suggest that brain-penetrant ACE1 inhibitors and ARBs may still hold potential for the prevention of AD.

    References:

    . Valsartan lowers brain beta-amyloid protein levels and improves spatial learning in a mouse model of Alzheimer disease. J Clin Invest. 2007 Nov;117(11):3393-402. PubMed.

  3. Epidemiologists have known for more than a decade that blood-pressure-lowering medicines, such as ACE inhibitors and angiotensin receptor blockers, are associated with reduced incidence and prevalence of Alzheimer’s disease. The recent discovery of an SNP in the ACE1 gene that is significantly associated with AD by GWAS drives home the relevance of the renin-angiotensin system for AD. However, the mechanism through which dysfunction of the ACE1 gene might contribute to AD has been unclear.

    This paper enables a mechanistic deep dive into this field. The initial findings are both surprising and striking. The renin-angiotensin system is best known for its role in controlling vascular function, and has given rise to multiple frontline medications used in treatment of cardiovascular disease. However, the angiotensin receptors I and II are not restricted to the vasculature and are distributed broadly throughout the body.

    In the current manuscript, Cuddy et al. generate a transgenic knock-in mouse harboring the ACE1 R1279Q gene, which is a polymorphism associated with elevated risk of AD. These mice exhibit a strong phenotype with enhanced neurodegeneration, memory impairment, and abnormal EEGs. Surprisingly, the mice do not exhibit blood pressure abnormalities, and crosses with 5XFAD mice show no changes in amyloid deposition. Rather, the major change appears to be increased expression in neurons.

    The group goes on to examine the effects of ACE inhibitors (ACEIs) and angiotensin receptor blockers (ARBs). They show significant neuroprotection by both. These results raise the possibility that the benefits of ACEIs and ARBs observed in studies of AD subjects might derive from direct actions of these compounds on neurons.

    Indeed, in a study my group published in 2010, we observed that brain-penetrant ARBs were associated with improved outcomes compared to ARBs that did not distribute into the brain (Li et al., 2010). These results suggested that the ARBs were acting on a target in the brain beyond the vasculature. The current work by Cuddy et al. complement these results nicely, although the exact cell type most relevant in AD to ACEIs and ARBs remains to be determined.

    The new ACE1 R1279Q knock-ins provide an outstanding mouse model to explore the role of the renin-angiotensin system in the pathophysiology of AD. I expect many additional insights to come from this valuable mouse model.

    References:

    . Use of angiotensin receptor blockers and risk of dementia in a predominantly male population: prospective cohort analysis. BMJ. 2010;340:b5465. PubMed.

Make a Comment

To make a comment you must login or register.

References

News Citations

  1. Anti-hypertensives for AD?—Remembrances of NSAIDs Past
  2. In Veterans, Blood Pressure Meds Delay Dementia
  3. Can Blood Pressure Drug Put the Squeeze on Brain Amyloid?
  4. GWAS, GWAX: bioRχiv Hosts Bonanza of Alzheimer’s Genetics
  5. Off Key—Aβ Detunes the Theta Rhythms of the Hippocampus
  6. Plaques, Tangles Throw Off the Brain’s Rhythms
  7. Down to Sex? Boy and Girl Microglia Respond Differently
  8. We Are What We Consume? Foods, Drugs Affect Amyloid, AD
  9. Minocycline Does Not Work in Mild Alzheimer’s Disease

Research Models Citations

  1. 5xFAD (B6SJL)

Therapeutics Citations

  1. Saracatinib

Paper Citations

  1. . Genetic meta-analysis of diagnosed Alzheimer's disease identifies new risk loci and implicates Aβ, tau, immunity and lipid processing. Nat Genet. 2019 Mar;51(3):414-430. Epub 2019 Feb 28 PubMed. Correction.
  2. . Genome-wide meta-analysis identifies new loci and functional pathways influencing Alzheimer's disease risk. Nat Genet. 2019 Mar;51(3):404-413. Epub 2019 Jan 7 PubMed.
  3. . ACE gene is associated with Alzheimer's disease and atrophy of hippocampus and amygdala. Neurobiol Aging. 2005 Aug-Sep;26(8):1153-9. PubMed.
  4. . Brain Renin-Angiotensin System and Microglial Polarization: Implications for Aging and Neurodegeneration. Front Aging Neurosci. 2017;9:129. Epub 2017 May 3 PubMed.
  5. . What Are the Odds of Finding a COVID-19 Drug from a Lab Repurposing Screen?. J Chem Inf Model. 2020 Sep 11; PubMed.

Further Reading

Primary Papers

  1. . Aβ-accelerated neurodegeneration caused by Alzheimer's-associated ACE variant R1279Q is rescued by angiotensin system inhibition in mice. Sci Transl Med. 2020 Sep 30;12(563) PubMed.