The identification of ACE as a possible AD gene and ACE inhibitors as potential risk factors for AD are both potentially very important observations. Similarly, the many reports on relatively nontoxic dietary factors modulating amyloidosis in animal models is reason to believe that we will be able to find ways to prevent AD. The fact that there are so many possible approaches should not jade people, or convince them that we can’t all be right.

It seems highly likely that Alzheimer’s, like most other late-onset diseases of aging, has multiple and usually weak genetic and environmental influences that modulate susceptibility. In contrast, diseases with strong single genetic or environmental risk factors will typically be more clearly causal, with earlier onset due to the potent genetic risk factors (e.g., autosomal dominant) or gross deficiencies of essential nutrients (scurvy, rickets, etc.). In this situation, common sense suggests a multifactorial approach to address these multiple risk factors for the prevention of late-onset AD. And common sense suggests that this is what...  Read more

The identification of ACE as a possible AD gene and ACE inhibitors as potential risk factors for AD are both potentially very important observations. Similarly, the many reports on relatively nontoxic dietary factors modulating amyloidosis in animal models is reason to believe that we will be able to find ways to prevent AD. The fact that there are so many possible approaches should not jade people, or convince them that we can’t all be right.

It seems highly likely that Alzheimer’s, like most other late-onset diseases of aging, has multiple and usually weak genetic and environmental influences that modulate susceptibility. In contrast, diseases with strong single genetic or environmental risk factors will typically be more clearly causal, with earlier onset due to the potent genetic risk factors (e.g., autosomal dominant) or gross deficiencies of essential nutrients (scurvy, rickets, etc.). In this situation, common sense suggests a multifactorial approach to address these multiple risk factors for the prevention of late-onset AD. And common sense suggests that this is what current research is setting the stage for.

This is not a new lesson. With intense cancer research over the last 50 years, we have fairly good epidemiology and preclinical data to support a spectrum of risk and protective factors. We have a well-developed list of oncogenes, and multiple carcinogens and anticarcinogens present in the environment and diet. We also have some idea about initiation and promotion phases and very good animal models. We can make sensible recommendations about diet, smoking, and pollutants to minimize mutagen and carcinogen exposure.

Unfortunately, though, apart from a few successes like sunscreens, cancer chemoprevention lags behind. We can only hope that with AD, researchers will fare better, as indeed they have with atherosclerosis. There, some 50 years of concerted research has also led to a well-established list of risk and protective factors that allows the formulation of “heart healthy” recommendations with some strong clinical trial support for efficacy. With the great overlap between risk factors for AD and heart disease, we can only hope that controlling the common risk factors will limit our risk for both diseases. With this in mind, the knowledge that ACE inhibitors may increase AD risk is certainly significant, but somewhat disconcerting news.

With Alzheimer’s and other neurodegenerative diseases of aging, 20 years of intense effort have brought us less far along, but, unsurprisingly, put us on a similar track. AD researchers at the bench have only recently developed suitable (albeit still imperfect) animal models. AD clinicians were at a comparative disadvantage in establishing biomarkers. They lacked large populations of clearly high-risk patients in “remission” or at high risk of second cardiovascular events to conduct prospective clinical trials. Now that AD researchers have the animal models and the MCI patient pools, we will begin to see what translates.

We will likely end up with a set of “alzogens” to limit our exposure and “anti-alzogens” whose intake we want to optimize, albeit being mindful of potential side effect profiles. The polyphenolic antioxidants including green tea catechins, resveratrol, and curcumin (and doubtlessly others as yet unexplored) have real potential as protective factors for several diseases of aging, including AD. While these polyphenols have several overlapping antioxidant and anti-inflammatory properties, each of them also has its own unique targets, issues, and merits that add up to a reasonable case for further investigation. The catechins seem to reduce Aβ production, while curcumin limits Aβ aggregation and resveratrol may be particularly useful in protecting DNA via sirtuin.

I am personally convinced that cocktails of protective agents, notably including the polyphenols, would be a logical way to go, but optimizing and clinically testing more than one drug is a formidable task. Our choice to pursue curcumin for AD came after an initial in-vivo Aβ infusion model drug screen made some 10 years ago, together with a broad consideration of multiple factors, including the advanced stage of preclinical and clinical drug development for other diseases, costs of production, remarkably benign toxicity profile, and long history of use. Another factor was curcumin’s “cocktail” of active products tetrahydrocurcumin, vanillin, and ferulic acid. Today we can point to multiple mechanisms of action and an ever-increasing list of diseases where curcumin and its natural products look useful in animal models. There is even data on lifespan extension.

View all comments by Gregory Cole

So much for ACE inhibitors being the "perfect pill" and protective as one ages. I find this upsetting, but clinically important information. I am hopeful that more data will be accumulated on this issue quickly.

View all comments by J. Lucy Boyd

I was very interested to see the Hemming and Selkoe study regarding the possibility that ACE inhibition may not be advisable in AD.

I refer to the recent ARF news article (1), which reports that Wolozin and colleagues find that the relative risk for AD in the CABG group was 1.7-fold that of the PTCA group.

It's interesting that Pell et al. (2) report that 22 percent of CABG patients were on angiotensin-converting enzyme inhibitors, compared with 15 percent of PTCA patients.

References:
1. ARF related news story.

2. Pell JP, Walsh D, Norrie J, Berg G, Colquhoun AD, Davidson K, Eteiba H, Faichney A, Flapan A, Hogg KJ, Jeffrey RR, Jennings K, McArthur J, Mankad P, Oldroyd K, Pell AC, Starkey IR. Outcomes following coronary artery bypass grafting and percutaneous transluminal coronary angioplasty in the stent era: a prospective study of all 9890 consecutive patients operated on in Scotland over a two year period. Heart. 2001 Jun;85(6):662-6. Abstract

View all comments by Mary Reid

It is difficult to know whether the anti-amyloid effect of the mentioned natural compounds (i.e., resveratrol or EGCG) observed in cell culture systems or even in mouse models may explain or support the beneficial effect of specific diets. This effect may represent only the tip of the iceberg. Indeed, wine contains more than 600 different components, including well-characterized antioxidant molecules. It is, therefore, difficult to narrow down the beneficial effect of wine or green tea intake to one specific compound. Furthermore, we cannot exclude the possibility that several compounds work in synergy to slow down the progression of the neurodegenerative process in human.

The oral bioavailability of resveratrol is almost null due to efficient metabolism by the kidney system (see Wenzel and Somoza, 2005). Therefore we do not believe that resveratrol could be used per se as an anti-amyloidogenic drug in vivo. Its potential biological activity in the brain after peripheral administration is, therefore, very questionable. However,...  Read more

It is difficult to know whether the anti-amyloid effect of the mentioned natural compounds (i.e., resveratrol or EGCG) observed in cell culture systems or even in mouse models may explain or support the beneficial effect of specific diets. This effect may represent only the tip of the iceberg. Indeed, wine contains more than 600 different components, including well-characterized antioxidant molecules. It is, therefore, difficult to narrow down the beneficial effect of wine or green tea intake to one specific compound. Furthermore, we cannot exclude the possibility that several compounds work in synergy to slow down the progression of the neurodegenerative process in human.

The oral bioavailability of resveratrol is almost null due to efficient metabolism by the kidney system (see Wenzel and Somoza, 2005). Therefore we do not believe that resveratrol could be used per se as an anti-amyloidogenic drug in vivo. Its potential biological activity in the brain after peripheral administration is, therefore, very questionable. However, this observation is a powerful starting point for screening analogues of resveratrol for more active and more stable compounds, a task in which our laboratory is actively involved.

View all comments by Philippe Marambaud

Regarding the impact of certain foods, beverages, and drugs on the development of AD, I think it is likely a situation similar to that in cancer epidemiology: A healthy lifestyle lowers risk; a bad life style enhances risk. Of course, genetic factors provide a background which may determine how effective these changes in risk prove to be. It doesn't much matter what you eat or drink if you have an aggressive PS1 mutation; you'll get AD. And it may not matter too much, either, if you have an ApoE2 allele, since you are well protected (this is less certain, but makes the point). For the rest of the population, there is probably a sliding scale of risk. Nothing is absolutely protective or absolutely causal.

We need to try to think in terms of risk/benefit ratios. I take vitamin C and vitamin E every day, and have done so for years, as it seems reasonably clear that the risk/benefit ratio is in favor of these compounds. They may lower my risk of AD more than they raise my risk of cardiovascular disease. It's hard to come up with real numbers, or any degree of certainty, with the...  Read more

Regarding the impact of certain foods, beverages, and drugs on the development of AD, I think it is likely a situation similar to that in cancer epidemiology: A healthy lifestyle lowers risk; a bad life style enhances risk. Of course, genetic factors provide a background which may determine how effective these changes in risk prove to be. It doesn't much matter what you eat or drink if you have an aggressive PS1 mutation; you'll get AD. And it may not matter too much, either, if you have an ApoE2 allele, since you are well protected (this is less certain, but makes the point). For the rest of the population, there is probably a sliding scale of risk. Nothing is absolutely protective or absolutely causal.

We need to try to think in terms of risk/benefit ratios. I take vitamin C and vitamin E every day, and have done so for years, as it seems reasonably clear that the risk/benefit ratio is in favor of these compounds. They may lower my risk of AD more than they raise my risk of cardiovascular disease. It's hard to come up with real numbers, or any degree of certainty, with the compounds Gabrielle Strobel discusses.

I've been thinking about our approach in this area. We test everything in short-term studies to see if it makes a difference to the cognitive performance of AD patients, but this is a high hurdle to climb, and most of the compounds discussed in this article are unlikely to succeed in these studies. Does this mean that they are useless? Probably not. A diet rich in fruit and veggies, moderate or no alcohol consumption, no smoking, lots of exercise, physical and mental—all this probably lowers your risk of AD.

Would switching patients who already have the disease to a healthy lifestyle improve cognitive function? Unlikely, and so we dismiss these factors in favor of the next "wonder drug." But I suspect that if we all adopted this kind of lifestyle, the incidence of AD would decline significantly. If you believe the epidemiology, adding vitamins C and E would cut the incidence of AD by about 50 percent. Add the other goodies, and perhaps we could do even better.

But we don't think this way because we don't care about populations; we care about individuals, especially ourselves. I want a drug that will stop me from getting AD, or treat AD if I do get it. I don't want my relative risk decreased—I want it eliminated! It isn't likely that any of the lifestyle factors will do this, and they will probably end up being largely ignored. We want cures that can be taken "just in time," not major, lasting changes to our lifestyles that may or may not work for us.

View all comments by Peter Davies

This study by Eckman and colleagues elegantly shows how instructive pharmacology and genetics can be in understanding, and perhaps preventing, a complex disease. Using vasopeptidase inhibitors already used or in development for the clinic, the authors demonstrate that acute inhibition of NEP and/or ECE, but not ACE, results in elevation of cerebral and plasma Aβ levels. Furthermore, mice lacking expression of ACE in the brain are not burdened by elevated levels of murine Aβ, while, interestingly, in doubly NEP and ECE-1- or -2-deficient mice there is an additive accumulation of the peptide.

Where does this leave us in understanding the human disease? The findings on NEP and ECE inhibitors suggest caution when using these potential drugs in patients at risk for AD. The story with ACE may be more complicated. From a genetic perspective, there is a tremendous amount of work supporting the involvement of ACE in Alzheimer disease (AD), but less so for NEP and ECE. From clinical studies, reports of ACE inhibitor use have ranged from showing improvement to no effect on the course...  Read more

This study by Eckman and colleagues elegantly shows how instructive pharmacology and genetics can be in understanding, and perhaps preventing, a complex disease. Using vasopeptidase inhibitors already used or in development for the clinic, the authors demonstrate that acute inhibition of NEP and/or ECE, but not ACE, results in elevation of cerebral and plasma Aβ levels. Furthermore, mice lacking expression of ACE in the brain are not burdened by elevated levels of murine Aβ, while, interestingly, in doubly NEP and ECE-1- or -2-deficient mice there is an additive accumulation of the peptide.

Where does this leave us in understanding the human disease? The findings on NEP and ECE inhibitors suggest caution when using these potential drugs in patients at risk for AD. The story with ACE may be more complicated. From a genetic perspective, there is a tremendous amount of work supporting the involvement of ACE in Alzheimer disease (AD), but less so for NEP and ECE. From clinical studies, reports of ACE inhibitor use have ranged from showing improvement to no effect on the course of AD. A recent study reported finding that, on average, antihypertensive medications lowered the incidence of AD (Khachaturian et al., 2006). Interestingly, ACE inhibitors were the only drug class to actually elevate AD risk, though the comparison to control did not reach significance. Clearly, more questions must be answered to understand the role of ACE genetics and ACE inhibition in Alzheimer’s disease.

One limitation of animal modeling of human diseases is genetic homology. Murine Aβ differs from human at amino acid positions 5, 10, and 13 (with the proposed ACE cleavage site between residues 7 and 8), and murine ACE shares 83 percent amino acid identity with the human form. Human ACE has been shown to cleave human Aβ in several in vitro studies. However, it is unclear if murine ACE can cleave murine Aβ, if murine ACE can cleave human Aβ, or if human ACE can cleave murine Aβ. While other proteases have been shown to actively degrade Aβ in murine and human forms, it is unclear if ACE is an exception.

One possibility is that ACE is a minor Aβ-degrading protease, due to kinetic parameters, substrate competition, localization, or some other factor. How would ACE’s role in Aβ metabolism change as Aβ levels increase, plaques form, and other Aβ-degrading proteases become saturated? Could chronic ACE inhibition reveal a role for the protease in Aβ metabolism? Given that hypertension is a chronic condition, often treated with ACE inhibitors, and that Alzheimer disease develops over a lifetime, these questions warrant further investigation, and the work reported here takes many steps toward answering them.

References:
Khachaturian AS, Zandi PP, Lyketsos CG, Hayden KM, Skoog I, Norton MC, Tschanz JT, Mayer LS, Welsh-Bohmer KA, Breitner JC. Antihypertensive medication use and incident Alzheimer disease: the Cache County Study. Arch Neurol. 2006 May;63(5):686-92. Epub 2006 Mar 13. Abstract

View all comments by Matthew Hemming

With this study, Eckman and colleagues have made an important contribution to our understanding of the possible roles of ACE and other Aβ-metabolizing enzymes in the pathogenesis of AD. It comes at a time of pressing need for in vivo data on the catabolism of Aβ and the mechanism of involvement of ACE in AD, following on from successive and consistently supportive meta-analyses of genetic association between ACE gene polymorphism and AD, and in vitro and cell-based overexpression studies demonstrating ACE-mediated degradation of Aβ.

The Eckman study provides interesting data and seems to argue against ACE being a significant player in vivo, and Matthew Hemming has also commented in Alzforum that “The story with ACE may be more complicated." Longitudinal and cross-sectional clinical studies have shown cognitive benefits from anti-hypertensive medications, including ACE inhibitors, but the picture remains unclear due to variability among these studies in the measurement and interpretation of cognitive performance and decline and the absence of neuropathological information. In...  Read more

With this study, Eckman and colleagues have made an important contribution to our understanding of the possible roles of ACE and other Aβ-metabolizing enzymes in the pathogenesis of AD. It comes at a time of pressing need for in vivo data on the catabolism of Aβ and the mechanism of involvement of ACE in AD, following on from successive and consistently supportive meta-analyses of genetic association between ACE gene polymorphism and AD, and in vitro and cell-based overexpression studies demonstrating ACE-mediated degradation of Aβ.

The Eckman study provides interesting data and seems to argue against ACE being a significant player in vivo, and Matthew Hemming has also commented in Alzforum that “The story with ACE may be more complicated." Longitudinal and cross-sectional clinical studies have shown cognitive benefits from anti-hypertensive medications, including ACE inhibitors, but the picture remains unclear due to variability among these studies in the measurement and interpretation of cognitive performance and decline and the absence of neuropathological information. In some studies, the long-term cognitive benefits of ACE inhibitors have been substantially less than those of other types of anti-hypertensive medications (e.g., Khachaturian et al., 2006). However, clinical trial data on patients with probable Alzheimer disease (Ohrui et al., 2004) and an observational study of patients with MCI (Rozzini et al., 2006) both suggest that the use of brain-penetrating ACE inhibitors is helpful in delaying cognitive decline in AD or stabilizing cognitive function in MCI.

The clinical picture as far as ACE is concerned is therefore far from clear, and in interpreting the findings of Eckman and colleagues we should be mindful of several limitations/implications of this study:

1. As Hemming has highlighted (and Eckman and colleagues acknowledge themselves), it would be premature to discount the possible effects of species differences in trying to extrapolate findings based upon interactions between murine forms of ACE and Aβ to the situation in humans.

2. The animals were young when tested. Aging affects a range of metabolic processes and compensatory pathways, including those involved in Aβ synthesis and clearance. Similar studies (as well as examination of other RAAS components) are needed on older animals.

3. A point the authors acknowledge is that their pharmacological modeling does not negate possible effects of chronic administration of ACE-Is on Aβ metabolism.

4. A fourth point and one partly suggested by Hemming is that ACE may be only “a minor Aβ-degrading protease, due to kinetic parameters, substrate competition, localization, or some other factor.” We would extend this point with respect to the localization of enzyme activity. The Eckman study did not examine the possible relevance of ACE-mediated Aβ degradation with respect to cerebrovascular Aβ (cerebral amyloid angiopathy). In animals this young there is little or no cerebrovascular Aβ deposition, and it is unlikely that differences in vascular Aβ or enzyme activity would have been detected in the crude homogenates analyzed in this study; however, even if ACE plays only a minor role in overall degradation of Aβ, the localization of this activity to blood vessels could have important clinical implications.

5. The final point we would make is that the findings in this study have implications that extend beyond Alzheimer disease. Considerable epidemiological evidence points to a relationship between hypertension, risk factors for atherosclerotic vascular disease and Alzheimer disease. It is important to consider the effects that intervention (such as enzyme inhibitors) directed at reducing the risk at one disease may have on the risk of others. We hope that this study will encourage more research on the metabolic processes and interactions that are involved in the development and progression of these overlapping diseases.

References:
Khachaturian AS, Zandi PP, Lyketsos CG, Hayden KM, Skoog I, Norton MC, Tschanz JT, Mayer LS, Welsh-Bohmer KA, Breitner JC. Antihypertensive medication use and incident Alzheimer disease: the Cache County Study. Arch Neurol. 2006 May;63(5):686-92. Abstract

Ohrui T, Tomita N, Sato-Nakagawa T, et al. Effects of brain-penetrating ACE inhibitors on Alzheimer disease progression. Neurology 2004;63(7):1324-5. Abstract

Rozzini L, Chilovi BV, Bertoletti E, et al. Angiotensin converting enzyme (ACE) inhibitors modulate the rate of progression of amnestic mild cognitive impairment. Int J Geriatr Psychiatry 2006;21(6):550-5. Abstract

View all comments by Patrick Kehoe
View all comments by Seth Love