TY - JOUR
AU - Holmes, Michael, V
AB - Mendelian randomization (MR) has unravelled many new insights into the causal mechanisms driving disease.1,2 Whilst elucidation of the aetiological role of risk factors and biomarkers in disease [e.g. blood lipids in coronary artery disease (CAD)3] is of clinical and public health importance, it is also valuable to know whether particular pharmacological targets, typically proteins, are of therapeutic relevance in specific diseases. ‘Drug target MR’ is an application of MR that uses genetic variants as proxies for drug target modulation.4,5 A well-studied example is that of 3-hydroxy-3-methylglutaryl-CoA reductase (HMGCR), where genetic variants in and close to the HMGCR gene are associated with both lower low-density lipoprotein cholesterol levels and a reduced risk of CAD,6 recapitulating the effects of therapeutic HMGCR inhibition by statins on risk of vascular disease in large-scale randomized trials. Drug target MR may identify novel disease associations of drug targets, thereby informing on potential repositioning opportunities of medicines.5 One such example of repurposing is that of interleukin-6 receptor (IL-6R) inhibition and cardiovascular disease, where genetic variants in IL6R that mimic the effect of therapeutic IL-6R blockade (e.g. reduced risk of rheumatoid arthritis) also associate with a lower risk of CAD and arterial aneurysm.7,8 This suggests that pharmacological inhibition of this target (e.g. with tocilizumab, an IL-6R inhibitor) may produce beneficial cardiovascular effects—a hypothesis supported by tocilizumab’s label extension for giant cell arteritis (a condition where aortic aneurysm is a common complication) in 20179 and by ongoing clinical trials in CAD.10 Investigations such as these may be undertaken in a phenotype- or disease-specific manner or using a hypothesis-free approach through phenome-wide approaches (e.g. applying a PheWAS, or phenome-wide association study, design). The latter approach allows for discovery of potentially unanticipated effects, albeit at the expense of lower statistical power. In this edition of IJE, Walker et al.11 use a disease-specific drug target MR approach to investigate the effects of blood pressure lowering therapies on risk of Alzheimer’s disease (AD). Previous MR studies have investigated whether genetically-mediated reductions in blood pressure may influence risk of AD,12,13 with inconclusive results. In addition, these studies did not investigate whether specific anti-hypertensive therapies may yield unique benefits (or, alternatively, adverse effects) with regards to AD. The approach by Walker et al. is one of several recent examples aiming to leverage gene expression datasets to identify novel effects of drug target modulation.14–16 The use of gene expression data aims to identify instruments with a potential functional link to the gene encoding the drug target (i.e. the exposure of interest), particularly when protein concentration data may not be available. However, there are challenges in using gene expression data.17 For example, it may be difficult to reconcile tissue-specific effects of gene expression with the target tissue(s) of a drug, and only a minority of gene expression quantitative trait loci (eQTL) signals have been confirmed to also affect circulating protein concentrations18 (with the latter being the downstream target for most currently available drugs). Walker et al.’s work, also illustrates the complexity that lies in instrumenting drug targets. They used an approach to select genetic variants that allowed a therapy to target more than one protein product, thereby making the relative contribution of each protein a potential challenge to unpick. Most drugs currently marketed modify only one biological target; however, this may change going forward.19,19,Furthermore, inclusion of variants representing a large number of genes to instrument the targets of a class of drugs may lead to unreliable estimates due to genetic pleiotropy, but potentially allows for the application of sensitivity analyses such as MR-Egger,20 which Walker et al. conducted. Walker et al. did not find evidence for any single antihypertensive target having a protective effect on risk of AD. However, associations of a genetic instrument of a therapeutic target may also signify potential on-target (i.e. target-mediated) adverse effects.21 Walker et al.’s investigation shows that lowering blood pressure through angiotensin-converting enzyme (ACE) inhibitors may lead to a higher risk of AD, an intriguing finding which they do not observe for other anti-hypertensive targets. The effect estimate for this association is relatively large (odds ratio of AD 13.3 per 10 mmHg reduction in systolic blood pressure (SBP), with wide confidence intervals (95% confidence interval 2.1-81.2) and, taking into account the multiple tests conducted, provides only weak evidence of an association. Since the estimate for ACE is larger than that for the effect of SBP overall (odds ratio of 1.04), the authors suggest that this may mean that the AD association is not mediated by SBP, but through other target-mediated pleiotropic effects. Nevertheless, it is worth considering whether the association between ACE modulation and AD may be real or indeed spurious, particularly as ACE inhibitors remain among the most prescribed drugs in the USA and the UK.22,23 In identifying other sources of data, we note that conventional studies in pharmacoepidemiology are subject to numerous sources of bias (for example, confounding by indication),5 and are therefore less well-suited to help us answer this question. How might we investigate this putative relationship identified by Walker et al. further? Let us consider the potential scenarios in which this association, if real, might arise, and the assumptions that such entails (Figure 1). First, we ought to consider whether the instrument selected by Walker et al. (rs4968783, a non-coding single nucleotide polymorphism ∼4 kb upstream from the ACE gene) is associated with the exposure of interest (the relevance assumption in MR2; in our case, modulation of ACE). Given that this is a non-coding variant, we might investigate whether it influences ACE gene expression. Evaluation of mRNA expression in the GTEx database suggests that the instrument is strongly associated with ACE gene expression in several tissues, including lung tissue (P = 1.4 × 10–15).24 Further examination of the existing literature also suggests that variants in high linkage disequilibrium (LD) with the instrument strongly associate with altered levels of ACE in cerebrospinal fluid (P = 3.9 × 10−12; reported lead variant: rs4968782, r2 = 1.00 with rs4968783 in European-ancestry populations).25 The instrument is also robustly associated with blood pressure (P = 1.8 × 10–20 for SBP in a recent large-scale genome-wide association study of blood pressure traits).26 These associations suggest that the variant regulates ACE expression and influences blood pressure—the blood pressure alteration through modulating ACE being commensurate with the effect of therapeutic ACE inhibition, and suggesting that at least some of the instrument’s associations (including, potentially, with AD) are attributable to ACE inhibition. Figure 1 Open in new tabDownload slide Diagram to illustrate potential (non-exhaustive) explanations for the observed association between the angiotensin-converting enzyme (ACE) genetic instrument and Alzheimer’s disease (AD). Target-mediated effects may include effects acting through systolic blood pressure (SBP; A in the diagram) or via non-SBP-mediated mechanisms (e.g. by exerting an effect on amyloid β-protein, here marked as B). The null association (P = 0.45) of the overall SBP instrument (using 135 SBP-associated genetic variants) with risk of AD suggests that the effect of the ACE variant on AD (if target-mediated) is mediated by a non-SBP-related mechanism. Spurious associations may arise if the instrument has effects beyond those mediated by ACE (i.e. horizontal pleiotropy, by e.g. influencing the expression of other nearby genes, marked as C), or if the instrument is in linkage disequilibrium (LD) with another variant that is causally related to AD (confounding through LD, marked as D). Currently available data suggest that both C and D may explain the association between the ACE genetic instrument and AD (see main text). Figure 1 Open in new tabDownload slide Diagram to illustrate potential (non-exhaustive) explanations for the observed association between the angiotensin-converting enzyme (ACE) genetic instrument and Alzheimer’s disease (AD). Target-mediated effects may include effects acting through systolic blood pressure (SBP; A in the diagram) or via non-SBP-mediated mechanisms (e.g. by exerting an effect on amyloid β-protein, here marked as B). The null association (P = 0.45) of the overall SBP instrument (using 135 SBP-associated genetic variants) with risk of AD suggests that the effect of the ACE variant on AD (if target-mediated) is mediated by a non-SBP-related mechanism. Spurious associations may arise if the instrument has effects beyond those mediated by ACE (i.e. horizontal pleiotropy, by e.g. influencing the expression of other nearby genes, marked as C), or if the instrument is in linkage disequilibrium (LD) with another variant that is causally related to AD (confounding through LD, marked as D). Currently available data suggest that both C and D may explain the association between the ACE genetic instrument and AD (see main text). Second, we should consider whether the instrument-outcome association is influenced by horizontal pleiotropy (the situation where a genetic instrument associates with a disease through mechanisms other than through the exposure of interest—i.e. violation of the exclusion restriction criterion17). In drug target MR studies, it is often difficult to apply typical MR sensitivity analyses (e.g. MR-Egger) due to the limited number of exposure-associated instruments. One approach is to examine the association of the instrument with the expression of other genes and/or protein levels, although such associations may be due to chance or may represent target-mediated effects (i.e. vertical pleiotropy). In the GTEx database, rs4968783 associates with the expression of multiple nearby genes other than ACE (e.g. KCNH6, P = 7.0 × 10–28 in lung tissue).24 Such co-expression suggests that the instrument-outcome associations may be driven by more than one gene, thereby possibly introducing pleiotropy which may limit specificity of the instrument for ACE, a potential violation of the exclusion restriction criterion17 (it is worth noting in this case that a recent study found an excess burden of rare deleterious variants associated with AD in KCNH6,27 suggesting that KCNH6 may be responsible for the instrument’s association with AD). Another challenge highlighted by gene expression data is that variants may associate with divergent directions of effect on expression of the same gene in different tissues, posing difficulties in reconciling such effects with that of therapeutic modulation. For example, the rs4968783 A-allele associates with higher ACE mRNA expression in lung tissue, but lower expression in various brain tissues (Figure 2). Figure 2 Open in new tabDownload slide Diagram illustrating the association of rs4968783 with ACE gene expression across multiple tissue types. All estimates are aligned to the A-allele. Diagram modified from GTEx v8.24 NES, normalized effect size; CI, confidence interval. Figure 2 Open in new tabDownload slide Diagram illustrating the association of rs4968783 with ACE gene expression across multiple tissue types. All estimates are aligned to the A-allele. Diagram modified from GTEx v8.24 NES, normalized effect size; CI, confidence interval. Third, it is important to investigate the local linkage disequilibrium (LD) structure for the traits of interest, to mitigate against the risk of ‘confounding through LD’.28,29 Such confounding arises when genetic association with two or more traits is attributable to distinct causal variants, i.e. the variant that influences the exposure is in LD with another variant that independently influences the outcome. One way of evaluating the possibility of such confounding is by visually comparing the genetic association plots for two or more traits and by performing colocalization analyses. If ACE inhibition did lead to a target-mediated increase in risk of AD, we would expect the genetic association signals for SBP and AD in the ACE locus to colocalize. However, assessment of genetic association signals for AD and SBP at the ACE locus suggests that these traits are unlikely to be driven by the same causal variant (Figure 3). Figure 3 Open in new tabDownload slide Genetic association plots for systolic blood pressure26 (SBP; panel A) and Alzheimer’s disease27 (AD; panel B, genome-wide association statistics only available for stage 1 meta-analysis), with the angiotensin-converting enzyme (ACE) instrument used by Walker et al. (rs4968783) highlighted in purple. Each point represents a genetic variant, with colour indicating a variant’s linkage disequilibrium (LD; as measured in r2) with rs4968783. The y-axis indicates the –log10(P-value of association) for each plotted point, with chromosomal coordinates (and gene locations) shown on the x-axis. The posterior probability of a shared causal variant driving the association with SBP and AD is 20% (derived using a Bayesian test for colocalization30 applied to a 400-kilobase window centred around rs4968783 and using default prior settings), suggesting that the association of the instrument with SBP and AD may be confounded through LD. To illustrate colocalizing traits, we also performed the analysis with SBP and self-reported anti-hypertensive medication use31 (panel C). This yielded a posterior probability of 99%. Plots were created using LocusZoom.32 Figure 3 Open in new tabDownload slide Genetic association plots for systolic blood pressure26 (SBP; panel A) and Alzheimer’s disease27 (AD; panel B, genome-wide association statistics only available for stage 1 meta-analysis), with the angiotensin-converting enzyme (ACE) instrument used by Walker et al. (rs4968783) highlighted in purple. Each point represents a genetic variant, with colour indicating a variant’s linkage disequilibrium (LD; as measured in r2) with rs4968783. The y-axis indicates the –log10(P-value of association) for each plotted point, with chromosomal coordinates (and gene locations) shown on the x-axis. The posterior probability of a shared causal variant driving the association with SBP and AD is 20% (derived using a Bayesian test for colocalization30 applied to a 400-kilobase window centred around rs4968783 and using default prior settings), suggesting that the association of the instrument with SBP and AD may be confounded through LD. To illustrate colocalizing traits, we also performed the analysis with SBP and self-reported anti-hypertensive medication use31 (panel C). This yielded a posterior probability of 99%. Plots were created using LocusZoom.32 Fourth, it remains a possibility, albeit remote, that the disease endpoint may be influencing the exposure (AD having a causal influence on SBP or ACE levels, i.e. reverse causality) and various methods now exist to investigate this (e.g. bi-directional MR and/or Steiger filtering).28,33 Finally, it is important to consider orthogonal sources of evidence in support of the hypothesis that inhibition of ACE may increase risk of AD (i.e. through the process of triangulation34). Observational evidence in humans is conflicting, with some studies suggesting that ACE inhibitors might be protective against AD,35 whereas others have shown that patients taking ACE inhibitors have a higher risk of AD.36 AD and brain amyloid β-protein burden are associated with altered brain37 and cerebrospinal fluid25,38 levels of ACE, and evidence in animals suggests that ACE overexpression might protect against cognitive decline in a murine model of AD.39 There are data to suggest that ACE has a role in degradation of amyloid β-protein,39–41 but it is worth noting that most therapeutic approaches to lowering amyloid β-protein (with a notable recent exception42) have thus far failed to modify AD progression in clinical trials43,44—thereby potentially calling into question the aetiological relevance of amyloid β-protein in AD.45,46 In conclusion, Walker et al.’s study illustrates how MR, coupled with a systematic investigation as laid out above, has utility in the genetically guided development of medicines, including repurposing to new indications and elucidating potential adverse effects. The refinement of drug target MR methodologies and availability of larger and more granular datasets, such as tissue-specific protein expression, coupled together with richly characterized biobanks linked to electronic medical records, will facilitate such analyses, providing opportunities to acquire insights into disease aetiology and identify new ways to treat disease. Funding J.B. is supported by funding from the Rhodes Trust, Clarendon Fund and the Medical Sciences Doctoral Training Centre, University of Oxford. J.C.C. is funded by the Oxford Medical Research Council Doctoral Training Partnership (Oxford MRC DTP) and the Nuffield Department of Clinical Medicine, University of Oxford. C.M.L. is supported by the Li Ka Shing Foundation, WT-SSI/John Fell funds, the National Institute for Health Research Biomedical Research Centre, Oxford, Widenlife and National Institutes of Health (5P50HD028138-27). M.V.H. works in a unit that receives funding from the Medical Research Council and is supported by a British Heart Foundation Intermediate Clinical Research Fellowship (FS/18/23/33512) and the National Institute for Health Research Oxford Biomedical Research Centre. Computation used the Oxford Biomedical Research Computing (BMRC) facility, a joint development between the Wellcome Centre for Human Genetics and the Big Data Institute supported by Health Data Research UK and the National Institute for Health Research Oxford Biomedical Research Centre. Financial support was provided by the Wellcome Trust Core Award Grant Number 203141/Z/16/Z. The views expressed are those of the author(s) and not necessarily those of the National Health Service, the National Institute for Health Research or the Department of Health. Conflict of Interest C.M.L. has collaborated with Novo Nordisk and Bayer in research, and in accordance with a university agreement, did not accept any personal payment. M.V.H. has collaborated with Boehringer Ingelheim in research, and in accordance with the policy of the Clinical Trial Service Unit and Epidemiological Studies Unit (University of Oxford), did not accept any personal payment. References 1 Holmes MV , Ala-Korpela M , Davey Smith G. Mendelian randomization in cardiometabolic disease: challenges in evaluating causality . Nat Rev Cardiol 2017 ; 14 : 577 – 90 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Davies NM , Holmes MV , Davey Smith G. Reading Mendelian randomisation studies: a guide, glossary, and checklist for clinicians . BMJ 2018 ; 362 : k601 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Holmes MV , Asselbergs FW , Palmer TM et al. ; on behalf of the UCLEB Consortium. Mendelian randomization of blood lipids for coronary heart disease . Eur Heart J 2015 ; 36 : 539 – 50 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Holmes MV. Human genetics and drug development . N Engl J Med 2019 ; 380 : 1076 – 79 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Walker VM , Davey Smith G , Davies NM , Martin RM. Mendelian randomization: a novel approach for the prediction of adverse drug events and drug repurposing opportunities . Int J Epidemiol 2017 ; 46 : 2078 – 89 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Ference BA , Robinson JG , Brook RD et al. Variation in PCSK9 and HMGCR and risk of cardiovascular disease and diabetes . N Engl J Med 2016 ; 375 : 2144 – 53 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Cai T , Zhang Y , Ho Y-L et al. ; for the VA Million Veteran Program. Association of interleukin 6 receptor variant with cardiovascular disease effects of interleukin 6 receptor blocking therapy: a phenome-wide association study . JAMA Cardiol 2018 ; 3 : 849 – 57 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Interleukin-6 Receptor Mendelian Randomisation Analysis (IL6R MR) Consortium, Swerdlow DI , Holmes MV , Kuchenbaecker KB. The interleukin-6 receptor as a target for prevention of coronary heart disease: a mendelian randomisation analysis . Lancet 2012 ; 379 : 1214 – 24 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Office of the Commissioner. FDA Approves First Drug to Specifically Treat Giant Cell Arteritis. U.S. Food and Drug Administration. 2017 . http://www.fda.gov/news-events/press-announcements/fda-approves-first-drug-specifically-treat-giant-cell-arteritis (22 August 2019, date last accessed). 10 Carroll MB. Tocilizumab in the treatment of myocardial infarction . Mod Rheumatol 2018 ; 28 : 733 – 35 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Walker VM , Kehoe PG , Martin RM , Davies NM. Repurposing antihypertensive drugs for the prevention of Alzheimer’s disease: a Mendelian randomization study . Int J Epidemiol 2019 , Jul 23. doi:10.1093/ije/dyz155. (Epub before print.) WorldCat 12 Larsson SC , Traylor M , Malik R et al. ; CoSTREAM Consortium, on behalf of the International Genomics of Alzheimer’s Project. Modifiable pathways in Alzheimer’s disease: Mendelian randomisation analysis . BMJ 2017 ; 359 : j5375 . Google Scholar Crossref Search ADS PubMed WorldCat 13 Østergaard SD , Mukherjee S , Sharp SJ et al. ; Alzheimer’s Disease Genetics Consortium. Associations between potentially modifiable risk factors and Alzheimer disease: a Mendelian randomization study . PLoS Med 2015 ; 12 : e1001841 ; discussion e1001841. Google Scholar Crossref Search ADS PubMed WorldCat 14 Gaspar HA , Gerring Z , Hübel C et al. ; Major Depressive Disorder Working Group of the Psychiatric Genomics Consortium. Using genetic drug-target networks to develop new drug hypotheses for major depressive disorder . Transl Psychiatry 2019 ; 9 : 117 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Taylor K , Davey Smith G , Relton CL , Gaunt TR , Richardson TG. Prioritizing putative influential genes in cardiovascular disease susceptibility by applying tissue-specific Mendelian randomization . Genome Med 2019 ; 11 : 6 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Fang H , De Wolf H , Knezevic B et al. ; The ULTRA-DD Consortium. A genetics-led approach defines the drug target landscape of 30 immune-related traits . Nat Genet 2019 ; 51 : 1082 – 91 . Google Scholar Crossref Search ADS PubMed WorldCat 17 Wainberg M , Sinnott-Armstrong N , Mancuso N et al. Opportunities and challenges for transcriptome-wide association studies . Nat Genet 2019 ; 51 : 592 – 99 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Sun BB , Maranville JC , Peters JE et al. Genomic atlas of the human plasma proteome . Nature 2018 ; 558 : 73 – 79 . Google Scholar Crossref Search ADS PubMed WorldCat 19 Ramsay RR , Popovic-Nikolic MR , Nikolic K , Uliassi E , Bolognesi ML. A perspective on multi-target drug discovery and design for complex diseases . Clin Transl Med 2018 ;7: 3 . WorldCat 20 Bowden J , Davey Smith G , Burgess S. Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression . Int J Epidemiol 2015 ; 44 : 512 – 25 . Google Scholar Crossref Search ADS PubMed WorldCat 21 Nguyen PA , Born DA , Deaton AM , Nioi P , Ward LD. Phenotypes associated with genes encoding drug targets are predictive of clinical trial side effects . Nat Commun 2019 ; 10 : 1579 . Google Scholar Crossref Search ADS PubMed WorldCat 22 Fuentes AV , Pineda MD , Venkata K. Comprehension of top 200 prescribed drugs in the US as a resource for pharmacy teaching, training and practice . Pharmacy (Basel) 2018 , May 14. doi: 10.3390/pharmacy6020043. WorldCat 23 Hicks BM , Filion KB , Yin H , Sakr L , Udell JA , Azoulay L. Angiotensin converting enzyme inhibitors and risk of lung cancer: population based cohort study . BMJ 2018 ; 363 : k4209 . Google Scholar Crossref Search ADS PubMed WorldCat 24 Aguet F , Barbeira AN , Bonazzola R et al. The GTEx Consortium Atlas of Genetic Regulatory Effects Across Human Tissues . bioRxiv 2019 . https://www.biorxiv.org/content/10.1101/787903v1 (13 November 2019, date last accessed). WorldCat 25 Kauwe JSK , Bailey MH , Ridge PG et al. ; the Alzheimer's Disease Neuroimaging Initiative. Genome-wide association study of CSF levels of 59 Alzheimer’s disease candidate proteins: significant associations with proteins involved in amyloid processing and inflammation . PLoS Genet 2014 ; 10 : e1004758 . Google Scholar Crossref Search ADS PubMed WorldCat 26 Evangelou E , Warren HR , Mosen-Ansorena D et al. Genetic analysis of over 1 million people identifies 535 new loci associated with blood pressure traits . Nat Genet 2018 ; 50 : 1412 – 25 . Google Scholar Crossref Search ADS PubMed WorldCat 27 Kunkle BW , Grenier-Boley B , Sims R et al. ; Genetic and Environmental Risk in AD/Defining Genetic, Polygenic and Environmental Risk for Alzheimer’s Disease Consortium (GERAD/PERADES). Genetic meta-analysis of diagnosed Alzheimer’s disease identifies new risk loci and implicates Aβ, tau, immunity and lipid processing . Nat Genet 2019 ; 51 : 414 – 30 . Google Scholar Crossref Search ADS PubMed WorldCat 28 Hemani G , Bowden J , Davey Smith G. Evaluating the potential role of pleiotropy in Mendelian randomization studies . Hum Mol Genet 2018 ; 27 : R195 – 208 . Google Scholar Crossref Search ADS PubMed WorldCat 29 Fortune MD , Guo H , Burren O et al. Statistical colocalization of genetic risk variants for related autoimmune diseases in the context of common controls . Nat Genet 2015 ; 47 : 839 – 46 . Google Scholar Crossref Search ADS PubMed WorldCat 30 Giambartolomei C , Vukcevic D , Schadt EE et al. Bayesian test for colocalisation between pairs of genetic association studies using summary statistics . PLoS Genet 2014 ; 10 : e1004383 . Google Scholar Crossref Search ADS PubMed WorldCat 31 UK Biobank—Neale Lab. Neale Lab. Available from: http://www.nealelab.is/uk-biobank (14 August 2019, date last accessed). 32 Kreisler RE, Cornell HN, Levy JK. Decrease in population and increase in welfare of community cats in a twenty-three year trap-neuter-return program in key largo, FL: The ORCAT program. Front Vet Sci. 2019;6:7. 33 Hemani G , Tilling K , Davey Smith G. Orienting the causal relationship between imprecisely measured traits using GWAS summary data . PLoS Genet 2017 ; 13 : e1007081 . Google Scholar Crossref Search ADS PubMed WorldCat 34 Lawlor DA , Tilling K , Davey Smith G. Triangulation in aetiological epidemiology . Int J Epidemiol 2016 ; 45 : 1866 – 86 . Google Scholar Crossref Search ADS PubMed WorldCat 35 Gao Y , O'Caoimh R , Healy L et al. Effects of centrally acting ACE inhibitors on the rate of cognitive decline in dementia . BMJ Open 2013 ; 3 : e002881 . Google Scholar Crossref Search ADS PubMed WorldCat 36 Liu S , Ando F , Fujita Y et al. A clinical dose of angiotensin-converting enzyme (ACE) inhibitor and heterozygous ACE deletion exacerbate Alzheimer’s disease pathology in mice . J Biol Chem 2019 ; 294 : 9760 – 70 . Google Scholar Crossref Search ADS PubMed WorldCat 37 Miners S , Ashby E , Baig S et al. Angiotensin-converting enzyme levels and activity in Alzheimer’s disease: differences in brain and CSF ACE and association with ACE1 genotypes . Am J Transl Res 2009 ; 1 : 163 – 77 . Google Scholar PubMed WorldCat 38 Jochemsen HM , Teunissen CE , Ashby EL et al. The association of angiotensin-converting enzyme with biomarkers for Alzheimer’s disease . Alzheimers Res Ther 2014 ; 6 : 27 . Google Scholar Crossref Search ADS PubMed WorldCat 39 Bernstein KE , Koronyo Y , Salumbides BC et al. Angiotensin-converting enzyme overexpression in myelomonocytes prevents Alzheimer’s-like cognitive decline . J Clin Invest 2014 ; 124 : 1000 – 12 . Google Scholar Crossref Search ADS PubMed WorldCat 40 Hemming ML , Selkoe DJ. Amyloid β-protein is degraded by cellular angiotensin-converting enzyme (ACE) and elevated by an ACE inhibitor . J Biol Chem 2005 ; 280 : 37644 – 50 . Google Scholar Crossref Search ADS PubMed WorldCat 41 Kugaevskaya EV , Veselovsky AV , Indeykina MI et al. N-domain of angiotensin-converting enzyme hydrolyzes human and rat amyloid-β(1-16) peptides as arginine specific endopeptidase potentially enhancing risk of Alzheimer’s disease . Sci Rep 2018 ; 8 : 298 . Google Scholar Crossref Search ADS PubMed WorldCat 42 Abbott A. Fresh Push for ‘Failed’ Alzheimer’s Drug. Nature Publishing Group, 2019. https://www.nature.com/articles/d41586-019-03261-5 (13 November 2019, date last accessed). 43 Mehta D , Jackson R , Paul G , Shi J , Sabbagh M. Why do trials for Alzheimer’s disease drugs keep failing? A discontinued drug perspective for 2010-2015 . Expert Opin Investig Drugs 2017 ; 26 : 735 – 39 . Google Scholar Crossref Search ADS PubMed WorldCat 44 Mullard A. Anti-amyloid failures stack up as Alzheimer antibody flops . Nat Rev Drug Discov 2019 . doi: 10.1038/d41573-019-00064-1. WorldCat 45 Morris GP , Clark IA , Vissel B. Questions concerning the role of amyloid-β in the definition, aetiology and diagnosis of Alzheimer’s disease . Acta Neuropathol 2018 ; 136 : 663 – 89 . Google Scholar Crossref Search ADS PubMed WorldCat 46 Knopman DS. Lowering of amyloid-beta by β-secretase inhibitors—some informative failures . N Engl J Med 2019 ; 380 : 1476 – 78 . Google Scholar Crossref Search ADS PubMed WorldCat © The Author(s) 2020; all rights reserved. Published by Oxford University Press on behalf of the International Epidemiological Association This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Using human genetics to guide the repurposing of medicines
JF - International Journal of Epidemiology
DO - 10.1093/ije/dyaa015
DA - 2004-03-01
UR - https://www.deepdyve.com/lp/oxford-university-press/using-human-genetics-to-guide-the-repurposing-of-medicines-UrwlOaPiLQ
SP - 1
VL - Advance Article
IS - 
DP - DeepDyve
ER -