- Email: [email protected]
Meta-analysis of the association between two neprilysin gene polymorphisms and Alzheimer's disease
Journal of the Neurological Sciences 346 (2014) 6–10
Contents lists available at ScienceDirect
Journal of the Neurological Sciences journal homepage: www.elsevier.com/locate/jns
Review article
Meta-analysis of the association between two neprilysin gene polymorphisms and Alzheimer's disease Guo Xingzhi a,b, Tang Peng a, Liu Peng a, Liu Yue a, Hou Chen a, Li Rui a,⁎ a b
Department of Neurology, Shaanxi Provincial People's Hospital, Xi'an 710068, China The Third Affiliated Hospital, Xi'an Jiaotong University School of Medicine, Xi'an 710061, China
a r t i c l e
i n f o
Article history: Received 8 May 2014 Received in revised form 23 July 2014 Accepted 29 July 2014 Available online 7 August 2014 Keywords: Neprilysin Polymorphism β-Amyloid Alzheimer's disease Meta-analysis Risk
a b s t r a c t Objective: The aim of this study is to evaluate the association between two neprilysin variants (rs989692 and rs3736187) and Alzheimer's disease (AD). Methods: All eligible studies were searched in PubMed and Embase from inception to July 2014. Data was extracted by two investigators independently. The complete overdominant model (CC + TT vs. CT) and codominant model (GG vs. AA and GA vs. AA) were used for rs989692 and rs3736187, respectively. A comparison of allele frequencies was also conducted. Results: Six studies containing 2555 AD patients and 1914 controls were included for rs989692 polymorphisms. The pooled odds ratio (OR) and confidence interval (CI) suggested that rs989692 polymorphisms were not associated with AD based on the current published studies (C vs. T, OR = 1.01, 95% CI = 0.85–1.19; CC + TT vs. CT, OR = 0.89, 95% CI = 0.78–1.01). Five studies containing 2438 AD patients and 1452 controls were identified for rs3736187 polymorphisms (G vs. A, OR = 0.77, 95% CI = 0.66–0.91; GG vs. AA, OR = 0.38, 95% CI = 0.19–0.77; GA vs. AA, OR = 0.81, 95% CI = 0.61–0.99). The result showed that rs3736187 polymorphisms were likely associated with the decreased risk of AD. Conclusions: This meta-analysis indicates that rs3736187 (A/G) polymorphisms may be a potential beneficial single nucleotide polymorphism (SNP), which are associated with a decreased risk in AD. Further larger scale studies are necessary to validate gene-to-gene interactions and to define the association of neprilysin polymorphisms with AD. © 2014 Elsevier B.V. All rights reserved.
Contents 1. 2.
Introduction . . . . . . . . . . . . . . . . Methods and materials . . . . . . . . . . . 2.1. Literature search and inclusion criteria . 2.2. Data extraction and outcome measures 2.3. Assessment of risk of bias . . . . . . . 2.4. Statistical analysis . . . . . . . . . . 3. Results . . . . . . . . . . . . . . . . . . 3.1. Selection of eligible studies . . . . . . 3.2. Study characteristics . . . . . . . . . 3.3. Quantitative data synthesis . . . . . . 3.4. Assessment of bias . . . . . . . . . . 4. Discussion . . . . . . . . . . . . . . . . . Conflicts of interest . . . . . . . . . . . . . . . Acknowledgments . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
6 7 7 7 7 7 7 7 7 7 8 8 9 9 9
1. Introduction ⁎ Corresponding author at: Department of Neurology, Shaanxi Provincial People's Hospital, No. 256, Youyi West Road, Xi'an 710068, China. Tel.: +86 13096923029. E-mail address: [email protected] (R. Li).
http://dx.doi.org/10.1016/j.jns.2014.07.064 0022-510X/© 2014 Elsevier B.V. All rights reserved.
Alzheimer's disease (AD) is the most common neurodegenerative disease in the elderly, affecting about 24 million people in the world
X. Guo et al. / Journal of the Neurological Sciences 346 (2014) 6–10
in 2012 [1]. The main characteristics of AD are senile plaque containing β-amyloid (Aβ), neurofibrillary tangles and neuronal loss. Numerous studies suggest that the accumulation of Aβ in the brain was the major cause of AD, which could induce inflammatory response and neuronal death [2–4]. Promoting the degradation of Aβ was supposed to be a potential strategy for the treatment of AD. Multiple Aβ-cleaving enzymes, including neprilysin (NEP), insulin-degrading enzyme (IDE), angiotensin-converting enzyme (ACE) and matrix metalloproteinase (MMP), play a vital role in the regulation Aβ level in the brain [5,6]. Neprilysin (NEP), also known as membrane metallo-endopeptidase (MME), is a type II metalloproteinase widely expressed in the neocortex and cerebral vessels. Studies showed that NEP could remove Aβ and inhibit the formation of Aβ deposition [7]. Deficit of NEP could result in the aggregation of Aβ [8]. The gene encoding human NEP is located on chromosome 3q25.1-q25.2 (MIM: 120520) and is highly polymorphic. Among the different polymorphisms in NEP, two SNPs, rs989692 and rs3736187, are extensively studied. Large number of studies indicated that both SNPs are considered to be associated with AD [9–13]; however, the results remain controversial, with some studies unable to identify any association between neprilysin polymorphisms and AD. Therefore, we conducted a meta-analysis to assess the inconsistent results from published studies and clarify the association between rs989692 or rs3736187 polymorphisms and AD. 2. Methods and materials 2.1. Literature search and inclusion criteria We performed a comprehensive search in the PubMed and Embase databases up to July 2014 and selected the original studies surveying the neprilysin polymorphisms and AD. We also searched the AlzGene database for additional articles. The following search terms were used: neprilysin, NEP, metallo-endopeptidase, MME, polymorphism*, variant*, genotype, allele* and Alzheimer's disease*, Alzheimer disease*, and AD. The language was limited to English and Chinese. The search was restricted to human subjects. Studies included in this meta-analysis must meet all the following inclusive selection criteria: a) evaluation of the association between rs989692 or rs3736187 and AD; b) conformance to Hardy–Weinberg equilibrium (HWE); c) case–control design and if there were studies having overlapped subjects, only the one with a larger sample size was selected; and d) useful data available for estimating the odds ratio (OR) with 95% confidence intervals (CIs). 2.2. Data extraction and outcome measures For each study, information was recorded including first author, year of publication, country of origin, ethnicity of population, study design, definition of Alzheimer's disease, number of patients and controls, genotyping methods and genotype distributions. Data were extracted independently by two investigators (Guo Xingzhi and Tang Peng) using a standardized data extraction form. Agreement was reached after discussion for conflicting data with a third author (Li Rui), and then extracted data were typed into an excel file by another author (Hou Chen). 2.3. Assessment of risk of bias The Newcastle–Ottawa Scale (NOS) was conducted to assess the risk of bias [14]. This scale evaluates quality of observational studies from three aspects, including selection (0–4 points), exposure (0–3 points) and comparability (0–2 points). A score of 1 is deducted when one point is unmatched and high scores indicate better quality. Two authors (Liu Peng and Liu Yue) independently estimated the risk of bias.
7
2.4. Statistical analysis The association between the NEP gene polymorphism at rs989692 or rs3736187 and AD was estimated by calculating pooled ORs with 95% CIs. According to the recently suggested guidelines for analysis of genetic association data [15], we adopted the most probably appropriate genetic model for both SNPs. The allele model and complete overdominant model were performed for rs989692, whereas allele model and codominant model were performed for rs3736187. The Cochran Q test and I-squared test were conducted to assess the statistical heterogeneity (P b 0.1 was considered significant). A fixed-effects model (Mantel–Haenszel method) [16] or random-effects model (DerSimonian and Laird method) [17] was applied according to the presence (P b 0.1) or absence (P N 0.1) heterogeneity. Sensitive analysis was used to assess the stability of the results by sequentially excluding one publication at a time. Begg's and Egger's tests were applied to assess the publication bias. All statistical analyses and graphics were conducted in Stata 12.0 and Revman 5.2 software. P b 0.05 was considered statistically significant. 3. Results 3.1. Selection of eligible studies A total of 94 potential studies were identified by the initial search in PubMed and Embase databases, and 25 duplicate studies were excluded After reviewing titles and abstracts, 53 studies were excluded for they were reviews or not relevant to our analysis. Sixteen potentially relevant articles were reviewed for full-text analysis (Fig. 1). Ten studies were excluded: 9 articles for other diseases or SNPs and one with overlapped data [18]. One additional study with available data was identified in the AlzGene database and included in the analysis [19]. Thus, a total of 7 studies were included in final meta-analysis and four of them evaluated both rs989692 and rs3736187 polymorphisms [9–13, 19,20]. 3.2. Study characteristics The detailed characteristics of each included studies were presented in Table 1. These studies were published from 2004 to 2014. All articles were case–control studies. Six studies defined AD according to the NINCDS–ADRDA criteria and the DSM-IV criteria, and only Miners et al. used the CERAD criteria for AD diagnosis [13]. Among the included studies, four of them evaluated both rs989692 and rs3736187 polymorphisms [9,10,13,20]. For the rs989692 polymorphism, all the six studies were performed in Caucasians and 2555 AD patients and 1914 normal controls were included [9,10,12,13,19,20]. The study of Helisalmi et al. screened the mutations in APP, PSEN-1 or PSEN-2 genes, but none of the mutations were found [9]. For the rs3736187 polymorphism, finally five studies were included in our meta-analysis with 2438 AD patients and 1452 normal controls [9–11,13,20]. Four studies were conducted in Caucasians, and only one was performed in Asia [11]. 3.3. Quantitative data synthesis The result for the rs989692 polymorphism was presented in Fig. 2. Compared to the T allele, the C allele had no significant association with AD susceptibility for both models (C vs. T, OR = 1.01, 95% CI = 0.85–1.19; CC + TT vs. CT, OR = 0.89, 95% CI = 0.78–1.01). A fixedeffects model was conducted for the complete overdominant model (CC + TT vs. CT) without significant heterogeneity, whereas a random-effects model was used for the allele model which had a significant heterogeneity. Thus, the sensitivity analyses were applied to explore the potential sources of heterogeneity. Exclusion of the
8
X. Guo et al. / Journal of the Neurological Sciences 346 (2014) 6–10
Fig. 1. Selection for eligible studies included in this meta-analysis.
study of Helisalmi et al. resolved the heterogeneity but not changed the results. Fig. 3 shows the result of the association between the rs3736187 polymorphism and AD. There was no evidence of heterogeneity, so a fixed-effects model was performed. The results were as follows: G vs. A, OR = 0.77, 95% CI = 0.66–0.91; GG vs AA, OR = 0.38, 95% CI = 0.19–0.77; and GA vs AA, OR = 0.81, 95% CI = 0.67–0.99. After combining the ORs and CIs, we found that the rs3736187 polymorphism was significantly associated with AD in all the genetic models. 3.4. Assessment of bias Both Egger's and Begg's tests were used to assess the publication bias of the literature. The results showed that there was no obvious publication bias in overall analysis. For rs989692 polymorphisms, the P values of Egger's test for the allele model and the complete overdominant model were 0.970 and 0.976, respectively. For rs3736187 polymorphisms, the P values of Egger's test for the allele model, the
homozygote codominant model, and the heterozygote codominant model were 0.932, 0.823 and 0.843, respectively. After evaluating the quality of the included studies based on NOS score, the median NOS score (6.5, range from 5 to 7) was obtained. 4. Discussion AD is the leading cause of neurodegenerative disease in the elderly. Deposition of Aβ is the main neuropathological hallmark in AD, which can lead to the neuronal death [21]. A host of enzymes are involved in the degradation of Aβ in the brain and neprilysin is one of the Aβdegrading enzymes [22]. Increasing the expression of NEP could promote the degradation of β-amyloid [23,24]. Studies had been conducted to evaluate the association of neprilysin polymorphisms and AD, but the results were inconsistent. Two SNPs in neprilysin, rs989692 and rs3736187, were widely studied, so we selected these two SNPs to explore the association between neprilysin polymorphisms and AD risks.
Table 1 Characteristics of studies included in this meta-analysis. Author
Country
(A) Studies for rs989692 polymorphism Blomqvist Sweden/USA Giedraitis Sweden Helisalmi Finland Millard USA Miners UK/Italy/Sweden Wood USA
Ethnicity
Year
Sample
Sample
Average age
Cases
Controls
Cases
Control
Genotyping
Caucasian Caucasian Caucasian Caucasian Caucasian Caucasian
2010 2009 2004 2014 2012 2007
Blood Blood Blood Blood Brain tissue and blood Blood?
665 85 390 99 1018 298
184 400 468 168 398 296
76.2 80.2 72 71 NA 72.6
73.2 81.8 70 67 NA 74.7
Hybridization Gene chip and minisequencing PCR PCR PCR OLA/PCR
(B) Studies for rs3736187 polymorphism Blomqvist Sweden/USA Caucasian Helisalmi Finland Caucasian Miners UK/Italy/Sweden Caucasian Wood USA Caucasian Zhou et China Asian
2010 2004 2012 2007 2010
Blood Blood Brain tissue and Blood Blood? Blood
650 390 984 297 111
195 468 375 297 117
76.2 72 NA 72.6 72
73.2 70 NA 74.7 71
Hybridization PCR PCR PCR PCR-RFLP
X. Guo et al. / Journal of the Neurological Sciences 346 (2014) 6–10
9
Fig. 2. Forest plot for the association of rs989692 polymorphisms and AD. Each square represents an OR for each specific study, and the area is proportional to the weight. (A) C vs. T; (B) CC + TT vs. CT. There was no association of the rs989692 polymorphisms and AD in both models.
Fig. 3. Forest plot for the association of rs3736187 polymorphisms and AD. Each square represents an OR for each specific study, and the area is proportional to the weight. (A) G vs A; (B) GG vs AA; (C) GA vs AA. The rs3736187 polymorphisms are associated with a decreased risk of AD in all genetic models.
Six studies for rs989692 and five studies for rs3736187 were included in this meta-analysis to conduct the estimation for the association of neprilysin polymorphisms and AD. The results showed that rs989692 polymorphisms were not associated with AD susceptibility. However, there was a significant association between neprilysin rs3736187 polymorphisms and AD, which suggested that carriage of G alleles was associated with a decreased risk of AD. The study from Fukami et al. found that the expression and activity of neprilysin in the hippocampus and neocortex decreased with aging in mice [8]. NEP-dysfunction could induce Aβ pathology in human brains. A possible explanation for the significant association between rs3736187 and AD was that G variant within the NEP gene may lead to subtle alterations of neprilysin structure or transcriptional activity, which resulted in an increased enzymatic activity and in the promotion of the degradation of Aβ. Our sensitive analysis showed that the heterogeneity in analysis of the association between rs989692 polymorphisms and AD may probably attribute to the study of Helisalmi and colleagues [9]. Some aspects of the study differed from others. For instance, the participants in this trial were all from Finland. Furthermore, thirty-six percent of AD patients had a positive family history. Thus, the results suggested that rs989692 may be a specific SNP in Finnish. Some potential limitations of this meta-analysis should be addressed. Firstly, only 7 studies are included in our meta-analysis. Secondly, there was considerable clinical heterogeneity among the included studies. The population structure of them was different, which would influence the distribution of gene frequency and morbidity of AD, resulting
in distinctive outcome as AD was prevalent in old women [25]. The experimental methods were different among the included studies. Finally, some unpublished trials may be missed and result in the bias in effect size. In summary, the current limited evidence suggests that rs3736187 polymorphisms are associated with a decreased risk of AD. However, given the heterogeneity among study designs and small sample size, the results of this meta-analysis should be interpreted with caution. Further larger scale studies are needed to investigate the association between rs989692 and rs3736187 polymorphisms and AD to confirm the present findings. Conflicts of interest The authors report no conflicts of interest. Acknowledgments This work was supported by a grant from the National Natural Science Foundation of China (No. 81072654). References [1] Mayeux R, Stern Y. Epidemiology of Alzheimer disease. Cold Spring Harbor Perspect Med 2012;2:a006239. [2] Karran E, Mercken M, De Strooper B. The amyloid cascade hypothesis for Alzheimer's disease: an appraisal for the development of therapeutics. Nat Rev Drug Discovery 2011;10:698–712.
10
X. Guo et al. / Journal of the Neurological Sciences 346 (2014) 6–10
[3] Herholz K, Ebmeier K. Clinical amyloid imaging in Alzheimer's disease. Lancet Neurol 2011;10:667–70. [4] Czeczor J, Connor T, Genders A, Walder K, McGee S. Amyloid β42 administration impairs energy metabolism in vivo and in vitro. J Neurol Sci 2013;333:e600-e. [5] Miners JS, Barua N, Kehoe PG, Gill S, Love S. Aβ-degrading enzymes: potential for treatment of Alzheimer disease. J Neuropathol Exp Neurol 2011;70:944–59. [6] Jayaraman A, Carroll JC, Morgan TE, Lin S, Zhao L, Arimoto JM, et al. 17β-Estradiol and progesterone regulate expression of β-amyloid clearance factors in primary neuron cultures and female rat brain. Endocrinology 2012;153:5467–79. [7] Miners JS, Van Helmond Z, Chalmers K, Wilcock G, Love S, Kehoe PG. Decreased expression and activity of neprilysin in Alzheimer disease are associated with cerebral amyloid angiopathy. J Neuropathol Exp Neurol 2006;65:1012–21. [8] Fukami S, Watanabe K, Iwata N, Haraoka J, Lu B, Gerard NP, et al. Abeta-degrading endopeptidase, neprilysin, in mouse brain: synaptic and axonal localization inversely correlating with Abeta pathology. Neurosci Res 2002;43:39–56. [9] Helisalmi S, Hiltunen M, Vepsalainen S, Iivonen S, Mannermaa A, Lehtovirta M, et al. Polymorphisms in neprilysin gene affect the risk of Alzheimer's disease in Finnish patients. J Neurol Neurosurg Psychiatry 2004;75:1746–8. [10] Blomqvist ME, McCarthy S, Blennow K, Andersson B, Prince JA. Evaluation of neprilysin sequence variation in relation to CSF (beta)-Amyloid levels and Alzheimer disease risk. Int J Mol Epidemiol Genet 2010;1:47–52. [11] Zhou XH, Xing SF, Luo XM. Association of the neprilysin and apolipoprotein E gene polymorphisms with Alzheimer's disease in Xinjiang Uygur population. Chin J Med Genet 2010;27:176–9. [12] Millard SP, Lutz F, Li G, Galasko DR, Farlow MR, Quinn JF, et al. Association of cerebrospinal fluid A(beta)42 with A2M gene in cognitively normal subjects. Neurobiol Aging 2014;35:357–64. [13] Miners S, Van Helmond Z, Barker R, Passmore PA, Johnston JA, Todd S, et al. Genetic variation in MME in relation to neprilysin protein and enzyme activity, Aβ levels, and Alzheimer's disease risk. Int J Mol Epidemiol Genet 2012;3:30.
[14] Wells G, Shea B, O'connell D, Peterson J, Welch V, Losos M, et al. The Newcastle– Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in metaanalyses; 2000. [15] Thakkinstian A, McElduff P, D'Este C, Duffy D, Attia J. A method for meta-analysis of molecular association studies. Stat Med 2005;24:1291–306. [16] Sutton AJ, Abrams KR, Jones DR, Sheldon TA, Song F. Methods for meta-analysis in medical research. J. Wiley; 2000. [17] DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986;7: 177–88. [18] Vepsäläinen S, Helisalmi S, Mannermaa A, Pirttilä T, Soininen H, Hiltunen M. Combined risk effects of IDE and NEP gene variants on Alzheimer disease. J Neurol Neurosurg Psychiatry 2009;80:1268–70. [19] Giedraitis V, Kilander L, Degerman-Gunnarsson M, Sundelöf J, Axelsson T, Syvänen A-C, et al. Genetic analysis of Alzheimer's disease in the Uppsala Longitudinal Study of Adult Men. Dement Geriatr Cogn Disord 2009;27:59–68. [20] Wood LS, Pickering EH, McHale D, Dechairo BM. Association between neprilysin polymorphisms and sporadic Alzheimer's disease. Neurosci Lett 2007;427:103–6. [21] Ittner LM, Götz J. Amyloid-β and tau—a toxic pas de deux in Alzheimer's disease. Nat Rev Neurosci 2011;12:67–72. [22] Nalivaeva NN, Beckett C, Belyaev ND, Turner AJ. Are amyloid-degrading enzymes viable therapeutic targets in Alzheimer's disease? J Neurochem 2012;120:167–85. [23] Liang K, Yang L, Yin C, Xiao Z, Zhang J, Liu Y, et al. Estrogen stimulates degradation of β-amyloid peptide by up-regulating neprilysin. J Biol Chem 2010;285:935–42. [24] Yamamoto N, Tanida M, Ono Y, Kasahara R, Fujii Y, Ohora K, et al. Leptin inhibits amyloid β-protein degradation through decrease of neprilysin expression in primary cultured astrocytes. Biochem Biophys Res Commun 2014;445:214–7. [25] Vest RS, Pike CJ. Gender, sex steroid hormones, and Alzheimer's disease. Horm Behav 2013;63:301–7.
Contents lists available at ScienceDirect
Journal of the Neurological Sciences journal homepage: www.elsevier.com/locate/jns
Review article
Meta-analysis of the association between two neprilysin gene polymorphisms and Alzheimer's disease Guo Xingzhi a,b, Tang Peng a, Liu Peng a, Liu Yue a, Hou Chen a, Li Rui a,⁎ a b
Department of Neurology, Shaanxi Provincial People's Hospital, Xi'an 710068, China The Third Affiliated Hospital, Xi'an Jiaotong University School of Medicine, Xi'an 710061, China
a r t i c l e
i n f o
Article history: Received 8 May 2014 Received in revised form 23 July 2014 Accepted 29 July 2014 Available online 7 August 2014 Keywords: Neprilysin Polymorphism β-Amyloid Alzheimer's disease Meta-analysis Risk
a b s t r a c t Objective: The aim of this study is to evaluate the association between two neprilysin variants (rs989692 and rs3736187) and Alzheimer's disease (AD). Methods: All eligible studies were searched in PubMed and Embase from inception to July 2014. Data was extracted by two investigators independently. The complete overdominant model (CC + TT vs. CT) and codominant model (GG vs. AA and GA vs. AA) were used for rs989692 and rs3736187, respectively. A comparison of allele frequencies was also conducted. Results: Six studies containing 2555 AD patients and 1914 controls were included for rs989692 polymorphisms. The pooled odds ratio (OR) and confidence interval (CI) suggested that rs989692 polymorphisms were not associated with AD based on the current published studies (C vs. T, OR = 1.01, 95% CI = 0.85–1.19; CC + TT vs. CT, OR = 0.89, 95% CI = 0.78–1.01). Five studies containing 2438 AD patients and 1452 controls were identified for rs3736187 polymorphisms (G vs. A, OR = 0.77, 95% CI = 0.66–0.91; GG vs. AA, OR = 0.38, 95% CI = 0.19–0.77; GA vs. AA, OR = 0.81, 95% CI = 0.61–0.99). The result showed that rs3736187 polymorphisms were likely associated with the decreased risk of AD. Conclusions: This meta-analysis indicates that rs3736187 (A/G) polymorphisms may be a potential beneficial single nucleotide polymorphism (SNP), which are associated with a decreased risk in AD. Further larger scale studies are necessary to validate gene-to-gene interactions and to define the association of neprilysin polymorphisms with AD. © 2014 Elsevier B.V. All rights reserved.
Contents 1. 2.
Introduction . . . . . . . . . . . . . . . . Methods and materials . . . . . . . . . . . 2.1. Literature search and inclusion criteria . 2.2. Data extraction and outcome measures 2.3. Assessment of risk of bias . . . . . . . 2.4. Statistical analysis . . . . . . . . . . 3. Results . . . . . . . . . . . . . . . . . . 3.1. Selection of eligible studies . . . . . . 3.2. Study characteristics . . . . . . . . . 3.3. Quantitative data synthesis . . . . . . 3.4. Assessment of bias . . . . . . . . . . 4. Discussion . . . . . . . . . . . . . . . . . Conflicts of interest . . . . . . . . . . . . . . . Acknowledgments . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
6 7 7 7 7 7 7 7 7 7 8 8 9 9 9
1. Introduction ⁎ Corresponding author at: Department of Neurology, Shaanxi Provincial People's Hospital, No. 256, Youyi West Road, Xi'an 710068, China. Tel.: +86 13096923029. E-mail address: [email protected] (R. Li).
http://dx.doi.org/10.1016/j.jns.2014.07.064 0022-510X/© 2014 Elsevier B.V. All rights reserved.
Alzheimer's disease (AD) is the most common neurodegenerative disease in the elderly, affecting about 24 million people in the world
X. Guo et al. / Journal of the Neurological Sciences 346 (2014) 6–10
in 2012 [1]. The main characteristics of AD are senile plaque containing β-amyloid (Aβ), neurofibrillary tangles and neuronal loss. Numerous studies suggest that the accumulation of Aβ in the brain was the major cause of AD, which could induce inflammatory response and neuronal death [2–4]. Promoting the degradation of Aβ was supposed to be a potential strategy for the treatment of AD. Multiple Aβ-cleaving enzymes, including neprilysin (NEP), insulin-degrading enzyme (IDE), angiotensin-converting enzyme (ACE) and matrix metalloproteinase (MMP), play a vital role in the regulation Aβ level in the brain [5,6]. Neprilysin (NEP), also known as membrane metallo-endopeptidase (MME), is a type II metalloproteinase widely expressed in the neocortex and cerebral vessels. Studies showed that NEP could remove Aβ and inhibit the formation of Aβ deposition [7]. Deficit of NEP could result in the aggregation of Aβ [8]. The gene encoding human NEP is located on chromosome 3q25.1-q25.2 (MIM: 120520) and is highly polymorphic. Among the different polymorphisms in NEP, two SNPs, rs989692 and rs3736187, are extensively studied. Large number of studies indicated that both SNPs are considered to be associated with AD [9–13]; however, the results remain controversial, with some studies unable to identify any association between neprilysin polymorphisms and AD. Therefore, we conducted a meta-analysis to assess the inconsistent results from published studies and clarify the association between rs989692 or rs3736187 polymorphisms and AD. 2. Methods and materials 2.1. Literature search and inclusion criteria We performed a comprehensive search in the PubMed and Embase databases up to July 2014 and selected the original studies surveying the neprilysin polymorphisms and AD. We also searched the AlzGene database for additional articles. The following search terms were used: neprilysin, NEP, metallo-endopeptidase, MME, polymorphism*, variant*, genotype, allele* and Alzheimer's disease*, Alzheimer disease*, and AD. The language was limited to English and Chinese. The search was restricted to human subjects. Studies included in this meta-analysis must meet all the following inclusive selection criteria: a) evaluation of the association between rs989692 or rs3736187 and AD; b) conformance to Hardy–Weinberg equilibrium (HWE); c) case–control design and if there were studies having overlapped subjects, only the one with a larger sample size was selected; and d) useful data available for estimating the odds ratio (OR) with 95% confidence intervals (CIs). 2.2. Data extraction and outcome measures For each study, information was recorded including first author, year of publication, country of origin, ethnicity of population, study design, definition of Alzheimer's disease, number of patients and controls, genotyping methods and genotype distributions. Data were extracted independently by two investigators (Guo Xingzhi and Tang Peng) using a standardized data extraction form. Agreement was reached after discussion for conflicting data with a third author (Li Rui), and then extracted data were typed into an excel file by another author (Hou Chen). 2.3. Assessment of risk of bias The Newcastle–Ottawa Scale (NOS) was conducted to assess the risk of bias [14]. This scale evaluates quality of observational studies from three aspects, including selection (0–4 points), exposure (0–3 points) and comparability (0–2 points). A score of 1 is deducted when one point is unmatched and high scores indicate better quality. Two authors (Liu Peng and Liu Yue) independently estimated the risk of bias.
7
2.4. Statistical analysis The association between the NEP gene polymorphism at rs989692 or rs3736187 and AD was estimated by calculating pooled ORs with 95% CIs. According to the recently suggested guidelines for analysis of genetic association data [15], we adopted the most probably appropriate genetic model for both SNPs. The allele model and complete overdominant model were performed for rs989692, whereas allele model and codominant model were performed for rs3736187. The Cochran Q test and I-squared test were conducted to assess the statistical heterogeneity (P b 0.1 was considered significant). A fixed-effects model (Mantel–Haenszel method) [16] or random-effects model (DerSimonian and Laird method) [17] was applied according to the presence (P b 0.1) or absence (P N 0.1) heterogeneity. Sensitive analysis was used to assess the stability of the results by sequentially excluding one publication at a time. Begg's and Egger's tests were applied to assess the publication bias. All statistical analyses and graphics were conducted in Stata 12.0 and Revman 5.2 software. P b 0.05 was considered statistically significant. 3. Results 3.1. Selection of eligible studies A total of 94 potential studies were identified by the initial search in PubMed and Embase databases, and 25 duplicate studies were excluded After reviewing titles and abstracts, 53 studies were excluded for they were reviews or not relevant to our analysis. Sixteen potentially relevant articles were reviewed for full-text analysis (Fig. 1). Ten studies were excluded: 9 articles for other diseases or SNPs and one with overlapped data [18]. One additional study with available data was identified in the AlzGene database and included in the analysis [19]. Thus, a total of 7 studies were included in final meta-analysis and four of them evaluated both rs989692 and rs3736187 polymorphisms [9–13, 19,20]. 3.2. Study characteristics The detailed characteristics of each included studies were presented in Table 1. These studies were published from 2004 to 2014. All articles were case–control studies. Six studies defined AD according to the NINCDS–ADRDA criteria and the DSM-IV criteria, and only Miners et al. used the CERAD criteria for AD diagnosis [13]. Among the included studies, four of them evaluated both rs989692 and rs3736187 polymorphisms [9,10,13,20]. For the rs989692 polymorphism, all the six studies were performed in Caucasians and 2555 AD patients and 1914 normal controls were included [9,10,12,13,19,20]. The study of Helisalmi et al. screened the mutations in APP, PSEN-1 or PSEN-2 genes, but none of the mutations were found [9]. For the rs3736187 polymorphism, finally five studies were included in our meta-analysis with 2438 AD patients and 1452 normal controls [9–11,13,20]. Four studies were conducted in Caucasians, and only one was performed in Asia [11]. 3.3. Quantitative data synthesis The result for the rs989692 polymorphism was presented in Fig. 2. Compared to the T allele, the C allele had no significant association with AD susceptibility for both models (C vs. T, OR = 1.01, 95% CI = 0.85–1.19; CC + TT vs. CT, OR = 0.89, 95% CI = 0.78–1.01). A fixedeffects model was conducted for the complete overdominant model (CC + TT vs. CT) without significant heterogeneity, whereas a random-effects model was used for the allele model which had a significant heterogeneity. Thus, the sensitivity analyses were applied to explore the potential sources of heterogeneity. Exclusion of the
8
X. Guo et al. / Journal of the Neurological Sciences 346 (2014) 6–10
Fig. 1. Selection for eligible studies included in this meta-analysis.
study of Helisalmi et al. resolved the heterogeneity but not changed the results. Fig. 3 shows the result of the association between the rs3736187 polymorphism and AD. There was no evidence of heterogeneity, so a fixed-effects model was performed. The results were as follows: G vs. A, OR = 0.77, 95% CI = 0.66–0.91; GG vs AA, OR = 0.38, 95% CI = 0.19–0.77; and GA vs AA, OR = 0.81, 95% CI = 0.67–0.99. After combining the ORs and CIs, we found that the rs3736187 polymorphism was significantly associated with AD in all the genetic models. 3.4. Assessment of bias Both Egger's and Begg's tests were used to assess the publication bias of the literature. The results showed that there was no obvious publication bias in overall analysis. For rs989692 polymorphisms, the P values of Egger's test for the allele model and the complete overdominant model were 0.970 and 0.976, respectively. For rs3736187 polymorphisms, the P values of Egger's test for the allele model, the
homozygote codominant model, and the heterozygote codominant model were 0.932, 0.823 and 0.843, respectively. After evaluating the quality of the included studies based on NOS score, the median NOS score (6.5, range from 5 to 7) was obtained. 4. Discussion AD is the leading cause of neurodegenerative disease in the elderly. Deposition of Aβ is the main neuropathological hallmark in AD, which can lead to the neuronal death [21]. A host of enzymes are involved in the degradation of Aβ in the brain and neprilysin is one of the Aβdegrading enzymes [22]. Increasing the expression of NEP could promote the degradation of β-amyloid [23,24]. Studies had been conducted to evaluate the association of neprilysin polymorphisms and AD, but the results were inconsistent. Two SNPs in neprilysin, rs989692 and rs3736187, were widely studied, so we selected these two SNPs to explore the association between neprilysin polymorphisms and AD risks.
Table 1 Characteristics of studies included in this meta-analysis. Author
Country
(A) Studies for rs989692 polymorphism Blomqvist Sweden/USA Giedraitis Sweden Helisalmi Finland Millard USA Miners UK/Italy/Sweden Wood USA
Ethnicity
Year
Sample
Sample
Average age
Cases
Controls
Cases
Control
Genotyping
Caucasian Caucasian Caucasian Caucasian Caucasian Caucasian
2010 2009 2004 2014 2012 2007
Blood Blood Blood Blood Brain tissue and blood Blood?
665 85 390 99 1018 298
184 400 468 168 398 296
76.2 80.2 72 71 NA 72.6
73.2 81.8 70 67 NA 74.7
Hybridization Gene chip and minisequencing PCR PCR PCR OLA/PCR
(B) Studies for rs3736187 polymorphism Blomqvist Sweden/USA Caucasian Helisalmi Finland Caucasian Miners UK/Italy/Sweden Caucasian Wood USA Caucasian Zhou et China Asian
2010 2004 2012 2007 2010
Blood Blood Brain tissue and Blood Blood? Blood
650 390 984 297 111
195 468 375 297 117
76.2 72 NA 72.6 72
73.2 70 NA 74.7 71
Hybridization PCR PCR PCR PCR-RFLP
X. Guo et al. / Journal of the Neurological Sciences 346 (2014) 6–10
9
Fig. 2. Forest plot for the association of rs989692 polymorphisms and AD. Each square represents an OR for each specific study, and the area is proportional to the weight. (A) C vs. T; (B) CC + TT vs. CT. There was no association of the rs989692 polymorphisms and AD in both models.
Fig. 3. Forest plot for the association of rs3736187 polymorphisms and AD. Each square represents an OR for each specific study, and the area is proportional to the weight. (A) G vs A; (B) GG vs AA; (C) GA vs AA. The rs3736187 polymorphisms are associated with a decreased risk of AD in all genetic models.
Six studies for rs989692 and five studies for rs3736187 were included in this meta-analysis to conduct the estimation for the association of neprilysin polymorphisms and AD. The results showed that rs989692 polymorphisms were not associated with AD susceptibility. However, there was a significant association between neprilysin rs3736187 polymorphisms and AD, which suggested that carriage of G alleles was associated with a decreased risk of AD. The study from Fukami et al. found that the expression and activity of neprilysin in the hippocampus and neocortex decreased with aging in mice [8]. NEP-dysfunction could induce Aβ pathology in human brains. A possible explanation for the significant association between rs3736187 and AD was that G variant within the NEP gene may lead to subtle alterations of neprilysin structure or transcriptional activity, which resulted in an increased enzymatic activity and in the promotion of the degradation of Aβ. Our sensitive analysis showed that the heterogeneity in analysis of the association between rs989692 polymorphisms and AD may probably attribute to the study of Helisalmi and colleagues [9]. Some aspects of the study differed from others. For instance, the participants in this trial were all from Finland. Furthermore, thirty-six percent of AD patients had a positive family history. Thus, the results suggested that rs989692 may be a specific SNP in Finnish. Some potential limitations of this meta-analysis should be addressed. Firstly, only 7 studies are included in our meta-analysis. Secondly, there was considerable clinical heterogeneity among the included studies. The population structure of them was different, which would influence the distribution of gene frequency and morbidity of AD, resulting
in distinctive outcome as AD was prevalent in old women [25]. The experimental methods were different among the included studies. Finally, some unpublished trials may be missed and result in the bias in effect size. In summary, the current limited evidence suggests that rs3736187 polymorphisms are associated with a decreased risk of AD. However, given the heterogeneity among study designs and small sample size, the results of this meta-analysis should be interpreted with caution. Further larger scale studies are needed to investigate the association between rs989692 and rs3736187 polymorphisms and AD to confirm the present findings. Conflicts of interest The authors report no conflicts of interest. Acknowledgments This work was supported by a grant from the National Natural Science Foundation of China (No. 81072654). References [1] Mayeux R, Stern Y. Epidemiology of Alzheimer disease. Cold Spring Harbor Perspect Med 2012;2:a006239. [2] Karran E, Mercken M, De Strooper B. The amyloid cascade hypothesis for Alzheimer's disease: an appraisal for the development of therapeutics. Nat Rev Drug Discovery 2011;10:698–712.
10
X. Guo et al. / Journal of the Neurological Sciences 346 (2014) 6–10
[3] Herholz K, Ebmeier K. Clinical amyloid imaging in Alzheimer's disease. Lancet Neurol 2011;10:667–70. [4] Czeczor J, Connor T, Genders A, Walder K, McGee S. Amyloid β42 administration impairs energy metabolism in vivo and in vitro. J Neurol Sci 2013;333:e600-e. [5] Miners JS, Barua N, Kehoe PG, Gill S, Love S. Aβ-degrading enzymes: potential for treatment of Alzheimer disease. J Neuropathol Exp Neurol 2011;70:944–59. [6] Jayaraman A, Carroll JC, Morgan TE, Lin S, Zhao L, Arimoto JM, et al. 17β-Estradiol and progesterone regulate expression of β-amyloid clearance factors in primary neuron cultures and female rat brain. Endocrinology 2012;153:5467–79. [7] Miners JS, Van Helmond Z, Chalmers K, Wilcock G, Love S, Kehoe PG. Decreased expression and activity of neprilysin in Alzheimer disease are associated with cerebral amyloid angiopathy. J Neuropathol Exp Neurol 2006;65:1012–21. [8] Fukami S, Watanabe K, Iwata N, Haraoka J, Lu B, Gerard NP, et al. Abeta-degrading endopeptidase, neprilysin, in mouse brain: synaptic and axonal localization inversely correlating with Abeta pathology. Neurosci Res 2002;43:39–56. [9] Helisalmi S, Hiltunen M, Vepsalainen S, Iivonen S, Mannermaa A, Lehtovirta M, et al. Polymorphisms in neprilysin gene affect the risk of Alzheimer's disease in Finnish patients. J Neurol Neurosurg Psychiatry 2004;75:1746–8. [10] Blomqvist ME, McCarthy S, Blennow K, Andersson B, Prince JA. Evaluation of neprilysin sequence variation in relation to CSF (beta)-Amyloid levels and Alzheimer disease risk. Int J Mol Epidemiol Genet 2010;1:47–52. [11] Zhou XH, Xing SF, Luo XM. Association of the neprilysin and apolipoprotein E gene polymorphisms with Alzheimer's disease in Xinjiang Uygur population. Chin J Med Genet 2010;27:176–9. [12] Millard SP, Lutz F, Li G, Galasko DR, Farlow MR, Quinn JF, et al. Association of cerebrospinal fluid A(beta)42 with A2M gene in cognitively normal subjects. Neurobiol Aging 2014;35:357–64. [13] Miners S, Van Helmond Z, Barker R, Passmore PA, Johnston JA, Todd S, et al. Genetic variation in MME in relation to neprilysin protein and enzyme activity, Aβ levels, and Alzheimer's disease risk. Int J Mol Epidemiol Genet 2012;3:30.
[14] Wells G, Shea B, O'connell D, Peterson J, Welch V, Losos M, et al. The Newcastle– Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in metaanalyses; 2000. [15] Thakkinstian A, McElduff P, D'Este C, Duffy D, Attia J. A method for meta-analysis of molecular association studies. Stat Med 2005;24:1291–306. [16] Sutton AJ, Abrams KR, Jones DR, Sheldon TA, Song F. Methods for meta-analysis in medical research. J. Wiley; 2000. [17] DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986;7: 177–88. [18] Vepsäläinen S, Helisalmi S, Mannermaa A, Pirttilä T, Soininen H, Hiltunen M. Combined risk effects of IDE and NEP gene variants on Alzheimer disease. J Neurol Neurosurg Psychiatry 2009;80:1268–70. [19] Giedraitis V, Kilander L, Degerman-Gunnarsson M, Sundelöf J, Axelsson T, Syvänen A-C, et al. Genetic analysis of Alzheimer's disease in the Uppsala Longitudinal Study of Adult Men. Dement Geriatr Cogn Disord 2009;27:59–68. [20] Wood LS, Pickering EH, McHale D, Dechairo BM. Association between neprilysin polymorphisms and sporadic Alzheimer's disease. Neurosci Lett 2007;427:103–6. [21] Ittner LM, Götz J. Amyloid-β and tau—a toxic pas de deux in Alzheimer's disease. Nat Rev Neurosci 2011;12:67–72. [22] Nalivaeva NN, Beckett C, Belyaev ND, Turner AJ. Are amyloid-degrading enzymes viable therapeutic targets in Alzheimer's disease? J Neurochem 2012;120:167–85. [23] Liang K, Yang L, Yin C, Xiao Z, Zhang J, Liu Y, et al. Estrogen stimulates degradation of β-amyloid peptide by up-regulating neprilysin. J Biol Chem 2010;285:935–42. [24] Yamamoto N, Tanida M, Ono Y, Kasahara R, Fujii Y, Ohora K, et al. Leptin inhibits amyloid β-protein degradation through decrease of neprilysin expression in primary cultured astrocytes. Biochem Biophys Res Commun 2014;445:214–7. [25] Vest RS, Pike CJ. Gender, sex steroid hormones, and Alzheimer's disease. Horm Behav 2013;63:301–7.