Abstract
Objective:
The association between urokinase-plasminogen activator (PLAU) gene rs2227564 polymorphism and Alzheimer’s disease (AD) risk has been widely reported across different ethnic populations, with inconsistent results. Thus, we performed a meta-analysis to assess the association between PLAU rs2227564 polymorphism and AD risk.
Methods:
Fixed or random effect model was used as the pooling method to assess the basis of homogeneity test among studies. Summarized estimation of odds ratio (OR) and 95% confidence interval (CI) were calculated. Heterogeneity among studies was evaluated using Q test and I 2. Publication bias was estimated using Harbord’s test.
Results:
A total of 27 studies (comprising 6100 AD cases and 5718 controls) were included in this meta-analysis. The present meta-analysis showed a significant increased effect of T allele on risk of AD in dominant model (fixed effect model [FEM] OR 1.123, 95% CI 1.025-1.231) and heterozygote comparison (CT vs CC; FEM OR 1.126, 95% CI 1.027-1.235). No publication bias was detected.
Conclusion:
This meta-analysis showed that T allele of rs2227564 polymorphism in PLAU gene could increase the effects on risk of AD, and this result needs to be confirmed by further studies.
Keywords: urokinase-plasminogen activator gene, polymorphism, Alzheimer’s disease, meta-analysis
Background
Alzheimer’s disease (AD) is the most common form of dementia in aging human. It is predicted that worldwide population of AD would grow to over 100 million by 2050. 1 Genetic variation has been postulated to influence the variable risk of AD observed both within and across populations. Linkage studies indicate that chromosome 10 contains several genetic risk loci for late-onset AD. 2,3 Urokinase-plasminogen activator (PLAU) gene that maps to chromosome 10q22.2 is located within this linkage peak and has been shown to block β-amyloid (Aβ) protein neurotoxicity and to enhance α-secretase cleavage of the amyloid precursor protein and Aβ degradation. 4,5 The PLAU protein is a serine protease whose action is to convert plasminogen to plasmin. Plasmin is involved in the clearance of secreted Aβ, reducing aggregated forms of Aβ and preventing Aβ neurotoxicity. Therefore, PLAU is a reasonable positional candidate for association with AD.
The rs2227564 (also called P141L) polymorphism in PLAU gene is a variant in exon 6 which causes a proline to leucine change at position 141 in the coded protein within the kringle domain of PLAU at the junction between 2 β-pleated sheets. 6,7 Additionally, this mutation (P141L) in PLAU gene was more common and was a better candidate than the other coding single-nucleotide polymorphisms (SNPs). 8 Although several articles 8 -21 have investigated the genetic associations between rs2227564 polymorphism in PLAU gene and AD risk, the results were inconsistent. Single studies may be underpowered to estimate the effects of loci conferring small changes in disease risk. So large-scale studies are required to reliably confirm or refute gene–disease associations. 22,23 Therefore, we conducted a meta-analysis including 27 case–control studies comprising 6100 AD cases and 5718 controls to derive a more precise estimation of the relationship between the PLAU SNP rs2227564 and the AD risk.
Methods
Search Strategy
A comprehensive search was conducted for available articles published in English or Chinese up to September 2012 for studies on the association between PLAU gene and AD from the following databases: (1) PubMed, (2) Embase, (3) ISI (Web of Science), (4) China National Knowledge Infrastructure (CNKI), (5) China Biology Medical literature database (CBM), and (6) Wan Fang Med Online. The search strategy used the following keywords: “AD,” “plasminogen activator urina” or “PLAU” or “rs2227564” or “P141L” or “PLAU_1,” and “polymorphism” or “mutation” or “variant”. We also reviewed the bibliographies of relevant articles and searched the studies not captured by our database as well as those of relevant studies.
Inclusion Criteria
In the current meta-analysis, all relevant studies reporting the association of PLAU gene rs2227564 polymorphism and AD risk were considered for inclusion. The inclusion criteria were as follows: (1) evaluation of the rs2227564 polymorphism in PLAU gene and AD risk; (2) using a case–control or cohort design; (3) English or Chinese language articles were included; (4) numbers for the genotype was reported in the article or could be obtain from authors or others; (5) the most recent or largest population was selected if the studies were published with the same or overlapping data by the same authors; (6) AD in each articles were diagnosed explicitly; and (7) each subpopulation was considered as a separate study in this meta-analysis if an article reported results with different ethnicity subpopulations. Accordingly, the following exclusion criteria were also used: (1) abstracts or reviews; (2) genotype frequency not reported; and (3) repeated or overlapped publications.
Data Extraction
Two investigators independently extracted the information needed from all the studies based on the inclusion criteria. The retrieved data were as follows: name of the first author, publication date, journal, country, mean age of the patients with AD and control, selection criteria for cases, total number of cases and controls, and genotype distributions in cases and controls. If a consensus could not be established, a third investigator was invited to the discussion.
Statistical Analysis
The chi-square (χ2) analysis was used to test for deviation from Hardy-Weinberg equilibrium (HWE) for the rs2227564 genotype distribution of PLAU gene in case groups and control groups, and P < .05 was considered as departure from HWE. Pooled measure was calculated as the inverse variance-weighted mean of the logarithm of odds ratio (OR) with 95% confidence interval (CI) to estimate the strength of association between rs2227564 polymorphism in PLAU gene and risk of AD for codominant model (T vs C), dominant model (TT + CT vs CC) and recessive model (TT vs CC + CT), heterozygote comparison (CT vs CC), and homozygote comparison (TT vs CC), respectively. Heterogeneity among studies was estimated using the I 2 of Higgins and Thompson. 24 The I 2 reflects the proportion of total variation attributable to between-study heterogeneity as opposed to random error or chance. We would use the DerSimonian and Laird random effect model (REM; if I 2 > 50%) 25 or fixed effect model (FEM; if I 2 < 50%) as the pooling method. Metaregression with restricted maximum likelihood estimation was performed to describe the potentially important covariates, including sample size (the sum of case and control numbers), age (ratio of age or median age in the case group to that in the control group), and publication year which might exert substantial impacts on between-study heterogeneity. If no significant covariates were found to be heterogeneous, the “leave-one-out” sensitive analysis 26 was carried out to evaluate the key studies with substantial impact on between-study heterogeneity. A study of influence analysis was conducted 27 to describe how robust the pooled estimator is to removal of individual studies. An individual study that has excessive influence should be removed if the point estimate of its omitted analysis lies outside the 95% CI of the combined analysis. Publication bias was assessed using Harbord’s test. 28 All the statistical analyses were performed with STATA version 10.0 (Stata Corporation, College Station, Texas). Two-tailed P ≤ .05 was accepted as statistically significant.
Results
Characteristics of Studies
According to the comprehensive search and selection based on the inclusion criteria, 14 articles (including 27 independent studies) with 6100 AD cases and 5718 controls were identified with data on association between rs2227564 polymorphisms and susceptibility to AD. All the eligible studies included in this meta-analysis were case–control designs. The characteristics of rs2227564 polymorphism genotype distributions in present meta-analysis are shown in Table 1.
Table 1.
The Characteristics of PLAU Gene rs2227564 Polymorphism Genotype Distributions in this Meta-Analysis.
| Author | Year | Country | Diagnostic criteria | Mean age (case/control) | Genotypes CC/CT/TT | T allele frequency, % | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Case | P h valuea | Control | P h valuea | Case | Control | P value | ORb | |||||
| Finckh et al 9 | 2003 | Germany | NINCDS-ADRDA | NA/NA | 134/64/12 | .29 | 107/85/14 | .73 | 21.0 | 27.4 | .030 | 0.701 |
| Finckh et al 9 | 2003 | Switzerland | NINCDS-ADRDA | NA/NA | 31/12/0 | .58 | 27/22/6 | .75 | 14.0 | 30.9 | .007 | 0.362 |
| Finckh et al 9 | 2003 | Italy | NINCDS-ADRDA | NA/NA | 69/23/2 | 1 | 18/11/1 | 1.00 | 14.4 | 21.7 | .183 | 0.606 |
| Myers et al 8 | 2003 | America | ADRC | 75.5/78.4 | 116/54/16 | .02 | 115/65/11 | .68 | 23.1 | 22.8 | .911 | 1.020 |
| Myers et al 8 | 2003 | America | ADRC | 82.7/75.4 | 180/131/17 | .31 | 282/170/30 | .53 | 25.2 | 23.9 | .552 | 1.072 |
| Myers et al 8 | 2003 | UK | ADRC | 77/75.6 | 80/52/6 | .63 | 79/67/5 | .052 | 23.2 | 25.5 | .519 | 0.882 |
| Papassotiropoulos et al 10 | 2004 | Switzerland | NINCDS-ADRDA | NA/NA | 71/49/4 | .31 | 158/100/19 | .52 | 23.0 | 24.9 | .557 | 0.900 |
| Papassotiropoulos et al 10 | 2004 | Greece | NINCDS-ADRDA | NA/NA | 120/57/4 | .27 | 65/32/2 | .52 | 18.0 | 18.2 | .947 | 0.985 |
| Bagnoli et al 11 | 2004 | Italy | DSM-IV | 72.3/84.2 | 171/62/5 | 1 | 151/50/6 | .42 | 15.1 | 15.0 | .950 | 1.012 |
| Ertekin-Taner et al 12 | 2005 | America | NINCDS-ADRDA | 75.1/79.2 | 67/34/12 | .03 | 96/59/5 | .35 | 25.7 | 21.6 | .264 | 1.256 |
| Ertekin-Taner et al 12 | 2005 | America | NINCDS-ADRDA | 80.4/80.3 | 114/99/14 | .25 | 142/71/14 | .24 | 28.0 | 21.8 | .032 | 1.393 |
| Ertekin-Taner et al 12 | 2005 | America | NINCDS-ADRDA | 78.3/77.8 | 80/74/10 | .25 | 87/71/6 | .10 | 28.7 | 25.3 | .333 | 1.186 |
| Ertekin-Taner et al 12 | 2005 | Sweden | NINCDS-ADRDA | 75.3/68.8 | 54/46/11 | .21 | 41/28/8 | .40 | 30.6 | 28.6 | .668 | 1.104 |
| Ertekin-Taner et al 12 | 2005 | Sweden | NINCDS-ADRDA | 80.2/76.6 | 49/37/7 | 1 | 46/39/10 | .64 | 27.4 | 31.1 | .439 | 0.839 |
| Ertekin-Taner et al 12 | 2005 | UK | NINCDS-ADRDA | 57.8/58.1 | 61/48/12 | .66 | 80/59/11 | 1.00 | 29.8 | 27.0 | .479 | 1.145 |
| Riemenschneider et al 13 | 2006 | Germany | NINCDS-ADRDA | 69.1/68.8 | 215/177/30 | .47 | 151/98/8 | .11 | 28.1 | 22.2 | .016 | 1.370 |
| Riemenschneider et al 13 | 2006 | Germany | NINCDS-ADRDA | 74.8/75.1 | 64/37/8 | .44 | 126/42/5 | .55 | 24.3 | 15.0 | .006 | 1.816 |
| Riemenschneider et al 13 | 2006 | Australia | NINCDS-ADRDA | 74.3/77.1 | 116/85/18 | .74 | 223/98/17 | .17 | 27.6 | 19.5 | .002 | 1.573 |
| Riemenschneider et al 13 | 2006 | Italy | NINCDS-ADRDA | 67.3/64.6 | 77/36/7 | .4 | 85/13/1 | .44 | 20.8 | 7.6 | .000 | 3.211 |
| Ozturk et al 14 | 2006 | America | NINCDS-ADRDA | 76.57/75.2 | 38/319/563 | .43 | 31/211/402 | .64 | 78.5 | 78.8 | .855 | 0.984 |
| Blomqvist et al 15 | 2006 | Sweden | NINCDS-ADRDA | 76.2/73.2 | 422/330/56 | .48 | 113/95/21 | .88 | 27.4 | 29.9 | .281 | 0.882 |
| Pesaresi et al 16 | 2007 | Italy | NINCDS-ADRDA | 74.5/67.9 | 129/57/6 | 1 | 93/27/6 | .08 | 18.0 | 15.5 | .413 | 1.196 |
| Ji et al 17 | 2007 | China | NINCDS-ADRDA | 76.1/62.1 | 61/86/10 | .01 | 56/53/19 | .33 | 33.8 | 35.5 | .655 | 0.924 |
| Xiang et al 18 | 2008 | China | NINCDS-ADRDA | 77.67/75.65 | 25/33/9 | .8 | 25/34/12 | 1.00 | 38.1 | 40.8 | .636 | 0.890 |
| Giedraitis et al 19 | 2008 | Sweden | NINCDS-ADRDA and DSM-IV | 80.2/81.8 | 44/36/5 | .59 | 227/139/34 | .07 | 27.1 | 25.9 | .750 | 1.063 |
| Cousin et al 20 | 2009 | France | NINCDS-ADRDA | 64.9/66.2 | 274/132/15 | 1 | 289/148/25 | .33 | 19.2 | 21.4 | .254 | 0.874 |
| Zhou et al 21 | 2010 | China | DSM-I | 67.2/65.5 | 121/84/4 | .01 | 115/98/7 | .01 | 22.0 | 25.5 | .236 | 0.826 |
Abbreviations: ADRC, Alzheimer Disease Research Center; NINCDS-ADRDA, National Institute of Neurological and Communicative Disorders and Stroke-Alzheimer’s Disease and Related Disorders Association criteria; DSM-IV, Diagnostic and Statistical Manual of Mental Disorders (Fourth Edition); DSM-I, Diagnostic and Statistical Manual of Mental Disorders (First Edition); HWE, Hardy-Weinberg equilibrium; NA, not available; PLAU, urokinase-plasminogen activator; OR, odds ratio.
a Exact P value for HWE test.
b OR ratio calculated by T versus C allele.
Quantitative Synthesis
Results of pooled analysis are summarized in Table 2. The meta-analysis showed a significant increased effect of T allele on risk of AD in heterozygote comparison (CT vs CC; FEM OR 1.096, 95% CI 1.006-1.193), but this was only marginally significant in the codominant model (REM OR 1.048, 95% CI 0.944-1.162) and dominant model (REM OR 1.085, 95% CI 0.956-1.232). However, no significant associations were found in the recessive model (FEM OR 0.963, 95% CI 0.842-1.103) and homozygote comparison (TT vs CC; FEM OR 1.031, 95% CI 0.869-1.224). After excluding 4 studies 8,12,17,21 that deviated from HWE in cases and/or in controls, significant associations were also found between the T allele and increased AD risk considering heterozygote comparison (CT vs CC; FEM OR 1.126, 95% CI 1.027-1.235), only marginally significant in the codominant model (REM OR 1.059, 95% CI 0.941-1.193) and dominant model (REM OR 1.108, 95% CI 0.958-1.281) but not in the recessive model (FEM OR 0.960, 95% CI 0.833-1.105) and homozygote comparison (TT vs CC; FEM OR 1.029, 95% CI 0.856-1.236).
Table 2.
Pooled Measures on the Relation of PLAU Gene rs2227564 Polymorphism With AD.
| Data | Inherited model | Before HETREDa analysis | After HETRED analysis | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Sample size cases/controls | Pooled OR (95% CI) | Q value | I 2, % | Sample size cases/controls | Pooled OR (95% CI) | Q value | I 2, % | ||||
| FEM | REM | ||||||||||
| All included articles | T vs C | 6100/5718 | 1.041 (0.978-1.109) | 1.048 (0.944-1.162) | 63.683 | 59.2 | 5761/5281 | 1.006 (0.943-1.074) | 41.419 | 42.1 | |
| TT + CT vs CC | 6100/5718 | 1.086 (1.000-1.178) | 1.085 (0.956-1.232) | 57.758 | 55.0 | 5980/5619 | 1.067 (0.983-1.159) | 46.738 | 46.5 | ||
| TT vs CC + CT | 6100/5718 | 0.963 (0.842-1.103) | 0.983 (0.806-1.200) | 37.698 | 31.0 | – | – | – | – | ||
| TT vs CC | 3846/3733 | 1.031 (0.869-1.224) | 1.040 (0.830-1.303) | 40.117 | 35.2 | – | – | – | – | ||
| CT vs CC | 5237/5013 | 1.096 (1.006-1.193) | 1.096 (0.968-1.241) | 50.599 | 48.6 | – | – | – | – | ||
| Excluded for DHWE | T vs C | 5453/5043 | 1.052 (0.984-1.126) | 1.059 (0.941-1.193) | 60.214 | 63.5 | 5114/4606 | 1.012 (0.944-1.085) | 38.482 | 48.0 | |
| TT + CT vs CC | 5453/5043 | 1.110 (1.016-1.212) | 1.108 (0.958-1.281) | 54.027 | 59.3 | 5123/4738 | 1.123 (1.025-1.231) | 34.726 | 42.4 | ||
| TT vs CC + CT | 5453/5043 | 0.960 (0.833-1.105) | 0.968 (0.811-1.155) | 24.934 | 11.8 | – | – | – | – | ||
| TT vs CC | 3439/3309 | 1.029 (0.856-1.236) | 1.036 (0.823-1.304) | 30.680 | 28.3 | – | – | – | – | ||
| CT vs CC | 4614/4356 | 1.126 (1.027-1.235) | 1.127 (0.983-1.291) | 43.845 | 49.8 | – | – | – | – | ||
| Europeb | T vs C | 3528/3117 | 1.002 (0.918-1.093) | 1.015 (0.867-1.187) | 46.415 | 65.5 | 3299/2845 | 0.951 (0.869-1.041) | 23.780 | 41.1 | |
| TT + CT vs CC | 3528/3117 | 1.023 (0.920-1.138) | 1.036 (0.867-1.239) | 40.713 | 60.7 | 3408/3018 | 0.993 (0.892-1.105) | 28.390 | 47.2 | ||
| TT vs CC + CT | 3528/3117 | 0.914 (0.723-1.155) | 0.931 (0.704-1.231) | 20.460 | 21.8 | – | – | – | – | ||
| TT vs CC | 2255/2038 | 0.919 (0.723-1.167) | 0.947 (0.691-1.296) | 24.085 | 33.6 | – | – | – | – | ||
| CT vs CC | 3320/2911 | 1.038 (0.929-1.160) | 1.051 (0.891-1.240) | 32.187 | 50.3 | 3207/2813 | 1.010 (0.903-1.130) | 22.950 | 34.6 | ||
| America | T vs C | 1938/1868 | 1.090 (0.979-1.214) | 1.090 (0.979-1.214) | 4.727 | 0.0 | – | – | – | – | |
| TT + CT vs CC | 1938/1868 | 1.185 (1.010-1.390) | 1.185 (1.010-1.390) | 4.930 | 0.0 | – | – | – | – | ||
| TT vs CC + CT | 1938/1868 | 1.019 (0.852-1.220) | 1.167 (0.830-1.641) | 8.417 | 40.6 | – | – | – | – | ||
| TT vs CC | 1227/1221 | 1.266 (0.944-1.697) | 1.272 (0.941-1.720) | 5.190 | 3.7 | – | – | – | – | ||
| CT vs CC | 1306/1400 | 1.168 (0.988-1.380) | 1.151 (0.924-1.433) | 8.107 | 38.3 | – | – | – | – | ||
| Excluded for DHWE | T vs C | 1639/1517 | 1.085 (0.964-1.222) | 1.103 (0.955-1.275) | 4.091 | 26.7 | – | – | – | – | |
| TT + CT vs CC | 1639/1517 | 1.273 (1.058-1.531) | 1.273 (1.058-1.531) | 2.527 | 0.0 | – | – | – | – | ||
| TT vs CC + CT | 1639/1517 | 0.958 (0.794-1.155) | 0.958 (0.794-1.155) | 1.455 | 0.0 | – | – | – | – | ||
| TT vs CC | 1016/994 | 1.131 (0.814-1.571) | 1.131 (0.814-1.571) | 1.405 | 0.0 | – | – | – | – | ||
| CT vs CC | 1035/1065 | 1.307 (1.078-1.584) | 1.307 (1.078-1.584) | 2.744 | 0.0 | – | – | – | – | ||
| Asiac | T vs C | 433/419 | 0.873 (0.708-1.077) | 0.873 (0.708-1.077) | 0.225 | 0.0 | – | – | – | – | |
| TT + CT vs CC | 433/419 | 0.938 (0.714-1.233) | 0.938 (0.714-1.233) | 1.922 | 0.0 | – | – | – | – | ||
| TT vs CC + CT | 433/419 | 0.533 (0.308-0.921) | 0.533 (0.308-0.921) | 1.168 | 0.0 | – | – | – | – | ||
| TT vs CC | 230/234 | 0.570 (0.319-1.017) | 0.570 (0.319-1.017) | 0.426 | 0.0 | – | – | – | – | ||
| CT vs CC | 410/381 | 1.015 (0.765-1.345) | 1.042 (0.701-1.548) | 3.525 | 43.3 | – | – | – | – | ||
Abbreviations: AD, Alzheimer’s disease; CI, confidence interval; DHWE, deviated from HWE in cases and/or in controls; FEM, fixed effect model; HWE, Hardy-Weinberg equilibrium; OR, odds ratio; PLAU, urokinase-plasminogen activator; REM, random effect model.
a Sensitivity analysis for reducing heterogeneity by omitting study using the STATA module when I 2 ≥ 50%.
b All articles for Europe were DHWE in cases and/or in controls.
c There is only 1 study left after deviated from HWE in cases and/or in controls.
Sources of Heterogeneity and Sensitive Analysis
Before and after excluding the studies 8,12,17,21 deviating from HWE in the cases and/or controls, strong evidence of heterogeneity was demonstrated in the codominant- and dominant-inherited models for this meta-analysis. However, univariate metaregression, with the covariates of sample size (the sum of case and control numbers), age (ratio of mean age in the case group to that in the control group), and publication year for the rs2227564 polymorphism, showed that no covariates had a significant impact on between-study heterogeneity (data not shown). The key contributor of the study to between-study heterogeneity was assessed by the “leave-one-out” sensitive analysis 26 in the included studies. Low and moderate heterogeneities (I 2 < 50%) were found in any inherited models following the exclusion of certain studies. However, the association of PLAU gene rs2227564 polymorphism with AD risk was significant in the dominant model (FEM OR 1.123, 95% CI 1.025-1.231) and heterozygote comparison (CT vs CC; FEM OR 1.126, 95% CI 1.027-1.235), only marginally significant in the codominant model (FEM OR 1.012, 95% CI 0.944-1.085). We also conducted a subgroup analysis with studies conducted in Europe groups, America groups, and Asia groups (detailed data are shown in Table 2). Figure 1 shows the forest plot of OR for AD in dominant model of PLAU gene rs2227564 polymorphism in all articles after excluding studies deviating from HWE in cases and/or controls and sensitivity analysis.
Figure 1.
Forest plots of relationship between urokinase-plasminogen activator (PLAU) gene rs2227564 polymorphism and Alzheimer’s disease (AD) in dominant model (TT + CT vs CC) after excluding the studies that deviated from Hardy-Weinberg equilibrium (HWE) in cases and/or control groups and sensitivity analysis. White diamond denotes the pooled odds ratio (OR). Black squares indicate the OR in each study, with square sizes inversely proportional to the standard error of the OR. Horizontal lines represent 95% confidence interval (CI).
Influence Analysis
After exclusion of studies deviating from HWE in cases and/or controls and sensitivity analysis, there is 1 individual study 14 that should be removed, because the point estimate of its omitted analysis lies outside the 95% CI of the combined analysis in codominant model, dominant model, and heterozygote comparison. No independent study was found having excessive influence on the pooled effect in the above-mentioned inherited models after further excluding the individual study. 14
Publication Bias
Harbord’s test was used to assess the publication bias. There is no evidence showing publication bias for association between PLAU gene rs2227564 polymorphism and AD in the above-mentioned inherited models after exclusion of studies deviating from HWE in cases and/or controls and sensitivity analysis.
Discussion
Urokinase-plasminogen activator gene is a serine protease that converts plasminogen to plasmin, and plasmin also degrades plasma Aβ protein, affecting its circulating concentration that might be important in AD neuropathogenesis or diagnosis. 29,30 Since the first study that attempted to explore the association between PLAU rs2227654 polymorphism and AD risk in humans was reported in 2003, 9 many studies have tried to replicate the association. However, these results were inconsistent. Therefore, a meta-analysis should be performed to form a more precise estimation of those studies.
A total of 27 independent studies with 6100 AD cases and 5718 controls were included to assess the association between rs2227564 polymorphism and AD. To our knowledge, this is the first meta-analysis carried out to investigate the relationship between PLAU gene rs2227564 genetic polymorphism and AD. In the present meta-analysis, our pooled results showed that the PLAU rs2227564 polymorphism was significantly associated with AD risk in the dominant model, heterozygote comparison, and marginally significant in the codominant model after exclusion of studies deviating from HWE in cases and/or controls and sensitivity analysis. However, no significant associations were found in the recessive model and homozygote comparison. Evidence of heterogeneity was found for the rs2227564 polymorphism with AD risk. Between-study heterogeneity is common in meta-analysis for genetic association studies 31 and exploring the potential sources of between-study heterogeneity is the essential component of meta-analysis. 32 The between-study heterogeneity might arise from an indeterminate number of characteristics that vary among studies, such as study quality, characteristics of the subjects involved, genotyping quality, variation of the covariate, and deviation from HWE in some studies, and so on. Thus, we used metaregression to explore the causes of heterogeneity for covariates. However, sample size (the sum of case and control numbers), age (ratio of mean age in the case group to that in the control group), and publication year were not found to be the important sources of between-study heterogeneity in this meta-analysis. Thus, we used “leave-one-out” sensitive analysis, 26 which aims to reduce between-study heterogeneity and explore the potential important causes of between-study heterogeneity for both covariates and studies. The key contributor of the article to this low between-study heterogeneity assessed by the “leave-one-out” sensitive analysis was the one conducted by Finckh et al. 9 When we performed the data analysis, after the “leave-one-out” sensitive analysis, on the studies that obeyed the HWE in cases and controls, our results showed that the T allele of rs2227564 polymorphism in PLAU gene had significant effect on increased AD risk. No publication bias were found in all the inherited models after exclusion of studies deviating from HWE in cases and/or controls and sensitivity analysis.
Meta-analysis that can summarize and review previously published quantitative research has been recognized as an effective method to solve a wide variety of clinical questions. Nevertheless, some limitations have affected the objectivity of the conclusions and should be addressed. For example, lack of original information for included studies made it impracticable to stratify by other variables, such as smoking status, drinking status, family history, or other relevant diseases, which may affect AD. In spite of these limitations, our meta-analysis also showed some advantages. First, no publication biases were detected, indicating that the whole pooled results may be unbiased; second, substantial number of cases and controls were pooled from different studies, which significantly increased the statistical power of the analysis, and so on.
Conclusions
Our study suggested that the mutant genotype of TT + CT was significantly associated with AD risk from 27 case–control studies, and the T allele probably acts as an important AD risk factor. With regard to AD with multifactorial etiology, large well-designed epidemiological studies, especially considering different ethnic background, gene–gene, gene–environmental interactions, or other risk factors, should be performed in the future to clarify the possible roles of PLAU rs2227564 polymorphism in the pathogenesis of AD.
Footnotes
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The authors disclosed receipt of the following financial support for the research, authorship and/or publication of this article: This report was supported by the Fok Ying Tong Education Foundation (121034).
References
- 1. Brookmeyer R, Johnson E, Ziegler-Graham K, Arrighi HM. Forecasting the global burden of Alzheimer's disease. Alzheimers Dement. 2007;3(3):186–191. [DOI] [PubMed] [Google Scholar]
- 2. Ertekin-Taner N, Graff-Radford N, Younkin LH, et al. Linkage of plasma Abeta42 to a quantitative locus on chromosome 10 in late-onset Alzheimer's disease pedigrees. Science. 2000;290(5500):2303–2304. [DOI] [PubMed] [Google Scholar]
- 3. Myers A, Holmans P, Marshall H, et al. Susceptibility locus for Alzheimer's disease on chromosome 10. Science. 2000;290(5500):2304–2305. [DOI] [PubMed] [Google Scholar]
- 4. Ledesma MD, Da Silva JS, Crassaerts K, Delacourte A, De Strooper B, Dotti CG. Brain plasmin enhances APP alpha-cleavage and Abeta degradation and is reduced in Alzheimer's disease brains. EMBO Rep. 2000;1(6):530–535. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Tucker HM, Kihiko M, Caldwell JN, et al. The plasmin system is induced by and degrades amyloid-beta aggregates. J Neurosci. 2000;20(11):3937–3946. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Li X, Bokman AM, Llinas M, Smith RA, Dobson CM. Solution structure of the kringle domain from urokinase-type plasminogen activator. J Mol Biol. 1994;235(5):1548–1559. [DOI] [PubMed] [Google Scholar]
- 7. Yoshimoto M, Ushiyama Y, Sakai M, et al. Characterization of single chain urokinase-type plasminogen activator with a novel amino-acid substitution in the kringle structure. Biochim Biophys Acta. 1996;1293(1):83–89. [DOI] [PubMed] [Google Scholar]
- 8. Myers AJ, Marshall H, Holmans P, et al. Variation in the urokinase-plasminogen activator gene does not explain the chromosome 10 linkage signal for late onset AD. Am J Med Genet B Neuropsychiatr Genet. 2004;124B(1):29–37. [DOI] [PubMed] [Google Scholar]
- 9. Finckh U, van Hadeln K, Muller-Thomsen T, et al. Association of late-onset Alzheimer disease with a genotype of PLAU, the gene encoding urokinase-type plasminogen activator on chromosome 10q22.2. Neurogenetics. 2003;4(4):213–217. [DOI] [PubMed] [Google Scholar]
- 10. Papassotiropoulos A, Tsolaki M, Wollmer MA, et al. No association of a non-synonymous PLAU polymorphism with Alzheimer's disease and disease-related traits. Am J Med Genet B Neuropsychiatr Genet. 2005;132B(1):21–23. [DOI] [PubMed] [Google Scholar]
- 11. Bagnoli S, Tedde A, Cellini E, Rotondi M, Nacmias B, Sorbi S. The urokinase-plasminogen activator (PLAU) gene is not associated with late onset Alzheimer's disease. Neurogenetics. 2005;6(1):53–54. [DOI] [PubMed] [Google Scholar]
- 12. Ertekin-Taner N, Ronald J, Feuk L, et al. Elevated amyloid beta protein (Abeta42) and late onset Alzheimer's disease are associated with single nucleotide polymorphisms in the urokinase-type plasminogen activator gene. Hum Mol Genet. 2005;14(3):447–460. [DOI] [PubMed] [Google Scholar]
- 13. Riemenschneider M, Konta L, Friedrich P, et al. A functional polymorphism within plasminogen activator urokinase (PLAU) is associated with Alzheimer's disease. Hum Mol Genet. 2006;15(16):2446–2456. [DOI] [PubMed] [Google Scholar]
- 14. Ozturk A, Minster RL, DeKosky ST, Kamboh MI. Association of tagSNPs in the urokinase-plasminogen activator (PLAU) gene with Alzheimer's disease and associated quantitative traits. Am J Med Genet B Neuropsychiatr Genet. 2007;144B(1):79–82. [DOI] [PubMed] [Google Scholar]
- 15. Blomqvist ME, Reynolds C, Katzov H, et al. Towards compendia of negative genetic association studies: an example for Alzheimer disease. Hum Genet. 2006;119(1-2):29–37. [DOI] [PubMed] [Google Scholar]
- 16. Pesaresi M, Batelli S, Prato F, et al. The urokinase-type plasminogen activator polymorphism PLAU_1 is a risk factor for APOE-epsilon4 non-carriers in the Italian Alzheimer's disease population and does not affect the plasma Abeta(1-42) level. Neurobiol Dis. 2007;25(3):609–613. [DOI] [PubMed] [Google Scholar]
- 17. Ji X, Jia L, Jia J. Relationship between urokinase plasminogen activator gene polymorphism and sporadic Alzheimer's disease. Clin J Neurol. 2007;40(9):616–617. [Google Scholar]
- 18. Xiang S, Xu S, Li D. Relation between the P141L polymorphism of urokinase plasminogen activator gene and late-onset Alzheimer's disease. J Capital Med Univ. 2008;29(3):278–282. [Google Scholar]
- 19. Giedraitis V, Kilander L, Degerman-Gunnarsson M, et al. Genetic analysis of Alzheimer's disease in the Uppsala Longitudinal Study of Adult Men. Dement Geriatr Cogn Disord. 2009;27(1):59–68. [DOI] [PubMed] [Google Scholar]
- 20. Cousin E, Mace S, Rocher C, et al. No replication of genetic association between candidate polymorphisms and Alzheimer's disease. Neurobiol Aging. 2011;32(8):1443–1451. [DOI] [PubMed] [Google Scholar]
- 21. Zhou X, Zou T, Ku M, Xing S, Yue Y. The analysis of Alzheimer's disease and PLAU gene polymorphisms in Xinjiang Uygur and Chinese Han. Chin J Nerv Ment Dis. 2010;36(11):688–690. [Google Scholar]
- 22. Ioannidis JP, Trikalinos TA, Ntzani EE, Contopoulos-Ioannidis DG. Genetic associations in large versus small studies: an empirical assessment. Lancet. 2003;361(9357):567–571. [DOI] [PubMed] [Google Scholar]
- 23. Lohmueller KE, Pearce CL, Pike M, Lander ES, Hirschhorn JN. Meta-analysis of genetic association studies supports a contribution of common variants to susceptibility to common disease. Nat Genet. 2003;33(2):177–182. [DOI] [PubMed] [Google Scholar]
- 24. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med. 2002;21(11):1539–1558. [DOI] [PubMed] [Google Scholar]
- 25. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ. 2003;327(7414):557–560. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Patsopoulos NA, Evangelou E, Ioannidis JP. Sensitivity of between-study heterogeneity in meta-analysis: proposed metrics and empirical evaluation. Int J Epidemiol. 2008;37(5):1148–1157. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Tobias A. Assessing the influence of a single study in the meta-analysis estimate. Stata Tech Bull. 1999;47(8);15–17. [Google Scholar]
- 28. Harbord RM, Egger M, Sterne JA. A modified test for small-study effects in meta-analyses of controlled trials with binary endpoints. Stat Med. 2006;25(20):3443–3457. [DOI] [PubMed] [Google Scholar]
- 29. Melchor JP, Pawlak R, Strickland S. The tissue plasminogen activator-plasminogen proteolytic cascade accelerates amyloid-beta (Abeta) degradation and inhibits Abeta-induced neurodegeneration. J Neurosci. 2003;23(26):8867–8871. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Solfrizzi V, D'Introno A, Colacicco AM, et al. Circulating biomarkers of cognitive decline and dementia. Clin Chim Acta. 2006;364(1-2):91–112. [DOI] [PubMed] [Google Scholar]
- 31. Munafo MR, Flint J. Meta-analysis of genetic association studies. Trends Genet. 2004;20(9):439–444. [DOI] [PubMed] [Google Scholar]
- 32. Lau J, Ioannidis JP, Schmid CH. Summing up evidence: one answer is not always enough. Lancet. 1998;351(9096):123–127. [DOI] [PubMed] [Google Scholar]