- Email: [email protected]
Meta-Analysis on the Association between Brain-Derived Neurotrophic Factor Polymorphism rs6265 and Ischemic Stroke, Poststroke Depression
ARTICLE IN PRESS
Meta-Analysis on the Association between Brain-Derived Neurotrophic Factor Polymorphism rs6265 and Ischemic Stroke, Poststroke Depression Mei-Hua Bao, PhD,*,†,‡ Shu-Zhen Zhu, PhD,†,‡ Xin-Zheng Gao, Msc,†,‡ Hong-Shuo Sun, MD, MSc, PhD,†,‡ and Zhong-Ping Feng, MD, MSc, PhD†
Background: Ischemic stroke is a multifactorial neurologic injury that causes mortality and disability worldwide. Poststroke depression is the most important neuropsychiatric consequence of stroke. Brain-derived neurotrophic factor is a neurotrophin family member that plays key role in regulating neuron survival and differentiation. Studies found a polymorphism in brain-derived neurotrophic factor gene (rs6265) may associate with the ischemic stroke and poststroke depression risk. However, the results are inconclusive and inconsistent. Methods: In the present meta-analysis, the database PubMed, Embase, Cochrane Central Register of Controlled Trials, CNKI, and Chinese Biomedical Literature Database were searched until July 9, 2017. Results: Seven studies with 1287 cases and 1032 controls were included for the meta-analysis of ischemic stroke, and five studies with 272 cases and 503 controls were included for poststroke depression. The results indicated that the GG genotype of brain-derived neurotrophic factor is related to a significantly lower risk of ischemic stroke in the homozygous and dominant models (odds ratio = .57 and .80, respectively). No significant relation was found between rs6265 and poststroke depression. Conclusions: Thus, brain-derived neurotrophic factor rs6265 might be recommended as a predictor of susceptibility of ischemic stroke. However, the results of this meta-analysis should be interpreted with caution because of the heterogeneity between studies and low sample size. Further studies are needed to evaluate the associations between rs6265 and poststroke depression, especially in Caucasians, with large sample size. Key Words: Brain-derived neurotrophic factor—ischemic stroke—poststroke depression—rs6265. © 2018 National Stroke Association. Published by Elsevier Inc. All rights reserved.
From the *Department of Anatomy, Histology and Embryology, Science Research Center, Institute of Neuroscience, Changsha Medical University, Changsha, China; †Department of Physiology; and ‡Department of Surgery, Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada. Received November 8, 2017; revision received November 22, 2017 ; accepted January 12, 2018 . Grant support: The present study was supported by the Construct Program of the Key Discipline in Hunan Province, National Science Foundation of China (Grant No. 81670427), the Foundation of Hunan Educational Committee (Grant No. 15C0146) to M.-H.B. Heart and Stroke Foundation of Canada (G-13-0003069), a Canadian Institutes of Health Research (CIHR) China-Canada Joint Health Research Initiative (CIHR, FRN No. 132571) and Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery Grants (RGPIN-2016-04574) to H.S.S.; and the CIHR (PJT152155) to Z.-P.F. Author contributions: M.-H.B., H.-S.S., and Z.-P.F. conceived the project and designed the experiments; M.-H.B., S.-Z.Z., and X.-Z.G. performed the data extraction and analysis; M.-H.B. and Z.-P.F. wrote and M.-H.B., H.-S.S. and Z.-P.F. revised the manuscript. All authors read the manuscript. Address correspondence to Hong-Shuo Sun, MD, MSc, PhD, and Zhong-Ping Feng, MD, MSc, PhD, Department of Physiology, Faculty of Medicine, University of Toronto, 1 King’s College Circle, Toronto, Ontario M5S 1A8, Canada. E-mail: [email protected] and [email protected]. 1052-3057/$ - see front matter © 2018 National Stroke Association. Published by Elsevier Inc. All rights reserved. https://doi.org/10.1016/j.jstrokecerebrovasdis.2018.01.010
Journal of Stroke and Cerebrovascular Diseases, Vol. ■■, No. ■■ (■■), 2018: pp ■■–■■
1
ARTICLE IN PRESS M.-H. BAO ET AL.
2
Introduction Ischemic stroke is a multifactorial neurologic injury that causes mortality and disability worldwide. According to the Global Burden of Disease (GBD) 2013 study, stroke was the second most common cause of death worldwide, which accounts for about 11.8% of all death in 2013. Moreover, the absolute number of people who died or disabled by ischemic stroke has increased significantly from 1990 to 2013.1 Poststroke depression (PSD) is the most important neuropsychiatric consequence of stroke. Almost one third of stroke survivors experience major depression. Understanding the risk factors, improving the stroke care is likely to be very helpful for ischemic stroke and PSD prediction and treatment. Brain-derived neurotrophic factor (BDNF) belongs to the neurotrophin family, which plays key roles in regulating neuron survival and differentiation during development.2,3 BDNF is widely expressed in hippocampus, cortex, and basal forebrain. The activation of BDNF facilitates the synaptic plasticity, neurogenesis, and memory process through Tropomyosin receptor kinase B (TrkB) signal.2,4,5 In stroke, BDNF is induced and activated, and is considered necessary for stroke recovery.6 The mutations of BDNF gene cause abnormal BDNF/TrkB signal and reduce BDNF production, therefore, leading to poor recovery after ischemic stroke.7 Single-nucleotide polymorphisms (SNPs) are mutations in a single nucleotide of a specific position in the genome. The SNPs may change the amino acid sequence of the protein or affect the gene expression. Many SNPs have been found in the gene sequence of BDNF. Among these SNPs, rs6265, also named Val66Met, changes the G allele to A at position 196 on BDNF gene. The G allele encodes Val, whereas the A allele encodes Met. The mutation affects the normal translation and intracellular trafficking of BDNF mRNA, which decreases BDNF production and secretion in neurons.7-9 In recent years, several studies demonstrated a higher risk of ischemic stroke for the allele A of rs6265,10,11 whereas other studies did not.12,13 In this study, we performed the meta-analysis to give a more precise and comprehensive estimation of the association between BDNF rs6265 and ischemic stroke, PSD.
Methods Publication Search Strategy and Inclusion Criteria A comprehensive literature search was conducted in the present meta-analysis. For the association between BDNF polymorphism and ischemic stroke, the following terms were used (“ischemic stroke”) and (“BDNF” or “brainderived neurotrophic factor”) and (“polymorphism” or “mutation” or “variant” or “SNP” or “single nucleotide polymorphism”). For the association between BDNF polymorphism and poststroke depression, the terms (“post-stroke depression” or “PSD”) and (“BDNF” or
“brain-derived neurotrophic factor”) and (“polymorphism” or “mutation” or “variant” or “SNP” or “single nucleotide polymorphism”) were used. The literature searching was without restriction on language. The deadline for publication was July 9, 2017. We searched the published studies in the electronic databases: PubMed, Embase, Cochrane Central Register of Controlled Trials (CENTRAL), Chinese National Knowledge Infrastructure (CNKI), and Chinese Biomedical Literature Database (CBM). After removing of duplications, all the results from the databases were screened by the title and abstract first. If the title and abstract reach our criteria, the full text will be retrieved and read. The references of all eligible studies were retrieved manually for other potentially relevant studies. Inclusion criteria include (1) case–control studies; (2) studies demonstrated the association between BDNF polymorphisms and ischemic stroke or post-stroke depression; and (3) sufficient data within the individual publication. Studies were excluded if any of the following applies: (1) repeat publications, abstracts, dissertations or reviews; and (2) studies not meeting all of the inclusion criteria.
Data Extraction Information from each eligible publication was extracted manually by 2 investigators independently (M.H.B. and S.-Z.Z.). The following information was extracted from each eligible study: first author’s name, published year, country, ethnicity, genotype method, sex, age, the number of subjects. The discrepancies during data extraction were solved by consensus achieved by the third author (X.-Z.G.).
Quality Assessment Quality assessment was conducted according to the predefined scale provided by Jiang et al.28 The score scale includes source of cases, source of controls, specimens used for genotype, total sample size, and Hardy– Weinberg equilibrium (HWE). The scores range from 0 to 15. The low-quality studies were defined as quality score less than 10, whereas others were treated as high quality. The quality evaluation was performed by 2 authors independently (M.-H.B. and X.-Z.G.). Discrepancies were solved by holding a consensus.
Statistical Methods The HWE of the control group was evaluated by χ2 test, with P less than .05 considered as HWE deviated. The pooled odds ratio (OR), with 95% confidence interval (CI), was used to calculate the association between BDNF rs6265 and ischemic stroke, PSD risk. Genetic models such as homozygous (GG versus AA), heterozygous (GG versus GA), dominant (GG versus GA + AA), and allelic models (G versus A) were evaluated. Z test with P less than .05 was used to determine the statistical signifi-
ARTICLE IN PRESS META-ANALYSIS FOR RS6265 AND ISCHEMIC STROKE, PSD
3
cance of the associations. We also conducted subgroup analysis by ethnicity and quality score. To evaluate the heterogeneity of the included studies, I2 test was performed. I2 greater than 50% was considered to be with obvious heterogeneity, and the random-effects model (the DerSimonian and Laird method) was used for the metaanalysis. Otherwise, the fixed-effects model (the Mantel– Haenszel method) was used. The sensitivity analysis was performed by evaluating the influence of each study on the pooled ORs and 95% CIs. One single study at a time was omitted using STATA 12.0 software in the sensitivity analysis. We also evaluated the publication bias using Begg’s funnel plot and Egger’s linear regression method. All statistical analysis was performed using the STATA 12.0 software (StataCorp, College Station, TX) and RevMan 5.3 (Cochrane Community, London, UK).
Results Characteristics of Eligible Studies For the rs6265 and ischemic stroke analysis, a total of 55 studies were obtained without duplications after a comprehensive literature search. Among them, 37 studies were removed because of being irrelevant, reviews, dissertations, not case–control design, without enough data, or other languages than English and Chinese. At last, 7 studies were found eligible, with 1287 cases and 1032 controls. For the rs6265 and PSD, a total of 30 studies were obtained without duplications. Among them, 24 studies were excluded because of irrelevance, review, or dissertations. At last, 5 studies with 272 cases and 503 controls were included for the rs6265 and PSD analysis. The PRISMA flowcharts were shown in Figures 1 and 2, and the information about the included studies was presented in Table 1.
Figure 1. PRISMA flowchart for the association between BDNF rs6265 and ischemic stroke.
Results of Meta-Analysis As shown in Table 2 and Figure 3, 7 studies with 1287 cases and 1032 controls were included in the present metaanalysis. A significantly lower risk of ischemic stroke was found for the GG genotype in the homozygous model and dominant model in overall analysis (GG versus AA, OR = .57, 95% CI = .43-0.75, P < .0001; GG versus GA + AA, OR = .80, 95% CI = .65-0.98, P = .03). Subgroup analysis by ethnicity found lower susceptibility of ischemic stroke in Asians in all genetic models except heterozygous (GG versus AA, OR = .58, 95% CI = .43-0.77, P = .0002; GG versus GA + AA, OR = .78, 95% CI = .62-0.97, P = .03; G versus A, OR = .77, 95% CI = .67-0.89, P = .0003). We also conducted subgroup analysis by quality score. The results indicated there was no significant difference between highquality study group and low-quality group in homozygous, heterozygous, and dominant models. For the allelic model, we only found a significantly lower risk of ischemic stroke in the low-quality group.
Figure 2. PRISMA flowchart for the association between BDNF rs6265 and PSD.
4
Table 1. Characteristics of eligible studies included in the meta-analysis BDNF rs6265 and ischemic stroke
Year
Country
Ethnicity
Genotyping methods
Gao et al10 Keshavarz et al14 Lu et al15 Vilyanov et al16 Wang et al13 Zhao et al11 Zhou et al12
2015 2016 2013 2012 2010 2013 2011
China Iran China Russia China China China
Asian Caucasian Asian Caucasian Asian Asian Asian
PCR-RFLP PCR-RFLP PCR-RFLP NA PCR-RFLP PCR-RFLP PCR-RFLP
171/119:127/73 107/99:129/71 83/71:25/21 NA 24/25:26/20 300/194:185/152 NA
Age (case/control)
Quality score
61.4 ± 8.6:60.8 ± 7.9 64.5 ± 11.7:62.1 ± 6.7 70.0 ± 12.2:68.0 ± 12.4 NA 59.9 ± 9.8:59.7 ± 9.5 69.8 ± 11.3:68.9 ± 9.99 NA
9 12 12 6 11 13 11
G/G
G/A
A/A
Sample size (case/control)
(Case/ Control)
(Case/ Control)
(Case/ Control)
HWE of control
290/200 206/200 154/46 64/110 49/46 494/337 30/93
61/57 0/2 71/22 43/72 10/11 119/96 10/34
145/112 50/18 55/18 17/35 31/29 247/174 14/41
84/31 156/180 28/6 4/3 8/6 128/67 6/18
.049 .058 .457 .606 .061 .459 .378
G/G
G/A
A/A
(Case/ Control)
(Case/ Control)
ARTICLE IN PRESS
Author
Sex ratio (male/female) (case/control)
BDNF rs6265 and poststroke depression
Author Golimbet17 Kim et al18 Lu et al15 Rezaei et al19 Zhou et al12
Year
Country
Ethnicity
Genotyping methods
2010 2012 2013 2015 2011
Russia Korea China Iran China
Caucasian Asian Asian Caucasian Asian
RFLP-PCR RFLP-PCR PCR RFLP-PCR RFLP-PCR
Sex ratio (male:female) (case/control) NA 43/34:121/78 48/38:35/33 14/34:93/65 19/16:34/24
Age (case/control)
Quality score
Sample size (case/control)
(Case/ Control)
NA 66.5 ± 9.5:63.5 ± 9.4 71 ± 11.1:69 ± 13.3 74.5 ± 8.8:61.5 ± 10.8 61.7 ± 8.5:63.5 ± 12.5
5 9 11 12 11
24/20 77/199 86/68 48/158 35/58
15/16 13/51 41/30 0/0 13/21
GA + AA: 11/4 41/113 23/35 29/26 16/12 19/31 29/127 14/27 8/10
HWE of control NA .044 .143 .172 .794
M.-H. BAO ET AL.
Abbreviations: BDNF, brain-derived neurotrophic factor; RFLP-PCR, Restriction Fragment Length Polymorphism-Polymerase Chain Reaction; HWE, Hardy–Weinberg equilibrium; NA, not available.
ARTICLE IN PRESS META-ANALYSIS FOR RS6265 AND ISCHEMIC STROKE, PSD
5
Table 2. Pooled ORs and 95% CIs of the association between BDNF rs6265 and ischemic stroke, PSD BDNF rs6265 with ischemic stroke Genetic model GG versus AA
GG versus GA
GG versus GA + AA
G versus A
Overall Asian Caucasian High quality Low quality Overall Asian Caucasian High quality Low quality Overall Asian Caucasian High quality Low quality Overall Asian Caucasian High quality Low quality
N
OR [95% CI]
P value
I2 (%)
7 5 2 5 2 7 5 2 5 2 7 5 2 5 2 7 5 2 5 2
.57 [.43, .75] .58 [.43, .77] .38 [.10, 1.47] .66 [.47, .93] .40 [.24, .67] .89 [.71, 1.11] .88 [.69, 1.10] 1.01 [.53, 1.93] .87 [.66, 1.14] .93 [.64, 1.34] .80 [.65, .98] .78 [.62, .97] .97 [.52, 1.81] .81 [.63, 1.05] .77 [.54, 1.09] .93 [.71, 1.23] .77 [.67, .89] 1.50 [.61, 3.71] 1.04 [.71, 1.51] .70 [.56, .88]
<.0001 .0002 .16 .02 .0005 .29 .26 .98 .31 .69 .03 .03 .92 .11 .14 .63 .0003 .38 .84 .003
0 0 0 0 0 0 0 68 0 0 0 0 16 0 32 69 0 82 72 26
BDNF rs6265 with PSD Genetic model GG versus AA
GG versus GA
GG versus GA + AA
G versus A
Overall Asian Caucasian High quality Low quality Overall Asian Caucasian High quality Low quality Overall Asian Caucasian High quality Low quality Overall Asian Caucasian High quality Low quality
N
OR [95% CI]
P value
I2 (%)
4 3 1 3 1 4 3 1 3 1 5 3 2 3 2 4 3 1 3 1
.63 [.37, 1.07] .63 [.37, 1.07] NA .92 [.46, 1.87] .39 [.17, .87] .97 [.63, 1.51] .97 [.63, 1.51] NA 1.21 [.69, 2.14] .70 [.35, 1.42] .82 [.56, 1.21] .87 [.58, 1.30] .34 [.09, 1.31] 1.11 [.67, 1.86] .53 [.29, .97] 1.05 [.64, 1.73] .82 [.61, 1.12] 2.27 [1.22, 4.23] 1.27 [.76, 2.13] .66 [.45, .95]
.09 .09 NA .83 .02 .90 .90 NA .5 .33 .31 .5 .12 .68 .04 .83 .22 .01 .36 .03
25 25 NA 0 N.A 0 0 NA 0 NA 7 9 NA 0 0 74 24 NA 61 NA
Abbreviations: BDNF, brain-derived neurotrophic factor; CI, confidence interval; NA, not applicable; OR, odds ratio; PSD, poststroke depression.
As shown in Table 2 and Figure 4, 5 studies, with 272 cases and 503 controls were included in the present metaanalysis for the association between BDNF rs6265 and PSD. No significant association between rs6265 and PSD was found. As Golimbet17 only provided the data for GA + AA, the data were only pooled in the dominant genetic model.
Source of Heterogeneity No obvious heterogeneity was found for all the included studies in the homozygous, heterozygous, and dominant models. However, in the allelic model for both ischemic stroke and PSD, high heterogeneity was found
ARTICLE IN PRESS M.-H. BAO ET AL.
6
Figure 3.
Forest plots of odds ratio for the association between BDNF rs6265 and risks of ischemic stroke.
(I2 % = 69% for ischemic stroke and 74% for PSD). Subgroup analysis by ethnicity showed big heterogeneity among Caucasians. Specifically, Iranians14,19 are much more diverse as compared with Russians.10,11,16 Therefore, the ethnicity might be the source of heterogeneity.
Sensitivity We evaluated the influence of each individual on pooled ORs and 95% CIs by sensitivity analysis. No obvious influence was found after omitting any individual study
ARTICLE IN PRESS META-ANALYSIS FOR RS6265 AND ISCHEMIC STROKE, PSD
Figure 4.
7
Forest plots of odds ratio for the association between BDNF rs6265 and risks of PSD.
in all genetic models. Figure 5 shows the influence of individual study on pooled ORs and 95% CI in the dominant genetic model for ischemic stroke and PSD.
Publication Bias Egger’s and Begg’s tests were used to evaluate the publication bias of all included studies. As shown in Table 3 and Figure 6, we observed an asymmetry in the funnel plot. A significant bias also existed in Egger’s test (P = .047).
Discussion In this meta-analysis, we found a significantly lower risk of ischemic stroke for BDNF GG genotype in the homozygous and dominant models (OR = .57 and OR = .80, respectively). Subgroup analysis by ethnicity discovered a drastically decrease in risk of ischemic stroke in Asian in the homozygous, dominant, and allelic models (OR = .58, .78, and .77, respectively). No significant relations were found between rs6265 and PSD.
ARTICLE IN PRESS M.-H. BAO ET AL.
8
Figure 5. Sensitivity analysis of the influence of each study on pooled odds ratios and 95% confidence interval (CIs) in GG versus GA + AA model for associations of ALDH2 rs6265 and ischemic stroke, poststroke depression (PSD).
BDNF plays important roles in the neural system. It supports the survival of existing neurons and helps to stimulate neurogenesis.20,21 In patients with ischemic stroke, a decrease in serum BDNF concentration has been observed. The decrease in BDNF is also observed in ischemic stroke–related diseases, such as atherosclerosis, diabetes, and metabolic syndrome.22,23 Moreover, decreased BDNF is associated with poor recovery of stroke, poorer memory, and abnormal hippocampal activation.7,9,23 The mechanisms for BDNF include at least 2 receptors, TrkB and p75.24 The rs6265 mutation changes G allele to A allele in position 196 of BDNF gene, which switches the amino acid valine to methionine at codon 66, Val66Met. BDNF rs6265 is relatively common in humans, with a prevalence of 20%-30% for heterozygous and about ~4% for homozygous.25 The Met allele decreases the activitydependent secretion and intracellular distribution of BDNF by affecting intracellular trafficking of BDNF.7,9 The trafficking of BDNF from the biosynthetic pathway appears complex and highly regulated; however, the detailed process remains unclear. BDNF is secreted by neurons as a 32-kDa pro-BDNF that is changed to mature form rapidly after secretion.9 The pro-BDNF is secreted by vesicle-mediated processes, where the secretory vesicles are transported from postsynaptic dendrites to the
somatodendritic compartment.26 The rs6265 located in the 5′ pro-BDNF sequence blocks the trafficking of BDNF to secretory granules but does not affect the function of the mature protein.7 Transfection of rodent hippocampal neurons with val- or met-BDNF found val-BDNF secretory granules spread to the synapses, whereas metBDNF aggregates are accumulated in the cell body.7 This may hamper the local and synapse-specific regulation by BDNF, leading to impairment of the function of hippocampal neurons and the neurons in other parts of the brain. Several studies have found significant relations between rs6265 and ischemic stroke occurrence.10,11 Our metaanalysis found a significantly lower ischemic stroke risk for GG genotype in both the homozygous and the dominant models. This might be interpreted by the impaired secretion of BDNF for the A allele (met-BDNF). PSD is a common consequence of stroke. Biological mechanisms, such as neural circuits involved in mood regulation, and psychological stressors are considered to be the etiology of PSD.27 Genetic variants are thought to be involved in the development of PSD. BDNF is related to the repair of the neuronal system after stroke. Therefore, the Val66Met mutation of BDNF might relate to the PSD. However, our meta-analysis did not find any relations between BDNF rs6265 and PSD.
Table 3. Begg’s and Egger’s test for funnel plot asymmetries BDNF rs6265 and ischemic stroke, PSD (P value) Group
GG versus AA
GG versus GA
GG versus GA + AA
G versus A
Begg’s test Egger’s test
.283 .094
.283 .271
.436 .236
.213 .047
Abbreviations: BDNF, brain-derived neurotrophic factor; PSD, poststroke depression.
ARTICLE IN PRESS META-ANALYSIS FOR RS6265 AND ISCHEMIC STROKE, PSD
9 4
types are limited in most of the involved studies. Ischemic stroke and PSD are influenced by both genetic and environmental factors. The gene–gene and gene–environment interactions may play important roles in the functions of these polymorphisms, but most studies lack information on environmental exposure and multiple SNPs in haplotypes.
Conclusions
Figure 6. Begg’s funnel plot for the included studies for G versus A genotype of rs6265.
Intriguingly, the BDNF allele distribution in Iranians is quite different from those of Asians and other Caucasians. Iranians have more A allele than G allele. The frequency of G allele was 12.1% in Iranians patients with ischemic stroke and 5.5% in healthy Iranians,14,19 which are quite low than those in Asians (G allele greater than 46%)10 and in the U.S. population (G allele greater than 80.0%).7 Moreover, the G allele in Iranian is associate with higher risk of ischemic stroke and PSD.14,19 This difference causes a big heterogeneity for the studies of Iranians in respect to other ethnicities. Increased heterozygosity, selected patients with ischemic stroke, and presence of zero G allele in the studies are thought to contribute to the heterogeneity.14 Considering the difference of Iranian from other people, our meta-analysis results should be interpreted carefully for Iranians. We found a significant publication bias for all the involved studies in allelic model in Egger’s test (G versus A, P = .047). An asymmetry has also been found in funnel plots (Fig 6). As the Iranians have an obvious heterogeneity as compared with other ethnicities, we omitted the study on Iranians. We found the funnel plots become symmetric when the study was excluded. Therefore, we assume that this asymmetry was caused by a true heterogeneity existing in the involved studies. One of the limitations of the present meta-analysis is that the number of cases and controls were limited.1 In the present meta-analysis, only 7 studies with 1287 cases and 1032 controls were included for the meta-analysis for ischemic stroke, and 5 studies with 272 cases and 503 controls were included for the analysis of PSD. The small sample size may cause an overestimation of real relations.2 Most of the studies were on Asians. In the present metaanalysis, 77.9% of cases and 68.2% of controls are Asians. Moreover, the Iranians are with a different distribution of rs6265 genotype; and with different ischemic stroke or PSD risk for each genotype. Therefore, our results should be interpreted according to ethnicity.3 The disease sub-
In conclusion, the current meta-analysis suggests a lower risk of ischemic stroke for the GG genotype of BDNF rs6265 in both overall subjects and Asians. No significant association was found between rs6265 and PSD. However, the results of this meta-analysis should be interpreted with caution because of the heterogeneity between the involved studies. Further studies are needed to evaluate the associations between BDNF rs6265 and PSD, especially in Caucasians, with a large sample size.
References 1. Feigin VL, Norrving B, Mensah GA. Global burden of stroke. Circ Res 2017;120:439-448. 2. Bramham CR, Messaoudi E. BDNF function in adult synaptic plasticity: the synaptic consolidation hypothesis. Prog Neurobiol 2005;76:99-125. 3. Davies AM. The role of neurotrophins in the developing nervous system. J Neurobiol 1994;25:1334-1348. 4. Yamada K, Nabeshima T. Brain-derived neurotrophic factor/TrkB signaling in memory processes. J Pharmacol Sci 2003;91:267-270. 5. Bath KG, Akins MR, Lee FS. BDNF control of adult SVZ neurogenesis. Dev Psychobiol 2012;54:578-589. 6. Ploughman M, Windle V, MacLellan CL, et al. Brainderived neurotrophic factor contributes to recovery of skilled reaching after focal ischemia in rats. Stroke 2009;40:1490-1495. 7. Egan MF, Kojima M, Callicott JH, et al. The BDNF val66met polymorphism affects activity-dependent secretion of BDNF and human memory and hippocampal function. Cell 2003;112:257-269. 8. Baj G, Carlino D, Gardossi L, et al. Toward a unified biological hypothesis for the BDNF Val66Met-associated memory deficits in humans: a model of impaired dendritic mRNA trafficking. Front Neurosci 2013;7:188. 9. Chen ZY, Patel PD, Sant G, et al. Variant brain-derived neurotrophic factor (BDNF) (Met66) alters the intracellular trafficking and activity-dependent secretion of wild-type BDNF in neurosecretory cells and cortical neurons. J Neurosci 2004;24:4401-4411. 10. Gao LH, Zheng F, Wang JQ, et al. Correlation of brainderived neurotrophic factor gene polymorphism with prognosis in patients with acute ischemic stroke. Chin J Modern Med 2015;25:63-67. 11. Zhao J, Wu H, Weng YF. Study of correlation between genetic polymorphism of brain-derived neurotrophic factor and 90-day outcome of ischemic stroke in Chinese Han population. Chin J Clin Med 2015;503-509. 12. Zhou Z, Lu T, Xu G, et al. Decreased serum brain-derived neurotrophic factor (BDNF) is associated with post-stroke depression but not with BDNF gene Val66Met polymorphism. Clin Chem Lab Med 2011;49:185-189.
ARTICLE IN PRESS 10 13. Wang DX, Huang D, Zhang GH, et al. Association of brain-derived neurotrophic factor polymorphisms with vascular cognitive impairment. Chin J Geriatr Heart Brain Vessel Dis 2010;37-39. 14. Keshavarz P, Saberi A, Sharafshah A, et al. Association of BDNF G196A gene polymorphism with ischemic stroke occurrence and its 6-month outcome in an Iranian population. Top Stroke Rehabil 2016;23:254-260. 15. Lu LP, Zhu YB, Chen JY. The correlation between Val66Met polymorphism in brain-derived neurotrophic factor and post-stroke depression. Chin J Clin Neurosci 2013;21:413-417. 16. Vilyanov VB, Skipetrova LA, Remennik AY, et al. Dynamics of BDNF blood serum content and polymorphism of BDNF VAL166MET in patients with consequences of ischemic, hemorrhagic stroke and craniocerebral trauma. Eur J Nat History 2012;19-20. 17. Golimbet VE. Effects of the interaction of variants of the serotonin transporter and brain-derived neurotrophic factor on platelet serotonin levels in stroke patients. 41(5). United States, 2011. 18. Kim JM, Stewart R, Kang HJ, et al. A longitudinal study of BDNF promoter methylation and genotype with poststroke depression. J Affect Disord 2013;149:93-99. 19. Rezaei S, Asgari MK, Saberi A, et al. Brain-derived neurotrophic factor (BDNF) Val66Met polymorphism and post-stroke dementia: a hospital-based study from northern Iran. Neurol Sci 2016;37:935-942. 20. Acheson A, Conover JC, Fandl JP, et al. A BDNF autocrine loop in adult sensory neurons prevents cell death. Nature 1995;374:450-453.
M.-H. BAO ET AL. 21. Huang EJ, Reichardt LF. Neurotrophins: roles in neuronal development and function. Annu Rev Neurosci 2001;24:677-736. 22. Pikula A, Beiser AS, Chen TC, et al. Serum brain-derived neurotrophic factor and vascular endothelial growth factor levels are associated with risk of stroke and vascular brain injury: Framingham Study. Stroke 2013;44:2768-2775. 23. Stanne TM, Åberg ND, Nilsson S, et al. Low circulating acute brain-derived neurotrophic factor levels are associated with poor long-term functional outcome after ischemic stroke. Stroke 2016;47:1943-1945. 24. Patapoutian A, Reichardt LF. Trk receptors: mediators of neurotrophin action. Curr Opin Neurobiol 2001;11:272280. 25. Neves-Pereira M, Mundo E, Muglia P, et al. The brainderived neurotrophic factor gene confers susceptibility to bipolar disorder: evidence from a family-based association study. Am J Hum Genet 2002;71:651-655. 26. Kojima M, Takei N, Numakawa T, et al. Biological characterization and optical imaging of brain-derived neurotrophic factor-green fluorescent protein suggest an activity-dependent local release of brain-derived neurotrophic factor in neurites of cultured hippocampal neurons. J Neurosci Res 2001;64:1-10. 27. Spalletta G, Bossù P, Ciaramella A, et al. The etiology of poststroke depression: a review of the literature and a new hypothesis involving inflammatory cytokines. Mol Psychiatry 2006;11:984-991. 28. Jiang DK, Wangg WZ, Ren WH, et al. TP53 Arg72Pro polymorphism and skin cancer risk:a meta-analysis. J Invest Dermatol 2011;131:220-228.
Meta-Analysis on the Association between Brain-Derived Neurotrophic Factor Polymorphism rs6265 and Ischemic Stroke, Poststroke Depression Mei-Hua Bao, PhD,*,†,‡ Shu-Zhen Zhu, PhD,†,‡ Xin-Zheng Gao, Msc,†,‡ Hong-Shuo Sun, MD, MSc, PhD,†,‡ and Zhong-Ping Feng, MD, MSc, PhD†
Background: Ischemic stroke is a multifactorial neurologic injury that causes mortality and disability worldwide. Poststroke depression is the most important neuropsychiatric consequence of stroke. Brain-derived neurotrophic factor is a neurotrophin family member that plays key role in regulating neuron survival and differentiation. Studies found a polymorphism in brain-derived neurotrophic factor gene (rs6265) may associate with the ischemic stroke and poststroke depression risk. However, the results are inconclusive and inconsistent. Methods: In the present meta-analysis, the database PubMed, Embase, Cochrane Central Register of Controlled Trials, CNKI, and Chinese Biomedical Literature Database were searched until July 9, 2017. Results: Seven studies with 1287 cases and 1032 controls were included for the meta-analysis of ischemic stroke, and five studies with 272 cases and 503 controls were included for poststroke depression. The results indicated that the GG genotype of brain-derived neurotrophic factor is related to a significantly lower risk of ischemic stroke in the homozygous and dominant models (odds ratio = .57 and .80, respectively). No significant relation was found between rs6265 and poststroke depression. Conclusions: Thus, brain-derived neurotrophic factor rs6265 might be recommended as a predictor of susceptibility of ischemic stroke. However, the results of this meta-analysis should be interpreted with caution because of the heterogeneity between studies and low sample size. Further studies are needed to evaluate the associations between rs6265 and poststroke depression, especially in Caucasians, with large sample size. Key Words: Brain-derived neurotrophic factor—ischemic stroke—poststroke depression—rs6265. © 2018 National Stroke Association. Published by Elsevier Inc. All rights reserved.
From the *Department of Anatomy, Histology and Embryology, Science Research Center, Institute of Neuroscience, Changsha Medical University, Changsha, China; †Department of Physiology; and ‡Department of Surgery, Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada. Received November 8, 2017; revision received November 22, 2017 ; accepted January 12, 2018 . Grant support: The present study was supported by the Construct Program of the Key Discipline in Hunan Province, National Science Foundation of China (Grant No. 81670427), the Foundation of Hunan Educational Committee (Grant No. 15C0146) to M.-H.B. Heart and Stroke Foundation of Canada (G-13-0003069), a Canadian Institutes of Health Research (CIHR) China-Canada Joint Health Research Initiative (CIHR, FRN No. 132571) and Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery Grants (RGPIN-2016-04574) to H.S.S.; and the CIHR (PJT152155) to Z.-P.F. Author contributions: M.-H.B., H.-S.S., and Z.-P.F. conceived the project and designed the experiments; M.-H.B., S.-Z.Z., and X.-Z.G. performed the data extraction and analysis; M.-H.B. and Z.-P.F. wrote and M.-H.B., H.-S.S. and Z.-P.F. revised the manuscript. All authors read the manuscript. Address correspondence to Hong-Shuo Sun, MD, MSc, PhD, and Zhong-Ping Feng, MD, MSc, PhD, Department of Physiology, Faculty of Medicine, University of Toronto, 1 King’s College Circle, Toronto, Ontario M5S 1A8, Canada. E-mail: [email protected] and [email protected]. 1052-3057/$ - see front matter © 2018 National Stroke Association. Published by Elsevier Inc. All rights reserved. https://doi.org/10.1016/j.jstrokecerebrovasdis.2018.01.010
Journal of Stroke and Cerebrovascular Diseases, Vol. ■■, No. ■■ (■■), 2018: pp ■■–■■
1
ARTICLE IN PRESS M.-H. BAO ET AL.
2
Introduction Ischemic stroke is a multifactorial neurologic injury that causes mortality and disability worldwide. According to the Global Burden of Disease (GBD) 2013 study, stroke was the second most common cause of death worldwide, which accounts for about 11.8% of all death in 2013. Moreover, the absolute number of people who died or disabled by ischemic stroke has increased significantly from 1990 to 2013.1 Poststroke depression (PSD) is the most important neuropsychiatric consequence of stroke. Almost one third of stroke survivors experience major depression. Understanding the risk factors, improving the stroke care is likely to be very helpful for ischemic stroke and PSD prediction and treatment. Brain-derived neurotrophic factor (BDNF) belongs to the neurotrophin family, which plays key roles in regulating neuron survival and differentiation during development.2,3 BDNF is widely expressed in hippocampus, cortex, and basal forebrain. The activation of BDNF facilitates the synaptic plasticity, neurogenesis, and memory process through Tropomyosin receptor kinase B (TrkB) signal.2,4,5 In stroke, BDNF is induced and activated, and is considered necessary for stroke recovery.6 The mutations of BDNF gene cause abnormal BDNF/TrkB signal and reduce BDNF production, therefore, leading to poor recovery after ischemic stroke.7 Single-nucleotide polymorphisms (SNPs) are mutations in a single nucleotide of a specific position in the genome. The SNPs may change the amino acid sequence of the protein or affect the gene expression. Many SNPs have been found in the gene sequence of BDNF. Among these SNPs, rs6265, also named Val66Met, changes the G allele to A at position 196 on BDNF gene. The G allele encodes Val, whereas the A allele encodes Met. The mutation affects the normal translation and intracellular trafficking of BDNF mRNA, which decreases BDNF production and secretion in neurons.7-9 In recent years, several studies demonstrated a higher risk of ischemic stroke for the allele A of rs6265,10,11 whereas other studies did not.12,13 In this study, we performed the meta-analysis to give a more precise and comprehensive estimation of the association between BDNF rs6265 and ischemic stroke, PSD.
Methods Publication Search Strategy and Inclusion Criteria A comprehensive literature search was conducted in the present meta-analysis. For the association between BDNF polymorphism and ischemic stroke, the following terms were used (“ischemic stroke”) and (“BDNF” or “brainderived neurotrophic factor”) and (“polymorphism” or “mutation” or “variant” or “SNP” or “single nucleotide polymorphism”). For the association between BDNF polymorphism and poststroke depression, the terms (“post-stroke depression” or “PSD”) and (“BDNF” or
“brain-derived neurotrophic factor”) and (“polymorphism” or “mutation” or “variant” or “SNP” or “single nucleotide polymorphism”) were used. The literature searching was without restriction on language. The deadline for publication was July 9, 2017. We searched the published studies in the electronic databases: PubMed, Embase, Cochrane Central Register of Controlled Trials (CENTRAL), Chinese National Knowledge Infrastructure (CNKI), and Chinese Biomedical Literature Database (CBM). After removing of duplications, all the results from the databases were screened by the title and abstract first. If the title and abstract reach our criteria, the full text will be retrieved and read. The references of all eligible studies were retrieved manually for other potentially relevant studies. Inclusion criteria include (1) case–control studies; (2) studies demonstrated the association between BDNF polymorphisms and ischemic stroke or post-stroke depression; and (3) sufficient data within the individual publication. Studies were excluded if any of the following applies: (1) repeat publications, abstracts, dissertations or reviews; and (2) studies not meeting all of the inclusion criteria.
Data Extraction Information from each eligible publication was extracted manually by 2 investigators independently (M.H.B. and S.-Z.Z.). The following information was extracted from each eligible study: first author’s name, published year, country, ethnicity, genotype method, sex, age, the number of subjects. The discrepancies during data extraction were solved by consensus achieved by the third author (X.-Z.G.).
Quality Assessment Quality assessment was conducted according to the predefined scale provided by Jiang et al.28 The score scale includes source of cases, source of controls, specimens used for genotype, total sample size, and Hardy– Weinberg equilibrium (HWE). The scores range from 0 to 15. The low-quality studies were defined as quality score less than 10, whereas others were treated as high quality. The quality evaluation was performed by 2 authors independently (M.-H.B. and X.-Z.G.). Discrepancies were solved by holding a consensus.
Statistical Methods The HWE of the control group was evaluated by χ2 test, with P less than .05 considered as HWE deviated. The pooled odds ratio (OR), with 95% confidence interval (CI), was used to calculate the association between BDNF rs6265 and ischemic stroke, PSD risk. Genetic models such as homozygous (GG versus AA), heterozygous (GG versus GA), dominant (GG versus GA + AA), and allelic models (G versus A) were evaluated. Z test with P less than .05 was used to determine the statistical signifi-
ARTICLE IN PRESS META-ANALYSIS FOR RS6265 AND ISCHEMIC STROKE, PSD
3
cance of the associations. We also conducted subgroup analysis by ethnicity and quality score. To evaluate the heterogeneity of the included studies, I2 test was performed. I2 greater than 50% was considered to be with obvious heterogeneity, and the random-effects model (the DerSimonian and Laird method) was used for the metaanalysis. Otherwise, the fixed-effects model (the Mantel– Haenszel method) was used. The sensitivity analysis was performed by evaluating the influence of each study on the pooled ORs and 95% CIs. One single study at a time was omitted using STATA 12.0 software in the sensitivity analysis. We also evaluated the publication bias using Begg’s funnel plot and Egger’s linear regression method. All statistical analysis was performed using the STATA 12.0 software (StataCorp, College Station, TX) and RevMan 5.3 (Cochrane Community, London, UK).
Results Characteristics of Eligible Studies For the rs6265 and ischemic stroke analysis, a total of 55 studies were obtained without duplications after a comprehensive literature search. Among them, 37 studies were removed because of being irrelevant, reviews, dissertations, not case–control design, without enough data, or other languages than English and Chinese. At last, 7 studies were found eligible, with 1287 cases and 1032 controls. For the rs6265 and PSD, a total of 30 studies were obtained without duplications. Among them, 24 studies were excluded because of irrelevance, review, or dissertations. At last, 5 studies with 272 cases and 503 controls were included for the rs6265 and PSD analysis. The PRISMA flowcharts were shown in Figures 1 and 2, and the information about the included studies was presented in Table 1.
Figure 1. PRISMA flowchart for the association between BDNF rs6265 and ischemic stroke.
Results of Meta-Analysis As shown in Table 2 and Figure 3, 7 studies with 1287 cases and 1032 controls were included in the present metaanalysis. A significantly lower risk of ischemic stroke was found for the GG genotype in the homozygous model and dominant model in overall analysis (GG versus AA, OR = .57, 95% CI = .43-0.75, P < .0001; GG versus GA + AA, OR = .80, 95% CI = .65-0.98, P = .03). Subgroup analysis by ethnicity found lower susceptibility of ischemic stroke in Asians in all genetic models except heterozygous (GG versus AA, OR = .58, 95% CI = .43-0.77, P = .0002; GG versus GA + AA, OR = .78, 95% CI = .62-0.97, P = .03; G versus A, OR = .77, 95% CI = .67-0.89, P = .0003). We also conducted subgroup analysis by quality score. The results indicated there was no significant difference between highquality study group and low-quality group in homozygous, heterozygous, and dominant models. For the allelic model, we only found a significantly lower risk of ischemic stroke in the low-quality group.
Figure 2. PRISMA flowchart for the association between BDNF rs6265 and PSD.
4
Table 1. Characteristics of eligible studies included in the meta-analysis BDNF rs6265 and ischemic stroke
Year
Country
Ethnicity
Genotyping methods
Gao et al10 Keshavarz et al14 Lu et al15 Vilyanov et al16 Wang et al13 Zhao et al11 Zhou et al12
2015 2016 2013 2012 2010 2013 2011
China Iran China Russia China China China
Asian Caucasian Asian Caucasian Asian Asian Asian
PCR-RFLP PCR-RFLP PCR-RFLP NA PCR-RFLP PCR-RFLP PCR-RFLP
171/119:127/73 107/99:129/71 83/71:25/21 NA 24/25:26/20 300/194:185/152 NA
Age (case/control)
Quality score
61.4 ± 8.6:60.8 ± 7.9 64.5 ± 11.7:62.1 ± 6.7 70.0 ± 12.2:68.0 ± 12.4 NA 59.9 ± 9.8:59.7 ± 9.5 69.8 ± 11.3:68.9 ± 9.99 NA
9 12 12 6 11 13 11
G/G
G/A
A/A
Sample size (case/control)
(Case/ Control)
(Case/ Control)
(Case/ Control)
HWE of control
290/200 206/200 154/46 64/110 49/46 494/337 30/93
61/57 0/2 71/22 43/72 10/11 119/96 10/34
145/112 50/18 55/18 17/35 31/29 247/174 14/41
84/31 156/180 28/6 4/3 8/6 128/67 6/18
.049 .058 .457 .606 .061 .459 .378
G/G
G/A
A/A
(Case/ Control)
(Case/ Control)
ARTICLE IN PRESS
Author
Sex ratio (male/female) (case/control)
BDNF rs6265 and poststroke depression
Author Golimbet17 Kim et al18 Lu et al15 Rezaei et al19 Zhou et al12
Year
Country
Ethnicity
Genotyping methods
2010 2012 2013 2015 2011
Russia Korea China Iran China
Caucasian Asian Asian Caucasian Asian
RFLP-PCR RFLP-PCR PCR RFLP-PCR RFLP-PCR
Sex ratio (male:female) (case/control) NA 43/34:121/78 48/38:35/33 14/34:93/65 19/16:34/24
Age (case/control)
Quality score
Sample size (case/control)
(Case/ Control)
NA 66.5 ± 9.5:63.5 ± 9.4 71 ± 11.1:69 ± 13.3 74.5 ± 8.8:61.5 ± 10.8 61.7 ± 8.5:63.5 ± 12.5
5 9 11 12 11
24/20 77/199 86/68 48/158 35/58
15/16 13/51 41/30 0/0 13/21
GA + AA: 11/4 41/113 23/35 29/26 16/12 19/31 29/127 14/27 8/10
HWE of control NA .044 .143 .172 .794
M.-H. BAO ET AL.
Abbreviations: BDNF, brain-derived neurotrophic factor; RFLP-PCR, Restriction Fragment Length Polymorphism-Polymerase Chain Reaction; HWE, Hardy–Weinberg equilibrium; NA, not available.
ARTICLE IN PRESS META-ANALYSIS FOR RS6265 AND ISCHEMIC STROKE, PSD
5
Table 2. Pooled ORs and 95% CIs of the association between BDNF rs6265 and ischemic stroke, PSD BDNF rs6265 with ischemic stroke Genetic model GG versus AA
GG versus GA
GG versus GA + AA
G versus A
Overall Asian Caucasian High quality Low quality Overall Asian Caucasian High quality Low quality Overall Asian Caucasian High quality Low quality Overall Asian Caucasian High quality Low quality
N
OR [95% CI]
P value
I2 (%)
7 5 2 5 2 7 5 2 5 2 7 5 2 5 2 7 5 2 5 2
.57 [.43, .75] .58 [.43, .77] .38 [.10, 1.47] .66 [.47, .93] .40 [.24, .67] .89 [.71, 1.11] .88 [.69, 1.10] 1.01 [.53, 1.93] .87 [.66, 1.14] .93 [.64, 1.34] .80 [.65, .98] .78 [.62, .97] .97 [.52, 1.81] .81 [.63, 1.05] .77 [.54, 1.09] .93 [.71, 1.23] .77 [.67, .89] 1.50 [.61, 3.71] 1.04 [.71, 1.51] .70 [.56, .88]
<.0001 .0002 .16 .02 .0005 .29 .26 .98 .31 .69 .03 .03 .92 .11 .14 .63 .0003 .38 .84 .003
0 0 0 0 0 0 0 68 0 0 0 0 16 0 32 69 0 82 72 26
BDNF rs6265 with PSD Genetic model GG versus AA
GG versus GA
GG versus GA + AA
G versus A
Overall Asian Caucasian High quality Low quality Overall Asian Caucasian High quality Low quality Overall Asian Caucasian High quality Low quality Overall Asian Caucasian High quality Low quality
N
OR [95% CI]
P value
I2 (%)
4 3 1 3 1 4 3 1 3 1 5 3 2 3 2 4 3 1 3 1
.63 [.37, 1.07] .63 [.37, 1.07] NA .92 [.46, 1.87] .39 [.17, .87] .97 [.63, 1.51] .97 [.63, 1.51] NA 1.21 [.69, 2.14] .70 [.35, 1.42] .82 [.56, 1.21] .87 [.58, 1.30] .34 [.09, 1.31] 1.11 [.67, 1.86] .53 [.29, .97] 1.05 [.64, 1.73] .82 [.61, 1.12] 2.27 [1.22, 4.23] 1.27 [.76, 2.13] .66 [.45, .95]
.09 .09 NA .83 .02 .90 .90 NA .5 .33 .31 .5 .12 .68 .04 .83 .22 .01 .36 .03
25 25 NA 0 N.A 0 0 NA 0 NA 7 9 NA 0 0 74 24 NA 61 NA
Abbreviations: BDNF, brain-derived neurotrophic factor; CI, confidence interval; NA, not applicable; OR, odds ratio; PSD, poststroke depression.
As shown in Table 2 and Figure 4, 5 studies, with 272 cases and 503 controls were included in the present metaanalysis for the association between BDNF rs6265 and PSD. No significant association between rs6265 and PSD was found. As Golimbet17 only provided the data for GA + AA, the data were only pooled in the dominant genetic model.
Source of Heterogeneity No obvious heterogeneity was found for all the included studies in the homozygous, heterozygous, and dominant models. However, in the allelic model for both ischemic stroke and PSD, high heterogeneity was found
ARTICLE IN PRESS M.-H. BAO ET AL.
6
Figure 3.
Forest plots of odds ratio for the association between BDNF rs6265 and risks of ischemic stroke.
(I2 % = 69% for ischemic stroke and 74% for PSD). Subgroup analysis by ethnicity showed big heterogeneity among Caucasians. Specifically, Iranians14,19 are much more diverse as compared with Russians.10,11,16 Therefore, the ethnicity might be the source of heterogeneity.
Sensitivity We evaluated the influence of each individual on pooled ORs and 95% CIs by sensitivity analysis. No obvious influence was found after omitting any individual study
ARTICLE IN PRESS META-ANALYSIS FOR RS6265 AND ISCHEMIC STROKE, PSD
Figure 4.
7
Forest plots of odds ratio for the association between BDNF rs6265 and risks of PSD.
in all genetic models. Figure 5 shows the influence of individual study on pooled ORs and 95% CI in the dominant genetic model for ischemic stroke and PSD.
Publication Bias Egger’s and Begg’s tests were used to evaluate the publication bias of all included studies. As shown in Table 3 and Figure 6, we observed an asymmetry in the funnel plot. A significant bias also existed in Egger’s test (P = .047).
Discussion In this meta-analysis, we found a significantly lower risk of ischemic stroke for BDNF GG genotype in the homozygous and dominant models (OR = .57 and OR = .80, respectively). Subgroup analysis by ethnicity discovered a drastically decrease in risk of ischemic stroke in Asian in the homozygous, dominant, and allelic models (OR = .58, .78, and .77, respectively). No significant relations were found between rs6265 and PSD.
ARTICLE IN PRESS M.-H. BAO ET AL.
8
Figure 5. Sensitivity analysis of the influence of each study on pooled odds ratios and 95% confidence interval (CIs) in GG versus GA + AA model for associations of ALDH2 rs6265 and ischemic stroke, poststroke depression (PSD).
BDNF plays important roles in the neural system. It supports the survival of existing neurons and helps to stimulate neurogenesis.20,21 In patients with ischemic stroke, a decrease in serum BDNF concentration has been observed. The decrease in BDNF is also observed in ischemic stroke–related diseases, such as atherosclerosis, diabetes, and metabolic syndrome.22,23 Moreover, decreased BDNF is associated with poor recovery of stroke, poorer memory, and abnormal hippocampal activation.7,9,23 The mechanisms for BDNF include at least 2 receptors, TrkB and p75.24 The rs6265 mutation changes G allele to A allele in position 196 of BDNF gene, which switches the amino acid valine to methionine at codon 66, Val66Met. BDNF rs6265 is relatively common in humans, with a prevalence of 20%-30% for heterozygous and about ~4% for homozygous.25 The Met allele decreases the activitydependent secretion and intracellular distribution of BDNF by affecting intracellular trafficking of BDNF.7,9 The trafficking of BDNF from the biosynthetic pathway appears complex and highly regulated; however, the detailed process remains unclear. BDNF is secreted by neurons as a 32-kDa pro-BDNF that is changed to mature form rapidly after secretion.9 The pro-BDNF is secreted by vesicle-mediated processes, where the secretory vesicles are transported from postsynaptic dendrites to the
somatodendritic compartment.26 The rs6265 located in the 5′ pro-BDNF sequence blocks the trafficking of BDNF to secretory granules but does not affect the function of the mature protein.7 Transfection of rodent hippocampal neurons with val- or met-BDNF found val-BDNF secretory granules spread to the synapses, whereas metBDNF aggregates are accumulated in the cell body.7 This may hamper the local and synapse-specific regulation by BDNF, leading to impairment of the function of hippocampal neurons and the neurons in other parts of the brain. Several studies have found significant relations between rs6265 and ischemic stroke occurrence.10,11 Our metaanalysis found a significantly lower ischemic stroke risk for GG genotype in both the homozygous and the dominant models. This might be interpreted by the impaired secretion of BDNF for the A allele (met-BDNF). PSD is a common consequence of stroke. Biological mechanisms, such as neural circuits involved in mood regulation, and psychological stressors are considered to be the etiology of PSD.27 Genetic variants are thought to be involved in the development of PSD. BDNF is related to the repair of the neuronal system after stroke. Therefore, the Val66Met mutation of BDNF might relate to the PSD. However, our meta-analysis did not find any relations between BDNF rs6265 and PSD.
Table 3. Begg’s and Egger’s test for funnel plot asymmetries BDNF rs6265 and ischemic stroke, PSD (P value) Group
GG versus AA
GG versus GA
GG versus GA + AA
G versus A
Begg’s test Egger’s test
.283 .094
.283 .271
.436 .236
.213 .047
Abbreviations: BDNF, brain-derived neurotrophic factor; PSD, poststroke depression.
ARTICLE IN PRESS META-ANALYSIS FOR RS6265 AND ISCHEMIC STROKE, PSD
9 4
types are limited in most of the involved studies. Ischemic stroke and PSD are influenced by both genetic and environmental factors. The gene–gene and gene–environment interactions may play important roles in the functions of these polymorphisms, but most studies lack information on environmental exposure and multiple SNPs in haplotypes.
Conclusions
Figure 6. Begg’s funnel plot for the included studies for G versus A genotype of rs6265.
Intriguingly, the BDNF allele distribution in Iranians is quite different from those of Asians and other Caucasians. Iranians have more A allele than G allele. The frequency of G allele was 12.1% in Iranians patients with ischemic stroke and 5.5% in healthy Iranians,14,19 which are quite low than those in Asians (G allele greater than 46%)10 and in the U.S. population (G allele greater than 80.0%).7 Moreover, the G allele in Iranian is associate with higher risk of ischemic stroke and PSD.14,19 This difference causes a big heterogeneity for the studies of Iranians in respect to other ethnicities. Increased heterozygosity, selected patients with ischemic stroke, and presence of zero G allele in the studies are thought to contribute to the heterogeneity.14 Considering the difference of Iranian from other people, our meta-analysis results should be interpreted carefully for Iranians. We found a significant publication bias for all the involved studies in allelic model in Egger’s test (G versus A, P = .047). An asymmetry has also been found in funnel plots (Fig 6). As the Iranians have an obvious heterogeneity as compared with other ethnicities, we omitted the study on Iranians. We found the funnel plots become symmetric when the study was excluded. Therefore, we assume that this asymmetry was caused by a true heterogeneity existing in the involved studies. One of the limitations of the present meta-analysis is that the number of cases and controls were limited.1 In the present meta-analysis, only 7 studies with 1287 cases and 1032 controls were included for the meta-analysis for ischemic stroke, and 5 studies with 272 cases and 503 controls were included for the analysis of PSD. The small sample size may cause an overestimation of real relations.2 Most of the studies were on Asians. In the present metaanalysis, 77.9% of cases and 68.2% of controls are Asians. Moreover, the Iranians are with a different distribution of rs6265 genotype; and with different ischemic stroke or PSD risk for each genotype. Therefore, our results should be interpreted according to ethnicity.3 The disease sub-
In conclusion, the current meta-analysis suggests a lower risk of ischemic stroke for the GG genotype of BDNF rs6265 in both overall subjects and Asians. No significant association was found between rs6265 and PSD. However, the results of this meta-analysis should be interpreted with caution because of the heterogeneity between the involved studies. Further studies are needed to evaluate the associations between BDNF rs6265 and PSD, especially in Caucasians, with a large sample size.
References 1. Feigin VL, Norrving B, Mensah GA. Global burden of stroke. Circ Res 2017;120:439-448. 2. Bramham CR, Messaoudi E. BDNF function in adult synaptic plasticity: the synaptic consolidation hypothesis. Prog Neurobiol 2005;76:99-125. 3. Davies AM. The role of neurotrophins in the developing nervous system. J Neurobiol 1994;25:1334-1348. 4. Yamada K, Nabeshima T. Brain-derived neurotrophic factor/TrkB signaling in memory processes. J Pharmacol Sci 2003;91:267-270. 5. Bath KG, Akins MR, Lee FS. BDNF control of adult SVZ neurogenesis. Dev Psychobiol 2012;54:578-589. 6. Ploughman M, Windle V, MacLellan CL, et al. Brainderived neurotrophic factor contributes to recovery of skilled reaching after focal ischemia in rats. Stroke 2009;40:1490-1495. 7. Egan MF, Kojima M, Callicott JH, et al. The BDNF val66met polymorphism affects activity-dependent secretion of BDNF and human memory and hippocampal function. Cell 2003;112:257-269. 8. Baj G, Carlino D, Gardossi L, et al. Toward a unified biological hypothesis for the BDNF Val66Met-associated memory deficits in humans: a model of impaired dendritic mRNA trafficking. Front Neurosci 2013;7:188. 9. Chen ZY, Patel PD, Sant G, et al. Variant brain-derived neurotrophic factor (BDNF) (Met66) alters the intracellular trafficking and activity-dependent secretion of wild-type BDNF in neurosecretory cells and cortical neurons. J Neurosci 2004;24:4401-4411. 10. Gao LH, Zheng F, Wang JQ, et al. Correlation of brainderived neurotrophic factor gene polymorphism with prognosis in patients with acute ischemic stroke. Chin J Modern Med 2015;25:63-67. 11. Zhao J, Wu H, Weng YF. Study of correlation between genetic polymorphism of brain-derived neurotrophic factor and 90-day outcome of ischemic stroke in Chinese Han population. Chin J Clin Med 2015;503-509. 12. Zhou Z, Lu T, Xu G, et al. Decreased serum brain-derived neurotrophic factor (BDNF) is associated with post-stroke depression but not with BDNF gene Val66Met polymorphism. Clin Chem Lab Med 2011;49:185-189.
ARTICLE IN PRESS 10 13. Wang DX, Huang D, Zhang GH, et al. Association of brain-derived neurotrophic factor polymorphisms with vascular cognitive impairment. Chin J Geriatr Heart Brain Vessel Dis 2010;37-39. 14. Keshavarz P, Saberi A, Sharafshah A, et al. Association of BDNF G196A gene polymorphism with ischemic stroke occurrence and its 6-month outcome in an Iranian population. Top Stroke Rehabil 2016;23:254-260. 15. Lu LP, Zhu YB, Chen JY. The correlation between Val66Met polymorphism in brain-derived neurotrophic factor and post-stroke depression. Chin J Clin Neurosci 2013;21:413-417. 16. Vilyanov VB, Skipetrova LA, Remennik AY, et al. Dynamics of BDNF blood serum content and polymorphism of BDNF VAL166MET in patients with consequences of ischemic, hemorrhagic stroke and craniocerebral trauma. Eur J Nat History 2012;19-20. 17. Golimbet VE. Effects of the interaction of variants of the serotonin transporter and brain-derived neurotrophic factor on platelet serotonin levels in stroke patients. 41(5). United States, 2011. 18. Kim JM, Stewart R, Kang HJ, et al. A longitudinal study of BDNF promoter methylation and genotype with poststroke depression. J Affect Disord 2013;149:93-99. 19. Rezaei S, Asgari MK, Saberi A, et al. Brain-derived neurotrophic factor (BDNF) Val66Met polymorphism and post-stroke dementia: a hospital-based study from northern Iran. Neurol Sci 2016;37:935-942. 20. Acheson A, Conover JC, Fandl JP, et al. A BDNF autocrine loop in adult sensory neurons prevents cell death. Nature 1995;374:450-453.
M.-H. BAO ET AL. 21. Huang EJ, Reichardt LF. Neurotrophins: roles in neuronal development and function. Annu Rev Neurosci 2001;24:677-736. 22. Pikula A, Beiser AS, Chen TC, et al. Serum brain-derived neurotrophic factor and vascular endothelial growth factor levels are associated with risk of stroke and vascular brain injury: Framingham Study. Stroke 2013;44:2768-2775. 23. Stanne TM, Åberg ND, Nilsson S, et al. Low circulating acute brain-derived neurotrophic factor levels are associated with poor long-term functional outcome after ischemic stroke. Stroke 2016;47:1943-1945. 24. Patapoutian A, Reichardt LF. Trk receptors: mediators of neurotrophin action. Curr Opin Neurobiol 2001;11:272280. 25. Neves-Pereira M, Mundo E, Muglia P, et al. The brainderived neurotrophic factor gene confers susceptibility to bipolar disorder: evidence from a family-based association study. Am J Hum Genet 2002;71:651-655. 26. Kojima M, Takei N, Numakawa T, et al. Biological characterization and optical imaging of brain-derived neurotrophic factor-green fluorescent protein suggest an activity-dependent local release of brain-derived neurotrophic factor in neurites of cultured hippocampal neurons. J Neurosci Res 2001;64:1-10. 27. Spalletta G, Bossù P, Ciaramella A, et al. The etiology of poststroke depression: a review of the literature and a new hypothesis involving inflammatory cytokines. Mol Psychiatry 2006;11:984-991. 28. Jiang DK, Wangg WZ, Ren WH, et al. TP53 Arg72Pro polymorphism and skin cancer risk:a meta-analysis. J Invest Dermatol 2011;131:220-228.