Association of Variants in KCNK17 Gene with Ischemic Stroke and Cerebral Hemorrhage in a Chinese Population Lingbin He, MD,*†1 Qingfeng Ma, MD,‡1 Yongqin Wang, MS,†x1 Xin Liu, PhD,† Yuan Yuan, MD,*† Yongzhi Zhang, MD,*† Wenjing Ou, MPH,†k Lisheng Liu, MD,† Xuerui Tan, MD, PhD,** and Xingyu Wang, PhD*†{

Background: KCNK17 (potassium channel, subfamily K, member17) has a role in the pathogenesis of stroke. We reported previously that rs10947803 singlenucleotide polymorphism (SNP) in KCNK17 is associated with cerebral hemorrhage in a Chinese population. The aim of the present study was to examine other SNPs in the KCNK17 gene that are associated with cerebral hemorrhage and other subtypes of stroke in the Chinese population. Methods: A total of 1356 subjects with stroke and 1225 control patients were examined by a case-control methodology. The SNPs (rs12214600, rs12195376, rs2758912, and rs10807204) in KCNK17 gene were genotyped with the TaqMan real-time polymerase chain reaction assay. Results: rs12214600 SNP in KCNK17 was significantly associated with cerebral hemorrhage (unadjusted odds ratio 5 .55, 95% confidence interval 5 .35-.86, P 5.008, q 5 .0328) under the allele model. After adjusting for age, sex, and hypertension, we found that the association remained significant (odds ratio 5 .56, 95% confidence interval 5 .35-.90, P 5 .0158). There was no association detected for other SNPs in KCNK17 with cerebral hemorrhage, and none of the SNPs in KCNK17 had an association with ischemic stroke. Conclusions: The T carrier of an SNP (rs12214600) is associated with reduced risk of cerebral hemorrhage in the Chinese population, together with previous findings that SNPs rs10947803 and rs12214600 in the KCNK17 gene are associated with hemorrhagic stroke, but none of the SNPs tested had an association with ischemic stroke. KCNK17 may be important in the pathogenesis of cerebral hemorrhage. Key Words: Ischemic stroke—cerebral hemorrhage—KCNK17—SNP. Ó 2014 by National Stroke Association

From the *The First Affiliated Hospital, Medical College of Shantou University, Shantou, Guangdong, China; †Laboratory of Human Genetics, Beijing Hypertension League Institute, Beijing, China; ‡Department of Neurology, Xuanwu Hospital, Beijing, China; xSchool of Basic Courses, Baotou Medical College, Baotou, Neimenggu, China; kPublic Health School; Harbin Medical University, Harbin, Heilongjiang, China; and {National Research Institute for Family Planning, Beijing, China. Received March 18, 2014; revision received April 10, 2014; accepted April 16, 2014. Supported by the Beijing Hypertension League Institute, in part through an unrestricted educational grant from F. Hoffmann-La Roche, the National Infrastructure Program of Chinese Genetic Re-

2322

sources (2005DKA21300), Ministry of Science and Technology International Collaboration Project (No.2012DFB30130). and Specialized Research Fund for the Doctoral Program of Higher Education (No.20114404110006). Address correspondence to Xingyu Wang, PhD, the First Affiliated Hospital, Medical College of Shantou University, Shantou, Guangdong, 515041 China, or Laboratory of Human Genetics, Beijing Hypertension League Institute, 24 Shijingshan Road, Beijing, 100043 China. E-mail: [email protected]; [email protected]. 1 L. He and O. Ma contributed equally to this manuscript. 1052-3057/$ - see front matter Ó 2014 by National Stroke Association http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2014.04.029

Journal of Stroke and Cerebrovascular Diseases, Vol. 23, No. 9 (October), 2014: pp 2322-2327

ASSOCIATIONS OF KCNK17 WITH STROKE

Introduction Stroke is one of the most common causes of death worldwide and contributes to a major burden in health care.1,2 But the prevalence, prophylaxis, and treatment in stroke may be different in various regions or race. Some researchers have found that there are some differences between Japanese and European Guidelines for the management of ischemic stroke published between 2008 and 2011.3 In China, stroke is the leading type of cardiovascular disease.4 Ischemic stroke has the dominant proportion, which is approximately 43.7%-78.9% of overall strokes in Chinese population.5 Cerebral hemorrhage is a significant subtype of stroke and plays an important role in the Chinese population.6 Recent epidemiologic studies in the Chinese population reveal that approximate 30% of stroke cases, even much more, can be accounted for by hemorrhagic stroke, which indicates that hemorrhagic stroke is more frequent in Chinese than in the white patients.7,8 As a complex disease, stroke is assumed to result from both genetic and environmental factors as well as their interactions9; however, its responsible genetic and molecular mechanisms have remained largely unknown. KCNK17 (potassium channel, subfamily K, member 17) is located in chromosome 6p21.1 in humans and belongs to 1 of the 2-pore-domain potassium channels superfamily of background K1 channels, which generate the negative membrane potential in excitable and nonexcitable cells and contribute to the resting membrane potential in various tissues.10,11 The mutations of KCNK17 gene could bring about abnormal opening of potassium channels and then increase the concentration of extracellular potassium ion, which might produce marked dilation of cerebral blood vessels. It could play an important role in cardiovascular diseases. Since 2007, Matarin et al12 have identified 3 significant single-nucleotide polymorphisms (SNPs) in KCNK17 that are associated with ischemic stroke in white patients. After that, Domingues-Montanari et al13 replicated the finding in a Spanish cohort; however, different results, which indicate that the SNP in KCNK17 is not associated with ischemic stroke, have been reported in Chinese population.14 Ma et al15 reported that the rs10947803 SNP in KCNK17 is associated with cerebral hemorrhage in a Chinese population; however other SNPs in KCNK17 were been tested in that report. It is necessary to investigate whether the other SNPs in KCNK17 gene are associated with cerebral hemorrhage in a Chinese population. Therefore, we selected 4 representative SNPs (rs12214600, rs12195376, rs2758912, and rs10807204) by their minor allele frequency greater than 10% in Chinese with Haploview software (version 4.0)16 in the block located from 39376714 bp to 39383260 bp in chromosome 6 and tested the susceptible variants in KCNK17 associated with cerebral hemorrhage and other subtypes of stroke in the Chinese population.

2323

Materials and Methods Subjects The Stroke Hypertension Investigation in Genetics (SHINING) study was conducted by the Beijing Hypertension League Institute. Patients and control subjects from 6 geographical regions in China were recruited for the case–control study between 1997 and 2000, and 70% subjects came from or nearby Beijing.17 The SHINING study consisted of the subjects of Chinese Han ethnicity only.18 Any stroke patients who suffered a stroke within the past 5 years were eligible to participate in the study.17,18 All patients had medical records that indicated with a diagnosis of stroke confirmed by brain computed tomography/magnetic resonance imaging. In the initial study, patients with ischemic stroke, cerebral hemorrhage, ischemic/hemorrhage stroke, subarachnoid hemorrhage, transient ischemic attacks (TIA), and cryptogenic stroke were included.15 In considering the sample size of each subtype of stroke, we chose only 3 subtypes for the present study, namely, ischemic stroke, cerebral hemorrhage, and TIA. Control subjects were selected according to the case–control study criteria (control subjects matched to cases by sex, age within 3 years, geographic location, and blood pressure category [,140/90 mmHg, $140/90 mmHg, and #180/105, .180/105 mmHg]).18 Data collected included age, sex, body mass index, systolic blood pressure, diastolic blood pressure, and hypertension. Hypertension was defined as having current or past antihypertensive medication, systolic blood pressure $140 mmHg, or diastolic blood pressure $90 mmHg.17,18 Written informed consent was obtained from all study participants, and the study was approved by ethics committees of the Beijing Hypertension League Institute.

Selection of SNPs After excluding the SNPs that are not representative and with less than 10% MAF in Chinese, we selected 4 SNPs (rs12214600, rs2758912, rs12195376, and rs10807204) within the position of chromosome six from 39376714 bp to 39383260 bp, which were representative SNPs in Chinese.

DNA Extraction and Genotyping Genomic DNA had been extracted from the whole blood with salting out procedure. DNA concentration was measured by NanoDrop 2000. DNA was diluted to a concentration of 5 ng/mL before we genotyped the samples. A total of 3119 participants (1559 stroke cases and 1560 controls) were recruited for the SHINING study, and of these, 2581 participants (1356 stroke cases and 1225 controls) had DNA samples available in the present study. All the selected SNPs in

L. HE ET AL.

2324

.527 .118 85 .453 94 .314 16 6 Abbreviations: HWE, Hardy–Weinberg equilibrium; MAF, minor allele frequency; SNP, single-nucleotide polymorphism; TIA, transient ischemic attacks.

45 47 24 41 .920 .507 85 .406 94 .532 14 25 41 50 30 19 .584 .617 85 .165 94 .128 3 1 22 22 60 71 .842 .549 1 1 15 23 69 70

85 .100 94 .133

.308 .399 174 .339 182 .390 17 25 84 92 73 65 .195 .458 174 .537 182 .500 46 43 95 96 33 43 .842 .004 174 .138 182 .137 3 8 42 34 129 140 .647 .138 4 0 50 36 120 146

174 .167 182 .099

.484 .522 354 470 142 966 .390 417 516 147 1080 .375 .791 .522 253 479 234 966 .490 287 549 243 1079 .479 .233 .578 966 .139 1079 .150 23 22

C/C C/T T/T

721 222 777 280

sum MAF

P

.888 .488 966 .135 1080 .149 17 21 723 226 780 279

Ischemic stroke Control Case Cerebral hemorrhage Control Case TIA Control Case

P C/C C/G G/G sum MAF A/A A/G G/G sum MAF

HWE

P sum MAF C/C C/T T/T

Clinical data about continuous variables were expressed as mean 6 SD and compared by t-tests. Univariate associations were explored with frequency tables and Pearson’s c2 tests for independent proportions. The presence of Hardy-Weinberg equilibrium per SNP was tested by an exact c2 good-of-fit test. Concerning that the entire study group comprised 3 major types of stroke (ischemic stroke, cerebral hemorrhage, and TIA), in each stratum, cases were compared with those of the corresponding control groups. Unconditional logistic regression was used to estimate odds ratio (OR) and 95% confidence interval (CI) and adjusted OR for sex, age, and hypertension under the allelic model. The false discovery rate adjustment for multiple testing probability value was given by the q value method using the PROC MULTTEST procedure in SAS (SAS Institute, Cary, NC). Linkage disequilibrium analysis was conducted with the online software SNPStats.19 Power calculation by the Quanto software program (version 1.2.4; University of Southern California, Los Angeles, CA) indicated that we have more than 75% of detecting positive association under dominant model, assuming the effect size is .55 for rs12214600 with MAF 5 .167 in the subtype of cerebral hemorrhage in SHINING cohort. All other statistical analyses were performed with SAS statistical software (version 9.2; SAS Institute Inc.). P ,.05 was used to indicate statistical significance.

HWE

Statistical Analysis

HWE

KCNK17 were genotyped by TaqMan Real-Time PCR (Assay ID: C_1650898_20 for rs12214600, C_31340453_20 for rs12195376, C_16289539_20 for rs2758912, C_1650893_10 for rs10807204; Applied Biosystems, Inc., Foster City, CA) using the ABI PRISM 7900HT Sequence Detection System (Applied Biosystems, Inc.). The average genotyping call rate was 99.96%.

rs2758912 (A/G)

Abbreviations: BMI, body mass index; DBP, diastolic blood pressure; SBP, systolic blood pressure. Age, BMI, DBP, and SBP values are represented by mean 6 SD. Hypertension indicates systolic blood pressure $140 mmHg or diastolic blood pressure $90 mmHg (or both), or taking antihypertensive medication.

P

rs10807204 (C/G)

.029 .680 .000 .716 .146 .012

rs12195376 (C/T)

Number of 1356 1225 subjects Age, y 60.5 6 10.2 61.9 6 10.5 Sex, % male 60.1 60.9 BMI, kg/m2 24.4 6 3.0 25.1 6 3.3 SBP, mmHg 144.9 6 23.2 144.6 6 23.3 DBP, mmHg 87.1 6 12.6 86.4 6 12.8 Hypertension, % 79.9 75.8

P-value

rs12214600 (C/T)

Controls

Table 2. The distribution of genotypes of all the SNPs in KCNK17 in the present study

Stroke patients

HWE

Table 1. Characteristics of the study participants

ASSOCIATIONS OF KCNK17 WITH STROKE

2325

.96 (.85-1.08) .5001

.94 (.83-1.06) .3155

G

G

Abbreviations: CI, confidence interval; FDR, false discovery rate; OR, odds ratio; TIA, transient ischemic attacks. *The q value is equal to the P value after multi-testing correction under the unadjusted model. yThe P-value was assessed by logistic regression after we adjusted for sex, year, and hypertension.

.60 (.39-0.94) .0253 .4207

.93 (.81-1.06) .2616 1.25 (.92-1.69) .1577 .3154 1.26 (.92-1.75) .1527 .55 (.36-0.85) .0070 .0280

.84 (.61-1.15) .2720 1.66 (1.09-2.53) .0174 .0348 1.63 (1.06-2.52) .0265

1.10 (.92-1.31) .3011 T

.5001

1.12 (.94-1.34) .2006

rs12214600 (C/T) rs12195376 (C/T) rs2758912 (A/G) rs10807204 (C/G)

T

.4207

.4207

.98 (.86-1.11) .7161 .86 (.64-1.16) .3190 .4253

1.08 (.90-1.29) .4178 1.00 (.65-1.53) .9824 .9874 1.05 (.66-1.66) .8488 .74 (.41-1.34) .3213 .3344

.64 (.35-1.19) .1593

.56 (.35-0.90) .0158 1.38 (.72-2.66) .3344 .3344 1.51 (.75-3.07) .2520 1.13 (.94-1.36) .1934 .55 (.35-0.86) .0082 .0328

FDR q* OR （95%CI) P FDR q* OR 95% CI) P OR 95% CI) OR 95% CI) FDR q*

Adjusted

Py

Unadjusted

Adjusted

Py

OR 95% CI)

Adjusted Unadjusted

TIA Cerebral hemorrhage Ischemic stroke

P OR 95% CI) Minor allele

We carried out the present study in a Chinese Han population to test the association between the SNPs (rs12214600, rs12195376, rs2758912, and rs10807204) of KCNK17 gene and subtypes of stroke. Our results revealed that the rs12214600 SNP in KCNK17 was significantly associated with reduced risk for cerebral hemorrhage and that the association remained significant after adjusting for multi-testing (Table 3). In a previous study, Ma et al15 found that the SNP (rs10947803) in KCNK17 was associated with an increased risk of cerebral hemorrhage in a Chinese population, although

Unadjusted

Discussion

Table 3. The associations between the SNPs in KCNK17 and ischemic stroke, cerebral hemorrhage, and TIA under allelic models

The demographic characteristics of the study participants were shown in Table 1. The means of age and body mass index were greater in control patients than those in cases, and the proportion of hypertension was lower in control patients; however, systolic blood pressure and diastolic blood pressure were not significantly different in between cases and controls because the original goal of SHINING study was to identify SNPs that predispose to stroke independent of blood pressure.17,18 In the present study, the percentage of ischemic stroke (1080 patients), cerebral hemorrhage (182 patients), and TIA (94 patients) in all stroke cases were 79.7%, 13.4%, and 6.9%, respectively. To confirm the genetic susceptibility of the SNPs in KCNK17 with stroke in Chinese population, the SNPs (rs12214600, rs12195376, rs2758912, and rs10807204) were genotyped in 2581 participants, including 1356 cases and 1225 controls. All the SNPs conformed to the Hardy-Weinberg equilibrium in control subjects. The Hardy-Weinberg equilibrium information for the four SNPs was listed in Table 2. We further analyzed the association between the SNPs of KCNK17 with stroke subtypes, shown in Table 3. Our results revealed that the susceptive variant of the rs12214600 (C/T) SNP was significantly associated with cerebral hemorrhage (OR 5 .55, 95% CI 5 .35-.86, P 5 .0082, q 5 .0328). In addition, the SNPs rs2758912 (A/G) and rs10807204 (C/G) were associated with TIA (OR 5 1.66, 95% CI 5 1.09-2.53, P 5 .0174, q 5 .0348 and OR 5 .55, 95% CI 5 .36-.85, P 5 .0070, q 5 .0280, respectively). After we adjusted for age, sex, and hypertension, the significant associations still remained in cerebral hemorrhage (for rs12214600, OR 5 .56, 95% CI 5 .35-.90, P 5 .0158) and TIA (for rs2758912 and rs10807204, OR 5 1.63, 95% CI 5 1.06-2.52, P 5 .0265 and OR 5 .60, 95% CI 5 .39-.94, P 5 .0253, respectively). Linkage disequilibrium analysis is shown in Figure 1. There was no significant linkage disequilibrium existing among the SNPs in KCNK17, containing the 4 SNPs in the present study and the rs10947803 SNP in study of Ma et al.15

Py

Results

L. HE ET AL.

2326

Figure 1. The linkage disequilibrium analysis was conducted with the online software, SNPStats. All the data came from the present study except that of the rs10947803 SNP, which derived from the study by Ma et al. Abbreviation: SNP, single-nucleotide polymorphisms.

this association hadn’t been investigated in the other cohorts or in different ethnic groups. Considering the adjacent rs12214600 and rs10947803 SNPs in chromosome 6 (located within 39376714 bp and 39378588 bp, respectively) showed the associations with cerebral hemorrhage, these results indicate that the KCNK17 gene may indeed play an important role in cerebral hemorrhage in the Chinese population. We also performed a linkage disequilibrium analysis to determinate linkage correlation of these two SNPs (Fig 1). However, these 2 SNPs were not in linkage disequilibrium (D0 5 .036 and r 5 .018). The effect of 2 SNPs on the risk of cerebral hemorrhage are in a different direction; to our knowledge, there is no direct evidence of how rs10947803 and rs12214600 affect the function of KCNK17. Therefore, the underlying mechanism of the association is unclear. In the present study, none of the SNPs tested in KCNK17 were associated with ischemic stroke (Table 3), which was consistent with the previous studies performed in the Chinese population but inconsistent with studies conducted in the white population.12-15 After we adjusted sex, age, and hypertension, the associations remained insignificant (Table 3). The absence of the association in the Chinese Han population may result from the heterogeneity among different ethnic groups. The current study could not be replicated the findings derived from Caucasian population albeit having enough power to detect the associations (data not shown). In addition, we also found that the rs2758912 and rs10807204 SNPs were respectively associated with TIA (Table 3). Although the results were derived from a smaller sample sizes (94 cases and 85 controls), and

passed the multitesting adjustment, it would be worth the effort to replicate the results with larger cohort with TIA phenotype. All the SNPs selected in the present study are located in the intron region of NM_001135111.1 in KCNK17 gene. A recent study has revealed that the KCNK17 gene is abundantly expressed in liver, lungs, placenta, pancreas, small intestine, and aorta. In addition, intermediate levels also are detected in brain, heart, colon, ovary, and so on, but not in skeletal muscle.20 But by now, the molecular mechanism of how these SNPs’ variants affect the functions of KCNK17 still remains unclear. The KCNK17 gene codes the protein TALK-2, or TASK4, which belongs to a member of the 2-pore domain superfamily of K1 channels.10,20 As an open rectifier K1 channel that primarily passes outward current under physiological K1 concentrations, KCNK17 shares many common features with other family members, such as participating in ion fluxes necessary for cell volume regulation as well as in metabolic acidosis and hypertension caused by secretion of HCO32.20-23 Recently, some researchers have found that the expression of KCNK3 potassium channel gene, which codes TASK-1, lessens the injury from ischemic stroke, most likely by a general influence on blood pressure.24 According to the characteristics of the functions and distributions of KCNK genes, they may play important roles in stroke in humans, especially KCNK17 in cerebral hemorrhage in the Chinese population.

Conclusions Our study shows that T carrier of the rs12214600 SNP is associated with reduced risk of cerebral hemorrhage, but none of the 4 SNPs in KCNK17 is associated with ischemic stroke in the Chinese population. Together with our previous findings, the KCNK17 gene may play an important role in cerebral hemorrhage in the Chinese population. To validate these results, further investigations, including more samples and different ethnic groups with ischemic stroke, cerebral hemorrhage, and TIA, and more functional studies are needed. Acknowledgments: We thank Jian Li, Wei Zhang, Beiying Fu, and Yawei Li for their technical assistance.

References 1. Bonow RO, Smaha LA, Smith SC Jr, et al. World Heart Day 2002: the international burden of cardiovascular disease: responding to the emerging global epidemic. Circulation 2002;106:1602-1605. 2. World Health Organization. World health statistics 2008. Geneva, Switzerland: World Health Organization 2008. 3. Kern R, Nagayama M, Toyoda K, et al. Comparison of the European and Japanese guidelines for the management of ischemic stroke. Cerebrovasc Dis 2013;35:402-418.

ASSOCIATIONS OF KCNK17 WITH STROKE 4. Reddy KS. Cardiovascular disease in non-Western countries. N Engl J Med 2004;350:2438-2440. 5. Zhang L, Ding H, Yan J, et al. Genetic variation in cytochrome P450 2J2 and soluble epoxide hydrolase and risk of ischemic stroke in a Chinese population. Pharmacogenet Genomics 2008;18:45-51. 6. Yang QD, Niu Q, Zhou YH, et al. Incidence of cerebral hemorrhage in the Changsha community. A prospective study from 1986 to 2000. Cerebrovasc Dis 2004; 17:303-313. 7. Jiang B, Wang WZ, Chen H, et al. Incidence and trends of stroke and its subtypes in China: results from three large cities. Stroke 2006;37:63-68. 8. Zhang LF, Yang J, Hong Z, et al. Proportion of different subtypes of stroke in China. Stroke 2003;34:2091-2096. 9. Hassan A, Markus HS. Genetics and ischaemic stroke. Brain 2000;123:1784-1812. 10. Girard C, Duprat F, Terrenoire C, et al. Genomic and functional characteristics of novel human pancreatic 2P domain K(1) channels. Biochem Biophys Res Commun 2001;282:249-256. 11. Bayliss DA, Barrett PQ. Emerging roles for two-poredomain potassium channels and their potential therapeutic impact. Trends Pharmacol Sci 2008;29:566-575. 12. Matarin M, Brown WM, Scholz S, et al. A genome-wide genotyping study in patients with ischaemic stroke: initial analysis and data release. Lancet Neurol 2007; 6:414-420. 13. Domingues-Montanari S, Fernandez-Cadenas I, Del RioEspinola A, et al. KCNK17 genetic variants in ischemic stroke. Atherosclerosis 2010;208:203-209. 14. Ding H, Xu Y, Bao X, et al. Confirmation of genomewide association signals in Chinese Han population reveals risk loci for ischemic stroke. Stroke 2010;41:177-180. 15. Ma Q, Wang Y, Shen Y, et al. The rs10947803 SNP of KCNK17 is associated with cerebral hemorrhage but

2327

16.

17.

18.

19.

20.

21.

22.

23.

24.

not ischemic stroke in a Chinese population. Neurosci Lett 2013;539:82-85. Barrett JC, Fry B, Maller J, et al. Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics 2005;21:263-265. Wang X, Cheng S, Brophy VH, et al. A meta-analysis of candidate gene polymorphisms and ischemic stroke in 6 study populations: association of lymphotoxinalpha in nonhypertensive patients. Stroke 2009; 40:683-695. Wu L, Shen Y, Liu X, et al. The 1425G/A SNP in PRKCH is associated with ischemic stroke and cerebral hemorrhage in a Chinese population. Stroke 2009;40:2973-2976. Sole X, Guino E, Valls J, et al. SNPStats: a web tool for the analysis of association studies. Bioinformatics 2006; 22:1928-1929. Decher N, Maier M, Dittrich W, et al. Characterization of TASK-4, a novel member of the pH-sensitive, two-pore domain potassium channel family. FEBS Lett 2001; 492:84-89. Niemeyer MI, Gonzalez-Nilo FD, Zuniga L, et al. Neutralization of a single arginine residue gates open a two-pore domain, alkali-activated K1 channel. Proc Natl Acad Sci U S A 2007;104:666-671. Duprat F, Girard C, Jarretou G, et al. Pancreatic two P domain K1 channels TALK-1 and TALK-2 are activated by nitric oxide and reactive oxygen species. J Physiol 2005;562(Pt 1):235-244. Barriere H, Belfodil R, Rubera I, et al. Role of TASK2 potassium channels regarding volume regulation in primary cultures of mouse proximal tubules. J Gen Physiol 2003;122:177-190. Muhammad S, Aller MI, Maser-Gluth C, et al. Expression of the kcnk3 potassium channel gene lessens the injury from cerebral ischemia, most likely by a general influence on blood pressure. Neuroscience 2010;167:758-764.