Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2014 Jun 1.
Published in final edited form as: Stroke. 2013 May 14;44(6):1609–1615. doi: 10.1161/STROKEAHA.113.679936

17q25 Locus Is Associated With White Matter Hyperintensity Volume in Ischemic Stroke, But Not With Lacunar Stroke Status

Poneh Adib-Samii 1,*, Natalia Rost 1,*, Matthew Traylor 1, William Devan 1, Alessandro Biffi 1, Silvia Lanfranconi 1, Kaitlin Fitzpatrick 1, Steve Bevan 1, Allison Kanakis 1, Valerie Valant 1, Andreas Gschwendtner 1, Rainer Malik 1, Alexa Richie 1, Dale Gamble 1, Helen Segal 1, Eugenio A Parati 1, Emilio Ciusani 1, Elizabeth G Holliday 1, Jane Maguire 1, Joanna Wardlaw 1, Bradford Worrall 1, Joshua Bis 1, Kerri L Wiggins 1, Will Longstreth 1, Steve J Kittner 1, Yu-Ching Cheng 1, Thomas Mosley 1, Guido J Falcone 1, Karen L Furie 1, Carlos Leiva-Salinas 1, Benison C Lau 1, Muhammed Saleem Khan 1; Australian Stroke Genetics Collaborative1,‡; Wellcome Trust Case-Control Consortium-2 (WTCCC2)1,‡; METASTROKE1, Pankaj Sharma 1, Myriam Fornage 1, Braxton D Mitchell 1, Bruce M Psaty 1, Cathie Sudlow 1, Christopher Levi 1, Giorgio B Boncoraglio 1, Peter M Rothwell 1, James Meschia 1, Martin Dichgans 1, Jonathan Rosand 1,†, Hugh S Markus 1,†, on behalf of the International Stroke Genetics Consortium
PMCID: PMC3771337  NIHMSID: NIHMS502894  PMID: 23674528

Abstract

Background and Purpose

Recently, a novel locus at 17q25 was associated with white matter hyperintensities (WMH) on MRI in stroke-free individuals. We aimed to replicate the association with WMH volume (WMHV) in patients with ischemic stroke. If the association acts by promoting a small vessel arteriopathy, it might be expected to also associate with lacunar stroke.

Methods

We quantified WMH on MRI in the stroke-free hemisphere of 2588 ischemic stroke cases. Association between WMHV and 6 single-nucleotide polymorphisms at chromosome 17q25 was assessed by linear regression. These single-nucleotide polymorphisms were also investigated for association with lacunar stroke in 1854 cases and 51 939 stroke-free controls from METASTROKE. Meta-analyses with previous reports and a genetic risk score approach were applied to identify other novel WMHV risk variants and uncover shared genetic contributions to WMHV in community participants without stroke and ischemic stroke.

Results

Single-nucleotide polymorphisms at 17q25 were associated with WMHV in ischemic stroke, the most significant being rs9894383 (P=0.0006). In contrast, there was no association between any single-nucleotide polymorphism and lacunar stroke. A genetic risk score analysis revealed further genetic components to WMHV shared between community participants without stroke and ischemic stroke.

Conclusions

This study provides support for an association between the 17q25 locus and WMH. In contrast, it is not associated with lacunar stroke, suggesting that the association does not act by promoting small-vessel arteriopathy or the same arteriopathy responsible for lacunar infarction.

Keywords: genetics, Genome-wide Association Study, leukoaraiosis, small-vessel disease, stroke

A recent report by the Cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE) consortium identified a novel genetic locus at chromosome 17q25 associated with white matter hyperintensities (WMH) on MRI in stroke-free community individuals.1,2 WMH are common in stroke-free adults, and increase with age. Prospective studies show WMH predicts increased risks of cognitive decline, stroke, and death.3

One might expect risk factors for WMH in community populations to also confer increased risk of WMH in stroke patients. However, the underlying pathology of WMH is heterogeneous; small punctate lesions are associated with mixed causes, whereas more confluent areas correspond primarily to small-vessel disease.4 In stroke patients, WMH are more frequent and extensive compared with healthy age-matched individuals and are usually associated with small-vessel disease.4

We hypothesized that risk factors for WMH in community populations without stroke also confer increased risk of WMH in stroke patients. Therefore, we assessed whether the 17q25 locus is associated with WMH as a quantitative trait in ischemic stroke. If the 17q25 locus acted to increase WMH through promotion of the small-vessel arteriopathy, one might expect it to also increase risk of lacunar stroke, another small-vessel disease phenotype. To test this hypothesis, its association with lacunar stroke status was also examined in a case-control analysis.

Methods and Materials

Association of 17q25 Locus With WMH Volume

Subjects

Stroke cases were recruited from 1 community-based and 8 hospital-based cohorts of ischemic strokes with genome-wide association study (GWAS) data available as well as MRI scans for WMH volume (WMHV) measurement. Details of populations are in Table 1. Inclusion criteria were: >18 years of age, self-reported European ancestry, and a diagnosis of ischemic stroke of any subtype. Exclusion criteria were CADASIL (Cerebral Autosomal Dominant Arteriopathy with Subcortical Infarcts and Leukoencephalopathy), vasculitis, and demyelinating and mitochondrial disorders.

Table 1.

Clinical Characteristics of Ischemic Stroke Cohorts Included in WMH Analysis

Number Mean (SD); age (y) Male (%) Hypertension (%) Hypercholesterolemia (%) Ever Smoker (%) Diabetes (%)
St George’s* 381 70.6 (13.5) 241 (63.2) 300 (78.7) 286 (75.1) 263/380 (69.2) 78 (20.5)
Oxford* 170 67.0 (13.7) 99 (58.2) 114 (67.1) 83 (48.8) 91 (53.5) 22 (12.9)
Edinburgh* 72 68.6 (13.9) 38 (52.7) 34 (47.2) Unknown 51/68 (75.0) 5 (6.9)
Munich* 756 66.6 (12.3) 467 (61.7) 525 (69.4) 346 (45.7) 273 (36.1) 166 (21.9)
Milan 153 57.5 (14.3) 92 (60.1) 87 (56.8) 94 (61.4) 64 (41.8) 21 (13.7)
ISGS 209 67.9 (13.8) 129 (61.7) 127 (60.7) 49/128 (38.2) 51 (24.4) 8/26 (30.7)
ASGC 104 64.8 (13.3) 59 (56.7) 80 (76.9) 52 (50.0) 27/100 (27.0) 18 (17.3)
MGH 975 65.7 (14.2) 606 (62.2) 618 (63.4) 408 (41.8) 606 (62.2) 199 (20.4)
SWISS 115 66.3 (11.4) 56 (48.7) 85 (73.9) Unknown Unknown Unknown
Total 2935 66.4 (14.3) 1734 (60.9) 1911 (65.1) 1318/2667 (49.4) 1426/2811 (50.7) 535/2628 (20.4)
Open in a new tab

ASGC indicates Australian Stroke Genetics Collaborative; ISGS, Ischemic Stroke Genetics Study; MGH, Massachusetts General Hospital; SWISS, Siblings with Ischemic Stroke Study; and WMH, white matter hyperintensities.

*

Part of the Wellcome Trust Case Control Consortium 2 (WTCCC-2).

Clinical Characteristics

Hypertension was defined as antihypertensive prescription before stroke or systolic blood pressure >140 or diastolic >90 mm Hg >1 week after stroke. Hypercholesterolemia was defined as lipid-lowering treatment before stroke or elevated serum cholesterol (>5.2 mmol/L) on stroke admission. Eversmoker was defined as current and exsmokers. Diabetes was defined as a previous diagnosis of diabetes mellitus. Ischemic stroke cases were subtyped according to the Trial of Org 10172 in Acute Stroke Treatment (TOAST)5 or similar criteria (Table I in the online-only Data Supplement).

MRI Analysis

MRI was acquired as part of routine clinical practice in stroke evaluation using different scanners at individual centers. Fluid-attenuated inversion recovery (FLAIR) sequences were primarily used for WMH analysis; however, in their absence, T2 sequences were used (Table II in the online-only Data Supplement).

WMHV was measured in the hemisphere contralateral to the acute infarction to avoid confounding by T2 hyperintense signal attributable to acute stroke. All supratentorial white matter and deep gray matter lesions were included in WMHV, with the exception of WMH corresponding to lacunar infarcts. MRI scans with excessive movement artifact, incomplete brain coverage, or bihemispheric infarcts (other than lacunar) were excluded. To account for normal interindividual variability in head size, an estimate of total intracranial volume (TICV) was derived, using site-specific volumetric methodology.

Scans were anonymized and analyzed blind to genetic information. MRIs from the Massachusetts General Hospital (MGH), Ischemic Stroke Genetics Study (ISGS), and Australian Stroke Genetics Collaborative (ASGC) studies were analyzed in Boston. FLAIR sequences were analyzed using an MRIcro (www.mricro.com) semiautomated method as previously described.6 Using operator-mediated quality assurances, overlapping regions of interest corresponding to WMH produced the final maps for WMHV calculation.6 To adjust for head size, intracranial area was used as a validated marker of TICV.7 The intracranial area constituted the average of 2 midsagittal slices traced, using anatomic landmarks on T1 sequences.7 Siblings with Ischemic Stroke Study (SWISS) scans were analyzed in the same way at the University of Virginia by the Boston-trained rater.

The Wellcome Trust Case Control Consortium-2 (WTCCC2) and Milan cohorts were analyzed in London using DISPunc semiautomated lesion drawing software.8 A “seed” at the lesion border was manually marked, after which the program automatically outlined the lesion based on the signal-intensity gradient. Each region of interest was visually inspected and manually corrected as required. To estimate TICV, T2, and, in its absence, FLAIR, sequences were analyzed using an automated segmentation program, Structural Image Evaluation Using Normalization of Atrophy, X-sectional (SIENAX),9 and by summing the cerebrospinal fluid, gray, and white matter volumes.

WMH quantification agreement across the 2 main reading centers was performed for 50 randomly selected scans; agreement was very good (intraclass correlation coefficient, 0.95; confidence interval [CI], 0.91–0.97; n=50).

Genotyping

WTCCC2-UK and German cases were genotyped on the Illumina Human660W-Quad. Milan cases were genotyped using Illumina Human610-Quad or Human660W-Quad. ASGC samples were genotyped using the Human610-Quad. Both ISGS and SWISS samples were genotyped using the Illumina650K-Quad. MGH were genotyped on the Affymetrix 6.0, Illumina Human610-Quad, or Illumina OmniExpress beadchips.

For all cohorts, individuals were removed if their inferred sex was discordant with the recorded sex or if >5% genotype data were missing. Autosomal single-nucleotide polymorphisms (SNPs) were excluded for minor allele frequency <1%, >5% missing data, or Hardy-Weinberg equilibrium P<1×10−6. Each center performed checks for relatedness and population stratification. After quality-control procedures, there was an effective sample size of 2588.

Imputation was performed in all centers using IMPUTE2.10 All centers were imputed to HapMap3 and 1000 Genomes Project Phase pilot (June 2010) with the exception of MGH Omni, which was imputed to 1000 Genomes Integrated Release (June 2011). Postimputation, poorly imputed SNPs (r2<0.3) were removed, resulting in 2706548 SNPs common to all centers.

Genome-Wide Analyses and Meta-Analyses

To account for differences in MRI image acquisitions and population characteristics, we used a joint-analysis strategy; each cohort was analyzed independently and then meta-analyzed. Stroke cases with T2 images were analyzed separately from the FLAIR images. MGH samples were also subgrouped based on genotyping platform (Table II in the online-only Data Supplement).

Single-hemisphere WMHV was doubled to obtain whole-brain values and adjusted for normal interindividual variation in head size by multiplying by the ratio of mean TICV to individual TICV. Values were natural log–transformed to a normal distribution. Within each group, rank-transformed residuals were derived from a linear regression model predicting WMHV with age, sex, and the first 2 ancestry principle components as covariates in GenABEL.11 Thus, the phenotype was adjusted for age because WMHV is highly age dependent.3,4 Principle components, derived using EIGENSTRAT (http://genepath. med.harvard.edu/≈reich/Software.htm), were included to correct for potential population stratification. Association analysis was undertaken in PLINK (http://pngu.mgh.harvard.edu/≈purcell/plink/) using pseudodosages, a fractional count of 0 to 1 alleles for each genotype weighted by imputation probability, within a linear regression (additive) model. A meta-analysis of all the groups was performed under a fixed-effects inverse-variance model within METAL.12 SNPs showing heterogeneity (P<0.05) were removed.

Two further analyses were performed, in which WMHV was additionally adjusted for hypertension alone or combined cardiovascular risk factors (hypertension, hypercholesterolemia, eversmoker, and diabetes). Individuals with missing cardiovascular risk factors were removed from the latter analysis, resulting in an effective sample size of 1932.

In their discovery meta-analysis, the CHARGE consortium reported 62 SNPs significant at P<1×10−5; 61 of these were typed or imputed in ≥9 of 13 groups. A Z-score meta-analysis was performed of the association of these SNPs in CHARGE and the WMH cohorts. P values were weighted by sample size consistent with the CHARGE report.

SNPs at Locus 17q25

Six SNPs in high-linkage disequilibrium at locus 17q25 were examined (Figure 1). All SNPs were typed or imputed to high quality (r2≥0.9) in all cohorts, with the exception of MGH-Affymetrix in which rs936393 was typed and only rs11869977 was well imputed (Table III in the online-only Data Supplement).

Figure 1.

Figure 1

Open in a new tab

Linkage disequilibrium map of 6 single-nucleotide polymorphisms (SNPs). Numbers indicate r2 expressed as percentiles. Most SNPs are highly correlated (r2 >0.85), and rs1055129 is moderately correlated (r2 ≈ 0.5).

Allelic dosage for the most significant SNP was extracted and used in a conditional analysis of SNPs within a 100 kb window, using ProbABEL,13 followed by meta-analysis within METAL. Also, a meta-analysis was performed of the association results of these SNPs in discovery CHARGE report,1 Rotterdam replication report,2 and our WMH cohorts.

Genetic Risk Score

A genetic risk score (GRS) is a composite metric derived from a number of informative genetic variants associated with a phenotype of interest. We used SNPs highly associated with WMHV (P<1×10−5) in stroke-free populations1 to construct scores for ischemic strokes and test association with WMHV. SNPs were divided into 13 unique loci in relative linkage equilibrium (r2<0.2) as determined by SNP Annotation and Proxy Search (www.broadinstitute.org/mpg/snap/). One SNP per locus was chosen based on the largest effect size across WMH cohorts, and the lowest P value in the original CHARGE report. SNPs had to be typed or well imputed (r2>0.8) in all WMH cohorts. Twelve SNPs were ultimately included in the GRS because rs10012573 was not adequately imputed in 3 centers, and there were no other associated SNPs at this locus (Table III in the online-only Data Supplement). Individuals with missing genotype(s) at these SNPs were excluded, resulting in an effective sample size of 2564. The GRS was calculated by summing the risk allele doses and was not weighted because overall effect sizes are not available from the Z-score meta-analysis used in the CHARGE report.

Statistical Analysis

Statistical analyses were performed in R(http://R-project.org), and P<0.05 was considered significant. Within each group, log-transformed WMHV normalized for TICV was adjusted for sex and age by obtaining standardized residuals from a linear regression model. The GRS both including and excluding the 17q25 locus was assessed as a predictor of adjusted WMHV by linear regression. The analyses were repeated, with WMH additionally adjusted for hypertension, hypercholesterolemia, smoking, and diabetes. Individuals with missing cardiovascular risk factors were removed from these analyses. The analyses were performed per center and meta-analyzed using inverse-variance method.

Association of 17q25 Locus With Lacunar Stroke

Association between the 17q25 SNPs and lacunar stroke status was tested within METASTROKE, a consortium of ischemic stroke case-control GWA studies.14 A total of 1854 lacunar strokes and 51 939 controls free of symptomatic stroke from 12 cohorts were included (Table 2). Details of genotyping, quality control, and imputation are available in the online-only Data Supplement. All cases had brain imaging (computed tomography, MRI, or both) except in Atherosclerosis Risk in Communities Study (ARIC; 98% of cases), Australian Stroke Genetics Collaboration (ASGC; 97.4%), Cardiovascular Health Studies (CHS; 94.5%), and Heart and Vascular Health study (HVH; 96.5%).

Table 2.

Demographics for Lacunar Stroke Cases and Stroke-Free Controls and Effective Sample Size in METASTROKE

Continent Study Cases
Controls
n Mean Age (SD) % Male n Mean Age (SD) % Male
Europe WTCCC2-UK 474 75.4 (12.5) 52.3 5175 ≈52* 49.5
WTCCC2-Munich 106 65.1 (12.9) 72.6 797 62.7 (10.9) 51.4
BRAINS 97 73.9 (15.4) 52.5 407 ≥65† 35.8
deCode 240 68.8 (10.2) 56.5 26 970 57.3 (21.4) 38.0
Milan 25 56.2 (17.3) 56.7 407 50.8 (8.1) 87.7
North GEOS 54 44.3 (4.1) 72.2 498 39.5 (6.7) 56.6
America GASROS 38 65.7 (14.2) 60.3 1202 47.5 (8.5) 59.1
HVH 173 67.6 (9.3) 32.9 1290 66.6 (9.1) 47.7
ARIC 63 55.3 (6.2) 58.7 8803 54.1 (5.7) 46.5
CHS 73 74.3 (6.1) 34.2 2817 85.8 (5.6) 45.0
ISGS/SWISS 201 64.6 (13.6) 60.3 2329 64.8 (12.6) 48.0
Australia ASGC 310 77.5 (13.1) 57.4 1244 70.2 (12.1) 50.2
Total 1854 70.2 (14.0) 54.2 51 939 N/A 43.2
Open in a new tab

ARIC indicates Atherosclerosis Risk in Communities; ASGC, Australian Stroke Genetics Collaborative; BRAIN, Bio-repository of DNA in stroke; CHS, Cardiovascular Health Study; GASROS, Genes Affecting Stroke Risk and Outcome Study; GEOS, Genetics of Early Onset Stroke Study; HVH, Heart and Vascular Health; ISGS, Ischemic Stroke Genetics Study; SWISS, Siblings with Ischemic Stroke Study; and WTCCC2, Wellcome Trust Case-Control Consortium-2.

*

Approximate age at genotyping of the 2738 controls from the 1958 birth cohort. Age was not available for the remaining controls.

†

All controls were ≥65 years of age at the time of genotyping. No further information was available.

Lacunar strokes were designated based on compatible clinical and neuroimaging findings using the TOAST or similar criteria.5 TOAST requires normal neuroimaging or relevant brainstem/subcortical hemispheric lesions (<1.5 cm), whereas HVH/CHS required normal computed tomography with a typical lacunar syndrome or subcortical infarction (≤2 cm). ARIC defined strokes as lacunar if the infarct was in a typical location (basal ganglia, brain stem, thalamus, internal capsule, or cerebral white matter) and of unstated size or ≤2 cm. All cohorts excluded cases with recognized sources of emboli or large-vessel atherosclerosis. Samples were not entirely independent of the WMH cohorts, with 16% of cases included in WMHV analysis.

For Genetics of Early Onset Stroke Study, rs3744028 and rs11869977 were not imputed. For Genes Affecting Stroke Risk and Outcome Study only rs936393 and rs1055129 were typed; rs3744017 was poorly imputed (r2=0.64) and therefore not included in the meta-analysis of this SNP. All 6 SNPs were typed or adequately imputed (r2≥0.8) in the remaining cohorts (Table V in the online-only Data Supplement). Association analyses were performed in each center by logistic regression assuming an additive model. Results were adjusted by genomic control factor, followed by meta-analysis under an inverse-variance weighted model.

Results

Association of 17q25 Locus With WMHV

All SNPs at the 17q25 locus were significantly associated with WMHV as a quantitative trait in a direction and magnitude of effect consistent with previous reports. Most associations remained significant after Bonferroni adjustment for multiple comparisons (P<0.008), except for rs1055129 (P=0.015; Table 3). The correction may be overconservative, given these SNPs are highly correlated (r2>0.85), with the exception of rs1055129, which is moderately correlated with the others (r2 ≈ 0.5; Figure 1). The most significant SNP was rs9894383 (B=0.13; SE=0.04; P=0.0006): the risk allele was associated with a 13% (CI, 6%–22%) increase in the geometric mean of WMHV adjusted for age and sex (Figure 2). There was no evidence of between-study heterogeneity. The results remained significant when adjusting for hypertension alone or cardiovascular risk factors (Table VI in the online-only Data Supplement).

Table 3.

Association Statistics of SNPs at Locus 17q25 With WMH Volume and Lacunar Stroke

SNP Risk Allele WMH Volume
Lacunar Stroke
Effect Size (SE) P Value Effect Size (SE) P Value
rs3744028 C 0.12 (0.04) 0.0030 −0.005 (0.048) 0.92
rs9894383 G 0.13 (0.04) 0.00064 0.005 (0.046) 0.91
rs11869977 G 0.12 (0.04) 0.00069 −0.001 (0.047) 0.98
rs936393 G 0.11 (0.04) 0.0012 0.014 (0.045) 0.77
rs3744017 A 0.12 (0.04) 0.0032 −0.001 (0.046) 0.98
rs1055129 G 0.08 (0.03) 0.015 0.031 (0.039) 0.43
Open in a new tab

SNP indicates single-nucleotide polymorphism; and WMH, white matter hyperintensities.

Figure 2.

Figure 2

Open in a new tab

Forest plot showing inverse-variance weighted meta-analysis of the association of rs9894383 with adjusted white matter hyperintensities volume. ASGC indicates Australian Stroke Genetics Collaborative; FLAIR, fluid-attenuated inversion recovery; ISGS, Ischemic Stroke Genetics Study; MGH, Massachusetts General Hospital; and SWISS, Siblings with Ischemic Stroke Study.

Conditioning on rs9894383 did not reveal any other significant variants within a 100 kb window. Meta-analysis of the 6 17q25 SNPs with the original CHARGE and Rotterdam replication reports revealed rs9894383 as most significantly associated with WMHV (P=1.0×10−11). Meta-analysis of moderately significant SNPs (P<1×10−5) reported by CHARGE only revealed a further genome-wide significant SNP at locus 17q25 (Table VII in the online-only Data Supplement).

The GRS was a significant predictor of WMHV adjusted for age and sex (P=0.001; B=0.031; CI, 0.012–0.050) and after additional adjustment for hypertension (P=0.002; B=0.030; CI, 0.011–0.049) and combined cardiovascular risk factors (P=0.003; B=0.031; CI, 0.011–0.051). A GRS without the 17q25 locus was also significantly associated with WMHV (P=0.023; B=0.022; CI, 0.003–0.042) and after adjustment for hypertension (P=0.020; B=0.023; CI, 0.004–0.042) and cardiovascular risk factors (P=0.025; B=0.026; CI, 0.003–0.048). Figure 3 shows an approximate linear relationship between mean WMHV residuals and quintiles of the GRS with and without 17q25 locus. Table VIII in the online-only Data Supplement gives the mean GRS for each quintile.

Figure 3.

Figure 3

Open in a new tab

Mean-adjusted white matter hyperintensities volume (WMHV) residuals and 95% confidence intervals for quintiles of genetic risk scores (GRS). Analyses are shown both including (A) and excluding (B) the 17q25 locus.

Association of 17q25 Locus With Lacunar Stroke

There was no association between the 6 SNPs and lacunar stroke status in the Metastroke case-control analysis (Table 3). The 17q25 SNPs were not associated with cardioembolic or large artery stroke (data not shown).

Discussion

This study provides further support for an association between the 17q25 locus and WMH, and replication in ischemic stroke supports the hypothesis of shared genetic contribution to WMHV in community participants without stroke and ischemic stroke cases. The most significant SNP was rs9894383, and meta-analysis with 2 previous reports1,2 revealed a combined P value of 1.0×10−11. Most SNPs within the 17q25 region were highly correlated, and conditional analysis did not reveal more than 1 independent signal.

In contrast, the 17q25 locus was not associated with another manifestation of small-vessel disease, lacunar stroke. This may suggest that any causal variant linked to this locus does not act by directly promoting small-vessel arteriopathy or the type of arteriopathy primarily underlying lacunar infarction. The study was powered and detect an association with an odds ratio between 1.1 and 1.15. Importantly, however, controls did not have brain imaging to exclude silent brain infarction, which is found in ≈20% of healthy elderly,15 and this could have limited study power.

We performed a meta-analysis with the top SNPs reported by CHARGE; however, no novel loci reached genome-wide significance. To determine whether there were additional genetic variants shared by WMH occurring in community participants without stroke and ischemic stroke, we calculated GRS excluding the 17q25 locus. This was a significant predictor of WMHV consistent with additional shared genetic variants.

There are several genes in the 17q25 region; however, cis-expression quantitative trait loci, primarily of HapMap lymphoblastoid cells, implicate TRIM47. TRIM47 could modulate brain responses to ischemic injury because its RING domain confers protein ubiquitination, which promotes proteolysis and cellular homestasis.16 Imbalance in ubiquitin-proteasome pathways is integral to cerebral ischemic injury mechanisms17 and also evident in WMH-expression profiles.18

This study used volumetric MRI techniques, which have been demonstrated to be reliable and accurate,7–9 with good agreement across reading centers. A limitation is the variability in MRI protocols, resulting from the use of clinical imaging in these GWAS databases. However, studies applying volumetric techniques have shown high reproducibility across acquisition protocols8 and scanner models.19 To minimize effects of MRI heterogeneity, centers were analyzed separately and WMHV Z-scores were derived before association testing. We measured whole-brain rather than regional WMHV and therefore could not investigate genetic differences between periventricular and subcortical WMH, which are suggested to have differing pathological, risk factor, and functional associations.4 Another limitation is that genotyping was performed on multiple platforms. However, top SNPs were imputed to a high quality with consistent allele frequencies.

In conclusion, our data provide further support for an association between the 17q25 locus and WMHV in patients with ischemic stroke. Future studies are warranted to explore these genetic associations to understand biological underpinnings of this complex cerebrovascular phenotype. The lack of association with lacunar stroke suggests the 17q25 locus does not act via promoting small-vessel arteriopathy.

S1. Stroke Subtype (according to TOAST criteria) of ischemic stroke cases included in WMH analysis.

S2. Post-quality control numbers, magnetic resonance imaging (MRI) sequences and MRI models for cohorts in WMH GWAS. Note sub-grouping based on availability of MRI sequences.

S3. Imputation Quality (r-squared) and minor allele frequency (MAF) of 17q25 single nucleotide polymorphisms (SNPs) in ischemic stroke cohorts included WMH analysis.

S4. Imputation Quality (r-squared) and case minor allele frequency (MAF) of SNPs in Genetic Risk Score in ischemic stroke cohorts included WMH analysis.

S5. Imputation Quality (r-squared) and minor allele frequency (MAF) of 17q25 SNPs in twelve SVD stroke-subtype case-control analyses from Metastroke

S6. Association results for SNPs at locus 17q25 with WMHV in Ischaemic Stroke.

S7. Z-score based meta-analysis with CHARGE SNPs with p<1 × 10−5.

S8. Table showing the mean GRS and standard deviation per quintile used in Figure 3.

Supplementary Material

DATA SUPPLEMENT

Acknowledgments

The authors thank Sólveig Grétarsdóttir, PhD, (deCODE Genetics, Iceland) for samples contributed to Metastroke. SWISS study acknowledges Aparna Baheti, Rodney Smith (University of Virginia), and Kathryn Powell (University of South Florida) for their contributions.

Sources of Funding

The principal funding for this study was provided by the Stroke Association. The authors are supported by MRC (Training Fellowship, Dr Adib-Samii) and NINDS (K23NS064052, Dr Rost). Funding for collection, genotyping, and analysis of stroke samples was provided by Wellcome Trust (WTCCC2), the Intramural Research Program (NIA; MGH, ISGS), National Institute for Neurological Disorders and Stroke (SWISS, GASROS, ISGS, CHS, HVH, GEOS), Bugher Foundation of the American Heart Association, MGH Deane Institute for Integrative Study of Atrial Fibrillation and Stroke (GASROS), National Institutes of Health Genes, Environment and Health Initiative, Medical Research Service of the Department of Veterans Affairs and Centers for Disease Control (GEOS), National Health & Medical Research Council (ASGC), Italian Ministry of Health (Milan), National Human Genome Research Institute (GASROS), ARIC), National Heart, Lung, and Blood Institute (ARIC, CHS, HVH), Henry Smith Charity, and the British Council (BRAINS).

Footnotes

Disclosures

None.

References

  • 1.Fornage M, Debette S, Bis JC, Schmidt H, Ikram MA, Dufouil C, et al. Genome-wide association studies of cerebral white matter lesion burden: the CHARGE consortium. Ann Neurol. 2011;69:928–939. doi: 10.1002/ana.22403. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Verhaaren BF, de Boer R, Vernooij MW, Rivadeneira F, Uitterlinden AG, Hofman A, et al. Replication study of chr17q25 with cerebral white matter lesion volume. Stroke. 2011;42:3297–3299. doi: 10.1161/STROKEAHA.111.623090. [DOI] [PubMed] [Google Scholar]
  • 3.Debette S, Markus HS. The clinical importance of white matter hyper-intensities on brain magnetic resonance imaging: systematic review and meta-analysis. BMJ. 2010;341:c3666. doi: 10.1136/bmj.c3666. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Schmidt R, Schmidt H, Haybaeck J, Loitfelder M, Weis S, Cavalieri M, et al. Heterogeneity in age-related white matter changes. Acta Neuropathol. 2011;122:171–185. doi: 10.1007/s00401-011-0851-x. [DOI] [PubMed] [Google Scholar]
  • 5.Adams HP, Jr, Bendixen BH, Kappelle LJ, Biller J, Love BB, Gordon DL, et al. Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. TOAST. Trial of Org 10172 in Acute Stroke Treatment. Stroke. 1993;24:35–41. doi: 10.1161/01.str.24.1.35. [DOI] [PubMed] [Google Scholar]
  • 6.Rost NS, Rahman RM, Biffi A, Smith EE, Kanakis A, Fitzpatrick K, et al. White matter hyperintensity volume is increased in small vessel stroke subtypes. Neurology. 2010;75:1670–1677. doi: 10.1212/WNL.0b013e3181fc279a. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Nandigam RN, Chen YW, Gurol ME, Rosand J, Greenberg SM, Smith EE. Validation of intracranial area as a surrogate measure of intracranial volume when using clinical MRI. J Neuroimaging. 2007;17:74–77. doi: 10.1111/j.1552-6569.2006.00069.x. [DOI] [PubMed] [Google Scholar]
  • 8.Grimaud J, Lai M, Thorpe J, Adeleine P, Wang L, Barker GJ, et al. Quantification of MRI lesion load in multiple sclerosis: a comparison of three computer-assisted techniques. Magn Reson Imaging. 1996;14:495–505. doi: 10.1016/0730-725x(96)00018-5. [DOI] [PubMed] [Google Scholar]
  • 9.Smith SM, Zhang Y, Jenkinson M, Chen J, Matthews PM, Federico A, et al. Accurate, robust, and automated longitudinal and cross-sectional brain change analysis. Neuroimage. 2002;17:479–489. doi: 10.1006/nimg.2002.1040. [DOI] [PubMed] [Google Scholar]
  • 10.Howie BN, Donnelly P, Marchini J. A flexible and accurate genotype imputation method for the next generation of genome-wide association studies. PLoS Genet. 2009;5:e1000529. doi: 10.1371/journal.pgen.1000529. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Aulchenko YS, Ripke S, Isaacs A, van Duijn CM. GenABEL: an R library for genome-wide association analysis. Bioinformatics. 2007;23:1294–1296. doi: 10.1093/bioinformatics/btm108. [DOI] [PubMed] [Google Scholar]
  • 12.Willer CJ, Li Y, Abecasis GR. METAL: fast and efficient meta-analysis of genomewide association scans. Bioinformatics. 2010;26:2190–2191. doi: 10.1093/bioinformatics/btq340. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Aulchenko YS, Struchalin MV, van Duijn CM. ProbABEL package for genome-wide association analysis of imputed data. BMC Bioinformatics. 2010;11:134. doi: 10.1186/1471-2105-11-134. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Traylor M, Farrall M, Holliday EG, Sudlow C, Hopewell JC, Cheng YC, et al. Australian Stroke Genetics Collaborative, Wellcome Trust Case Control Consortium 2 (WTCCC2); International Stroke Genetics Consortium. Genetic risk factors for ischaemic stroke and its subtypes (the METASTROKE collaboration): a meta-analysis of genome-wide association studies. Lancet Neurol. 2012;11:951–962. doi: 10.1016/S1474-4422(12)70234-X. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Vermeer SE, Longstreth WT, Jr, Koudstaal PJ. Silent brain infarcts: a systematic review. Lancet Neurol. 2007;6:611–619. doi: 10.1016/S1474-4422(07)70170-9. [DOI] [PubMed] [Google Scholar]
  • 16.Meroni G, Diez-Roux G. TRIM/RBCC, a novel class of ‘single protein RING finger’ E3 ubiquitin ligases. Bioessays. 2005;27:1147–1157. doi: 10.1002/bies.20304. [DOI] [PubMed] [Google Scholar]
  • 17.Di Napoli M, McLaughlin B. The ubiquitin-proteasome system as a drug target in cerebrovascular disease: therapeutic potential of proteasome inhibitors. Curr Opin Investig Drugs. 2005;6:686–699. [PMC free article] [PubMed] [Google Scholar]
  • 18.Simpson JE, Hosny O, Wharton SB, Heath PR, Holden H, Fernando MS, et al. Medical Research Council Cognitive Function and Ageing Study Neuropathology Group. Microarray RNA expression analysis of cerebral white matter lesions reveals changes in multiple functional pathways. Stroke. 2009;40:369–375. doi: 10.1161/STROKEAHA.108.529214. [DOI] [PubMed] [Google Scholar]
  • 19.Jovicich J, Czanner S, Han X, Salat D, van der Kouwe A, Quinn B, et al. MRI-derived measurements of human subcortical, ventricular and intracranial brain volumes: Reliability effects of scan sessions, acquisition sequences, data analyses, scanner upgrade, scanner vendors and field strengths. Neuroimage. 2009;46:177–192. doi: 10.1016/j.neuroimage.2009.02.010. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

DATA SUPPLEMENT
NIHMS502894-supplement-DATA_SUPPLEMENT.pdf (137.2KB, pdf)

RESOURCES