A Neuroprotective Locus Modulates Ischemic Stroke Infarction Independent of Collateral Vessel Anatomy

Although studies with inbred strains of mice have shown that infarct size is largely determined by the extent of collateral vessel connections between arteries in the brain that enable reperfusion of the ischemic territory, we have identified strain pairs that do not vary in this vascular phenotype, but which nonetheless exhibit large differences in infarct size. In this study we performed quantitative trait locus (QTL) mapping in mice from an intercross between two such strains, WSB/EiJ (WSB) and C57BL/6J (B6). This QTL mapping revealed only one neuroprotective locus on Chromosome 8 (Chr 8) that co-localizes with a neuroprotective locus we mapped previously from F2 progeny between C3H/HeJ (C3H) and B6. The allele-specific phenotypic effect on infarct volume at the genetic region identified by these two independent mappings was in the opposite direction of the parental strain phenotype; namely, the B6 allele conferred increased susceptibility to ischemic infarction. Through two reciprocal congenic mouse lines with either the C3H or B6 background at the Chr 8 locus, we verified the neuroprotective effects of this genetic region that modulates infarct volume without any effect on the collateral vasculature. Additionally, we surveyed non-synonymous coding SNPs and performed RNA-sequencing analysis to identify potential candidate genes within the genetic interval. Through these approaches, we suggest new genes for future mechanistic studies of infarction following ischemic stroke, which may represent novel gene/protein targets for therapeutic development.


Figure S1
Infarct volume in F2 animals shows relatively wide distribution. Each graph describes the distribution of infarct volume for B6, WSB, F1, F2 animals shown in Figure 1B with scatter plots.
The range of the infarct volume is shown on the x-axis, and the y-axis indicates animal numbers.
A neuroprotective locus for stroke 3 Figure S2 Non-parametric mapping test of the genome-wide QTL mapping analysis. The nonparametric mapping test was performed to validate the genome-wide QTL mapping for infarct volume that was analyzed with the parametric mapping test shown in Figure 2. The non-parametric test indicated by a red line is superimposed on the parametric test indicated by a black line. The two independent mapping plots are nearly identical.
A neuroprotective locus for stroke 4 Figure S3 Characterization of the allelic effects of the locus mapping to Chromosome 17. The linkage peak on Chr 17 did not reach a significance threshold (p < 0.05) determined by 1,000 permutation test. However, allelic effects on infarct volume of the F2 cohort at gUNC27989312 (LOD score: 3.51) on Chr 17 shows that the B6 allele is protective for infarction. This follows the same trend as the parental strains. Statistial analysis was performed with one-way ANOVA followed by Tukey's multiple comparison test (p < 0.05; p < 0.001).
A neuroprotective locus for stroke Chromosomes 1 through X are represented numerically on the x-axis, and the y-axis represents the LOD score. The significant (p < 0.05) level of linkage was determined by 1,000 permutation tests. Three regions of the genome mapping to Chr 5, 8, and 18 display significant QTL mapping to the infarct volume trait with LOD scores of 4.13, 10.31, and 4.71, respectively. (B) The graph shows the major locus mapping across Chr 8 using 14 informative SNP markers. The LOD score 7 at the peak is 10.31 (rs13479735), and the 1.5-LOD support interval is from 36.02 to 50.26 Mb, indicated by the red bar on the graph. (C) Allelic effects on infarct volume of the F2 cohort at rs13479735. The B6 allele confers increased susceptibility to infarction and the C3H allele confers protection. Statistical analysis was performed with one-way ANOVA followed by Tukey's multiple comparison test (p < 0.01; p < 0.001).
A neuroprotective locus for stroke 8 Table S1 Raw data for all Figures. The Excel spreadsheet contains the numerical values with detailed statistical information and array information used to generate all of the figures.

Table S2
Genes mapping within the Civq4 that harbor coding SNP differences between protective allele and risk allele. The table shows non-synonymous coding SNPs between the protective allele (i.e., C3H and WSB) and the risk allele (i.e., B6). The functional consequences on protein function for each coding SNP was predicted using three independent in silico algorithms, SIFT, PolyPhen-2, and PROVEAN. Coding SNP predicted to be either "damaging" or "deleterious" are highlighted in red.

Table S3
Genes show ing strain-specific differential gene expression between WSB and B6. For each of the 1,716 genes, the table displays the p-value and fold change, with either a positive value or negative value, corresponding to increased or decreased expression of the WSB allele relative to the B6 allele, respectively.

Table S4
Genes show strain-specific differential gene expression between C3H and B6. For each of the 2,180 genes, the table displays the p-value and fold change, with either a positive value or negative value, corresponding to increased or decreased expression of the C3H allele relative to the B6 allele, respectively.