Common coding variant in SERPINA1 increases the risk for large artery stroke

Abstract

Large artery atherosclerotic stroke (LAS) shows substantial heritability not explained by previous genome-wide association studies. Here, we explore the role of coding variation in LAS by analyzing variants on the HumanExome BeadChip in a total of 3,127 cases and 9,778 controls from Europe, Australia, and South Asia. We report on a nonsynonymous single-nucleotide variant in serpin family A member 1 (SERPINA1) encoding alpha-1 antitrypsin [AAT; p.V213A; P = 5.99E-9, odds ratio (OR) = 1.22] and confirm histone deacetylase 9 (HDAC9) as a major risk gene for LAS with an association in the 3'-UTR (rs2023938; P = 7.76E-7, OR = 1.28). Using quantitative microscale thermophoresis, we show that M1 (A213) exhibits an almost twofold lower dissociation constant with its primary target human neutrophil elastase (NE) in lipoprotein-containing plasma, but not in lipid-free plasma. Hydrogen/deuterium exchange combined with mass spectrometry further revealed a significant difference in the global flexibility of the two variants. The observed stronger interaction with lipoproteins in plasma and reduced global flexibility of the Val-213 variant most likely improve its local availability and reduce the extent of proteolytic inactivation by other proteases in atherosclerotic plaques. Our results indicate that the interplay between AAT, NE, and lipoprotein particles is modulated by the gate region around position 213 in AAT, far away from the unaltered reactive center loop (357-360). Collectively, our findings point to a functionally relevant balance between lipoproteins, proteases, and AAT in atherosclerosis.

Keywords: antitrypsin; genetics; ischemic stroke; large artery stroke; variation.

Fig. 1.

Flow chart of the exome chip analysis. Samples from Germany, the United Kingdom, and Australia were pooled with samples from South Asia for the transethnic metaanalysis.

Fig. 2.

Main association results. Shown are the results for common (A, MAF > 5%), low-frequency (B, 1% < MAF < 5%), and rare (C, MAF < 1%) variants. Genomic position is plotted on the x axis, and the −log₁₀ of the P value is displayed on the y axis. The threshold for exome-wide significance after Bonferroni correction for the number of variants studied is displayed as a dashed red line. (D) LocusZoom plot [−log₁₀ (p-value)] for the SERPINA1 region and LAS. Shown is the region for the top signal ± 500 kb. Common, low-frequency, and rare variants are displayed as filled circles (●), squares (■), and diamonds (♦), respectively. Blue peaks represent estimated recombination rates. (E) Forest plot of associations with rs6647 in individual samples and in the resulting fixed-effects metaanalysis. Sample sizes are reflected by the size of each square. Gray bars show the [CI_95] of the point estimate.

Fig. S1.

Outlier analysis by PC analysis. German, UK, Australian, and South Asian samples were checked to remove potential outliers with respect to genetic background. Green with black outline, German cases; white with green outline, Kooperative Gesundheitsforschung in der Region Augsburg (KORA) controls; blue with black outline, UK/Australian cases; white with blue outline, Wellcome Trust Case Control Consortium 2 controls; orange with black outline, Pakistan cases; white with orange outline, Pakistan controls. Samples in gray were removed from the analysis after visual inspection. PC1 and PC2 are displayed on the x axis and y axis, respectively.

Fig. S2.

Quantile–quantile (QQ) plots for single-variant analysis. Shown are the observed versus expected −log₁₀ P value distributions for common (A), low-frequency (B), and rare (C) variants, along with the genomic inflation factors (λ). The red line shows the expected (null) distribution of the statistic.

Fig. 3.

Functional differences between M1 AAT variants found in lipid-containing, but not lipid-free, plasma. Shown are the fitted binding curves of the microscale thermophoresis assay under equilibrium conditions in lipid-containing (A) and lipid-free (B) plasma. A fixed concentration of fluorescently labeled and catalytically inactive NE was titrated with the M1(A213) or M1(V213) variant in AAT-deficient plasma. (A) In the presence of lipids, the K_D (KD) observed with the M1(V213) variant [KD(A213) = 6,800 ± 1000 nM] was substantially higher than the K_D observed with the M1(A213) variant [KD(V213) = 3,200 ± 500 nM]. (B) In the absence of lipids, there was no discernible difference between the two variants [KD(A213) = 540 ± 50 nM; KD(V213) = 550 ± 60 nM]. Fitted binding curves and K_D values (mean ± SD) were derived from global fitting of three measurements (three independent protein expressions). The measurements were performed in 7.5% plasma and with 5 nM labeled NE.

Fig. 4.

Differences in hydrogen/deuterium exchange kinetics between M1(A213) and M1(V213) AAT. (A) Shown is the cumulative difference across five time points (10 s to 120 min) in deuterium uptake between M1(A213) and M1(V213) mapped onto the crystal structure of AAT (Protein Data Bank ID code 1QLP). Peptides displaying a significant increase in deuterium uptake in M1(A213) compared with M1(V213) are colored red (two- to threefold increase) and pink (one- to twofold increase). Otherwise, there was no significant difference in deuterium uptake between the two variants, such as the stretch between positions 100 and 112 (gray arrow). Substitution of Ala for Val at position 213 results in subtle but widespread destabilization of AAT. (B) Example kinetic curves for the deuterium uptake with two peptides. Peptide 134–143 (Lower) displays a significantly higher deuterium uptake in the M1(A213) AAT variant compared with M1(V213), whereas no significant difference is seen with peptide 100–112 (Upper). Values represent means from three different measurements.