Selenoprotein deficiency disorder predisposes to aortic aneurysm formation

Abstract

Aortic aneurysms, which may dissect or rupture acutely and be lethal, can be a part of multisystem disorders that have a heritable basis. We report four patients with deficiency of selenocysteine-containing proteins due to selenocysteine Insertion Sequence Binding Protein 2 (SECISBP2) mutations who show early-onset, progressive, aneurysmal dilatation of the ascending aorta due to cystic medial necrosis. Zebrafish and male mice with global or vascular smooth muscle cell (VSMC)-targeted disruption of Secisbp2 respectively show similar aortopathy. Aortas from patients and animal models exhibit raised cellular reactive oxygen species, oxidative DNA damage and VSMC apoptosis. Antioxidant exposure or chelation of iron prevents oxidative damage in patient's cells and aortopathy in the zebrafish model. Our observations suggest a key role for oxidative stress and cell death, including via ferroptosis, in mediating aortic degeneration.

Fig. 1. Thoracic aortic aneurysms with cystic medial necrosis.

A CT imaging of P1 showing aortic valve and ascending aorta replacement at age 23 years (orange arrow) and distal arch dilatation (white arrow) at age 44 years. MR imaging of P2 and P3 showing dilatation of the aortic root (red star), with effacement of the sinotubular junction (red arrows). Echocardiogram of P4 showing positions (1, annulus; 2, root; 3, sinotubular junction; 4, tract) of abnormal aortic dimensions (upper panel) and aortic valvular insufficiency (lower panel). B Histochemistry and TUNEL/immunohistochemistry of control thoracic aorta (C), P1a (ascending aorta), P1b (aortic arch) or P2 (ascending aorta), using Movat’s pentachrome, Elastin van Gieson (EVG), Alcian Blue, and von Kossa stains, and terminal deoxynucleotidyl transferase dUTP nick end labelling (TUNEL) with anti-α-smooth muscle actin (SMA/Acta2). Outlined areas in Movat’s sections are shown in higher power in subsequent stains. Insets in TUNEL/SMA sections are high-power views of TUNEL-positive cells (arrows). Scale bars: 1 mm in Movat’s, 250 μm in other panels. C Immunohistochemistry of control ascending aorta (C), or P1b and P2 thoracic aorta for 8-oxoguanine (8-Oxo-G, left panels) or γ-H2AX (right panels), co-stained with anti-SMA. Insets are high-power views of outlined areas. Scale bars: 250 μm. B and C n = 1 independent experiment.

Fig. 2. Increased oxidative stress, apoptosis and susceptibility to ferroptosis in aortic vascular smooth muscle cells.

A H₂O₂ production in aortic VSMCs from patients P1, P2 (blue bars) and controls (C1, C2, C3: grey bars). Statistics: Ordinary one-way ANOVA with adjusted P values (Tukey’s multiple comparison test), each bar represents the mean (n ≥ 8 independent experiments), error bars represent SEM, p-values compare P1 and P2 with controls. B Membrane lipid peroxidation in aortic VSMCs from patients P1, P2 (blue bars) and controls (C1–C3: grey bars), with exposure to the vehicle, MitoQ or α-tocopherol. Statistics: Two-way ANOVA with adjusted P values (Tukey’s multiple comparison test), each bar represents the mean (n ≥ 4 independent experiments), error bars represent SEM, p-values compare P1 and P2 with controls, or vehicle-treated data from P1 and P2 with antioxidant-treated cells. C Representative microscopic images of primary, aortic vascular smooth muscle cells from controls (C1, C3, C3) or patients (P1, P2) following 1 h exposure to MitoPerOx, a fluorescent probe assessing lipid peroxidation within mitochondria, with insets below showing high power views of single cells. Each image was acquired independently ten times, with similar results. Scale bars: 50 μm. D Quantitation of relative fluorescence intensity (RFI) in primary VSMCs from patients (blue lines; P1 solid line; P2 broken line) and three controls (C1, C3, C3: black lines), treated with MitoPerOx over the time period indicated. Statistics: Unpaired, two-tailed t-test, error bars represent SEM, p-values compare patient versus controls (n = 10 independent experiments). E Apoptosis (Annexin V positive cells) of aortic VSMCs from patients P1, P2 (blue bars) and controls (C1–C3: grey bars), with exposure to the vehicle, MitoQ or α-tocopherol. Statistics: Two-way ANOVA with adjusted p values (Tukey’s multiple comparison test), each bar represents the mean (n ≥ 5 independent experiments), error bars represent SEM, p-values compare P1 and P2 with controls, or vehicle-treated data from P1 and P2 with antioxidant-treated cells. F and G. Membrane lipid peroxidation (F) and cell viability (G) of aortic VSMCs from patients (P1, P2: blue bars) and controls (C1–C3: grey bars), with exposure to vehicle, erastin or desferrioxamine (DFO). Statistics: Two-way ANOVA with adjusted p values (Tukey’s multiple comparison test), each bar represents the mean (F:n ≥ 4; G:n = 6 independent experiments), error bars represent SEM, p-values in F compare untreated/treated P1 and P2 with control cells treated similarly (ns: not significant). H and I H₂O₂-induced peroxidation of membrane lipid in PBMCs (J) or serum (K) from patient P2 at different days of dietary supplementation with α-tocopherol (D0: day 0, light blue), 40 (D40: day 40, shaded blue) or 152 (D152: day 152, dark blue) or untreated control subjects (C: grey). Statistics: H Two-way ANOVA, each line represents the mean for controls (n = 5) and single values for P2, p-value compares P2 D0 with controls. I Unpaired, two-tailed t-test, each bar represents the mean for controls (n = 5) and untreated P2 (n = 6), or single values for P2 treated for different numbers of days as indicated. Source data are provided as a Source Data file.

Fig. 3. Aortic abnormalities in Secisbp2 mutant zebrafish.

A Representative aortic outflow tract images, comprising ventral aorta (VA) and bulbus arteriosus (BA), from wild-type (Secisbp2^wt/wt; +/+), heterozygous (Secisbp2^Q333X/wt; +/−) and homozygous (Secisbp2^Q333X/Q333X; −/−) mutant adult (age 6 months) zebrafish. The dotted lines show the positions of VA and BA measurements quantified in (B and C). Each image was repeated independently 5 times with similar results. Scale bar: 250 μm. B and C VA and BA diameters, adjusted for body size, measured at dotted line positions in (A), in wild-type (+/+: grey bar), heterozygous (+/−: yellow bar) and homozygous (−/−: blue bar) Secisbp2 mutant fish. Statistics: Unpaired, two-tailed t-test, each bar represents the mean (n = 5), and error bars represent SD. D–F H₂O₂ (D, n = 8), lipid peroxidation (E, n = 8) and p53/casp3 mRNA expression (F, n = 4) in VA/BA tissue from wild type (+/+: grey bar), heterozygous (+/−: yellow bar) and homozygous (−/−: blue bar) Secisbp2 mutant fish. Statistics: unpaired, two-tailed t-test, each bar represents the mean, and error bars represent SD. G, H and L. Representative images, of the ventral aorta (VA) and bulbus arteriosus (BA) from wild-type (Secisbp2^wt/wt; +/+), heterozygous (Secisbp2^Q333X/wt; +/−) and homozygous (Secisbp2^Q333X/Q333X; −/−) zebrafish embryos at 5 dpf. Images show vascular endothelial cell-specific green fluorescent protein, VA diameter and BA diameter was measured at positions (1–3) marked with broken line (G, +/+). Zygotes were treated (from 3 to 120 hpf) with 0.5 mM H₂O₂ or H₂O₂ + 1 mM α-tocopherol (αToc) or H₂O₂ + 1 mM Decyltriphenylphosphonium (TPP) or H₂O₂ + 1 mM MitoQ (H) or treated (from 6 to 120 hpf) with 10 mM Erastin or 10 mM Erastin + 100 mM desferrioxamine (DFO) (L). Each image was acquired independently fifteen times, with similar results. Scale bars: 250 μm. I and J Mean VA diameter, measured at positions (1–3) marked in (A) (I), or BA diameter (J) quantified in different groups (5 dpf, n = 15) in wild-type (Secisbp2^wt/wt; +/+), heterozygous (Secisbp2^Q333X/wt; +/−), homozygous (Secisbp2^Q333X/Q333X; −/−) fish, untreated or treated with 0.5 mM H₂O₂ in combination with 1 mM α-tocopherol (αToc) or 1 mM Decyltriphenylphosphonium (TPP) or 1 mM MitoQ. Statistics: unpaired, two-tailed t-test, each bar represents the mean and error bars represent SD. K Percent survival of embryos over time in different groups described in (G–J). Statistics: Logrank test for trend (Chi-square), p-value for H₂O₂ treated Secisbp2^Q333X/Q333X (−/−) versus all comparisons, each line represents the % of surviving fish (n = 100–200). M and N Mean VA diameter, measured at positions (1–3) marked in (A) (M), or BA diameter (N) quantified in different groups (5 dpf, n = 15) in wild-type (Secisbp2^wt/wt; +/+), heterozygous (Secisbp2^Q333X/wt; +/−), homozygous (Secisbp2^Q333X/Q333X; −/−) fish, all treated with 10 mM Erastin or with 10 mM Erastin + 100 mM desferrioxamine (DFO). Statistics: unpaired, two-tailed t-test, each bar represents the mean, error bars represent SD. O Percent survival of embryos over time in different groups described in (L–I). Statistics: Logrank test for trend (Chi-square), the p-value for elastin-treated Secisbp2^Q333X/Q333X (−/−) versus all comparisons, each line represents the percentage of surviving fish (n = 100–200). Source data are provided as a Source Data file.

Fig. 4. Aortic abnormalities in Secisbp2 knockdown zebrafish.

A, G and J Representative aortic images, visualized by green fluorescent protein expression as described in Fig. 3 legend, of ventral aorta (VA) and bulbus arteriosus (BA) of zebrafish embryos (5dpf), following zygote injection with control (C), or Secisbp2 morpholinos (MO) or co-injection of Secisbp2 morpholino with Secisbp2 mRNA (Rescue; R). Zygotes were treated (3–120 hpf) with 0.5 mM H₂O₂ or H₂O₂ + 1 mM a-tocopherol (aToc) or H₂O₂ + 1 mM Decyltriphenylphosphonium (TPP) or H₂O₂ + 1 mM MitoQ (G) or treated (from 6 to 120 hpf) with vehicle (DMSO), 100 mM desferrioxiamine (DFO), 10 mM Erastin or 10 mM Erastin + 100 mM DFO (J). The dotted lines A show the positions of VA and BA measurements quantified in (B and C). Each image was repeated independently 15 times with similar results. Scale bar: 50 μm. B, C and F H₂O₂ production (B), membrane lipid peroxidation (C) and p53 and casp3 mRNA expression (F) in VA of 5 pools of 50 embryos at 4 dpf following zygote injection with control (C: grey), or Secisbp2 morpholinos (MO: blue) or co-injection of Secisbp2 morpholino with Secisbp2 mRNA (Rescue; R: yellow). Statistics: One-way ANOVA with adjusted p values (Tukey’s multiple comparison test), each bar represents the mean of n = 5 independent experiments, and error bars represent SEM. D, E, H, I, K and L. VA diameter, measured at positions (1–3) marked in A (D, H, K) or BA diameter (E, I, L) quantified in different groups (5 dpf, n = 15 independent experiments) following zygote injection with control (C), or Secisbp2 morpholinos (MO) or co-injection of Secisbp2 morpholino with Secisbp2 mRNA (Rescue; R) and treated as indicated (0.5 mM H₂O₂; 1 mM α-tocopherol (αToc); 1 mM Decyltriphenylphosphonium (TPP); 1 mM MitoQ). Statistics: Ordinary one-way ANOVA with adjusted p values (Tukey’s multiple comparison test), each bar represents the mean, error bars represent SEM, ns = not significant. Source data are provided as a Source Data file.

Fig. 5. Thoracic aortic aneurysms in VSMC-targeted, Secisbp2 knockout mice.

A Secisbp2 mRNA expression in aortic media or adventitia tissue isolated from wild-type (+/+: grey), heterozygous (+/−: orange) and homozygous (−/−: blue) VSMC-specific Secisbp2 knockout mice. Statistics: ordinary one-way ANOVA with adjusted p values (Tukey’s multiple comparison test), each bar represents the mean (n = 4–7 animals per genotype), error bars represent SEM. B Expression of selenoproteins (Gpx4, SelenoO, Sephs2) or β-actin (loading control) in aortic media tissue isolated from wild-type (+/+), heterozygous (+/−) and homozygous (−/−) VSMC-specific Secisbp2 knockout mice (n = 2 animals per genotype). C and D Percent survival of animals (C) or aortic aneurysm frequency (D), following angiotensin II infusion of wild type (+/+: grey), heterozygous (+/−: orange) or homozygous (−/−: blue) VSMC-specific, Secisbp2 knockout mice. Statistics: Two sample Z test of proportions, two-tail for −/− vs. +/− plus +/+ group comparison (n = 10–14 animals per genotype). E Histology of haematoxylin and eosin (H&E)-stained TAA (from angiotensin II-treated homozygous VSMC-specific Secisbp2 knockout mouse showing laminated thrombus within the vessel wall (asterisk) and cystic medial necrosis (arrowed inset). n = 10 independent experiments with similar results, scale bars = 300 μm. F Histology of thoracic aorta stained with H&E, EVG, Alcian Blue and von Kossa stains, or immunohistochemistry for 8-oxo-G, γ-H2AX or TUNEL with SMA from wild-type (+/+), heterozygous (+/−) and homozygous (−/−) VSMC-specific Secisbp2 knockout mice exposed to angiotensin II (as detailed in Supplementary Fig. 9B). Insets show high power views of outlined areas. Arrows indicate γ-H2AX or TUNEL-positive, αSMA-expressing cells, n = 9–14 independent experiments with similar results, scale bars = 300 μm. G–I Mean number of 8-oxo-G (G), γ-H2AX (H) and TUNEL (I)-positive cells in thoracic aorta sections of wild-type (+/+: grey), heterozygous (+/−: orange) and homozygous (−/−: blue) VSMC-specific Secisbp2 knockout mice. *p < 0.05 1-Way ANOVA. n = 9–14 mice 5 sections/mouse. Statistics: ordinary one-way ANOVA with adjusted p values (Tukey’s multiple comparison test), each bar represents the mean (n = 9–14 mice, 5 sections/mouse), error bars represent SEM. The genetic background of the Myh11-Cre^ERt2/Secisbp2^flox/flox mice generates only male, VSMC-specific knockout animals (C57BL/6N, 13 weeks old). Source data are provided as a Source Data file.