FOXE3 mutations predispose to thoracic aortic aneurysms and dissections

Abstract

The ascending thoracic aorta is designed to withstand biomechanical forces from pulsatile blood. Thoracic aortic aneurysms and acute aortic dissections (TAADs) occur as a result of genetically triggered defects in aortic structure and a dysfunctional response to these forces. Here, we describe mutations in the forkhead transcription factor FOXE3 that predispose mutation-bearing individuals to TAAD. We performed exome sequencing of a large family with multiple members with TAADs and identified a rare variant in FOXE3 with an altered amino acid in the DNA-binding domain (p.Asp153His) that segregated with disease in this family. Additional pathogenic FOXE3 variants were identified in unrelated TAAD families. In mice, Foxe3 deficiency reduced smooth muscle cell (SMC) density and impaired SMC differentiation in the ascending aorta. Foxe3 expression was induced in aortic SMCs after transverse aortic constriction, and Foxe3 deficiency increased SMC apoptosis and ascending aortic rupture with increased aortic pressure. These phenotypes were rescued by inhibiting p53 activity, either by administration of a p53 inhibitor (pifithrin-α), or by crossing Foxe3-/- mice with p53-/- mice. Our data demonstrate that FOXE3 mutations lead to a reduced number of aortic SMCs during development and increased SMC apoptosis in the ascending aorta in response to increased biomechanical forces, thus defining an additional molecular pathway that leads to familial thoracic aortic disease.

Figures

Figure 1. Identification of FOXE3 mutations as a cause of FTAADs.

(A) Pedigree of family TAA337, with the legend showing the designation of the disease and mutation status of the family members. The age at diagnosis of aortic root enlargement and/or dissection (dx.) is shown in years, and “d.” indicates age at death. The diagonal line across the symbol indicates that the individual is deceased. DNA from the individuals circled in red was used for exome sequencing. (B) Ascending aortic pathology in individual III:12 from family TAA337 and control aorta. H&E and Movat pentachrome staining of ascending aortic media from III:12 demonstrated loss of SMCs (red in Movat stain) between intact elastin fibers (black; original magnification, ×400). (C) Pedigree of MS300 with a FOXE3 p.Gly137Ala mutation. (D) Schematic representation of FOXE3 protein domain structure and location of rare variants. The FOXE3 rare variants identified in this study are on the top of the protein diagram. Red color indicates variants likely to be pathogenic for FTAAD. The FOXE3 mutations that were reported to be causative for ocular lens abnormalities are on the bottom of protein diagram. Blue color indicates dominant mutations; green color indicates recessive mutations.

Figure 2. Knockdown of foxe3 disrupts aortic arch development in zebrafish.

(A) Representative images of control and foxe3 MO–injected Tg(flk1:EGFP) zebrafish at 105 hours post fertilization (hpf). The majority of foxe3 morphants displayed an incomplete or missing aortic arch assembly when compared with control MO–injected embryos. Red letters (AA1–AA6) indicate the location of normal and abnormal aortic arches in control and foxe3 MO–injected zebrafish. Original magnification, ×10. (B) Frequency of aortic arch abnormalities associated with embryos injected with the following: 4 ng scramble MO, foxe3 MO; human FOXE3; WT human mRNA plus foxe3 MO; mutant (Mut) FOXE3 mRNA; mutant FOXE3 mRNA (harboring p.Asp153His); mutant FOXE3 mRNA plus foxe3 MO; p53MO; and p53 MO plus foxe3 MO. Aortic arch development was disrupted in 70% of foxe3 morphants. Coinjection of the foxe3 MO with human FOXE3 mRNA decreased the percentage of embryos with abnormal arch defects from 70% to 21% (P = 0.0003). However, coinjection of FOXE3 mRNA harboring the mutation p.Asp153His rescued significantly fewer embryos than did WT mRNA (52%, P = 0.001). Injection of WT or mutant FOXE3 mRNA alone had no significant effect on aortic arch morphology. Coinjection of the p53 MO also rescued foxe3 MO–induced aortic arch development. **P < 0.01, by 1-way ANOVA.

Figure 3. Mice express Foxe3 in PAs during development, and Foxe3–/– mice have reduced medial cell density and total numbers of medial cells in the ascending aorta and aortic arch.

(A) Expression of the Foxe3 transcription factor. Whole-mount ISH of WT mouse embryos (E9.5–E10.5) showed Foxe3 expression in the lens of the eye, in the forebrain/midbrain, and in the PAs; no expression was detected in the heart or its outflow tract. The higher-magnification image shows Foxe3 expression in PA1 and PA2, with weak expression in PA4. (B) Mouse aorta illustrates the regions of the aorta that were analyzed, indicating the ascending aorta, aortic arch, and DES1 and DES2. Original magnification, ×100. (C) Representative H&E-stained cross sections of ascending aorta, aortic arch, and descending aortae from male WT and Foxe3–/– mice at age P1, 4 weeks, and 6 months. (D) Bar graphs show that medial cell density (mean number of medial cell nuclei divided by the total medial area) and total number of medial cells per cross section of the ascending aorta, aortic arch, or descending aorta at day 1, 4 weeks, or 6 months of age were lower in the Foxe3–/– mice compared with WT mice. (E) Medial cell density and total number of medial cells were significantly lower in the ascending aortae and aortic arches in Foxe3–/– mice than in WT mice, but there was no statistical significant difference for the descending aorta. n = 5 per group. Error bars indicate SD from the average. *P < 0.05, by Student’s t test. Scale bars: 100 μm. ASC, ascending aorta.

Figure 4. Foxe3–/– ascending aortae have reduced expression of SMC genes encoding contractile protein but show no difference in proliferating or TUNEL-positive cell levels compared with WT controls.

(A) Anti–p-H3 staining (p-H3, for proliferative cells) of representative cross sections of ascending aortae from WT and Foxe3–/– mice at 4 weeks and 6 months of age. The cellular proliferation marker showed no differences between WT and Foxe3–/– aortae from mice of the same age. Scale bars: 100 μm. (B) TUNEL staining in ascending aortic sections from WT and Foxe3–/– mice at 4 weeks and 6 months of age. No differences in the percentage of TUNEL-positive cells were observed in aortae from WT and Foxe3–/– mice of the same age. Scale bars: 100 μm. Error bars indicate SD from the average. (C) Real-time qPCR analysis of SMC differentiation marker genes (Acta2, Myh11, Cnn1, and Tagln) using RNA from ascending and descending aortic tissues from WT and Foxe3–/– mice (expression levels were normalized to Gapdh). n = 5 per group. **P < 0.01, by Student’s t test. (D) Western blots for the SMC contractile proteins α-SMA, smooth muscle myosin heavy chain (SM-MHC), and calponin (CNN1) in ascending aortae, thoracic descending aortae, and explanted ascending aortic SMCs from WT and Foxe3–/– mice. Bar graphs show densitometric quantification for α-SMA, SM-MHC, and CNN1 in ascending aortae, descending aortae, and SMCs from WT and Foxe3–/– mice. Protein levels were significantly lower in the ascending aortae and explanted SMCs from Foxe3–/– mice compared with levels detected in WT mice, but there was no difference in levels in the descending aortae. n = 5 per group. *P < 0.05, by Student’s t test. Error bars indicate SD from the average. DES, descending aorta.

Figure 5. Foxe3–/– mice have altered aortic remodeling and rupture of the aorta after TAC.

(A) qPCR analysis of Foxe3 expression in WT ascending aortae 2 weeks after TAC indicated that Foxe3 expression was induced with TAC. n = 5 per group. **P < 0.01, by Student’s t test. (B) Cross sections of ascending aortae from mice before and after TAC, immunostained with anti-FOXE3 Ab. Foxe3 staining was not detected in aortic medial layer tissue from WT mice at baseline or in tissue from Foxe3–/– mice, but there was nonspecific staining in the aortic adventitial layer from Foxe3–/– and WT mice. FOXE3 staining was detected in the medial layer of ascending aortae from WT mice after TAC; adventitial layer staining was also present but may have been nonspecific. (C) Echocardiograms show the maximum (Max) systolic velocities and heart rate (bpm) in ascending aortae after sham operation or TAC. (D) Whole mounts of aortae from mice after TAC; arrows show banding location. Half of the aortae from Foxe3–/– mice were similar to those from WT mice (upper 2 panels), but half of the Foxe3–/– mice had complications including aortic hematoma formation (left lower panel, red arrow) or death due to aortic rupture (right lower panel; green arrow). Original magnification, ×15. (E) Echocardiographic assessment of the ascending aortic lumen in WT and Foxe3–/– mice after sham operation or TAC. n = 6 per group. *P < 0.05, by 1-way ANOVA. (F) Cross sections of ascending aortae from WT and Foxe3–/– mice 2 weeks after TAC with H&E, TUNEL, and anti–p-H3 staining. Blue line shows the boundary between the media and adventitia. Bar graphs show the quantification of aortic wall thickness, total medial area, total adventitial area, and the percentage of TUNEL-positive and anti–p-H3–positive medial cells. n = 5 per group. *P < 0.05, by Student’s t test.

Figure 6. Aortic developmental defects and aberrant aortic remodeling with TAC in Foxe3–/– mice is rescued by blocking p53.

(A) Annexin V assay by flow cytometric analysis. After a 24-hour exposure to staurosporine, the percentage of annexin V–positive cells was significantly greater in Foxe3–/– SMCs compared with that seen in WT SMCs. Inhibition of p53 with PFT significantly decreased the percentage of annexin V–positive Foxe3–/– SMCs to the level detected in WT SMCs. *P < 0.05, by Student’s t test. (B) Echocardiographic analysis of the ascending aortic diameter shows that PFT treatment rescued the increased aortic diameter with TAC in Foxe3–/– mice. n = 5 per group. *P < 0.05, by 1-way ANOVA. (C) TUNEL staining in ascending aortic sections from WT and Foxe3–/– mice with TAC and PFT treatment. Bar graphs show that the proportion of TUNEL-positive medial cells in the Foxe3–/– mice decreased with PFT treatment. n = 5 per group. * P < 0.05, by Student’s t test. Original magnification, ×400. (D) Representative images of H&E-stained cross sections of ascending aortae from 4-week-old male WT, Foxe3–/–, p53–/–, and Foxe3–/–p53–/– mice. Bar graphs show their medial cell densities (mean number of medial cell nuclei divided by the total medial area) and the total number of medial cells per cross section at the ascending aorta. Deletion of p53 effectively restored medial cell density and cell numbers in the ascending aortic media of Foxe3–/– mice and eliminated the differences among Foxe3–/–p53–/–, p53–/–, and WT mice of the same age. n = 6 per group. *P < 0.05, by 1-way ANOVA. Original magnification, ×200. (E) qPCR analysis of SMC-specific gene (Acta2, Myh11, and Cnn1) expression in ascending aortic tissues from WT, Foxe3–/–, and Foxe3–/–p53–/– mice. Gene expression levels were normalized to Gapdh. n = 5 per group. *P < 0.05 and **P < 0.01, by 1-way ANOVA.

Figure 7. Loss of p53 rescues TAC-induced aortic rupture and SMC apoptosis in Foxe3-deficient aortae.

(A) Representative images of ascending aortic cross sections from WT, Foxe3–/–, and Foxe3–/–p53–/– mice 2 weeks after TAC. Cross sections of ascending aortae were stained with H&E, VVG elastin, Movat pentachrome, TUNEL , and anti–p-H3. Scale bars: 100 μm. (B) Echocardiographic analysis of the aortic diameter of ascending aortae from WT, Foxe3–/–, and Foxe3–/–p53–/– mice 2 weeks after TAC. The diameters of the ascending aortae from Foxe3–/–p53–/– mice 2 weeks after TAC were similar to diameters of aortae from WT mice. n = 5 per group. *P < 0.05, by 1-way ANOVA. (C) Bar graphs show the quantification of aortic wall thickness, total medial area, total adventitial area, medial cell density (mean number of medial cell nuclei divided by medial area [mm2]), percentage of TUNEL-positive aortic medial cells, and percentage of medial cells stained with anti–p-H3 Abs. p53 deficiency rescued TAC-induced aberrant aortic remodeling in Foxe3–/– mice to an extent indistinguishable from that observed in WT mice, as shown by the increased medial and adventitial areas and number of TUNEL-positive medial cells as well as the increased number of p-H3–positive medial cells. n = 5 per group. *P < 0.05 and **P < 0.01, by 1-way ANOVA. Error bars indicate SD from the average.

Figure 8. Loss of p53 rescues SMC loss with Foxe3 deficiency through Cdkn1a.

(A) qPCR analysis of Cdkn1a mRNA expression in ascending aortae from WT, Foxe3–/–, and Foxe3–/–p53–/– mice at baseline and 2 weeks after TAC. Cdkn1a was significantly induced with TAC, and expression levels were significantly higher in the TAC aortae from Foxe3–/– mice when compared with levels in aortae from WT mice. Deletion of p53 significantly restored Cdkn1a expression levels to those detected in WT aortae. n = 5 per group. **P < 0.01, by 1-way ANOVA. (B) Quantification of aortic arch abnormalities associated with embryos injected with the control MO, foxe3 MO, p53 MO, or cdkna1 MO. Knockdown of foxe3 with the foxe3 MO disrupted development of the aortic arches in zebrafish. Coinjection of the foxe3 MO with the p53 MO significantly rescued the aortic defect. Injection of the cdkna1 MO also partially rescued foxe3 MO–induced abnormal aortic arch development in zebrafish. **P < 0.01, by 1-way ANOVA.