A case of congenital heart defects and familial exudative vitreoretinopathy caused by activation of a cryptic splice donor in NOTCH1

Abstract

Background: NOTCH1 is associated with two disorders of vascular development, Adams-Oliver Syndrome 5 (AOS5) and aortic valve disease 1 (AOVD1). Here we report a disease-causing variant in NOTCH1 that has a previously undemonstrated effect on splicing. Additionally, we found that the proband has the optic phenotype of familial exudative vitreoretinopathy (FEVR) which has been reported for probands with pathogenic variants in genes in the notch signaling pathway, but never for NOTCH1.

Case presentation: The proband presented with a ventricular septal defect, pulmonic stenosis, and ocular findings consistent with familial exudative vitreoretinopathy (FEVR), which NOTCH1 has not been associated with to date. Trio exome sequencing identified a paternally inherited variant of uncertain significance in NOTCH1:c.2153 A > G. We assessed the variant's effect using RT-PCR, finding an increased use of a cryptic donor compared to the control. On this basis, we were able to re-classify this variant as pathogenic.

Conclusions: We expand the phenotypic spectrum of NOTCH1 and contribute to the building evidence that variants in NOTCH1 cause a spectrum of disorders of vascular development.

Keywords: NOTCH1; Adams-oliver syndrome; Familial exudative vitreoretinopathy; Rare disease; Splice variant.

Fig. 1

Pedigree of the proband. The proband had a family history of congenital heart defects with a father self-reporting a self-repairing septal defect. His brother had died in infancy in India of an unspecified heart condition

Fig. 2

NOTCH1:c.2153 A > G causes an ocular phenotype consistent with familial exudative vitreoretinopathy. (A) Color fundus photograph of the right eye shows a normal appearing optic nerve, macula, and posterior pole. (B-C) Color fundus photographs of the left eye demonstrate mild pallor of a tilted optic nerve; there are several vessels with abnormal courses (asterisks). There is a chorioretinal scar affecting the fovea (arrow). Fibrosis is seen along a prominent ridge (arrowheads) between the vascular and avascular retina at the equator temporally and inferotemporally. (D) Fundus fluorescein angiography of the left eye highlights late staining of the central macular scar (arrow) and the abnormal vasculature (asterisks), including along the temporal ridge. A small area of superior neovascularization is not captured in this image. (E) FA of the right is included for comparison

Fig. 3

NOTCH1:c.2153A > G causes activation of a cryptic splice donor. (A) Splice-AI predictions for NOTCH1:c.2153A > G visualized in MobiDetails (CHU Montpelier) [18, 19]. The blue bars represent the strength of a predicted donor. The green highlight represents the wildtype allele, while the red highlight represents the mutant allele. The variant is predicted to cause activation of a cryptic donor 4 bp upstream of the variant with a prediction strength of 0.91 (max score = 1). (B) RT-PCR product from the proband (done in duplicate) and a healthy control. (C) Eight EcoRI digested plasmids from the proband are included as a reference, with lanes 1–6 and 8 showing canonical splicing, and lane 7 showing the use of the cryptic donor. (D) RT-PCR demonstrates allele-specific expression and use of the cryptic donor. A sample wildtype sequence and presumed variant sequence are shown. We cannot definitively say that all use of the frameshifting donor (shown in blue) was due to the variant (position shown in red) because it is used at low levels in the general population. 18 of 22 colonies sequenced (81.8%) showed the wildtype sequence, while 4 (18.2%) showed the presumed variant sequence. The 21 colonies sequenced from the healthy controls all showed canonical splicing. The image was created in BioRender