Genetic predisposition to porto-sinusoidal vascular disorder: A functional genomic-based, multigenerational family study

Abstract

Background and aims: Porto-sinusoidal vascular disorder (PSVD) is a group of liver vascular diseases featuring lesions encompassing the portal venules and sinusoids unaccompanied by cirrhosis, irrespective of the presence/absence of portal hypertension. It can occur secondary to coagulation disorders or insult by toxic agents. However, the cause of PSVD remains unknown in most cases. Hereditary cases of PSVD are exceptionally rare, but they are of particular interest and may unveil genetic alterations and molecular mechanisms associated with the disease.

Approach and results: We performed genome sequencing of four patients and two healthy individuals of a large multigenerational Lebanese family with PSVD and identified a heterozygous deleterious variant (c.547C>T, p.R183W) of FCH and double SH3 domains 1 ( FCHSD1 ), an uncharacterized gene, in patients. This variant segregated with the disease, and its pattern of inheritance was suggestive of autosomal dominant with variable expressivity. RNA structural modelling of human FCHSD1 suggests that the C-to-T substitution at position 547, corresponding to FCHSD1R183W , may increase both messenger RNA (mRNA) and protein stability and its interaction with MTOR-associated protein, LST8 homolog, a key protein of the mechanistic target of rapamycin (mTOR pathway). These predictions were substantiated by biochemical analyses, which showed that FCHSD1R183W induced high FCHSD1 mRNA stability, overexpression of FCHSD1 protein, and an increase in mTORC1 activation. This human FCHSD1 variant was introduced into mice through CRISPR/Cas9 genome editing. Nine out of the 15 mice carrying the human FCHSD1R183W variant mimicked the phenotype of human PSVD, including splenomegaly and enlarged portal vein.

Conclusions: Aberrant FCHSD1 structure and function leads to mTOR pathway overactivation and may cause PSVD.

Graphical abstract

FIGURE 1

Pedigree chart of the family and the proband's liver histopathology analysis. (A) Autosomal dominant inheritance pattern of family members with idiopathic noncirrhotic portal hypertension. *The participants had their genomic DNA analyzed by whole‐genome sequencing. (B) Proband's (IV‐2) liver histopathology. I: Masson stain showing septal fibrosis (×200); II: Aberrant vessels in portal tracts (hematoxylin and eosin, ×400); III: Reticulin stain showing features of nodular regenerative hyperplasia (×100). (C) Chromatogram of Sanger sequencing results of chr5: 141028548‐141028559. The red box labels the chromatogram at chr5: 141028553.

FIGURE 2

The effect of C>T mutation on FCH and double SH3 domains 1 (FCHSD1) expression. (A) The protein level of FCHSD1 in HepG2 cells (control) and HepG2 cells with overexpressed wild‐type (WT) FCHSD1 (FCHSD1^WT) and R183W mutant FCHSD1 (FCHSD1^R183W). The FCHSD1 protein was revealed by anti‐FCHSD1 antibody. (B) The protein level of FCHSD1 in FCHSD1^WT and FCHSD1^R183W HepG2 cells with/without overnight treatment of actinomycin D (ActD; 1 μm) or MG132 (1 μm). (C) The mRNA levels of FCHSD1 in FCHSD1^WT and FCHSD1^R183W HepG2 cells with/without ActD treatment for 6 h. ***p < 0.001; ****p < 0.0001. (D) Predicted morphological differences of the short RNA sequence of WT and R183W FCHSD1. The nucleobases are represented as green sticks, and the folded single strand is represented as gray transparent surface and orange cartoons. 3′ and 5′ ends are colored in red. The triplet of nucleobases containing the point mutation is framed in red, and the position is indicated with asterisks. (A–C) Representative data from one of at least two independent experiments. The relative expression of FCHSD1 to GAPDH is the normalized ratio of integrated density measured by ImageJ.

FIGURE 3

FCH and double SH3 domains 1 (FCHSD1) binds to MTOR‐associated protein, LST8 homolog (mLST8) and activates mechanistic target of rapamycin (mTOR) signaling. Protein–protein docking analysis shows that mLST8 interacts with mTOR1 and FCHSD1 and form two specific protein–protein interfaces. (A) Crystal structure of mLST8 (magenta surface) in its conformation activating mTOR1 (green surface, Protein Data Bank identifier: 4JSN). Protein–protein interface is represented as yellow surface. (B) Proposed docking model between FCHSD1 and mLST8 represented in green and magenta surfaces respectively. Yellow residues binding mTOR have the same coloring scheme as in (A). (C) Coimmunoprecipitation (IP) analysis of mLST8 and wild‐type (WT) FCHSD1 (FCHSD1^WT) or FCHSD1^R183W. The gray box indicates an unspecific band in all four IP lanes. The relative quantification (RQ) of immunoprecipitated mLST8 to FCHSD1 is the normalized ratio of integrated density measured by ImageJ. (D) The phosphorylation of mTOR Ser2448 and p70S6 kinase Thr389 in HepG2 (CRTL) and FCHSD1 overexpressed (FCHSD1^OE) HepG2 cells. The phosphorylation ratio equals (phosphor‐signal/GAPDH‐signal)/(total‐signal/GAPDH‐signal). (C, D) Representative data from one of at least two independent experiments. IB, immunoblot.

FIGURE 4

The characterization of R181W knock‐in mice. (A) The body weight comparison between wild‐type (WT) and R181W knock‐in mice at age 24 weeks. Het: heterozygote; Hom: Homozygote. *p < 0.05; **p < 0.01. (B) Representative images of liver and spleen of R181W knock‐in mice and WT mice. (C) Dissected livers and spleens of WT mouse and R181W carrier sacrificed at 44 weeks. (D) Hematoxylin and eosin staining of liver, portal vein, liver artery, and spleen of WT mouse and R181W carrier with enlarged portal vein sacrificed at 44 weeks.