Pathogenic variants in HTRA2 cause an early-onset mitochondrial syndrome associated with 3-methylglutaconic aciduria

Abstract

Mitochondrial diseases collectively represent one of the most heterogeneous group of metabolic disorders. Symptoms can manifest at any age, presenting with isolated or multiple-organ involvement. Advances in next-generation sequencing strategies have greatly enhanced the diagnosis of patients with mitochondrial disease, particularly where a mitochondrial aetiology is strongly suspected yet OXPHOS activities in biopsied tissue samples appear normal. We used whole exome sequencing (WES) to identify the molecular basis of an early-onset mitochondrial syndrome-pathogenic biallelic variants in the HTRA2 gene, encoding a mitochondria-localised serine protease-in five subjects from two unrelated families characterised by seizures, neutropenia, hypotonia and cardio-respiratory problems. A unifying feature in all affected children was 3-methylglutaconic aciduria (3-MGA-uria), a common biochemical marker observed in some patients with mitochondrial dysfunction. Although functional studies of HTRA2 subjects' fibroblasts and skeletal muscle homogenates showed severely decreased levels of mutant HTRA2 protein, the structural subunits and complexes of the mitochondrial respiratory chain appeared normal. We did detect a profound defect in OPA1 processing in HTRA2-deficient fibroblasts, suggesting a role for HTRA2 in the regulation of mitochondrial dynamics and OPA1 proteolysis. In addition, investigated subject fibroblasts were more susceptible to apoptotic insults. Our data support recent studies that described important functions for HTRA2 in programmed cell death and confirm that patients with genetically-unresolved 3-MGA-uria should be screened by WES with pathogenic variants in the HTRA2 gene prioritised for further analysis.

Fig. 1

Molecular genetics and biochemical studies of HTRA2 variants. a, b Pedigrees of the two affected families with variants in the HTRA2 gene. The probands S1 (family 1) and S2 (family 2) are indicated by a red arrow and filled symbols denote affected individuals. c Amplification of cDNA across HTRA2 exons 2–5 showed an aberrant splicing pattern for HTRA2 RNA transcripts in S1 as two abnormal splice products were detected compared to the wild type control (C1). d, e Western blot analysis of HTRA2 levels in protein lysates isolated from control (C1, C2) and HTRA2 S1 skeletal muscle and fibroblasts and from HTRA2 S2 mitochondrial muscle homogenate. β-actin and SDHA antibodies were used as loading controls

Fig. 2

Steady-state levels and assembly of OXPHOS complexes. a, b SDS-PAGE (12 %) and immunoblot analysis of steady-state levels of OXPHOS complex subunits in protein extracts (30ug) isolated from age-matched control (C1, C2) and a HTRA2 S1 fibroblasts (upper panel) and skeletal muscle (lower panel). b Mitochondrial lysates (40ug) from control and HTRA2 subject’s (S2) skeletal muscle were used to analyse the levels of OXPHOS complex subunits. In a and b β-actin and SDHA served as loading controls. c, d One-dimensional BN-PAGE analysis of the assembly of individual OXPHOS complexes in DDM-solubilised mitochondrial extracts from control (C1, C2) and HTRA2 subject’s c fibroblasts (S1) and d skeletal muscle (S2). Complex II (SDHA) was used as a loading control

Fig. 3

Analysis of mitochondrial morphogenesis and apoptosis. a The effects of loss of HTRA on OPA1 protein levels in control (C1, C2) and subject (S1) fibroblasts and skeletal muscle were analysed by SDS-PAGE (7 %) and immunoblotting against the mitochondrial fusion protein OPA1. Differences in the OPA1 proteolytic cleavage pattern between control and subject samples were detected, with increased levels of short OPA1 cleavage products present in the HTRA2 subject (S1-S3). b The panels on the left show representative images of TMRM staining in control C1 (upper), control C2 (middle) and subject S1 (lower) fibroblasts, revealing a well-connected tubular mitochondrial network in subject 1 consistent with both controls. The panels on the right show representative images of nucleoid staining by PicoGreen in control C1 (upper), control C2 (middle) and subject S1 (lower) fibroblasts. Nucleoid size and distribution in S1 is comparable to both controls. Scale bar = 10 μm. c Quantitative analysis of mitochondrial network demonstrated a non-significant difference in aspect ratio (left panel) and form factor (right panel) of S1 compared to the controls. All data shown are represented as mean ± SEM from two independent experiments. Statistical analysis was performed using a two-tailed unpaired Student’s t-test, ns = not significant. d Representative SBF-SEM images of mitochondria from control (C1, C2) and HTRA2 subject (S1) fibroblasts showing morphologically normal mitochondria with no obvious defects in cristae structure. Scale bar = 1 μm. e Quantitative analysis of the percentage of apoptotic cells in population (n = 10 000) in control (C1) and subject (S1) fibroblasts treated with or without 0.3 μM Staurosporine (ST) for 8 h. Results are shown as mean STDEV (n = 2); paired Student’s t-test: C1 vs S1 p = n.s., C1 vs C1 + ST p = 0.01; S1 vs S1 + ST p = 0.01 and C1 + ST vs S1 + ST p = 0.03