Abstract

Introduction

Leigh syndrome (OMIM: #256000) is a rare and severe autosomal recessive mitochondrial disease characterized by the progressive deterioration of the central nervous system (Rahman, 2023). This disorder, also called subacute necrotizing encephalomyelopathy, was first reported by British neuropathologist Denis A. Leigh in 1951 (Leigh, 1951). In general, the estimated incidence rate of the disease is 1 per 40,000 newborn individuals worldwide (Darin, 2001); however, there are regions with a higher incidence, such as the Saguenay Lac-Saint-Jean region of Quebec, Canada (1 in 2000 newborns) and the Faroe Islands located between Iceland and Scotland (1 in 1700 newborns). Symptoms include global developmental delay, failure to thrive, ataxia, seizure, respiratory failure, hypotonia, dystonia, and ophthalmologic abnormalities, such as nystagmus, optic atrophy, ptosis, and strabismus (Rahman et al., 1996; Finsterer, 2008). Aside from these prevalent features, problems in other organ systems (cardiac and gastrointestinal) have also been reported in some cases (Finsterer, 2008). Clinical symptoms usually manifest in infancy between three and 24 months of age but later-onset forms of disease have also been described (Rahman et al., 2023).

Leigh syndrome usually manifests from a defect in mitochondrial energy production caused by deleterious changes in genes encoding enzymes of the mitochondrial respiratory chain or pyruvate dehydrogenase complex. Perhaps surprisingly, early studies indicated that many such genes are encoded by the nuclear genome (Miranda et al., 1989). Decades of research have revealed that this complex neurodegenerative disease comprises more than 100 distinct monogenic disorders with substantial clinical heterogeneity (Rahman et al., 1996). These include maternally inherited mitochondrial genes such as ND1-6, COX III, ATPase tRNA, and nuclear genome encoded genes including NDUFV1-2, NDUFS1-4, 7, 8, SDHA, SDH, COX15, SURF1, CoQ, PDSS2, PDHX1, LRP130, and PDHc (reviewed in Finsterer, 2008). Among them, SURF1 (SURF1 Cytochrome C Oxidase Assembly Factor), LRP130 (leukin-rich protein 130), and MT-ATP6 (Mitochondrially Encoded ATP Synthase Membrane Subunit) are the most frequently mutated genes.

Ten subgroups of methylglutaconic aciduria (Type I, II, III, IV, V, VI, VIIA, VIIB, VIII, and IX) have been identified to date (Gorman et al, 2016b). Type VIII (abbreviated MGCA8 by OMIM), associated with the HTRA2 gene (MIM: #606441), is one of the least-studied forms (Mandel et al., 2016). Only twelve MGCA8 patients from seven unrelated families have been reported in the literature as of 2023. (Mandel et al., 2016; Olahova et al., 2017; Kovacs-Nagy et al., 2018; Sreedhara et al., 2020). The paucity of genetic and phenotypic data makes diagnosis of this disorder especially challenging. Here, we describe a two-year-old male proband who demonstrated the clinical and molecular characteristics of Leigh-like syndrome. Despite extensive genetic and biochemical testing, the cause of his disorder remained elusive. The proband and his parents were consented for further studies under an IRB-approved research protocol at our hospital. Exome sequencing revealed that the proband was compound heterozygous for two novel variants in the HTRA2 gene, conferring a diagnosis of 3-methylglutaconic aciduria type VIII (MGCA8). These findings not only ended a long diagnostic odyssey for the family, but also shed crucial new light into the variable clinical manifestations of this ultra-rare disorder.

Case description

Variant detection and diagnostic assessment

The proband was compound heterozygous for two novel missense variants in the HTRA2 gene (transcript ID ref: NM_013247.5; see Figure 1B) located on chromosome 2p13.1. The maternally-inherited variant (c.1037A>T in exon 5) is absent from public databases (gnomADv3), making it extremely rare. The predicted amino acid change (p.Glu346Val) is benign by most computational tools (REVEL = 0.145). However, this variant creates a GT dinucleotide 8 bp upstream of the exon-intron boundary and is predicted by multiple computational tools to create a novel splice donor site (SpliceAI = 0.85, Pangolin = 0.76). Patient RNA was not available to confirm disrupted splicing. The paternally-inherited variant (c.1172T>A in exon 7) is also absent from public databases (gnomADv3). This amino acid change (p.Val391Glu) is predicted to be damaging by virtually all in silico tools including meta-predictors (REVEL = 0.868, MetaRNN = 0.9511). There were no alternate candidate genes or significant variants identified in our investigation.

Open in a separate window

FIGURE 1

HTRA2 protein structure with disease-causing variants and the family pedigree of the proband (A). The VanDyPlot v0.1.1 depicts the ClinVar database pathogenic variants plotted on the HTRA2 protein (UniProt ID: 043464). Variants are shown at their predicted protein position as colored circles reflecting the effect type, with the two missense variants (black lollipops) reported here shown in orange arrows, and the two blue lollipops at position 221 and 315 were the missense variants reported from the literature. The HTRA2 protein domain annotations are retrieved from Pfam database of curated protein families (B). Family pedigree showing the inheritance of two pathogenic compound heterozygous variants of the HTRA2 gene in the proband.

Under current ACMG guidelines (Richards et al., 2015), both HTRA2 variants are classified as Variants of Uncertain Significance (Table 1), but given the significant clinical correlation, the patient care team considers this a diagnostic result. Both variants were independently validated by Sanger sequencing in a CLIA/CAP certified clinical laboratory.

TABLE 1

Genomic findings and variant interpretation. Variant interpretation was performed using ACMG 2015 guidelines (Richards et al., 2015). Interpretation evidence codes include PM2 (rare or absent in populations), PP3 (computational studies show deleterious effect). For the maternal variant, evidence code PP3 is applied due to the variant’s proximity to the intron-exon boundary and a computationally predicted effect on splicing (SpliceAI score = 0.85, Pangolin score = 0.76). For the paternal variant, evidence code PP3 is applied based on computational predictions of damaging effect for the amino acid change (REVEL = 0.868, MetaRNN = 0.9511). We considered but did not apply PP4 (phenotype highly specific for gene) given the paucity of data about HTRA2-associated disease.

Gene	Variant (GRCh38)	Location	Inheritance	Amino acid change	Interpretation
HTRA2	2-74531694-A-T	Exon 5	Maternal	(p.Glu346Val)	VUS (PM2, PP3)
HTRA2	2-74532675-T-A	Exon 7	Paternal	(p.Val391Glu)	VUS (PM2, PP3)

Open in a separate window

At present, the ClinVar database (https://www.ncbi.nlm.nih.gov/clinvar) contains 9 pathogenic/likely pathogenic variants in HTRA2 (excluding large events and structural variants), including 6 truncating variants (3 frameshift, 2 nonsense, 1 splice site) and 3 nonsynonymous variants (2 missense, 1 inframe indel), see Figure 1A. The relatively small number of disease-causing variants reported for this gene makes variant interpretation challenging.

Discussion

Mitochondrial diseases are the most heterogenous and commonly inherited metabolic disorders, often presenting with a broad spectrum of clinical manifestations (Gorman et al., 2016a). One such disorder, 3-methylglutaconic aciduria type VIII (MGCA8, MIM: #617248), is a rare, heritable, autosomal recessive inborn error of metabolism associated with variants in the HTRA2 gene. Human HTRA2 is a 49 kDa nuclear-encoded mitochondrial protein composed of 458 amino acids. It belongs to the HTRA family of stress-induced serine proteases (Mandel et al., 2016) and it is essential for maintaining mitochondrial homeostasis. HTRA2 is also involved in apoptosis via caspase-dependent and independent pathways (Simón-Sánchez and Singleton, 2008; Mandel et al., 2016). Yet the precise mechanism by which HTRA2 deficiency causes mitochondrial disease remains unclear.

Among the relatively small number of MGCA8 patients reported in the literature to date, core clinical features include hypotonia, seizure, abnormal movements, encephalopathy with depressed sensorium, respiratory insufficiency, and developmental delay. The proband report here shared many of these features. However, he did not exhibit the markedly elevated 3-methylglutaconic and 3-methylglutaric acids in urine, which until recently was considered a unifying feature of this clinically variable, ultra-rare disorder (Olahova et al., 2017; Kovacs-Nagy et al., 2018), was also a unifying feature. However, we note that our patient survived longer (10 months) and underwent urine organic acid testing later (7.5 months of age) than all previously reported patients. Furthermore, a recently published MGCA8 patient from India exhibited the common clinical manifestations but showed only borderline elevated levels of 3-MGA (Sreedhara et al., 2020). It is therefore possible that this biochemical signature is common but not diagnostic for HTRA2 deficiency, similar to other forms of genetic 3-methylglutaconic aciduria such as Barth syndrome and 3-MGA2 (MIM: #302060) (Cantlay et al., 1999; Schmidt et al., 2004). Because our patient passed away before the genetic diagnosis was made, we were unable to repeat urine organic testing to confirm the absence of 3-MGA. In any case, the preponderance of evidence including the genetic findings, consistent clinical features, and rapid progression of symptoms in our patient support the finding that MGCA8 is the diagnosis.

Conclusion

In summary, we used research WES to identify compound-heterozygous missense variants in HTRA2 in a patient with Leigh-like mitochondrial disease who had extensive negative prior testing. Our findings end a long diagnostic odyssey for the patient family and shed further light on the clinical and genetic spectra of this ultra-rare recessive disorder. Although extremely rare, MGCA8 has proven a severe and consistently lethal disease, underscoring the need for rapid genomic testing in symptomatic patients with or without 3-MGA.

Acknowledgments

Funding Statement

The author(s) declare financial support was received for the research, authorship, and/or publication of this article. This work was supported by the Abigail Wexner Research Institute at the Nationwide Children’s Hospital.

References