Consolidating biallelic SDHD variants as a cause of mitochondrial complex II deficiency

Isolated mitochondrial complex II deficiency is a rare cause of mitochondrial respiratory chain disease. To date biallelic variants in three genes encoding mitochondrial complex II molecular components have been unequivocally associated with mitochondrial disease (SDHA/SDHB/SDHAF1). Additionally, variants in one further complex II component (SDHD) have been identified as a candidate cause of isolated mitochondrial complex II deficiency in just two unrelated affected individuals with clinical features consistent with mitochondrial disease, including progressive encephalomyopathy and lethal infantile cardiomyopathy. We present clinical and genomic investigations in four individuals from an extended Palestinian family with clinical features consistent with an autosomal recessive mitochondrial complex II deficiency, in which our genomic studies identified a homozygous NM_003002.3:c.[205 G > A];[205 G > A];p.[(Glu69Lys)];[(Glu69Lys)] SDHD variant as the likely cause. Reviewing previously published cases, these findings consolidate disruption of SDHD function as a cause of mitochondrial complex II deficiency and further define the phenotypic spectrum associated with SDHD gene variants.


Introduction
The mitochondrial oxidative phosphorylation (OXPHOS) system is composed of five multi-subunit transmembrane protein complexes (I-V) encoded by the mitochondrial and nuclear genomes, and is the primary mechanism for adenosine triphosphate production in eukaryotic cells. OXPHOS defects result in mitochondrial disease, with an estimated prevalence of 1:4300 [1,2].
Mitochondrial complex II (succinate dehydrogenase) is composed of two catalytic subunits (SDHA/SDHB) anchored to the inner mitochondrial membrane by two smaller subunits (SDHC/SDHD) [3,4]. Complex II differs from other mitochondrial respiratory chain complexes, in that the four structural subunits and their two assembly factors (SDHAF1/SDHAF2) are solely encoded by the nuclear genome. Complex II is also unique in being involved in both the mitochondrial respiratory chain and the Krebs cycle [2].

Materials and methods
Model 3abv (porcine heart mitochondrial complex II) [12] was selected from X-ray diffraction/NMR-derived structures of SDHD (O14521) and its homologues (RCSB Protein Data Bank). Amino acid residues were visualised with their polar bonds, and annotated using Pymol 2.3 [13,14].

Clinical findings
We describe four affected Palestinian patients (three male, one female) aged 4-20 years, comprising of two sibships from an extended interconnecting family (Fig. 1A). All four children presented with developmental delay in infancy and variable clinical and laboratory findings suggestive of a mitochondrial disorder including elevated serum lactate/ urinary Krebs cycle metabolites, nystagmus, optic atrophy, progressive microcephaly, generalised hypotonia, epileptic seizures, severe/profound intellectual disability/developmental impairment and cardiomyopathy. The affected children were not dysmorphic (Fig. 1C), though individuals V:2 and V:4 were noted to have significant hypertrichosis, particularly over their back and limbs. MRI neuroimaging was unremarkable for one child at 8 months (V:2; Fig. 1D), however, his sister's scan revealed delayed myelination at age 6 months (V:4). Hirschsprung disease, confirmed by aganglionic rectal biopsy, was noted in a single individual (V:2). A full description of the clinical features and disease progression is summarised in Table 1.

Genetic findings and homology modelling
Genome-wide SNP genotyping and WES were undertaken assuming that a homozygous founder variant was responsible, although also considering other genetic mechanisms. SNP genotyping (individuals V:2, V:4 and V:6) identified four notable (>1 Mb) shared homozygous regions, the two largest identified on chromosome 11; a~7.00 Mb region (rs6485795-rs11246414, Chr11:g.47908294-54905443 [hg38]) and a~2.42 Mb region (rs120436-rs12794326, Chr11:g.110826521-113248134) (Fig. 1F). WES was performed on DNA from affected individual V:4, to identify rare functional candidate variants. Variants were prioritised by call quality and frequency (gnomAD v2.1.1/1000 Genomes Project, MAF ≤ 0.0001) and cross referenced with SNP data, identifying only a single candidate homozygous variant of relevance to the phenotype in SDHD NM_003002. ) in a single individual with autosomal recessive encephalomyopathy and isolated mitochondrial complex II deficiency [6] (ClinVar accession:VCV000156153.8 and SCV001424558). Dideoxy sequencing confirmed cosegregation as appropriate for an autosomal recessive disorder (Fig. 1A, B ). Protein modelling positions the p.(Glu69Lys) substitution within the first transmembrane alpha helix, where it likely disrupts tertiary structure through interrupting a predicted hydrogen bond with Gln109 on the adjacent helix ( Supplementary Fig. S1A). Conversely, a previously described variant, p.(Asp92Gly) is located at the apex of transmembrane alpha helices one and two, in close proximity to the membrane and inter-membrane space ( Supplementary Fig. S1A, B).

Discussion
Here we define a homozygous SDHD c.[205G > A]; [205G > A];p.[(Glu69Lys)];[(Glu69Lys)] missense variant as the likely cause of isolated mitochondrial complex II deficiency in three affected children from an extended Palestinian family. DNA was unavailable for V:5 (deceased age 10 years), whose clinical history overlapped that of his sibling (V:6). Tissues and organs heavily dependent on robust OXPHOS processes tend to be most affected by mitochondrial disease [15], explaining why common findings include optic atrophy, leukoencephalopathy, myopathy, cardiomyopathy and Leigh syndrome. These clinical features overlap those described in the two individuals with SDHD-related mitochondrial disease reported to date (Table 1) Progressive ocular (visual impairment, nystagmus, optic disc pallor) and neurological (epileptic seizures, ataxia, dystonia and continuous intractable myoclonic movement) involvement were described, and the child died aged 10 years. Urinalysis revealed lactic aciduria, ketonuria and Krebs cycle intermediates. Complex II activity was deficient in skeletal muscle and complementation studies in patient fibroblasts showed restoration of complex II assembly and function with expression of wildtype, but not mutant, SDHD cDNA [6]. Subsequently an Irish male infant was described [7] homozygous for a novel SDHD c. The four affected individuals described here show phenotypic overlap with both these individuals (Table 1). Our study extends the clinical spectrum and highlights the wide range of phenotypical features and severity across affected individuals, even those with the same SDHD genotype (Table 1). Hypertrichosis, a recognised feature of some forms of mitochondrial disease (most notably SURF1-Leigh syndrome [16]), was a noted feature in two Palestinian children. Hirschsprung disease diagnosed in a single affected individual (V:2) has not been previously reported in association with SDHD variants, and it remains unclear whether this is an associated or unrelated feature. Neurodevelopmental regression is a common characteristic of mitochondrial disease, particularly during physiologic stress through intercurrent infection, prolonged fasting or dehydration [17]. It is thus unsurprising that this appears to be a common feature of complex II deficiency due to biallelic SDHD variants (Table 1). An accurate molecular diagnosis for complex II deficient patients would support avoidance of prolonged fasting and dehydration. Homology modelling of the two putative pathogenic SDHD missense variants thus far associated with isolated autosomal recessive mitochondrial complex II deficiency [c.
[205G > A];p.(Glu69Lys) and c.[275A > G];p.(Asp92Gly)] predicts disruption of non-covalent bonds between transmembrane helices and changes to complex II positioning in the inner mitochondrial membrane as likely outcomes ( Supplementary Fig. S1a). In addition to their role in primary mitochondrial disease, heterozygous germline variants in other complex II subunits and assembly factors (including SDHA, SDHB, SDHC, SDHD and SDHAF2) are associated with paragangliomas, phaeochromocytomas and gastrointestinal stromal tumours [3]. None of the three SDHD variants associated with mitochondrial complex II deficiency have been previously linked to tumourigenesis, including in this extended Palestinian family, although a Dutch founder familial paraganglioma SDHD variant c.274G > T; p.(Asp92Tyr) has been described [18]. Additionally, SDHA and SDHB variants have been associated with both mitochondrial complex II deficiency in biallelic form, and hereditary cancer susceptibility in monoallelic form [5,19]. Therefore routine surveillance of heterozygous SDHD carriers is suggested for early detection of paragangliomas and phaeochromocytomas and appropriate intervention. Together the data presented here consolidate biallelic SDHD variants as a cause of mitochondrial disease due to mitochondrial complex II malfunction, and extend the variable associated clinical features.