The p.M292T NDUFS2 mutation causes complex I-deficient Leigh syndrome in multiple families

Abstract

Isolated complex I deficiency is the most frequently observed oxidative phosphorylation defect in children with mitochondrial disease, leading to a diverse range of clinical presentations, including Leigh syndrome. For most patients the genetic cause of the biochemical defect remains unknown due to incomplete understanding of the complex I assembly process. Nonetheless, a plethora of pathogenic mutations have been described to date in the seven mitochondrial-encoded subunits of complex I as well as in 12 of the nuclear-encoded subunits and in six assembly factors. Whilst several mitochondrial DNA mutations are recurrent, the majority of these mutations are reported in single families. We have sequenced core structural and functional nuclear-encoded subunits of complex I in a cohort of 34 paediatric patients with isolated complex I deficiency, identifying pathogenic mutations in 6 patients. These included a novel homozygous NDUFS1 mutation in an Asian child with Leigh syndrome, a previously identified NDUFS8 mutation (c.236C>T, p.P79L) in a second Asian child with Leigh-like syndrome and six novel, compound heterozygous NDUFS2 mutations in four white Caucasian patients with Leigh or Leigh-like syndrome. Three of these children harboured an identical NDUFS2 mutation (c.875T>C, p.M292T), which was also identified in conjunction with a novel NDUFS2 splice site mutation (c.866+4A>G) in a fourth Caucasian child who presented to a different diagnostic centre, with microsatellite and single nucleotide polymorphism analyses indicating that this was due to an ancient common founder event. Our results confirm that NDUFS2 is a mutational hotspot in Caucasian children with isolated complex I deficiency and recommend the routine diagnostic investigation of this gene in patients with Leigh or Leigh-like phenotypes.

Figure 1

NDUFS8 homozygous mutation, c.236C>T (p.P79L). (A) Family 34 pedigree, (B) sequence electropherograms of NDUFS8 gene with c.236 position highlighted in patient and family members and (C) amino acid alignment of NDUFS8 orthologues. Alignments were generated with ClustalW using GenBank sequences from Mus musculus (NP_659119.2), Rattus norvegicus (NP_001099792.1), Homo sapiens (NP_002487.1), Pan troglodytes (NP_001065248), Canis lupus familiaris (XP_852209.1), Bos taurus (NP_777243.2), Danio rerio (NP_998304.1), Drosophila melanogaster (NP_524719.1), Arabidopsis thaliana (NP_178022.1), Oryza sativa (NP_001051420.1) and Caenorhabditis elegans (NP_498595.1). Position p.79 (Homo sapiens) is indicated. Conserved sites are shown (asterisk).

Figure 2

NDUFS1 homozygous mutation, c.1222C>T (p.R408C). (A) Family 5 pedigree, (B) sequence electropherograms of NDUFS1 gene with c.1222 position highlighted in patient and family members, (C) confirmation of homozygous c.1222C>T mutation by PCR–RFLP. U = uncut; C1 = Control 1; C2 = Control 2. RFLP products were separated through a 12% non-denaturing polyacrylamide gel. The wild-type PCR product contains a single NlaIII site, which cuts the 218-bp amplicon into two fragments of 193 and 25 bp. In amplicons harbouring the c.1222C>T mutation, a second NlaIII site is created that cuts the 193-bp fragment into two smaller fragments of 148 and 45 bp. Fragment sizes (bp) are shown on the left; (D) amino acid alignment of NDUFS1 orthologues. Alignments were generated with ClustalW using GenBank sequences from Mus musculus (NP_663493.1), Rattus norvegicus (NP_001005550.1), Homo sapiens (NP_004997.4), Pan troglodytes (XP_516047.2), Canis lupus familiaris (XP_859697.1), Bos taurus (NP_777245.1), Danio rerio (NP_001007766.1), Drosophila melanogaster (NP_727255.1), Arabidopsis thaliana (NP_568550.1), Oryza sativa (NP_001051072.1) and Caenorhabditis elegans (NP_503733.1). Position p.408 (Homo sapiens) is indicated. Conserved sites are indicated (asterisk).

Figure 3

NDUFS2 heterozygous mutations, c.353G>A (p.R118Q), c.413G>A (p.R138Q), c.442G>A (p.E148 K), c.875T>C (p.M292T), c.998G>A (p.R333Q) and c.1328 T>A (p.M443 K). (A) Sequence electropherograms of NDUFS2 gene with c.353, c.413, c.442, c.875, c.998 and c.1328 positions highlighted in Patients (P) 3, 19, 22 and 27; (B) amino acid alignments of NDUFS2 orthologues. Alignments were generated with ClustalW using GenBank sequences from Mus musculus (NP_694704.1), Rattus norvegicus (NP_001011907.1), Homo sapiens (NP_004541.1), Pan troglodytes (XP_001173728.1), Canis lupus familiaris (XP_536138.2), Bos taurus (NP_001068605.1), Danio rerio (NP_001018481.1), Drosophila melanogaster (NP_651926.1), Arabidopsis thaliana (NP_085511.1), Oryza sativa (YP_514643.1) and Caenorhabditis elegans (NP_498423.1). Positions p.118, p.138, p.148, p.292, p.333 and p.443 (Homo sapiens) are indicated. Conserved sites are indicated (asterisk).

Figure 4

Complex I in-gel activity and assembly in patient fibroblasts. (A) Complex I in-gel activity. Mitochondria-enriched protein fractions (75 μg) extracted from patient (P) and control (C) fibroblasts were separated on 5–13% 1D blue-native PAGE gels. Gels were histochemically stained for 2 h with 2 mM Tris–HCl (pH 7.4), 0.1 mg/ml NADH and 2.5 mg/ml nitrotetrazolium blue to assess complex I activity. (B) Complex I assembly. Mitochondria-enriched protein fractions (75 μg) extracted from patient and control fibroblasts were separated on 5–13% 1D blue-native PAGE gels. Proteins were transferred to a polyvinylidine fluoride membrane and probed with antibodies raised against complex I NDUFS3 and complex II 70 kDa subunits. (C) Complex I subunit expression. Total cellular protein lysates (10 μg) of patient and control fibroblasts were separated on 10% sodium dodecyl sulphate–PAGE gels then transferred to a polyvinylidine fluoride membrane. Western blotting was performed with antibodies raised against complex I subunits NDUFS9, NDUFS3 and NDUFB8, complex II 70 kDa subunit and β-actin.

Figure 5

Reactive oxygen species levels and mitochondrial mass in patient fibroblasts. Mitochondrial superoxide levels were determined using MitoSOX (white bars) and mitochondrial mass was assessed with nonyl acridine orange (black bars) in fibroblasts of Patients 3, 5, 19, 22 and 27. Values are mean and standard error from at least three independent experiments. a.U. = arbitrary units; C1 = Control 1 (human neonatal skin fibroblasts); C2 = Control 2 (healthy paediatric skin fibroblasts); C3 = Control 3 (MRC-5 lung fibrobasts); NAO = nonyl acridine orange; *Bonferroni corrected P < 0.05.