BIOCHEMICAL AND GENETIC APPROACHES TO UNRAVEL MITOCHONDRIAL COMPLEX I DEFICIENCY

Using biochemical and genetic approaches we have been able to identify the genetic defect underlying mitochondrial complex I deficiency in 3 patients out of the 12 patients in the cohort. The patient cohort investigated in this study is genetically heterogeneous, originating from several different geographical areas: England, France, the Middle East, Israel, India and Pakistan. They presented with different clinical phenotypes: Leigh syndrome, congenital lactic acidosis, hypertrophic cardiomyopathy, encephalopathy, developmental delay, Alpers’ disease. After having excluded the mitochondrial DNA molecule for carrying mutations in muscle (analysis carried out by the diagnostic laboratories’ staff, at the National Hospital for Neurology, Queen Square, London), a biochemical investigation was undertaken in patients’ fibroblasts: the complex I activity defect was confirmed in this tissue as well for all the patients, and Blue Native studies were carried out. Antibodies against several subunits of complex I identified various subassemblies of the ~1MDa holoenzyme in several patients; in some others various degrees of reduction in holo-complex I content were observed. Analysis of the complex I pattern on Blue Native gels led the genetic screening towards a subset of genes, already known to be involved in complex I deficiency. Two families were also run on the Affymetrix 10K SNP chip array and then a homozygosity mapping approach was undertaken on the assumption that the affected individuals inherited two copies of the same ancestral mutated allele from a common ancestor (autozygosity). A subsequent bioinformatics analysis (also involving the implementation of the Maestro and MitoCarta databases) allowed the selection of a subgroup of genes that could possibly bear the genetic defect; this was done taking into account the Blue Native complex I pattern as well. Genetic screening identified a novel 8bp frameshift deletion (c.377_384del; Q126fsX2) in the NDUFS4 gene as cause of the disease in a patient from the first pedigree analyzed by homozygosity mapping. Her cousin was heterozygous for the same defect, but no other mutation has been identified, leaving this complex I deficiency case unsolved. In the second pedigree analyzed by homozygosity mapping approach, a homozygous mutation in a novel complex I assembly factor never previously linked to human disease has been identified. The c.1054C>T; R352W mutation in FOXRED1 segregated with disease in the family and was not found in 268 healthy control alleles. Western blot analysis showed a reduced steady-state level of FOXRED1 in patient fibroblasts and restoration of complex I activity after lentiviral transduction of patient fibroblasts with wild-type FOXRED1 cDNA. Finally, by candidate gene sequencing, two novel compound heterozygous mutations in NDUFAF1 were identified in a third patient: c.631C>T; R211C and c.733G>A; G245R. In summary, this study allowed the identification of mutations in (i) one complex I subunit, NDUFS4, already associated with complex I deficiency; (ii) one novel complex I assembly factor: FOXRED1; (iii) and NDUFAF1, a known complex I assembly factor whose mutations give a similar complex I Blue Native pattern to the one observed in our patient. In conclusion our combined approach proved to be efficient in the identification of the genetic defect in patients affected with complex I deficiency.