Modelling ND subunits of complex I: Leber’s Hereditary Optic Neuropathy (LHON) pathogenic mutations and non-synonymous population variants in genotype-phenotype correlation

Complex I (NADH:ubiquinone oxidoreductase) is the first site of respiratory chain and is built by 45 subunits, 7 of which mtDNA-encoded (ND subunits). The hydrophilic arm, protruding into the mitochondrial matrix, has been resolved by protein crystallization and X-ray diffraction, whereas the structure of the hydrophobic intramembrane portion, mainly composed by ND subunits, is still unclear. Since 1988, when the first point mutation in the ND4 subunit was associated with LHON, an increasing number of mutations has been reported to affect complex I, the most represented in mtDNA-encoded genes. Similarly, variation of mtDNA in worldwide population also involves complex I with numerous non-synonymous changes clustered in different haplogroups. These variants have been shown to influence LHON expression, as well as aging and other disorders. The relevance of non-synonymous nucleotide changes in these subunits is frequently ambiguous, in terms of function and pathology. We here undertook a systematic computational analysis of conservation patterns in three subsets of complete ND sequences (eukaryotes, vertebrates, mammals), producing global alignments and calculating the percentage of amino acid conservation. Furthermore, we generated new topological models for ND subunits using three different hydropathy scales and tools based on machine-learning approaches. In order to identify the possible role of non-synonymous variants and pathogenic mutations, the corresponding amino acid changes have been located onto the topological model and amino acid conservation has been assessed. All the sequences were characterized by long hydrophobic segments, probably folded in membrane spanning α-helices mainly connected by short hydrophilic loops. The most conserved is invariably ND1, whereas ND6 is always the most divergent. Our analysis revealed that some highly conserved domains, either transmembrane α-helices or extra-membrane loops, are LHON mutational hot spots and are probably involved in protein function. On contrary, ancient polymorphisms are scattered along the polypeptide chains and usually affects non-conserved residues.