Characterization of clinically identified mutations in NDUFV1, the flavin-binding subunit of respiratory complex I, using a yeast model system

Human Molecular Genetics, 2015, Vol. 24, No. 22 6350–6360 doi: 10.1093/hmg/ddv344 Advance Access Publication Date: 7 September 2015 Original Article ORIGINAL ARTICLE Characterization of clinically identified mutations in NDUFV1, the flavin-binding subunit of respiratory complex I, using a yeast model system Medical Research Council Mitochondrial Biology Unit, Wellcome Trust/MRC Building, Hills Road, Cambridge CB2 0XY, UK *To whom correspondence should be addressed. Tel: +44 1223252810; Email: Abstract Dysfunctions in mitochondrial complex I (NADH:ubiquinone oxidoreductase) are both genetically and clinically highly diverse and a major cause of human mitochondrial diseases. The genetic determinants of individual clinical cases are increasingly being described, but how these genetic defects affect complex I on the molecular and cellular level, and have different clinical consequences in different individuals, is little understood. Furthermore, without molecular-level information innocent genetic variants may be misassigned as pathogenic. Here, we have used a yeast model system (Yarrowia lipolytica) to study the molecular consequences of 16 single amino acid substitutions, classified as pathogenic, in the NDUFV1 subunit of complex I. NDUFV1 binds the flavin cofactor that oxidizes NADH and is the site of complex I-mediated reactive oxygen species production. Seven mutations caused loss of complex I expression, suggesting they are detrimental but precluding further study. In two variants complex I was fully assembled but did not contain any flavin, and four mutations led to functionally compromised enzymes. Our study provides a molecular rationale for assignment of all these variants as pathogenic. However, three variants provided complex I that was functionally equivalent to the wild-type enzyme, challenging their assignment as pathogenic. By combining structural, bioinformatic and functional data, a simple scoring system for the initial evaluation of future NDUFV1 variants is proposed. Overall, our results broaden understanding of how mutations in this centrally important core subunit of complex I affect its function and provide a basis for understanding the role of NDUFV1 mutations in mitochondrial dysfunction. Introduction Mitochondrial complex I (NADH:ubiquinone oxidoreductase) catalyzes oxidation of NADH in the mitochondrial matrix, reduction of ubiquinone in the inner mitochondrial membrane and transfer of protons across the membrane (1). It thus regenerates NAD+ in the matrix, to sustain crucial metabolic processes including the tricarboxylic acid cycle and β-oxidation of fatty acids, supplies electrons to respiratory complex III, and contributes to the proton motive force that drives ATP synthesis and transport processes. Complex I is also a significant source of reactive oxygen species generation in mitochondria and thus contributes to cellular oxidative stress (2). Mammalian complex I consists of 45 subunits (1,3,4). Fourteen of them are the conserved core subunits that are sufficient for catalysis; detailed structural information on them is available from complex I from the bacterium Thermus thermophilus (5). The remaining 31 are supernumerary subunits that have been accumulated through evolution; the locations of some of them have been assigned in the structure of the mammalian complex I from Bos taurus (4). Dysfunctions in complex I, which account for around a third of all early-onset mitochondrial disorders, are both genetically Received: June 19, 2015. Revised: July 31, 2015. Accepted: August 18, 2015 © The Author 2015. Published by Oxford University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. 6350 Febin Varghese, Erwan Atcheson, Hannah R. Bridges and Judy Hirst* Human Molecular Genetics, 2015, Vol. 24, No. 22 experimentation. Thus, we broaden understanding of how mutations around the flavin site affect complex I, evaluate Y. lipolytica and other model systems for their suitability for studying human complex I mutations, and develop approaches for evaluating the pathogenic potential of NDUFV1 variants identified clinically in the future. Results Mutations identified as pathological in human NDUFV1 Nineteen single amino acid substitutions that have been classified as pathological in the human NDUFV1 subunit of complex I were identified by searching the literature and in the Human Gene Mutation Database (23) (see Table 1 and Supplementary Material, Table S1). Two further mutations that introduce stop codons to produce premature terminations [R59X (30) and Q344X (25)] are not discussed here. Aligning the sequence of NDUFV1 with NUBM, its Y. lipolytica homolog, revealed that 16 of the mutations affect conserved residues (see Fig. 1 and Table 2), within an overall sequence identity of 67% (85% similarity). Therefore, these 16 mutations can be studied in Y. lipolytica. The three non-conserved mutations are S56P, K111E and A211V. Figure 1 also shows the high sequence identity (98%) between the human NDUFV1 subunit and its homolog in Bos taurus, the 51kDa subunit, so the human mutations can be mapped to the B. taurus structure (see Table 2 and Fig. 2). Note that we use the nomenclature and numbering for the human subunit throughout for consistency, regardless of the species referred to. In Figure 2, the subunit has been separated into its four constituent domains for clarity: an N-terminal domain that ends in a glycine-rich loop, followed by a Rossmann-fold domain, an ubiquitin-like domain and a C-terminal four-helix bundle that ligates the iron–sulfur (FeS) cluster known as cluster N3 (34). Seven mutations result in loss of complex I expression The 16 mutants for study were expressed in NDUFV1 in Y. lipolytica complex I, by expressing NUBM (NDUFV1) on the pUB26 plasmid in a GB10 Δnubm deletion strain of Y. lipolytica (35). The strain also expresses a matrix-targeted NDH1 (alternative dehydrogenase), to support cell growth even when complex I is inactive (36). None of the variants displayed abnormal growth, although variants in which intact complex I could not be observed (see below) grew up to twice as slowly as the wild-type and other variants. Mitochondrial membranes were prepared from each variant and investigated using blue-native (BN)-PAGE (by visualizing the band from intact complex I and by detecting complex I with an in-gel assay for NADH oxidation), and by measuring the rate of NADH:ferricyanide (FeCN) oxidoreduction by the complex I flavin site in solution assays (see Fig. 3). Intact complex I could be detected in only nine cases and so enzyme assembly or stability is disrupted in the seven variants (R88G, A117T, Y204C, E246K, P252R, R257Q and T423M) that did not contain detectable complex I. The seven mutations ar (...truncated)