Energy-converting NADH:ubiquionone oxidoreductase, respiratory complex I, plays a major role in cellular energy metabolism. It couples NADH oxidation and quinone reduction with proton translocation across the membrane, thus contributing to the formation of the proton motive force. Complex I deficiencies were found to be the most common source of human mitochondrial dysfunction, manifesting in a variety of neurodegenerative diseases. Seven core subunits of human complex I are encoded by mitochondrial DNA (mtDNA) that carry an unexpectedly large number of mutations in patient tissues. The biochemical consequences of these mutations are largely unknown due to the difficulty of experimental access to mitochondrial complex I. To understand the effects of the mutations, they should be introduced into complex I of Escherichia coli as a model system. It consists of 13 subunits, named NuoA to NuoN, and contains nine iron-sulfur clusters and one flavin mononucleotide as cofactors.
Mutations found in patients’ mtDNA were introduced into E. coli complex I and the resulting variants were biochemically characterized. Mutation V259AL disturbed the assembly of complex I leading to an inactive variant, whereas mutation F123LL caused the assembly of a stable but almost inactive complex. Inversion of seven base pairs in mtDNA resulted in a triple mutation. In E. coli the homologous mutation D213G/Q214K/P215VH led to the assembly of a stable but completely inactive variant. The biochemical effects on complex I caused by these mutations might contribute to their deleterious effects in humans. Only Position D213GH within the triple mutation is conserved. The single Mutation D213GH resulted in the assembly of a stable variant as well, but with the capability to catalyze redox-driven proton translocation, albeit at reduced levels. UV/vis-spectroscopic re-oxidation kinetics of the variant demonstrated that the cause of the diminished activity is the impaired formation of a catalytic intermediate of the energy converting process in the complex. From these data, a mechanism for energy coupling, which has not yet been fully elucidated, was proposed.
Via MD simulations and structural studies of complex I in different states of the catalytic cycle, key positions for several postulated mechanisms of proton translocation were proposed. The introduction of the mutations H322AM, H348AM, and H322A/H348AM confirmed a mechanistic proposal for their role in catalysis. Mutation E407KM was used to generate a symmetrical distribution of the central charged amino acids. The variant catalyzed redox-driven proton translocation, albeit with lower stoichiometry. This contradicts the hypothesis that charge asymmetry is essential for proton translocation and revealed the so-called "ND5-only" mechanism to be unlikely. The results from these four mutations in NuoM support the mechanistic proposal of a model in which dipole changes propagate in a wave-like manner from one end of the membrane arm to the other and back again, thus initiating the translocation of a proton via each of the four postulated proton channels.
Further point-mutations were introduced into subunits NuoA, H, J, K and CD to investigate the postulated functions of the respective amino acids. Possible clinically relevant mutations in these positions were introduced as well. The pathogenicity of the mutations could be explained via their drastic effects on E. coli complex I. First characterization of the additional variants helped to improve the understanding of the mechanistic processes at the interface between the peripheral arm and the membrane arm.
Previous work by our group described an interaction of complex I variant ΔNuoL with a monomer of LdcI, encoded by the gene cadA. The role of the LdcI in complex I assembly could not be clarified by characterization of complex I nor the ΔNuoL variant from different cadA-deletion strains and remains unclear. To identify the LdcI binding site on complex I by cryo-EM, the ΔNuoL variant with associated LdcI was purified. However, electron microscopic examination of negative-stained single particles could not identify the binding side. The Attempt to improve protein quality resulted in the loss of complex I-LdcI interaction.
A chaperone-like function for complex I and possible involvement in a repair mechanism associated with LdcI and the ΔNuoL variant were proposed for the proteins RavA and ViaA from E. coli. Characterization of complex I and the ΔNuoL variant from different ravA-viaA-deletion strains showed no specific effect of the deletions on the assembly or activity of the enzyme.