Case-study m.10191T>C

Leigh Disease is a very rare and severe mitochondrial disorder that arises in the first months of life; it has been associated with at least 75 nuclear and mitochondrially encoded genes with functional implications in mitochondrial biogenesis and metabolism (1). The disease is characterized by neurological regression with specific clinical lesions at basal ganglia and brainstem levels.

A mutation in the ND3 gene, m.10191T>C, has been associated with severe Complex I deficiency and neurological degeneration in adult patients (2) and children with Leigh disease (3,4). The mutation encodes for a Serine-Proline substitution in position 45. This missense variant is classified as Confirmed pathogenic mutation in MITOMAP, with Leigh Disease and Leigh Disease-like associated phenotypes.

We asked ourselves if theoretical compensatory variants exist for such mutation, within the same protein or in the interaction interface with another mitochondrial protein.

As a first step, we obtained 87 possibly inter-protein co-evolving pairs involving the site 45 of ND3. We considered only the pairs mapping on the physical interface between ND3 and ND1 and retrieved three PDB structures that resolved their interaction interface. Then, we calculated the ∆∆G of binding energy through the FoldX suite. ∆∆G_A is the binding energy variation of the S45P mutant concerning the wild type ND3 protein; ∆∆G_B is the partner mutant's binding energy variation for the wild type ND1 protein; ∆∆G_AB is the binding energy variation when both proteins are mutated at their interaction interface.
Note that ∆∆G_A is below the conventional threshold of ±0.61 Kcal/mol that identifies the disruptive variants for two PDB complexes, 5lc5 and 5ldx. We evidenced that three amino acid changes in the ND1 position 126 can restore the binding energy to, approximately, wild type levels (row in bold characters in the table below). Indeed, ∆∆G_AB is proximal to 0, indicating that the interface with two co-occurring variations has similar binding energy to the wild-type protein complex.

The effect of ND3-ND1 interactions can also be observed in the Molecular Dynamics section of the MitImpact website. MD16, MD29, and MD31 collect results obtained by MD simulations of the amino acid pairs involving the positions 45 and 126 in ND3 and ND1, respectively.

Some interesting results about the S45P-N126S variant pairs regard:

• The RMSD plot that shows how the S45P mutant subunit remains stably compact over the simulation time (i.e., lower RMSD) with respect to the wild-type protein complex (in red);
• The RMSD of the double mutant that neatly overlaps that of the wild-type complex;
• The RMSF plot, where we observe high residue fluctuations from the positions 20 to 50 in the ND3 protein (chain B). The S45P mutation causes lower mobility of that region, while the occurrence of both mutations on the proteins interface partially restores the wild-type condition.