2.6.1. NDUFA10

‐ The gene for NDUFA10 is found on chromosome 2. The encoded protein has 355 amino acids and a mass of about 40.8 kDa, with 35 N-terminal amino acids cleaved upon entry to the mitochondrion. Structurally, NDUFA10 has been identified as a member of the deoxynucleoside kinase (dNK) family [8], with the dNK domain (also known as the PF01712 family in the Pfam database), although it is unlikely to be an active enzyme. The protein is compact with a four-stranded parallel beta-sheet surrounded by numerous alpha-helices (see Figure 13). The human structure (PDB id = 5xtd) [9] did not reveal any ligands, but adenosine nucleotides have been found at the C-terminal ends of the beta-sheet in a pocket surrounded by alpha-helices in the mouse (PDB id-6g2j) [37] and sheep [65] structures. NDUFA10 is found on the matrix side of the membrane and primarily interacts with core subunits ND2 and NDUFS2. These interactions with the extended N-terminus of NDUFS2, and the matrix-side peripheral helices of ND2 are all centered over the C-terminal sector of ND2, which contains the broken helix that

is part of the proton translocation pathway. NDUFA10 also makes limited contact with the N-terminal regions of NDUFC1, NDUFC2, and NDUFB11 on the matrix side. ‐

‐ ‐

‐ ‐ ‐

‐ ‐

‐

‐

‐

‐ ‐ ‐ ‐ **Figure 13.** Structural features of two subunits from the ND2-module. NDUFA10, colored red, and NDUFC2, colored gold, are portrayed in ribbons. The site of one mutation, His48Leu in NDUFC2, is shown in space-filling and colored blue. The sites of three mutations in NDUFA10—Gly99Glu, Gln142Arg, and Leu294Pro—are shown in space-filling and colored yellow. The site of phosphorylation by PINK1, Ser250, is shown in space-filling and colored cyan. NDUFA10 likely binds an adenosine nucleoside, not shown.

‐ In knockout human cell line BN gel electrophoresis, only faint bands of complexes containing primarily membrane subunits are seen [11]. Perhaps because of its contacts with a Q-module, core subunit NDUFS2, and a core membrane subunit ND2, along with subunits at the interfaces of ND2/ND4 (NDUFC1 and NDUFC2) and of ND4/ND5 (NDUFB11), it appears to be a key subunit in assembly. Residue Ser250, which is found on the matrix surface, was reported to be phosphorylated by the kinase PINK1 [66], a kinase that is known to be imported into the mitochondrial matrix space [66,67]. This is a conserved amino acid, and phosphorylation appears to be important for complex I activity.

Three reports of clinical mutations in NDUFA10 have been described. The first patient had compound heterozygous mutations including one allele in which the start codon was changed to GTG, and a second in which p.Gln142Arg occurred [68]. This individual showed developmental problems at 10 months and was eventually diagnosed with Leigh syndrome. It is likely that the first mutation would significantly impair the translation unless an alternative start codon were available. The second mutation is found in the interior of the protein, and so the Arg likely disrupts the packing of the protein, especially because of the positive charge.

The second patient was a boy with Leigh syndrome who, in a screen, was determined to have compound heterozygous mutations: p.Leu294Pro and c.383\_384insTAA (p.Ser218delinslS) [27]. These were found to be inherited from the father and mother, respectively. The latter mutation would likely lead to a degraded mRNA or protein. The former mutation p.Leu294Pro resides on an alpha-helix

at an interior location. The substitution of Leu by Pro might be disruptive, as this alpha-helix contacts two distinct regions of ND2. The immunoblotting of fibroblasts showed reduced levels of NDUFA10.

The third patient was initially diagnosed as having a nonlethal infantile mitochondrial disorder, and the mutation was discovered in a screen [69]. This individual developed brain lesions and was considered to have Leigh syndrome. His parents were third cousins, and DNA sequencing revealed him to be homozygous for p.Gly99Glu, while both parents were heterozygous for the same mutation. A deltoid muscle biopsy showed significant reduction of complex I activity (about 40–70% of normal levels). This amino acid is packed against the N-terminal amino acids of NDUFC1, and so the introduction of the negatively-charged Glu could disrupt the assembly of the membrane arm.

#### 2.6.2. NDUFC2

The gene for NDUFC2 is located on chromosome 11. The NDUFC2 protein is 119 amino acids, with a mass of 14.2 kDa. It is a double-pass transmembrane protein that contacts core subunit ND2, with both termini on the IMS side of the membrane. From the N-terminus, the protein enters the membrane in a non-helical stretch from Ser 19 to Arg29, and then residues 29–47 are alpha-helical. The second crossing of the membrane is alpha-helical from residues 56–97, extending into the IMS. NDUFC2 also contacts NDUFC1, NDUFB5, and NDUFA8 (see Figure 11), and it is part of the ND2 assembly module. NDUFC1 has a special role in sealing NDUFC2 from the lipid bilayer, with its single transmembrane helix parallel to the long helix of NDUFC2 and very limited contact to any other subunits in the membrane. The extension of the long alpha-helix into the IMS, residues 84–97, is the region that contacts NDUFA8 and NDUFB5 (see Figure 11). On the matrix side it contacts NDUFA10 (see Figure 13).

Mutations of NDUFC2 in three patients from two families were reported in 2020 [70]. The patients presented with symptoms of Leigh syndrome and reduced complex I activity but had different outcomes. In both families, the parents were healthy consanguineous first cousins and were heterozygous for the mutations. In one family, the mutation was a deletion of 22 base pairs near the C-terminus at residue His116, p.His116\_Arg119delins21. This would cause a frame shift and elimination of the normal stop codon at position 120. It was not clear how long the new reading frame would be. The transcript of this gene was identified at low levels, but the immunoblotting of fibroblast samples was negative, suggesting the degradation of the protein or a lack of expression. Little or no complex I was seen in native gels. Complexome profiling found evidence of Q-module assembly and ND4-module assembly but little else of complex I subunits. The daughter had no seizures and survived until at least age six, while her brother passed away at three years of age.

In the second family, the mutation was p.His58Leu. This patient had normal transcript levels but only slightly more evidence of complex I assembly. His58 is found on the matrix side of the membrane and starts the long helix that crosses the membrane. It sits between Val44 of NDUFC1 and Trp353 of NDUFA10 and is very near the phosphate group of a bound lipid. In the mouse structure (PDB id = 6g2j), the corresponding His59 is even more tightly packed between NDUFC1 and NDUFA10 residues, suggesting that a Leu substitution could be deleterious. This child had impaired growth and seizures and passed away at eight months.

NDUFC2 has a unique position in complex I. It is embedded in the membrane near the ND2-ND4 junction but contacts two subunits that are embedded at distant sites, with long extension to NDUFC2: NDUFA8 is found at the "heel" of the complex I boot, in the IMS, while NDUFB5 is found at the junction of ND4-ND5. Interactions in this network of subunits appear to be essential for complex I assembly.
