Background: Mitochondria have a central role in cellular functions, aging and in certain diseases. They possess their own genome, a vestige of their bacterial ancestor. Over the course of evolution, most of the genes of the ancestor have been lost or transferred to the nucleus. In humans, the mtDNA is a very small circular molecule with a functional repertoire limited to only 37 genes. Its extremely compact nature with genes arranged one after the other and separated by short non-coding regions suggests that there is little room for evolutionary novelties. This is radically different from bacterial genomes, which are also circular but much larger, and in which we can find genes inside other genes. These sequences, different from the reference coding sequences, are called alternative open reading frames or altORFs, and they are involved in key biological functions. However, whether altORFs exist in mitochondrial protein-coding genes or elsewhere in the human mitogenome has not been fully addressed.

Results: We found a downstream alternative ATG initiation codon in the +3 reading frame of the human mitochondrial nd4 gene. This newly characterized altORF encodes a 99-amino acids long polypeptide, MTALTND4, which is conserved in primates. Our custom antibody, but not the pre-immune serum, was able to immunoprecipitate MTALTND4 from HeLa cell lysates, confirming the existence of an endogenous MTALTND4 peptide. The protein is localized in mitochondria and cytoplasm and is also found in the plasma, and it impacts cell and mitochondrial physiology.

Conclusions: Many human-mitochondrial-translated ORFs might have so far gone unnoticed. By ignoring mtaltORFs, we have underestimated the coding potential of the mitogenome. Alternative mitochondrial peptides such as MTALTND4 may offer a new framework for the investigation of mitochondrial functions and diseases.

HeLa cells (n=3 replicates) were washed with PBS and harvested in a lysing buffer containing 10 mM HEPES, 150 mM NaCl and a cocktail of protease inhibitors (benzamidine, PMSF, aprotinin and leupeptin). Cells were then lysed with a sonic dismembrator sonicator (Fisher) 3 times for 30 seconds before adding 100 µl of Triton X-100 10%. Cell lysates were incubated on ice for 20 minutes and then centrifuged for 15 minutes at 15000 RPM at 4˚C. The supernatant was kept. 20 µl of protein A agarose was added to 1 ml of samples and incubated at 4˚C for 1 hour. Samples were then centrifuged at 15000 rpm for 1 minute at 4˚C and the supernatant was kept. 20 µl of rabbit anti-MTALTND4 antibody and 20 µl of protein A agarose were added to the samples before incubating at 4˚C overnight. 20 µl of rabbit pre-immune serum was used as control. Samples were washed 6 times with lysing buffer. Samples were then sent to the proteomics platform of the McGill University Health Center (MUHC) Research Institute for mass spectrometry analysis. Samples were loaded onto a single stacking gel band to remove lipids, detergents and salts. The single gel band containing all proteins was reduced with DTT, alkylated with iodoacetic acid and digested with trypsin. 2 µg of extracted peptides were re-solubilized in 0.1% aqueous formic acid and loaded onto a Thermo Acclaim Pepmap (Thermo, 75µM ID X 2 cm C18 3 µM beads) precolumn and then onto an Acclaim Pepmap Easyspray (Thermo, 75 µM X 15 cm with 2 µM C18 beads) analytical column separation using a Dionex Ultimate 3000 uHPLC at 250 nl·min^-1 with a gradient of 2–35% organic (0.1% formic acid in acetonitrile) over 3 hours. Peptides were analyzed using a Thermo Orbitrap Fusion mass spectrometer operating at 120,000 resolutions (FWHM in MS1) with HCD sequencing (15,000 resolution) at top speed for all peptides with a charge of 2+ or greater. The raw data were converted into *.mgf format (Mascot generic format) for searching using the Mascot 2.6.2 search engine (Matrix Science) against Human Uniprot sequences (2022). The database search results were loaded onto Scaffold Q+ Scaffold_4.9.0 (Proteome Sciences) for statistical treatment and data visualization (Table S4).