Research

Starting in 2022, the MitoTOL consortium aimed to build high-quality mitochondrial protein inventories (mitoproteomes) across diverse species by developing experimental and computational workflows.  After piloting many different experimental approaches (including density gradients, affinity purification via immunoprecipitation, and proximity biotinylation via TurboID, APEX2, and miniTurbo), the consortium converged on a three step methodology that was adapted for each organism:

1. Serial mitochondrial purification (using a combination of density centrifugation, density gradients, with or without Proteinase K treatment to liberate mitochondria from attached organelles, and with or without MitoIP)
2. Data-independent acquisition (DIA) mass spectrometry applied to all purification steps 
3. Data analysis by protein correlation profiling (PCP) whereby each protein’s relative abundance across purification steps was scored using Pearson correlation to a “bait profile” of three mitochondrial marker proteins.

For each species, this methodology resulted in an experimentally defined inventory called “MitoViaPCP”, which was then augmented by combining this list with microscopy validation and literature evidence into a final “MitoCarta” inventory. 

We also used these new experimentally-defined inventories to retrain a computational prediction algorithm and predicted mitoproteomes across several hundred eukaryotic species.