Hidden Markov Model-Based Prokaryotic Genome Space Mining Reveals the Widespread Pervasiveness of Complex I and Its Potential Evolutionary Scheme

Most cellular reactions are interdependent; however, a subset of reactions often associate more closely to form a defined reaction pathway. An extreme arrangement of interdependent reactions occurs when the cognate proteins physically associate to constitute a complex. Respiratory Complex I (C-I) is one of the largest membrane-resident protein assemblies. Besides being a hallmark of bioenergetics, this enzyme complex is critical for maintaining redox homeostasis and facilitating transport. However, its evolutionary origins are unclear due to challenges in identifying close homologs and subunit ancestry. Using custom hidden Markov model (HMM) profiles, we analyzed the prokaryotic genome space to trace the distribution of 14 core C-I or NADH-quinone oxidoreductase (Nuo) subunits. Our findings include (i) a sensitive HMMER-based workflow for comprehensively annotating and analyzing the Nuo subunits, adaptable for similar analyses; (ii) the first species-level distribution of Nuo subunits; (iii) multiple C-I variants across ∼11,000 species, with 51.2% having a complete complex; (iv) C-I variants on plasmids, aiding evolutionary spread; and (v) extending our workflow to study mitochondrial C-I accessory subunits in prokaryotes, revealing their evolutionary roots. We also developed a web application to share our resources. Together, we comprehensively account for the distribution and probable evolutionary scheme of C-I subunits among prokaryotes.