Phylogenetic Profiling of Mitochondrial Proteins and Integration Analysis of Bacterial Transcription Units Suggest Evolution of F1Fo ATP Synthase from Multiple Modules

Abstract

ATP synthase is a complex universal enzyme responsible for ATP synthesis across all kingdoms of life. The F-type ATP synthase has been suggested to have evolved from two functionally independent, catalytic (F1) and membrane bound (Fo), ancestral modules. While the modular evolution of the synthase is supported by studies indicating independent assembly of the two subunits, the presence of intermediate assembly products suggests a more complex evolutionary process. We analyzed the phylogenetic profiles of the human mitochondrial proteins and bacterial transcription units to gain additional insight into the evolution of the F-type ATP synthase complex. In this study, we report the presence of intermediary modules based on the phylogenetic profiles of the human mitochondrial proteins. The two main intermediary modules comprise the α₃β₃ hexamer in the F1 and the c-subunit ring in the Fo. A comprehensive analysis of bacterial transcription units of F1Fo ATP synthase revealed that while a long and constant order of F1Fo ATP synthase genes exists in a majority of bacterial genomes, highly conserved combinations of separate transcription units are present among certain bacterial classes and phyla. Based on our findings, we propose a model that includes the involvement of multiple modules in the evolution of F1Fo ATP synthase. The central and peripheral stalk subunits provide a link for the integration of the F1/Fo modules.

Keywords: ATP synthase; ATPase; Modular evolution; Phylogenetic profile; Transcription units.

Fig. 1

Groups of mitochondrial F1Fo ATP synthase from mitochondrial proteins. Hierarchical clusters of phylogenetic profiles using 758 (a) or 490 randomly chosen (b) species with independent distance measurements: Euclidean, Manhattan, Jaccard, Kendall tau, mutual information, and Pearson absolute anti-correlation distance. Group number is set to 10, 25, or 40, and distinct groups are marked with different colors. (Color figure online)

Fig. 2

Cluster analysis for human mitochondrial F1Fo ATP synthase subunits. a The phylogenetic profile for human 17 subunits of the F1Fo ATP synthase clustered both by species (columns) and genes (rows). The top color bar indicates the three domains of life, and the left color bar represents the corresponding F1 and Fo subunits arranged in a cluster dendrogram. The blue and gray bars in the phylogenetic profiles indicate presence or absence in a given species, respectively. b Distance tree of mitochondrial F1Fo ATP synthase subunits generated by Euclidean distance with the group number set to 6. Different colors are used to indicate identified clusters. c Correlation matrix of mitochondrial F1Fo ATP synthase subunits. The Pearson correlation coefficients of paired genes are marked in numbers and color-coded from blue to red. The left cluster dendrogram indicates the Euclidean distance between subunits of F1 and Fo marked with red and green, respectively. d Principal component analysis (PCA) plot of the 17 subunits of F1Fo ATP synthases with a cluster number of 4. Identified groups are shown with different colors and shapes. e The evolutionarily conserved modules (ECMs) of human F1Fo ATP synthase. Blue and gray blocks indicate present and absent homologs, respectively. (Color figure online)

Fig. 3

Transcription unit modules of bacterial F1Fo ATP synthase. a The standard transcription units with cut points. b The frequency of transcription unit modules mapped to the standard one in A. Red dots denote that transcription unit modules are significantly higher than the upper confidence interval. The green dots are the modules which did not reach statistical significance. The dashed gray line marks the upper confidence interval (0.03) with the threshold 0.05. c Representative patterns of transcription units detected in bacterial genomes. The F1Fo ATP synthase genes are marked with different colors and shapes. The schematic patterns do not indicate the length of the genome or the actual transcription direction. (Color figure online)

Fig. 4

Dominant patterns of transcription units in bacterial classes and phyla. a A maximum-likelihood tree based on the bacterial 16S rDNA sequences. Bacterial classes and phyla are shown with distinct colors. A gradient of red color is used to represent the repeated transcription unit (modules) in the same genus. b A class/phylum phylogenetic tree combined with the dominant patterns of transcription units in each bacterial class and phylum. The histogram indicates the percentage of dominant patterns. (Color figure online)

Fig. 5

Distribution of minor cTUs and chimera fusions supporting the modular evolution. a–d The reconstructed phylogenetic tree combined with predicted transcription units containing F1Fo ATP synthase subunits in Bacteroidetes, Bacilli, Deltaproteobacteria, and Betaproteobacteria, respectively. Species in one genus are marked with the same color in bars and tree branches. (Color figure online)