Divergent Mitochondrial Respiratory Chains in Phototrophic Relatives of Apicomplexan Parasites Pavel Flegontov,y,1,2 Jan Michalek, y,1,3 Jan Janouskovec, z,4,5 De-Hua Lai,x,1 Milan Jirku,˚ 1,3 Eva Hajduskov a, 1 AlesTomc ˇala,1 Thomas D. Otto,6 Patrick J. Keeling,4,5 Arnab Pain,7 Miroslav Obornık,*,1,3,8 and Julius Lukes* ,1,3,5 1Institute of Parasitology, Biology Centre, Czech Academy of Sciences, Cesk eBud ejovice, Czech Republic 2Life Science Research Centre, Department of Biology and Ecology, Faculty of Science, University of Ostrava, Ostrava, Czech Republic 3Faculty of Science, University of South Bohemia, Cesk e Budejovice, Czech Republic 4Department of Botany, University of BC, Vancouver, Canada 5Canadian Institute for Advanced Research, Toronto, ON, Canada 6 Wellcome Trust Sanger Institute,Hinxton,UnitedKingdom Downloaded from 7Biological and Environmental Sciences and Engineering Division, King Abdullah University of Science and Technology, Thuwal, Kingdom of Saudi Arabia 8Institute of Microbiology, Czech Academy of Sciences, Trebo n, Czech Republic yThese authors contributed equally to this work. http://mbe.oxfordjournals.org/ zPresent address: San Diego State University, San Diego, CA xPresent address: Center for Parasitic Organisms, School of Life Sciences, Sun Yat-Sen University, Guangzhou, People’s Republic of China *Corresponding author: E-mail: [email protected]; [email protected]. Associate editor: David Irwin Abstract Four respiratory complexes and ATP-synthase represent central functional units in mitochondria. In some mitochondria at The University of British Colombia Library on January 20, 2016 and derived anaerobic organelles, a few or all of these respiratory complexes have been lost during evolution. We show that the respiratory chain of Chromera velia, a phototrophic relative of parasitic apicomplexans, lacks complexes I and III, making it a uniquely reduced aerobic mitochondrion. In Chromera, putative lactate:cytochrome c oxidoreductases are predicted to transfer electrons from lactate to cytochrome c, rendering complex III unnecessary. The mitochondrial genome of Chromera has the smallest known protein-coding capacity of all mitochondria, encoding just cox1 and cox3 on heterogeneous linear molecules. In contrast, another photosynthetic relative of apicomplexans, Vitrella brassicaformis, retains the same set of genes as apicomplexans and dinoflagellates (cox1, cox3,andcob). Article Key words: respiratory chain, Apicomplexa, Chromera, anaerobic metabolism, evolution, Vitrella. Introduction was lost altogether from their highly reduced organelles Although all extant mitochondria are believed to originate (Muller€ et al. 2012). The most reduced mitochondrial ge- from a single endosymbiotic event, mitochondrial genomes nomes are found in the alveolates, or more specifically the have evolved remarkable diversity (Burger et al. 2003). Most myzozoans, which include apicomplexans, dinoflagellates, mitochondria retain a genome, a remnant of an ancestral and their relatives. Myzozoan mitochondrial genomes -proteobacterial chromosome.Itisusuallyorganizedona encode only subunits 1 and 3 of cytochrome c oxidase multicopy, circular DNA molecule or a circularly permutated (cox1 and cox3), one subunit of cytochrome c reductase Fast Track linear array that encodes genes for rRNAs, tRNAs, and pro- (cob), and truncated but apparently functional fragments of teins involved in the electron transport chain, ATP synthesis, small (SSU) and large subunit (LSU) mitoribosomal rRNA and a few housekeeping functions (Burger et al. 2003). The genes (Feagin et al. 1997; Nash et al. 2008; Waller and most gene-rich mitochondrial genomes identified to date Jackson 2009). The gene coding for cox2, universally present have been found in jakobids (Burger et al. 2013), whereas all in other mitochondrial genomes, has been split and trans- other eukaryotes have gene complement reduced to varying ferred to the nucleus (Waller and Keeling 2006). extent (Gray et al. 2004). In several cases, for instance in the In apicomplexan parasites, mitochondrial genes are kinetoplastid flagellates, the mitochondrial genome has ex- encoded on a linear molecule ranging in length from 5.8 kb panded into highly complex forms (Lukes et al. 2005; Gray in Parahaemoproteus to 11.0 kb in Babesia,whichcanbe et al. 2010). Moreover, in several eukaryotic lineages including either monomeric or is composed of circularly permutated diplomonads, parabasalids, and microsporidians, the genome linear arrays (Feagin et al. 1997). In dinoflagellates, the ß The Author 2015. Published by Oxford University Press on behalf of the Society for Molecular Biology and Evolution. All rights reserved. For permissions, please e-mail: [email protected] Mol. Biol. Evol. 32(5):1115–1131 doi:10.1093/molbev/msv021 Advance Access publication February 6, 2015 1115 Flegontov et al. doi:10.1093/molbev/msv021 MBE mitochondrial genome is broken into small linear fragments divergent cox3 gene, and numerous rRNA fragments. containing genes with extensively edited transcripts, as well as Although cox1 can be easily found by BLAST search, cox3 is many pseudogenes and apparently nonfunctional gene frag- too divergent to be detected by homology-based searching ments (Imanian and Keeling 2007; Jackson et al. 2007; algorithms. However, when the mitochondrial sequences Kamikawa et al. 2007, 2009; Nash et al. 2007, 2008; Waller were investigated for the presence of open reading frames and Jackson 2009). The deep-branching relatives of dinoflagel- (ORFs), one ORF with about 250 amino acids of cox3-like lates, Perkinsus and Oxyrrhis, lack RNA editing, and in the sequence was identified as a putative cox3 (see Materials former species translational frame shifting is required for ex- and Methods for details). This gene contains 42% hydropho- pression of its mitochondrial genes (Slamovits et al. 2007; bic residues and seven predicted transmembrane segments Masuda et al. 2010; Zhang et al. 2011). Third major group of (supplementary figs. S1 and S2, Supplementary Material alveolates are the ciliates, which usually harbor a gene-rich online), and has a very weak sequence similarity with the (~50 genes) mitochondrial genome (Gray et al. 2004), with cox3 genes when investigated with hidden Markov model their anaerobic members featuring a reduced gene set (HMM) searches. However, the percentage of hydrophobic Downloaded from (Akhmanova et al. 1998).Thenatureofmitochondrial residues, number of predicted transmembrane domains (six genome of Acavomonas peruviana, a heterotrophic alveolate in cox3 of Plasmodium), presence of short motifs conserved in branching on the root of myzozoans, demonstrates that the myzozoan cox3 proteins, and similar divergence of cox3 in ancestral state for myzozoan mitochondria was a linear chro- dinoflagellates (see alignment in supplementary fig. S1, mosome with telomeric sequences at both ends (Janouskovec, Supplementary Material online), all point to the fact that http://mbe.oxfordjournals.org/ Tikhonenkov, et al. 2013; Tikhonenkov et al. 2014). thegeneidentifiedinChromera is cox3. Since the diversity of mitochondrial genomes of dino- The cob gene, which was so far present in all other aerobic flagellates and myzozoans allows reconstruction of a pu- mitochondrial genomes, was surprisingly absent from all as- tative ancestral state, we wondered how informative the sembliesandrawreadsfromChromera. Phylogenetic analysis deep-branching relatives of apicomplexans will be in this of cox1 shows the Chromera gene on a long branch with low respect. Since the discovery in 2008 of the coral-associated support (bootstrap 38%, posterior probability [pp] 0.92) as a Chromera velia and the more recent description of related sister clade to the comparably divergent sequence from Vitrella brassicaformis, these photosynthetic algae have Perkinsus. The highly divergent cox1 gene from Vitrella ap- at The University of British Colombia Library on January 20, 2016 received much attention (Moore et al. 2008; Janouskovec pears to be weakly related (bootstrap 45%, pp 0.99) to its et al. 2010; Woehle et al. 2011; Burki et al. 2012; Obornık Oxyrrhis homolog, both branching inside the highly sup- et al. 2012; Janouskovec, Sobotka, et al. 2013; Obornık and ported myzozoan clade (supplementary fig. S3, Lukes2013 ; Petersen et al. 2014). While Chromera and Supplementary Material online). Analysis of total and mito- Vitrella have been used to solve a few riddles about api- chondrial DNA-enriched sequences also showed that the two complexan plastid genome evolution (Janouskovec et al. protein-coding genes are found in multiple genomic contexts 2010; Janouskovec, Sobotka, et al. 2013), it remained in the mitochondrion of Chromera. The full-length cox1 genes unclear how their mitochondrial genomes compare to are flanked by various sequences, and the same applies for a the equally unusual apicomplexan mitochondrion, which number of gene fragments (supplementary tables S1 and S2, lacks some signature elements of the canonical mitochon- Supplementary Material online), yet only a single full-length drial metabolism. Chromera has been proposed to harbor transcript is formed (fig 1A). The putative cox3 is fused with just the cox1 gene in its mitochondrion (Petersen et al. an upstream cox1 fragment (amino acids 1–192; supplemen- 2014). Here, we report the mitochondrial genomes and tary table S2, Supplementary Material online) and, as shown transcriptomes of Chromera and Vitrella, supplemented byquantitativepolymerasechainreaction(PCR),afusion