Data from: Deep mitochondrial origin outside the sampled alphaproteobacteria

Mitochondria are ATP-generating organelles whose endosymbiotic origin has been a key event in the evolution of eukaryotic cells. Despite overwhelming phylogenetic evidence for an alphaproteobacterial ancestry, efforts to pinpoint the closest relatives among the sampled alphaproteobacteria have been conflicting, complicating detailed inferences about the identity and nature of the mitochondrial ancestor. While most studies support that mitochondria evolved from an ancestor related to Rickettsiales, an order that includes several host-associated pathogenic and endosymbiotic lineages, others suggest that mitochondria evolved from a free-living group. Here we re-evaluate the phylogenetic placement of mitochondria. We employ genome-resolved binning of oceanic metagenome datasets and increase the genomic sampling of Alphaproteobacteria with twelve novel divergent clades, and one novel clade representing a sister group to all Alphaproteobacteria. Subsequent phylogenomic analyses that specifically address long branch attraction and compositional bias artefacts suggest that mitochondria did not evolve from Rickettsiales or any other currently recognized alphaproteobacterial lineage. Rather, our analyses indicate that mitochondria evolved from a proteobacterial lineage that branched off prior to the divergence of all sampled alphaproteobacteria. In light of this new result, previous hypotheses on the nature of the mitochondrial ancestor should be re-evaluated.线粒体（mitochondria）是合成三磷酸腺苷（ATP）的细胞器，其通过内共生形成的起源过程，是真核细胞演化史上的关键事件。尽管已有大量系统发育（phylogenetic）证据支持线粒体起源于α-变形菌门（alphaproteobacteria）的祖先类群，但学界针对已采样α-变形菌门中与线粒体亲缘关系最近类群的定位研究始终存在分歧，这使得对线粒体祖先类群的身份与本质的精准推演变得更为复杂。尽管多数研究认为线粒体起源于与立克次体目（Rickettsiales）亲缘关系较近的祖先类群——该类群包含多种宿主关联的致病及内共生谱系——但也有部分研究提出线粒体起源于自由生活的类群。本研究重新评估了线粒体的系统发育定位。我们通过对海洋宏基因组（metagenome）数据集开展基因组解析分箱（genome-resolved binning），新增12个全新的分歧演化支（clade）以及1个代表所有α-变形菌门姊妹群的新演化支，从而拓展了α-变形菌门的基因组采样覆盖度。后续开展的系统基因组学（phylogenomic）分析专门针对长枝吸引（long branch attraction）与组成偏倚（compositional bias）两类系统发育假象（artefact）进行校正，结果显示线粒体并非起源于立克次体目，亦非当前已被认知的任何其他α-变形菌门谱系。与之相反，本研究的分析结果表明，线粒体起源于一个在所有已采样α-变形菌门类群分化之前就已独立分支的变形菌门（proteobacteria）谱系。基于这一新的研究结果，学界此前关于线粒体祖先类群本质的各项假说均需重新进行评估。