The Syntrophy hypothesis for the origin of eukaryotes revisited.

The discovery of Asgard archaea, phylogenetically closer to eukaryotes than other archaea, together with improved knowledge of microbial ecology, impose new constraints on emerging models for the origin of the eukaryotic cell (eukaryogenesis). Long-held views are metamorphosing in favour of symbiogenetic models based on metabolic interactions between archaea and bacteria. These include the classical Searcy's and Hydrogen hypothesis, and the more recent Reverse Flow and Entangle-Engulf-Endogenize models. Two decades ago, we put forward the Syntrophy hypothesis for the origin of eukaryotes based on a tripartite metabolic symbiosis involving a methanogenic archaeon (future nucleus), a fermentative myxobacterial-like deltaproteobacterium (future eukaryotic cytoplasm) and a metabolically versatile methanotrophic alphaproteobacterium (future mitochondrion). A refined version later proposed the evolution of the endomembrane and nuclear membrane system by invagination of the deltaproteobacterial membrane. Here, we adapt the Syntrophy hypothesis to contemporary knowledge, shifting from the original hydrogen and methane-transfer-based symbiosis (HM Syntrophy) to a tripartite hydrogen and sulfur-transfer-based model (HS Syntrophy). We propose a sensible ecological scenario for eukaryogenesis in which eukaryotes originated in early Proterozoic microbial mats from the endosymbiosis of a hydrogen-producing Asgard archaeon within a complex sulfate-reducing deltaproteobacterium. Mitochondria evolved from versatile, facultatively aerobic, sulfide-oxidizing and, potentially, anoxygenic photosynthesizing alphaproteobacterial endosymbionts that recycled sulfur in the consortium. The HS Syntrophy hypothesis accounts for (endo)membrane, nucleus and metabolic evolution in a realistic ecological context. We compare and contrast the HS Syntrophy hypothesis to other models of eukaryogenesis, notably in terms of the mode and tempo of eukaryotic trait evolution, and discuss several model predictions and how these can be tested.