The cell nucleus as a barrier against horizontal gene transfer in microbial endosymbioses

Abstract The origin of eukaryotic cells remains a highly contested problem. While eukaryotes arose from the merger of a bacterial and an archaeal partner giving rise to mitochondria and the cell proper, the order of steps is not known, nor is it understood why it was a singular event. Prokaryotes engage in various cooperative interactions everywhere, yet there is no evidence that they could establish stable endosymbiotic relationships on their own. Many assume that mitochondria came first, and their critical presence and features enabled the complex cellular architecture, including the nucleus. Here we find support for the alternative, claiming that a nuclear compartment was a prerequisite for successful stable endosymbiosis. We review independent lines of evidence suggesting that the pre-existence of a nuclear membrane or equivalent mechanism to separate translation from transcription may have been essential to limit genetic inference owing to extensive horizontal gene transfer in the wake of pre-mitochondrial (endo)symbionts and to stabilize the host genome against foreign DNA, especially from (endo)symbiotic partners. We claim that an asymmetry in control potential between partners is required for successful integration of an endosymbiont. This would explain why there are no further prokaryotic endosymbioses known to us (extant or extinct). We propose predictions that can be tested to support the hypothesis. This article is part of the theme issue ‘Evolutionary genetics of mitochondria: on diverse and common evolutionary constraints across eukarya’.