In the software domain, self-adaptive systems are able to modify their behavior at run-time to respond to changes in the environment they run, to changes of the users' requirements or to changes occurring in the system it-self. In life science, biological cells are power entities able to adapt to the (unpredictable) situations they incur in, in a complete decentralized fashion. Learning adaptation mechanism from the cell life-cycle, we propose in this paper a new architectural paradigm for self-adaptive software systems.