Serum-regulated nuclear localization is signal specific.

Abstract

The activity of nuclear factors can be regulated by blocking their ability to enter the nucleus, but how the cell achieves this is not yet understood. We demonstrate herein the serum-responsive nuclear localization of adenovirus E1a protein and show that this serum dependence is a property of the nuclear localization signal itself. When E1a protein is microinjected into the cytoplasm of cultured cells, it is found in the nucleus 30 min later only if the cells are serum fed; in serum-starved (growth-arrested) cells, the E1a is still cytoplasmic. Substituting the simian virus-40 T-antigen nuclear localization signal in place of the normal E1a signal abolishes this serum effect, and transferring the E1a signal to a heterologous protein also transfers the serum dependence. The serum effect on signal function is first exerted within 40 min after serum addition, suggesting that this is one of the earliest cellular responses to serum feeding. We conclude that the nuclear accumulation (and probably the function) of a protein is influenced not only by the presence of a nuclear localization signal, but also by the nature of that signal.