When the nucleus of a hen erythrocyte is introduced into the cytoplasm of a human or mouse cell in culture, it resumes the synthesis of RNA. The reactivated erythrocyte nucleus undergoes great enlargement, but it does not, for at least 2 or 3 days, develop nucleoli which can be discerned under the light microscope. During this period, the heterokaryon, although it may contain several active erythrocyte nuclei, does not synthesize any hen-specific surface antigens; and the hen-specific antigens introduced into the surface of the heterokaryon by the process of cell fusion are eliminated. But when, later, the erythrocyte nuclei do develop nucleoli, hen-specific antigens reappear on the surface of the heterokaryon and progressively accumulate.
Before developing nucleoli, the erythrocyte nuclei synthesize little, if any, normal 28 S or 16 S RNA; but they do synthesize large amounts of the RNA which shows polydisperse sedimentation in conventional sucrose density gradients. Autoradiographic studies involving the use of a microbeam of ultraviolet light show, however, that this ‘polydisperse’ RNA is not transferred to the cytoplasm of the cell in detectable amounts so long as the erythrocyte nucleus lacks a definitive nucleolus. The inability of the erythrocyte nucleus at this stage to determine the synthesis of hen-specific surface antigens is thus attributable to the fact that it fails to transfer the RNA made on its chromosomes to the cytoplasm of the cell. When the erythrocyte nuclei develop nucleoli, however, the RNA which they make is transferred to the cytoplasm of the cell, and the synthesis of hen-specific surface antigens then begins. These experiments suggest that the nucleolus may play a decisive role in the transfer of information from nucleus to cytoplasm. The possible nature of this role is discussed.