Abstract
To examine the role of primary amino acid sequence in the localization of proteins within the nucleus, we studied the nucleolar protein NO38 of amphibian oocytes. We synthesized NO38 transcripts in vitro, injected them into the oocyte cytoplasm, and followed the distribution of the translation products. The injected RNA contained a short sequence encoding an epitope derived from the human c-myc protein. We used an mAb against this epitope to detect translation products from injected RNAs by Western blots and by immunofluoresent staining of cytological preparations. When full-length transcripts of NO38 were injected into oocytes, the translation products accumulated efficiently in the germinal vesicle, and a major fraction was localized in the multiple nucleoli. To identify protein domains involved in this nucleolus-specific accumulation, we prepared a series of carboxy-terminal deletions of the cDNA. Oocytes injected with RNA encoding truncated forms of NO38 were examined for altered patterns of protein accumulation. We defined a domain of about 24 amino acids near the carboxy terminus that was essential for nucleolar localization of NO38. This domain is separated by more than 70 amino acids from two putative nuclear localization signals near the middle of the molecule. Hybrid constructs were made which encoded part of Escherichia coli beta-galactosidase or pyruvate kinase fused to a long segment of NO38 containing the essential domain. Injection of RNA from these constructs showed that the essential domain was not sufficient to target the hybrid proteins to the nucleolus. We suggest that nucleolar accumulation of NO38 requires more than a single linear domain.