“The dogma was that tRNAs are transcribed in the nucleus and function in the cytoplasm, and that is the whole story,” says Hopper. “But the whole tRNA model has become much more complex, and almost anything one would have thought would be true has turned out to be wrong.”Previous work showed that, when yeast were starved, tRNAs lacking introns were more abundant in the nucleus. It was not clear though whether they were processed in the nucleus and held there, or whether they underwent retrograde transport and returned to the nucleus from the cytoplasm. With the recent demonstration that yeast tRNAs are spliced in the cytoplasm, evidence pointed to the transport model.
To check this idea, the groups each took advantage of mutations that prevent nuclear fusion after mating, generating cells that have two nuclei in a single cytoplasm. When such cells expressed a tRNA gene from another species—either S. pombe or Dictyostelium—in one nucleus but not the other, the researchers saw that the spliced exogenous tRNA turned up in both nuclei.
It is not yet clear why processed tRNAs return to the nucleus, but Hopper thinks their removal from the cytoplasm might limit translation when the amino acid supply is low. Her group found that tRNA retrograde transport was RAN-dependent under starvation conditions, whereas Takano et al. saw that retrograde transport was normally RAN-independent. The difference could indicate a regulatory change that induces retrograde transport during times of stress. The nuclear presence of tRNAs also brings up the old debates of whether mature ribosomes can reenter the nucleus and whether translation occurs there as well.