Many RNA-binding proteins possess domains with a biased amino acid content. A common property of these low complexity domains (LCDs) is that they assemble into an ordered amyloid form， juxtaposing RNA recognition motifs in a subcellular compartment in which RNA metabolism is focused. Yeast Nab3 is one such protein that contains RNA-binding domains and a low complexity， glutamine/proline-rich， prion-like domain that can self-assemble. Nab3 also contains a region of structural homology to human hnRNP-C that resembles a leucine zipper which can oligomerize. Here we show that the LCD and the human hnRNP-C homology domains of Nab3 were experimentally separable， as cells were viable with either segment， but... More
Many RNA-binding proteins possess domains with a biased amino acid content. A common property of these low complexity domains (LCDs) is that they assemble into an ordered amyloid form， juxtaposing RNA recognition motifs in a subcellular compartment in which RNA metabolism is focused. Yeast Nab3 is one such protein that contains RNA-binding domains and a low complexity， glutamine/proline-rich， prion-like domain that can self-assemble. Nab3 also contains a region of structural homology to human hnRNP-C that resembles a leucine zipper which can oligomerize. Here we show that the LCD and the human hnRNP-C homology domains of Nab3 were experimentally separable， as cells were viable with either segment， but not when both were missing. In exploiting the lethality of deleting these regions of Nab3， we were able to test if heterologous prion-like domains known to assemble into amyloid， could substitute for the native sequence. Those from the hnRNP-like protein Hrp1， the canonical prion Sup35， or the epsin-related protein Ent2， could rescue viability and enable the new Nab3 chimeric protein to support transcription termination. Other low complexity domains from RNA-binding， termination-related proteins or a yeast prion， could not. As well， an unbiased genetic selection revealed a new protein sequence that could rescue the loss of Nab3's essential domain via multimerization. This new sequence and Sup35's prion domain could also rescue the lethal loss of Hrp1's prion-like domain when substituted for it. This suggests there are different cross-functional classes of amyloid-forming LCDs and that appending merely any assembly-competent LCD to Nab3 does not restore function or rescue viability. The analysis has revealed the functional complexity of LCDs and provides a means by which the differing classes of LCD can be dissected and understood.
