“…The predicted molecular weight of PABP2 is ;33 kDa (Nemeth et al+, 1995) and therefore, if the protein is present in the cell as free monomers, it is expected to passively diffuse through the pores and therefore equilibrate between the nucleus and the cytoplasm+ However, our finding that PABP2 targets to the nucleus a fusion protein (luciferase-PABP2, predicted molecular weight ;90 kDa) that is too large to cross the pores by simple diffusion provides evidence for involvement of a carrier-mediated mechanism+ This is further supported by the result that PABP2 binds directly and specifically to the nuclear import receptor transportin+ Until a few years ago, nuclear import studies were centered on proteins carrying either a mono-or a bipartite NLS being recognized by a heterodimeric carrier named importin a/b+ However, recent work led to the discovery of a second pathway for protein import mediated by an importin-b-related protein, termed transportin in mammals (Pollard et al+, 1996) and Kap104p in yeast (Aitchison et al+, 1996)+ In mammals transportin mediates import of a subset of hnRNP proteins, the best characterized of which is hnRNP A1 (Pollard et al+, 1996;Siomi et al+, 1997)+ The signal within hnRNP A1 that mediates its nuclear import is a stretch of 38 C-terminal amino acids termed the M9 domain (Siomi & Dreyfuss, 1995)+ Because hnRNP proteins are very abundant (10 6 -10 7 copies per cell), it has been speculated that transportin has evolved as a specialized carrier for these high abundance substrates, contrasting with the importin-a/b carrier that recognizes a broad range of cargoes but with most members present at relatively low abundance (Nigg, 1997)+ The data presented here demonstrate that PABP2 binds directly to transportin, thus providing evidence that PABP2 is a novel substrate for this carrier+ Importantly, PABP2 dissociates from transportin in the presence of RanGTP, suggesting that, in vivo, the high concentration of RanGTP in the nucleus triggers release of the cargo (PABP2) into the nucleoplasm, as previously demonstrated for the importin-a/b-mediated import pathway Izaurralde et al+, 1997b;Siomi et al+, 1997)+ The data also suggest that the NLS of PABP2 is contained in its C-terminal domain+ However, the C-terminal sequence of PABP2 (Fig+ 3, bottom diagram) shows no apparent homology with the M9 signal, raising the possibility that it represents a novel type of NLS+ An interesting feature of the C-terminus of PABP2 is that it is very rich in arginine residues, almost all of which are dimethylated (Smith et al+, 1999)+ N G ,N Gdimethylarginine is also a common modified amino acid in hnRNP proteins (Beyer et al+, 1977;Boffa et al+, 1977), and more recently, arginine methylation was shown to be important for the nuclear export of shuttling hnRNP proteins in yeast (Shen et al+, 1998)+ Therefore, an important question to be addressed in future studies is whether dimethylated arginines in the C-terminal domain of PABP2 play a role in transport of the protein in and out of the nucleus+…”