RESEARCH HIGHLIGHTS NUCLEAR TRANSPORT How to get out Nuclear transport is coordinated by nuclear Ran•GTP,which releases cargo from importin proteins and promotes cargo binding to exportin proteins. The role of different to that seen for nuclear import study of other mutant proteins indicated that Ran•GTP in cargo release is reasonably well complexes — is that Cse1 surrounds Ran•GTP free Cse1 adopts a different conformation to understood. But, how does Ran•GTP and Kap60. Cse1 interacts with Ran•GTP at that seen in the nuclear export complex. It function in nuclear protein export to two distinct sites, effectively locking it in the seems that Kap60 binding distorts Cse1 into promote cargo binding? Matsuura and GTP-bound state. The Kap95-binding domain a high-energy, strained conformation, such Stewart now provide insights by describing of Kap60 is also able to bind to the NLS- that the nuclear export complex is spring- the 2.0-Å resolution crystal structure of a binding sites of Kap60 in an autoinhibitory loaded for disassembly following GTP nuclear export complex in Nature. manner and, in the nuclear export complex, hydrolysis in the cytoplasm. The nuclear export complex comprised this domain is clamped to the Kap60 NLS- Rachel Smallridge Ran•GTP,Cse1 (a yeast exportin) and Kap60. binding sites by Ran•GTP and Cse1. The fact Kap60 is an adaptor protein (yeast importin-α) that this intramolecular interaction in Kap60 References and links ORIGINAL RESEARCH PAPER Matsuura, Y. & Stewart, M. that binds to nuclear localization signal is required for Cse1 binding ensures that only Structural basis for the assembly of a nuclear export (NLS)-containing cargo proteins and links cargo-free Kap60 is exported. complex. Nature 432, 872–877 (2004) them to Kap95 (yeast importin-β) in the Mutagenesis studies confirmed the FURTHER READING Hoelz, A. & Blobel, G. Cell biology: popping out of the nucleus. Nature 432, 815–816 (2004) cytoplasm. Following protein import, Kap60 is importance of Kap95-binding-domain WEB SITE exported from the nucleus as Cse1 cargo. interactions in nuclear-export-complex Murray Stewart’s laboratory: A notable feature of the structure — which is assembly and Kap60 export. In addition, the http://www2.mrc-lmb.cam.ac.uk/groups/ms/ EPIGENETICS transcriptional activity, and so it is appropriate that a transcriptional repressor would erase this modification. Erasing the mark Because of its homology to the amine oxidases, the researchers predicted that an LSD1-mediated demethylation reaction would constitute a permanent epigenetic modification. generate formaldehyde. Indeed, when LSD1 Now, Yang Shi and co-workers have brought us was incubated with dimethylated substrate in a one step closer to resolving spectrophotometric assay, large amounts of this question by the identi- formaldehyde were detected. This finding fication of a histone provided further compelling evidence that lysine demethylase, and LSD1 was a bona fide demethylase. they report this finding But does LSD1 repress endogenous genes by in Cell. catalysing histone demethylation? The enzyme is Shi and his team initially set out to charac- found in the Co-REST complex — a transcrip- Methylation terize a protein, LSD1, that had been identified tional repressor that silences neuron-specific of the lysine as a component of several transcriptional- genes. LSD1-deficient HeLa cells that were residues of repressor complexes. First, they showed that generated by RNA interference aberrantly histone LSD1 repressed gene activity in a reporter expressed several neuron-specific target genes, N-terminal assay and that this transcriptional repression and chromatin-immunoprecipitation analyses tails is associated with local was largely dependent on a region of the pro- detected both a loss of association of LSD1 changes in chromatin structure tein that had high sequence homology with and an increase in H3-K4 methylation at the and correlates with both gene amine oxidases — metabolic enzymes that are promoters of these derepressed genes. activation and gene repression. The regulation of putative histone demethylases. This indicated So, it seems that a lysine demethylase has at genes is, however, a dynamic process, and cells that LSD1 might repress transcription by last been pinned down. It is important to note, must therefore possess the ability to rapidly catalysing the demethylation of core histones. however, that an enzyme that removes the generate and erase the ‘methyl mark’.Whereas Direct demethylation assays confirmed methyl marks that identify heritably silenced several histone methyltransferases have been that LSD1 removed the methyl group from chromatin still defies detection… documented, the identification of an enzyme dimethyl and monomethyl histone H3-K4, Shannon Amoils that removes the methyl group from histone but not from trimethyl H3-K4, H3-K9 or sev- References and links ORIGINAL RESEARCH PAPER Shi, Y. et al. Histone tails has remained elusive, which has fuelled eral other methylated lysine and arginine demethylation mediated by the nuclear amine oxidase homolog the debate that histone methylation might residues. H3-K4 methylation correlates with LSD1. Cell 119, 941–953 (2004) 94 | FEBRUARY 2005 | VOLUME 6 www.nature.com/reviews/molcellbio.