Xing-Wang Deng xingwang.deng@ yale.edu Wolfgang Dubiel Ning Wei, Kay Hofmann Kirsten Mundt John Colicelli Jun-ya Kato Michael Naumann Daniel Segal Michael Seeger Michael Glickman Daniel A. Chamovitz Anthony Carr and if the transposase displaces polymer subunits that are nearby on the DNA, transposition will occur most fre- quently at distant sites. Daniel Voytas (Iowa State University, Ames, USA) reported that the Sir 3 and 4 genes, involved in silencing at the telomeres, direct yeast Ty5 to silent chromatin. And Nancy Craig (Howard Hughes Medical Institute/Johns Hopkins University, Baltimore, USA) reported that the TnsE pathway of Tn7 is biased to insert near double strand breaks and at replication termination sites. We imagine this is part of a strategy for Tn7 to get itself transferred onto a conjugating plasmid. Several speakers discussed host suppression of transpos- ition. Among them, Ronald Plasterk (Netherlands Cancer Institute, Amsterdam, The Netherlands) spoke on suppres- sion of transposition in the germ-line of Bristol strains of Caenorhabditis elegans. Surprisingly, the suppression path- way overlaps with the RNAi pathway, a mysterious process by which exogenous double-stranded RNA blocks expression of the proteins it encodes. Silke Jensen (Institut Gustave Roussy, Villejuif, France) reported on suppression of I elements in Drosophila. Interestingly, transgenes con- taining fragments of I, whether in the sense or in the anti- sense direction, suppress invasion by a new I element. Joan Curcio (Wadsworth Center, Albany, USA) reported on inhibition of both Ty transposition and invasive growth by the yeast protein Fus3. Zooming in from evolution to mechanism, several ses- sions focused on the assembly of protein–DNA com- plexes. Assembly is coupled to activation, to ensure that recombination occurs at the proper time and place and on the proper substrates. The crystal structure of Cre bound to a Lox Holliday junction, presented by Greg Van Duyne (University of Pennsylvania School of Medicine, Philadelphia, USA), explained how, of the four DNA mol- ecules in a Holliday junction, only two at a time are acti- vated to be cleaved. Phoebe Rice (University of Chicago, USA) presented the structure of Flp recombinase, confirm- ing biochemical predictions that its catalytic tyrosine, unlike that of closely related Cre, is donated in trans; the tyrosine reaches into the active site of the neighboring sub- unit, to cleave the DNA bound by the neighbor. Similarly, William Reznikoff (University of Wisconsin, Madison, USA) reported that Tn5 transposase donates catalytic acidic residues (the DDE motif) in trans. This mechanism helps ensure that two distant DNA sites move together. Transposons have evolved means to recognize them- selves, to find a good insertion site, and to coordinate the cutting and pasting of their two ends: a complex series of tasks. And yet they must work with the tools they carry on their backs and the proteins they co-opt from their hosts. These elements are streamlined, and we see in them the beauty both of complexity and of simplicity. Outlook MEETING REPORTS Non-homologous recombination TIG May 2000, volume 16, No. 5 202 T he COP9 signalosome was first identified in Arabidopsis thaliana as an essential regulator of light signal transduction 1 . Subsequently, the COP9 signalosome was identified in animal systems, suggesting that this com- plex had a general role in developmental regulation 2,3 . Genetic analysis in Drosophila indicated that this complex is essential for animal development 4 . This role might include the control of the cell cycle and regulation of MAP-kinase signaling 3,5–10 . The biochemical purification of the COP9 signalosome from cauliflower 11 , mammals 9,16 and Arabidopsis 12 indicated that this complex comprises eight core subunits. The complex does not exist in Saccharomyces cerevisiae but appears to be present in Schizosaccharomyces pombe 5 . The COP9 signalosome is similar, both in size and composition, to two other regula- tory complexes: the lid of the 19S proteasome regulatory particle and the eukaryotic translational regulatory com- plex eIF3 (Refs 13, 14). Most subunits from all three complexes contain one of two structural motifs: the PCI/PINT domain and the MPN domain 14,15 . Reports have suggested interactions between these complexes and/or their subunits 3 . Because the genes encoding these subunits were iso- lated either through a variety of unrelated genetic screens, or by the biochemical purification of the complex from various organisms, the original names for the COP9 sig- nalosome subunits are largely unrelated; in most cases, the names do not imply that the protein product is a COP9 signalosome subunit. The fact that many of the subunits have several names adds to the confusion. In order to clarify this situation, a unified subunit nomencla- ture was agreed upon that conveys the fact that these pro- teins function cooperatively within a specific complex, and that will be applicable for all organisms. The sub- units from human, mouse and Drosophila have previ- ously been numbered in decreasing size according to their Unified nomenclature for the COP9 signalosome and its subunits: an essential regulator of development Outlook LETTERS 0168-9525/00/$ – see front matter © 2000 Elsevier Science Ltd. All rights reserved. PII: S0168-9525(00)01982-X