Báo cáo khoa học: INVITED LECTURES potx

INVITED LECTURES 1. Chromosome Architecture and Nuclear Dynamics 1A. Chromosome Structure, Movements and Insulators IL1A-1 Gene regulatory mechanisms in innate immunity T. Maniatis Harvard University, Department of Molecular and Cellular Biology, Cambridge, MA, USA The innate immune system has provided an excellent model for study- ing complex regulatory networks at the levels of signal transduction and transcription. Virus infection of mammalian cells triggers the coor- dinate activation of multiple signaling pathways leading to the activa- tion of specific sets of transcription factors that assemble on the interferon-b (IFN-b) gene enhancer to form an enhanceosome. The function of the enhanceosome is to recruit histone acetylases and chro- matin remodeling complexes that displace a phased nucleosome block- ing the IFN-b promoter. Expression of the IFN-b gene leads to the secretion of IFN-b, which binds to IFN receptors, thus activating the JAK/STAT signaling pathway. This results in the tyrosine phosphory- lation of the transcription factor STAT1 and the formation of the ISGF3 complex consisting of the transcription factors STAT1, STAT2 and IRF-9. STAT1 is further phosphorylated at serine 708 by the non- canonical IjB kinase IKK e, which is required for the activation of a subset of antiviral genes. In the absence of Serine 708 phosphorylation a STAT1 complex is formed, but is unable to bind to a subset of IFN- inducible promoters. DNA sequence comparisons have identified dis- tinct consensus sequences for IKK e-dependent and independent pro- moters. These observations suggest that the phosphorylation of STAT1 by IKK e regulates the structure and/or composition of the ISGF3 complex, which in turn determines the specificity of DNA binding and the nature of the anti-viral response. The results of recent studies bear- ing on the mechanisms of both IFN gene expression and the expression of IFN-inducible genes will be presented. IL1A-2 Chromatin assembly factors, histone H3 variants and cell cycle D. Ray-Gallet, S. Polo, J P. Quivy, A. Groth, D. Roche and G. Almouzni Institut Curie – Recherche, Paris Cedex 05, FRANCE The ordered assembly of chromatin produces a nucleoprotein template capable of regulating the expression and maintenance of the genome functions. Factors have been isolated from cell extracts that stimulate early steps in chromatin assembly in vitro. One such factor, chromatin assembly factor-1 (CAF-1), facilitates nucleosome formation coupled to DNA synthesis. It is thought to participate in a marking system at the crossroads of DNA replication and repair to monitor genome integrity and to define particular epigenetic states. We have begin to approach its critical importance during early development in Xenopus laevis and using mammalian cell systems. In addition, we have now identified a chromatin assembly pathway independent of DN A synthesis. The HIRA protein appears critical for this pathway in Xenopus egg extracts. Notab ly, CAF-1 was part of the the histone H3 complex, H3.1 complex (replica- tive form) and HIRA of the H3.3 complex (replacement form) (Tagami et al., 2004, Nakatani et al., 2004). A major goal in our laboratory is now to better integrate the function of these factors in vivo during develop- ment and also in connection with replication, repair and control of his- tone pools. We will discuss our recent findings on this topics and the interrelationships with other assembly factors. References: Groth A, Ray-Gallet D, Quivy JP, Lukas J, Bartek J and Almouzni G. Human Asf1 regulates the flow of S-phase histones during replica- tional stress. Mol. Cell. 2005; 17: 301–311. Polo S and Almouzni G. Chromatin assembly : a basic recipe with var- ious flavors. Curr. Opin. Genet. Dev. 2006; 16: 104–111. Ge ´ rard A, Koundrioukoff S, Ramillon V, Serge ` re JC, Mailand N, Quivy JP and Almouzni G. The replication kinase Cdc7-Dbf4 pro- motes the interaction of the p150 subunit of Chromatin Assembly Factor 1 with proliferating cell nuclear antigen. EMBO. Reports. 2006; 7: 817–823. Polo S, Roche D and Almouzni G. Evidence for new histone incorpo- ration marking sites of UV-repair in human cells. Cell 2006; 127: 481–493. Loyola A, Bonaldi T, Roche D, Imhof A and Almouzni G. PTMs on H3 variants before chromatin assembly potentiate their final epige- netic state. Mol. Cell. 2006; 24: 309–316. Groth A, Rocha W, Verreault A and Almouzni G. Chromatin chal- lenges during DNA replication and repair. Cell. 2007; 128: 721–733. Groth A, Corpet A, Cook A, Roche D, Bartek J, Lukas J and Almo- uzni G. Regulation of replication fork progression through histone supply/demand. Science; 318: 1928–1931. IL1A-3 Genome organization in the formation of cancer translocations and DNA repair T. Misteli National Cancer Institute, NIH, Bethesda, MD, USA Chromosomal translocations and genomic instability are universal hall- marks of tumor cells. While the molecular mechanisms involved in the formation of translocations and DNA repair are rapidly being eluci- dated, a cell biological understanding of is largely lacking. We have developed an experimental system in which a single double strand break (DSB) can be induced in a controlled manner at a defined geno- mic site and the fate of each of the two damaged DNA ends can be followed in real time in living cells. We have demonstrated that broken chromosome ends are positionally immobile suggesting that transloca- tions can only occur amongst proximally positioned chromosomes. These observations support our earlier findings demonstrating that the spatial proximity of genome regions directly correlates with their trans- location frequency and with tissue-prevalence of translocations. This experimental system also provides a unique tool to probe how DSBs are recognized in vivo and how DNA damage response (DDR) path- ways are elicited in the context of chromatin. Remarkably, we find that a full DNA damage response can be triggered in the absence of DNA damage by tethering DNA damage sensor proteins to chromatin. These observations strongly suggest that the cellular DNA damage response can be activated in the absence of DNA lesions. These results have implications for our conceptual and mechanistic understanding of how cancer translocations form and how DNA damage is detected in vivo. IL1A-4 A model for all genomes: gene activation and repression P. R. Cook The Sir William Dunn School of Pathology, Oxford, UK A model for the 3D structure of all genomes will be presented; it is based on the structure of the bacterial nucleoid, where active RNA polymerases cluster into ‘factories’ to loop the intervening DNA. Essentially all transcription takes place in factories where the local con- centration of polymerases is so much higher than in the soluble pool. Aspects of this organization in mammalian cells will be described: C Nucleoplasmic factories can be visualized by electron spectroscopic imaging after allowing permeabilized HeLa cells to extend transcripts by 40 nucleotides in BrUTP, and immuno labelling the resulting BrRNA with gold particles; nascent transcripts and their templates are found on the surface of huge protein-rich structures – the factories (diameter 87 nm, mass 10 MDa). C Results obtained using RNA ‘FISH’ and ‘3C’ show that factories specialize in transcribing different genes depending on promoter type and presence of an intron – a pro- moter must diffuse to the appropriate factory before it can initiate. C Monte Carlo simulations indicate how proximity to a factory affects the frequency that a promoter contacts a factory (and so initiates); bar- riers, silencers, and enhancers – which encode active transcription units and/or binding sites for transcription factors – regulate activity by teth- ering target genes at relevant distances from appropriate factories. Reference: Marenduzzo D, Faro-Trindade I, and Cook PR. What are the molecular ties that maintain genomic loops? Trends Genet. 2007; 23: 126–133. Abstracts 1A. Chromosome Structure, Movements and Insulators 4 FEBS Journal 275 (Suppl. 1) 4–64 (2008) ª 2008 The Authors Journal compilation ª 2008 FEBS IL1A-5 Analyzing chromatin during animal development D. J. McKay, M. Agelopoulos and R. S. Mann Department of Biochemistry and Molecular Biophysics, Columbia University, NY, USA Animal development requires the precise and coordinated activities of many transcription factors, which function combinatorially to generate unique cellular fates. Over the past decade, many experiments have linked transcription factor activities with affects on chromatin struc- ture. In general, these experiments have been limited to systems that use nearly pure populations of cells, such as unicellular organisms or cells grown in culture. In contrast to these more simple systems, addressing the role of chromatin structure in developing embryos is problematic because of the large number of cell types, which likely vary in chromatin status at specific loci. Monitoring chromatin struc- ture in whole embryos, for example, cannot reveal differences that con- tribute to gene regulation in a tissue-specific manner. To circumvent this problem, we have developed a new technique in Drosophila that we call, cell and gene specific Chromatin Immunoprecipitation (cgChIP). In this method, specific cis-regulatory elements, such as pro- moters and enhancers, are tagged with a high affinity DNA binding site for a bacterial DNA binding protein. Expression of this DNA binding protein in a cell type-specific manner followed by immunopre- cipitation allows the purification of the tagged chromatin. Subsequent immunoprecipitations with antibodies directed against histone modifi- cations, transcription factors, or remodeling complexes can reveal the chromatin status of the tagged cis-regulatory elements in specific cell types. cgChIP can also be used to examine the tissue-specific interac- tions between promoters and distant enhancers. Results using this method, applied to the developmentally regulated gene Distalless, will be described. IL1A-6 Chromosomal networks and epigenetic reprogramming R. Ohlsson Department of Development & Genetics, Evolution Biology Centre, Uppsala University, SWEDEN The epigenome is usually described with overlapping patterns of his- tone and DNA modifications at the primary chromatin fibre level to establish/manifest active and inactive states. However, it is becoming increasingly clear that epigenetic states also involve extensive crosstalk both within and between chromosomes. By using a circular chromo- somal conformation capture (4C) method we have earlier shown that this crosstalk is regulated by epigenetic marks for a selected set of sequences and that it can confer transcriptional regulation in trans. Here we describe a genome-wide perspective using high-resolution NimbleGen microarrays. More than 700 different regions were found to interact with the H19 imprinting control region (ICR) at one point or another. While the initial study, representing a smaller subset of the genome, demonstrated that many interactions were specific for the maternal allele, the genome-wide screen revealed that the CTCF binding sites of the maternal allele chisel out specific patterns of chro- mosomal networks by both promoting and preventing local and long- ranging interactions. Extensive chromosomal networks could also be observed in undifferentiated ES cells. Upon differentiation to embryoid bodies there is a dramatic and virtually complete exchange of sequences interacting with the H19 ICR, suggesting that chromosomal networks reflect and/or participate in epigenetic reprogramming steps. As the pattern of chromosomal networks in ES cells and embryoid bodies were linked with gene repression and activation, respectively, we propose that the H19 ICR participates in a larger transcriptional regu- lation network that physically connects a subset of the genome. 1A. Chromosome Structure, Movements and Insulators Abstracts FEBS Journal 275 (Suppl. 1) 4–64 (2008) ª 2008 The Authors Journal compilation ª 2008 FEBS 5 1B. Imprinting and Epigenetic Regulation IL1B-1 Epigenetic regulation of Polycomb group proteins mediated higher order structures during the cell cycle C. Lanzuolo and V. Orlando Dulbecco Telethon Institute, European Brain Research Institute and IRCCS Santa Lucia, Rome, ITALY During cell differentiation or metabolic switch, cells undergo pro- found changes in gene expression. These events are accompanied by complex modifications of chromosomal components and nuclear structures that contribute to the stability and heritability of tran- scription programmes. The proteins of the Polycomb group (PcG) proteins are part of a conserved cell memory system that conveys epigenetic inheritance of repressed transcriptional states by regulating multiple levels of chromatin structure. Defects in any of the compo- nents of this system results in a severe impairment of developmental programmes, stem cell renewal and in an increased rate of cell trans- formation. In Drosophila, PcG function is mediated by specialized epigenetic DNA modules, called Polycomb response elements (PREs). We have recently shown that, in the repressed state, all major ele- ments that have been shown to bind PcG proteins, including PREs and core promoters give rise to a topologically complex higher order structure, important for epigenetic silencing [1]. We went further to try to clarify the exact role of Polycomb proteins and higher order structures in epigenome reprogramming during the cell cycle. By combining ChIP and 3C approaches, we have been analysing i) the stability and regulation of PRE-mediated higher order struc- tures during the cell cycle, in particular during S-phase, ii) the dynamics and potential ordered recruitment of PRC2 and PRC1 components during the cell cycle and iii) the replication timing of PREs. Our data suggest that both PRE-mediated higher order struc- tures and PcG protein complexes dynamics appear to be regulated during specific moments of early and late S-phase. Reference: 1. Lanzuolo et al., Nat Cell Biol 2007 IL1B-2 4C Technology: uncovering the multi-dimensional organisation of the genome M. Simonis, D. Noordermeer, P. Klous, E. Splinter and W. De Laat Department of Cell Biology, Erasmus MC, Rotterdam, THE NETHER- LANDS The architect ure of DNA in the cell nucleus is an emerging key con- tributor to genome function. Based on microscopy studies, genes have been found to locate at different positions in the nucleus depending on their expression status. Such relocation is often mediated by regulatory DNA elements like enhancers that control the transcriptional output of these genes. When looking by 3C technology at higher resolution at single gene loci, regulatory DNA elements are found to physically con- tact target genes for their activation, thereby looping out intervening DNA. We recently developed 4C technology, a high-throughput tech- nique that combines 3C technology with tailored micro-arrays to uniquely allow for an unbiased genome-wide search for DNA loci that interact in the nuclear space. Here, we will present data based on 4C technology that investigate the nuclear environments that genes leave and enter when they get activated. We have targeted the prototype of a strong, tissue-specific transcription regulatory DNA element, the beta- globin locus control region (LCR), to a gene-dense region of the mouse genome. When active, the LCR repositions this locus in the nucleus. We will reveal the genomic environments this locus leaves and enters upon LCR activity and we will show how expression in cis and in trans is affected by this ectopic LCR. 4C technology not only allows investi- gating the three-dimensional structure of DNA in the nucleus, it also accurately reconstructs at least 5–10 megabases of the one-dimensional chromosome sequence map around the target sequence. Changes in this physical map as a result of genomic rearrangements are therefore iden- tified by 4C technology. We will demonstrate that 4C detects deletions, balanced inversions and translocations in patient samples at a resolu- tion (7 kb) that allowed immediate sequencing of the breakpoints. 4C technology therefore offers a high-resolution genomic approach that can identify both balanced and unbalanced genomic rearrangements. IL1B-3 Co-transcriptional recruitment of the dosage compensation complex to X linked target genes A. Akhtar EMBL, Gene Expression Program, Heidelberg, GERMANY Dosage compensation is a process required to balance the expression of X-linked genes between males and females. In Drosophila this is achieved by targeting the dosage compensation complex or the MSL complex to the male X chromosome. In order to study the mechanism of targeting, we have studied two X chromosomal genes, mof and CG3016, using chromatin immunoprecipitation as well as immuno- FISH analysis on transgenic flies. We show that MSL complex recruit- ment requires the genes to be in a transcriptionally active state. MSL complex recruitment is reversible because blocking transcription severely reduces MSL binding to its target genes. Furthermore, target- ing cues are found towards the 3’ end of the gene and depend on the passage of the transcription machinery through the gene, whereby the type of promoter and the direction of transcription are dispensable. We propose a model of dynamic MSL complex binding to active genes based on exposed DNA target elements. IL1B-4 Poised transcription complexes in epigenetics and genome function E. Brookes, S. Q. Xie, J. K. Stock, I. Jesus, L. Lavitas and A. Pombo Genome Function Group, MRC Clinical Sciences Centre, Imperial Col- lege School of Medicine, London, UK Embryonic stem (ES) cells are characterised by their pluripotency and ability to self-renew. A number of silent, developmentally-important genes display ‘bivalent’ histone modifications that are associated with both silent and active chromatin and are regulated by Polycomb repressor complexes (PRC1 and PRC2). We have recently identified the presence of RNA polymerase II (RNAPII) complexes at bivalent genes in an unusual poised conformation, associated with low-level transcript production. Poised RNAPII complexes are detected at biva- lent genes in amounts comparable to active genes, and are associated with high levels of phosphorylation of Ser 5 residues of the C-terminal domain of RNAPII, no detectable phosphorylation on Ser 2 residues, and little detectable reactivity to an antibody raised against unphos- phorylated CTD repeats. De-repression of bivalent genes after removal of Ring1A and B, two PRC1 enzymatic components responsible for H2A ubiquitination, is accompanied by a partial change in the confor- mation of poised RNAPII complexes, but occurs in the absence of Ser 2 phosphorylation which is typical of active genes. We currently investi- gate the contribution of poised RNAPII and PRC complexes to gen- ome function and epigenetic processes in ES cells, and in priming developmental regulator genes for activation upon onset of differentia- tion pathways. Our observations provide novel insight into the molecu- lar mechanisms that allow ES cells to self-renew and yet retain the ability to expedite multiple lineage outcomes. Abstracts 1B. Imprinting and Epigenetic Regulation 6 FEBS Journal 275 (Suppl. 1) 4–64 (2008) ª 2008 The Authors Journal compilation ª 2008 FEBS 1C. Nuclear Extrachromosomal Structures IL1C-1 Mapping the epigenetic landscape of the kinetochore W. C. Earnshaw 1 , M. Nakano 2 , S. Cardinale 1 , V. N. Noskov 2 , R. Gassmann 1 , P. Vagnarelli 1 , S. Kandels-Lewis 1 , V. Larionov 2 and H. Masumoto 2,3 1 Wellcome Trust Centre for Cell Biology, University of Edinburgh, Edinburgh, UK, 2 National Cancer Institute, National Institutes of Health, Bethesda, MD, USA, 3 Division of Biological Science, Nagoya University, Nagoya, JAPAN We describe the first human synthetic kinetochore, which can be inacti- vated by targeted modification of its epigenetic configuration in vivo. The kinetochore was assembled on a human artificial chromosome (HAC) in HT1080 cells using a DNA array based upon a synthetic alpha-satellite (alphoid) repeat containing one natural monomer, with a binding site for CENP-B (CENP-B box), and one completely syn- thetic monomer in which the region corresponding to the CENP-B box was replaced with a tetracycline operator (tetO). Targeting the syn- thetic kinetochore with a transcriptional activator (tTA) or repressor (tTS) inactivated it by changing the balance between centromeric chro- matin and heterochromatin. Interestingly, formation of heterochroma- tin with HP1 appeared to destabilize the HAC more efficiently than activating transcription of the kinetochore DNA. Inactivation of the kinetochore blocked segregation of the HAC, which formed extremely small micronuclei (nanonuclei) and was rapidly lost from the cells. In controls, binding of several other tetracycline-repressor fusion proteins had no significant effect on HAC stability. The opportunity to selec- tively target different proteins into an active kinetochore and thereby regulate centromere function opens the way for an unprecedented mechanistic and structural analysis of the human kinetochore as well as for development of new HAC-based conditional gene expression sys- tems. IL1C-2 Nuclear dynamics: Pathways, proteins and progress L. Wallrath 1 , S. Speese 2 , G. Dialynas 1 , B. Curio-Penny 1 ,D.E. Cryderman 1 , P. K. Geyer 1 , V. Budnick 2 and S. Schulze 3 1 Department of Biochemistry, University of Iowa, Iowa City, IA, USA, 2 Department of Neurobiology, University of Massachusetts, Worcester, MA, USA, 3 Department of Biology, Western Washington University, Bellingham, WA, USA Introduction: The nuclear envelope (NE) is lined with a network of intermediate filament proteins called lamins. Lamins are classified as A- or B-types based on structure and expression pattern. A-type lamins localize to the nuclear periphery and within extrachromosomal nuclear domains, in some cases forming novel nuclear bodies. A-type lamins provide structural support for the nucleus, organize the genome by making contacts with chromatin and transmit signals from the cyto- plasm. In humans, mutations in A-type lamins give rise to diseases with tissue-specific defects, such as Emery-Dreifuss muscular dystro- phy. Here, Drosophila is serving as a model to reveal tissue-specific functions of A-type lamins in development, with implications for understanding human disease mechanisms. Methods: Transgenic flies expressing wild type and mutant versions of Lamin C that mimic human disease forms were generated. Effects of expressing mutant versions of Lamin C ubiquitously or in specific tis- sues were analyzed. Results: Expression of wild type Lamin C does not alter nuclear archi- tecture or viability. Substitutions within the Lamin C rod domain cause extrachromosomal Lamin C aggregation, yet support viability. In con- trast, deletion of the head domain causes no apparent nuclear defects, but results in semi-lethality when expressed specifically in larval muscle. Adult ‘escapers’ possess malformed legs remarkably similar to those produced in flies with defects in ecdysone signaling. Conclusions: These studies demonstrate that Lamin C plays a devel- opmental role in muscle function, with implications for regulating gene expression through hormone signaling. IL1C-3 Tinkering with a cellular nanomachine by a minimalist approach R. Y. H. Lim 1 , B. Fahrenkrog 1 , J. Deng 2 , L. Kapinos 1 , J. Koeser 1 and U. Aebi 1 1 M.E. Mu ¨ ller Institute for Structural Biology, Biozentrum, University of Basel, Basel, SWITZERLAND, 2 Institute of Materials Research and Engineering, SINGAPORE Nucleocytoplasmic transport is fundamental to eukaryotic cells and describes the bidirectional exchange of molecular cargoes between the nucleus and the cytoplasm across distinct perforations in the nuclear envelope (NE), called nuclear pore complexes (NPCs) [1]. Vertebrate NPCs are cellular nanomachines consisting of 30 different proteins called nucleoporins (Nups). To come to understand the modus operandi of the NPC’s selective gating mechanism, i.e. its ability to restrict or promote bidirectional cargo translocation necessitates a detailed knowl- edge of both the biochemical ingredients and the corresponding physi- cal responses of the NPC machinery. To uncover this mechanism, we have developed a heuristic, interdisciplinary approach involving phenyl- alanine-glycine (FG)-molecules that are tethered to gold nanostructures designed to mimic the NPC geometry and dimensions. Atomic force- volume mapping yields the collective behavior of such surface-tethered FG-repeat domains, indicating that these are thermally mobile and exist in a polymer brush-like conformation, thus forming an entropic barrier [2]. Access to the NPC is provided by transport receptors caus- ing a transient collapse of the FG-molecules. We reason that the reversible, receptor-driven collapse of the FG-domains specifies the physical aspects of selective gating, and propose that the flux of col- lapsing and distending FG-domains serves to promote translocation of receptor-cargo complexes while simultaneously maintaining the entro- pic barrier [3]. Last, we will demonstrate how the principles of nano- mechanical selective gating can be applied to the construction of a de novo designed ‘minimalist’ NPC. References: 1. Lim and Fahrenkrog. Curr. Opin. Cell. Biol. 2006; 18: 1–6 2. Lim et al., Proc. Natl. Acad. Sci. USA. 2006; 103: 9512–17 3. Lim et al., Science. 2007; 318: 640–3 IL1C-4 Spliceosome assembly and recycling in the nucleus M. Carmo-Fonseca Instituto de Medicina Molecular, Faculdade de Medicin, Universidade de Lisboa, Lisboa, PORTUGAL In higher eukaryotes, the vast majority of protein-coding genes contain introns that must be removed with great precision by the spliceosome. Spliceosomes are massive RNA-protein macromolecular machines with over 200 distinct components that assemble onto nascent transcripts and are released from the RNA after splicing. A large body of evidence supports the view that the spliceosome is constructed de novo, piece- by-piece, for each round of splicing. However, how splicing factors are recruited to nascent transcripts in the nucleus is unknown. To address this question, our lab uses live-cell microscopy approache s (including photobleaching, photoactivation and FRET methodologies) combined with computer modelling. Our data argue that spliceosomal compo- nents are not sequestered in dedicated nuclear compartments until a signal triggers their recruitment to nascent transcripts. Rather, the results sugges t that splicing proteins are constantly diffusing through- out the entire nucleus and collide randomly and transiently with pre- mRNAs. Our data further suggest that spliceosomal components can interact with each other in complexes devoid of pre-mRNA. We pro- pose that the nucleus contains highly dynamic extra-spliceosomal com- plexes. These may ensure a high local concentration of spliceosomal subunits in close vicinity to active genes, thereby increasing the prob- ability of productive collisions with the pre-mRNA. 1C. Nuclear Extrachromosomal Structures Abstracts FEBS Journal 275 (Suppl. 1) 4–64 (2008) ª 2008 The Authors Journal compilation ª 2008 FEBS 7 IL1C-5 Protein dynamics in the cell nucleus A. Lamond 1 , F. M. Boivert 1 , M. Ono 1 , L. Trinkle-Mulcahy 1 , S. V. Koningsbruggen 1 , D. Lleres 1 , J. Andersen 2 and M. Mann 3 1 Division of Gene Regulation and Expression, Wellcome Trust Biocentre, MSI/WTB Complex, University of Dundee, Dundee, UK, 2 Center for Experimental Bioinformatics, Department of Biochemistry and Molecular Biology, University of Southern Denmark, Odense, DENMARK, 3 Department of Proteomics and Signal Transduction, Max-Planck Institute for Biochemistry, Martinsried, GERMANY We are studying the functional organization of mammalian cell nuclei using a dual strategy that combines mass spectrometry (MS) based pro- teomics with live cell fluorescence imaging (see www.LamondLab.com). This allows two distinct but complementary quantitative techniques to be applied to analyse the same biological problem, providing a rigorous approach where potential artifacts or limitations of one method are avoided in the complementary approach and vice versa. The quantita- tive proteomic methods involve metabolic labeling of cellular proteins in cultured cell lines with the amino acids lysine and arginine containing either the heavy isotope 13 C, or 13 C and 15 N. The quantitative imaging experiments, including time-lapse microscopy, FRAP, FLIP, FLIM and FLIM-FRET, are performed on mammalian cell lines stably expressing one or more fluorescent protein-tagged reporters. Both the MS and imaging methods are used to study the same stable cell lines, allowing a direct comparison the resulting data from both techniques. Using this dual strategy we have studied the protein composition of nucleoli under a range of different metabolic conditions and at specific stages of cell cycle progression (see http://lamondlab.com/nopdb/). Using dual pulse- labeling methods we have characterized also the turnover of nucleolar proteins. The data show that a large fraction of newly synthesized and imported r proteins are continually degraded within the nucleus. We have recently developed a MS-based proteomics strategy to perform quantitative analyses of subcellular protein localization and to map pro- tein-protein interactions. We have demonstrated that this approach can used to analyse the effects of genotype on protein localization, dynamics and interactions. This strategy thus offers a general approach for char- acterizing the composition, dynamic properties and interactions of either organelles or multi-protein complexes. IL1C-6 Epigenetic regulation of bone-phenotypic genes B. Henriquez, F. Cruzat, M. Hepp, J. Olate, S. Gutierrez and M. Montecino Department of Biochemistry and Molecular Biology, School of Biological Sciences, University of Concepcion, Concepcion, CHILE Introduction: Runx2 transcription factor is essential for skeletal devel- opment as it regulates the expression of key bone-related genes. Tran- scription of the RUNX2/p57 form is controlled by the P1 promoter sequence, although the mechanisms that regulate this expression are poorly understood. Changes in chromatin structure are mediated by SWI/SNF-type chromatin remodeling complexes, which alter chroma- tin in an ATP-dependent manner, and by nuclear complexes that cova- lently modify structural components of the chromatin. Here, we assess mechanisms that control chromatin remodeling and transcriptional activity of the RUNX2 gene in osteoblastic cells. Methods: Osteoblastic cells that inducibly express inactive SWI/SNF complexes were generated to evaluate the contribution of SWI/SNF activity to both chromatin remodeling and gene transcription. Histone modifications and interaction of regulatory factors at the RUNX2 gene P1 promoter were assessed by chromatin immunoprecipitation assays. Results: Chromatin-remodeling at the P1 promoter region accompa- nies transcription of the RUNX2 gene as reflected by the presence of two nuclease hypersensitive sites spanning key regulatory elements for RUNX2 transcription. Chromatin-remodeling and transcription are independent of SWI/SNF activity but are tightly associated with enhanced histone acetylation. A pattern of covalent modifications on histones H3 and H4 was detected in the nucleosomes organizing the regulatory regions of the RUNX2 P1 promoter during transcription. Among the modifications found are histone H3 acetylation (H3K9 and H3K14) and methylation (H3K4 and H3R2) as well as histone H4 acetylation (H4K16) and methylation (H4K20). Conclusions: Our results indicate that transcription of the RUNX2/ p57 in osteoblasts involves specific epigenetic mechanisms that are independent of SWI/SNF activity. Abstracts 1C. Nuclear Extrachromosomal Structures 8 FEBS Journal 275 (Suppl. 1) 4–64 (2008) ª 2008 The Authors Journal compilation ª 2008 FEBS 1D. Evolving Genomes and Synthetic Biology IL1D-1 Sequence similarity network reveals common ancestry of multidomain proteins D. Durand Departments of Biological Sciences and Computer Science, Pittsburgh, PA, USA We address the problem of homology identification in complex multi- domain families with varied domain architectures. The challenge is to distinguish sequence pairs that share common ancestry from pairs that share an inserted domain, but are otherwise unrelated. This distinction is essential for accuracy in gene annotation, function prediction and comparative genomics. There are two major obstacles to multidomain homology identification: lack of a formal definition and lack of curated benchmarks for evaluating the performance of new methods. We offer preliminary solutions to both problems: (i) an extension of the tradi- tional model of homology to include domain insertions; and (ii) a man- ually curated benchmark of well-studied families in mouse and human. We further present Neighborhood Correlation, a novel method that exploits the local structure of the sequence similarity network to iden- tify homologs with great accuracy, based on the observation that gene duplication and domain shuffling leave distinct patterns in the sequence similarity network. In a rigorous, empirical comparison using our curated data, Neighborhood Correlation outperforms sequence similar- ity, alignment length, and domain architecture comparison. Neighbor- hood Correlation is well-suited for automated, genome-scale analyses. It is easy to compute, does not require explicit knowledge of domain architecture, and classifies both single and multidomain homologs with high accuracy. Our study demonstrates the utility of mining network structure for evolutionary information, suggesting this is a fertile approach for investigating evolutionary processes in the post-genomic era. IL1D-2 Population genomics of human gene expression E. T. Dermitzakis Wellcome Trust Sanger Institute, Hinxton, Cambridge, UK The recent comparative analysis of the human genome has revealed a large fraction of functionally constrained non-coding DNA in mamma- lian genomes. However, our understanding of the function of non-cod- ing DNA is very limited. In this talk I will present recent analysis in my group and collaborators that aims at the identification of function- ally variable regulatory regions in the human genome by correlating SNPs and copy number variants with gene expression data. I will also be presenting some analysis on inference of trans regulatory interac- tions and evolutionary consequences of gene expression variation. IL1D-3 Frequent and widespread independent evolution of protein sequences A. Rokas Department of Biological Sciences, Vanderbilt University, Nashville, TN, USA Molecular homoplasy, the evolution of the same amino acids at orthol- ogous sites in unrelated taxa, is an important biological parameter. Deciphering the extent and causes of homoplasy are key to understand- ing the processes that have sculpted the protein record, but relatively little is known about its prevalence in eukaryotic proteomes. To quan- tify the extent of homoplasy among evolving proteins, we used phylo- genetic methodology to analyze eight data matrices from clades spanning the eukaryotic tree of life. We found that the frequency of homoplastic amino acid substitutions in eukaryotic proteins was more than 2-fold higher than expected under widely accepted neutral models of protein evolution. Detailed inspection revealed that approximately 43% of these homoplastic substitutions were between very frequently exchangeable amino acids that are only one mutational step away, with fewer than 1% of substitutions occurring between rarely exchangeable amino acids one mutational step away. We discuss plausible neutral and selective explanations for these results and their consequences for studies of molecular evolution and comparative genomics. IL1D-4 Genomics, evolution and evolution of genomics N. Kyrpides Genome Biology Program, DOE Joint Genome Institute, Walnut Creek, CA, USA Efficient analysis and interpretation of isolate microbial genomes is a growing area that is expected to lead to advances in healthcare, envi- ronmental cleanup, agriculture, industrial processes, and alternative energy. At the same time, the recent advancements in genome technol- ogy are mediating a transition from organismal to community genom- ics (metagenomics) revealing a new universe of microbial communities, involving previously uncultured organisms. This revolution is expected to lead to further advances in understanding the structure and function of entire microbial communities and ultimately describe the physical and biological context in which microorganisms coordinately operate in nature. In parallel, it already provide a window in the rapidly approaching future challenges of storing, processing, integrating, pre- senting and analyzing thousands of isolate genomes. I will discuss the current challenges in the field and present a vision of its future. 1D. Evolving Genomes and Synthetic Biology Abstracts FEBS Journal 275 (Suppl. 1) 4–64 (2008) ª 2008 The Authors Journal compilation ª 2008 FEBS 9 IL1D-5 How do restriction enzymes acquire specificities for different target sites? V. Siksnys 1 , G. Tamulaitis 1 , M. Zaremba 1 , R. Szczepanowski 2 and M. Bochtler 2 1 Institute of Biotechnology, Vilnius, LITHUANIA, 2 IICMB, Warsaw, POLAND Restriction enzymes recognize 4–8 bp DNA sequences and cut at their target site with exquisite specificity. Many orthodox restriction endonucleases interacting with symmetric recognition sites achieve sequence specificity making direct nucleobase-amino acid contacts with their target sites. Ecl18kI and EcoRII-C/PspGI restriction enzymes recognize interrupted pentanucleotide sites CCNGG and CCWGG, respectively. Structural studies demonstrate that Ecl18kI/ EcoRII-C/PspGI and the NgoMIV restriction enzyme specific for the non-interrupted GCCGGC site share striking structural similarities and interact identically with the symmetrical CC:GG half-sites. It turned out that Ecl18kI specific for the CCNGG and EcoRII-C/ PspGI specific for the CCWGG site flip central nucleotides from DNA double helix to interact with the conserved CC:GG parts of their target sequences and achieve cleavage. However, Ecl18kI and EcoRII-C/PspGI accept different base pairs at the center, raising a question whether the base pair strength/stability or a direct readout of the flipped out bases determine the differences in specificity. Using a combination of the X-ray structural analysis and biochemi- cal methods we show that EcoRII-C/PspGI achieve the specificity for the CCWGG sequence by a double check mechanism by probing both the stability of the central base pair and identity of the flipped out base in the pocket. Abstracts 1D. Evolving Genomes and Synthetic Biology 10 FEBS Journal 275 (Suppl. 1) 4–64 (2008) ª 2008 The Authors Journal compilation ª 2008 FEBS 1E. Pharmacogenomics IL1E-1 Gene specific therapeutic development and the application of new technological advances to gene discovery for human complex traits D. Housman Massachusetts Institute of Technology, Department of Biology and Koch Institute for Integrative Cancer Research, Cambridge, MA, USA In my presentation, I will discuss the evolution of new strategies for the discovery and evaluation of genetic variants responsible for com- plex genetic traits. In particular, I will discuss the impact of single mol- ecule DNA sequencing and transcriptional analysis to the problem of complex trait genetics. I will address specific examples from studies of modifier genes for age of onset for Huntington’s disease, cardiovascular disease phenotypes and cancer treatment response. Development of therapeutic intervention based on disease gene phenotype in these cases will be considered. IL1E-2 The era of genome wide scans – applications in pharmacogenomics P. Deloukas The Wellcome Trust Sanger Institute, Hinxton, Cambridge, UK Following SNP discovery and characterisation of patterns of common variation in multiple populations (HapMap) we have been able to select genome-wide marker sets to test for association to common com- plex traits including disease, susceptibility to pathogens and variable response to drugs. Last year witnessed the reporting of many disease susceptibility loci widely replicated in large independent samples. Well powered studies yield true associations as demonstrated by the Well- come Trust Case Control Consortium study in which we identified over 20 independent loci reaching genome wide significance for conditions such as type 1 and 2 diabetes, and Crohn’s disease. We interrogated 500 000 SNPs (Affymetrix) in 2000 British samples of each case series and a set of 3000 common controls. The use of robust genotype calling algorithms, stringent quality control metrics of genotypic data (call rates >95%) and exclusion of ethnic outliers are essential. Imputation methods which generate in silico genotypes of untyped SNPs based on HapMap, are emerging as a powerful tool to combine data from differ- ent genotype platforms. Pharmacogenetic studies have focused so far on candidate genes but the limitation of this approach is seen in the newly discovered disease loci that typically do not figure on candidate gene lists. Clinical use of the anticoagulant drug warfarin is hindered by wide inter-individual variation in dose required and by a serious bleeding side effect in a minority of patients. We will discuss finding of a genome scan we undertook in 1000 Swedish warfarin patients (WARG study) with Illumina’s 370K Infinium chip. IL1E-3 Application of pharmacogenetics and pharmacogenomics in pediatrics J. S. Leeder Division of Pediatric Pharmacology and Medical Toxicology, Children’s Mercy Hospitals and Clinics, Kansas City, MO, USA The 18-year time span between birth and attainment of legal status as an adult is a period of profound changes in physiology, and these physiolog- ical changes include maturational processes as well as those impacting drug disposition and response. As a consequence, the application of pharmacogenetic and pharmacogenomics principles to drug therapy in children is superimposed on the dynamic changes in gene expression and their functional consequences throughout the deve lopmental process. Thus, it is necessary to consider pharmacogenetics and pharmacogenom- ics in this context. For example, it is now becoming apparent that there are important developmental differences in the ontogeny of specific drug biotransformation enzymes for both phase 1 reactions (primarily cyto- chrome P450 (CYP) and flavin monooxygenase (FMO) genes) and phase 2 reactions involving glucuronosyltransferases (UGTs) and sulfotransfe- rases (SULTs), among others. In general terms, the unborn fetus and newborns possess rather limited drug biotransformation capacity but functional activity is acquired post-natally in gene- and isoform-depen- dent patterns. Inter-individual variability in the developmental trajectory of each pathway complicates therapeutic decisions at this developmental stage, and is accompanied by a high risk of drug toxicity in this age group. However, there are few, if any, means by which patients at high risk for drug toxicity or lack of efficacy can be identified a priori. The same is true of older ages, but recent in vitro drug biotransformation investigations and longitudinal in vivo phenotyping studies are beginning to shed light on the extent of population variability at different develop- mental stages and the genetic basis of the variability. Longitudinal stud- ies are also able to explore issues related to the consequences of maturation on patterns of drug bioactivation and detoxification that might underlie the increased susceptibility of children to specific adverse drug reactions (e.g. valproic acid hepatotoxicity). The phenomena of developmental trajectories and maturation of drug biotransformation in children will be illustrated using the ontogeny of cytochrome P450 2D6 (CYP2D6) and valproic acid biotransformation as examples. Despite advances in understanding developmental drug biotransformation phe- notypes, our ability to translate a developmental pharmacogenetic phe- notype into a personalized therapeutic strategy is modest at best. One major area of deficiency is the impact of development on the expression of drug targets throughout the newborn period, childhood and adoles- cence. To address this issue, strategies must move away from targeted investigations of single genes and toward modeling the function of path- ways and gene networks as phenotypes. Developmental pharmacoge- nomics, then, should encompass detailed characterization of the genetic basis of inter-individual variability in drug response as it changes throughout maturation. Opportunities exist in many areas, including asthma, autism and pervasive developmental disorders, epilepsy and other disease processes that have their origins in childhood and persist into adult life as improvement in therapeutic interventions early in life may have important consequences at later stages of life. IL1E-4 Cardiovascular drugs: integration of genomics personalized information and transcriptomics G. Siest Equipe Inserm ‘‘Ge ´ ne ´ tique Cardiovasculaire’’ Du CIC 9501, Universite ´ Henri Poincare ´ , Nancy, FRANCE Personalized medicine is based on a better knowledge of biological var- iability between individuals, considering the important part due to genetics. When trying to identify genes and products of genes involved in differential cardiovascular drug responses, a five-step strategy is to be followed: (1) Pharmacokinetic-related genes and phenotypes; (2) Pharmacodynamic targets, genes and products; (3) Cardiovascular diseases and risks depending on specific or large metabolic cycles; (4) Physiological variations of previously identified genes and proteins; and (5) Environment influences on them. Pharmacogenomics and pharmacoproteomics approaches of many car- diovascular drugs are very often essentially looking to the first step, the pharmacokinetic one with the main CYP involved in the oxidation of the drugs and the ABC transporters. But the majority of these enzymes and systems are membranous, localized in the endoplasmic reticulum and are rarely released into blood or plasma. By the way, the only enzymes mea- surable are the cytosolic ones: glutathione S-transferases and sul- fotransferases. Fortunately, blood carries cellular and non-cellular components that could be use for measuring: (i) Sulfotransferase (plate- lets); (ii) GSH-T (red and white cells); and (iii) CYP450 (WBC). WBC might be useful to follow through a transcriptomic approach the gene products of many drug metabolizing enzymes including the following CYP: 1A1, 1A2, 1B1, 2A6, 2B6, 2C, 2D6, 2E1, 3A3, 4B1, 4F. Phase III transporters i.e. the ABC genes also expressed in lymphocytes could be surrogate markers for testing individual responses to cardiovascular risks factors or to cardiovascular drugs such as statins or fibrates. Finally, we should not forget to look simultaneously to the transcription factors present in WBC, particularly Ah receptor, other nuclear factors (RXR, CXR, PPAR) which are regulating the expression of drug metabolizing enzymes. We will attempt to review the situation of all these enzymes and genes in pre-vision to their use in pharmacogenomics and to describe the biological variations of some messenger RNA.The same strategy should be followed for the other pharmacogenomics steps. References: 1. Siest G, Marteau J B, Maumus S, Berrahmoune H, Jeannesson E, Samara A, Batt A M and Visvikis-Siest S: Eur. J. Pharmacol. 2005; 527: 1–22. 2. Siest G, Jeannesson E, Marteau J B, Samara A, Marie B, Pfister M and Visvikis-Siest S.: Drug Metab. Dispos. 2008; 36: 182–189. 1E. Pharmacogenomics Abstracts FEBS Journal 275 (Suppl. 1) 4–64 (2008) ª 2008 The Authors Journal compilation ª 2008 FEBS 11 IL1E-5 Pharmacogenomics: from drug metabolism to disease susceptibility and beyond E. Sim Department of Pharmacology, University of Oxford, Oxford, UK The identification of polymorphisms in drug metabolising enzymes has been important in the development of pharmacogenetics. The geneti- cally determined ability to metabolise isoniazid was found to be due to the enzyme, arylamine N-acetyltransferase (NAT), which inactivates the drug through acetylation. This enzyme is now known to be one of two human NATs. NAT2 which is responsible for isoniazid acetylation is subject to genetic polymorphism where haplotypes of SNPs contrib- ute to the slow acetylator phenotype. Within different human popula- tions, the existence of different predominant NAT2 SNP profiles underlies the distinction in acetylator frequency in different ethnic groups. These studies have important implications for defining popula- tions in clinical trials in relation to treatment populations where acety- lation may be an important route of metabolism. SNPs have been identified associated with poor enzymic activity of the NAT2 isoen- zyme and also its sister enzyme NAT1, with each enzyme having a dis- tinct substrate specificity profile and tissue distribution. The effect of SNPs appears to be due to destabilisation of the protein such that the enzyme aggregates and is marked for degradation via ubiquitina- tion.The recently determined structures of the human NAT enzymes assist in understanding the basis of the substrate specificity of the highly homologous different human NAT isoenzymes and also allow rationalisation of the effects of the SNPs and amino acid substitutions on enzyme stability. The pharmacogenetic variation in enzymic activity may also relate to the level of expression in different individuals and the search for pharmacogenetic variation in the genomic regions associ- ated with control of expression are beginning to be explored, both in the human NAT genes and in animal models. Other studies relating polymorphisms in the human NAT genes to susceptibility to arylamine or environmentally induced disease has been the subject of much research. There is clear evidence that the slow acetylator type contrib- utes to the development of industrial and idiopathic bladder cancer but the contribution of this pharmacogenetic variation in other cancers is less clearly identified. Acknowledgement: The Wellcome Trust provides financial support for original work. IL1E-6 Variable response to opioid treatment: genetic predictors within sight? F. Skorpen Molecular Biology Section of Pain and Palliation Research Group, Department of Laboratory Medicine, Children’s and Women’s Health, Faculty of Medicine, Norwegian University of Science and Technology, Trondheim, NORWAY Control of pain and related symptoms is paramount to clinical success in caring for patients with advanced cancer and other terminal ill- nesses. Opioids are the mainstay of therapy for moderate to severe can- cer pain, with oral morphine being recommend ed as the conventional opioid of choice. In spite of expert recommendations, careful dose esca- lation and ‘optimi zation’ of the management regime, adequate pain relief is not attained in a substantial minority of patients, and unpleas- ant side effects are usually inevitable, and unpredictable. In recent years, research investigating the relationship between the genetic vari- ability among individuals and outcome of opioid therapy has grown considerably. Yet, the utility of genotyping as a valuable supplemental tool in pain management is so far limited to codeine, a weak l-opioid which should not be administered to individuals deficient in cyto- chrome P450 2D6 (CYP2D6) activity (poor metabolizers), and with care to individuals with CYP2D6 gene duplications (ultra-rapid metab- olizers). However, several interesting candidates have emerged as potentially relevant factors in l-opioid analgesia, including polymor- phisms affecting the l-opioid receptor (OPRM1), the efflux transporter P-glycoprotein encoded by the ABCB1 gene, uridine diphosphate-glu- curonosyltransferase 2B7 (UGT2B7), as well as less obvious candidates such as melanocortin-1 receptor (MC1R) and catechol-O-methyl trans- ferase (COMT). This lecture will give an overview of the available evi- dence for a relationship between common polymorphisms in human genes and variability in opioid analgesia and side effects among patients treated for cancer pain. Abstracts 1E. Pharmacogenomics 12 FEBS Journal 275 (Suppl. 1) 4–64 (2008) ª 2008 The Authors Journal compilation ª 2008 FEBS 2. Coupling and Regulation of Gene Expression Machines 2A.Transcription, RNA Processing and Export IL2A-1 RNA polymerase II transcriptional termination: a paradigm of the coupling of transcription and pre-mRNA processing M. Dye Oxford University, Oxford, UK One of the first demonstrations of the coupling of RNA Polymerase II (Pol II) transcription and pre-mRNA processing was the finding that a functional poly(A) signal is required for Pol II transcriptional termina- tion (1). Since then it has been shown that pre-mRNA splicing is also required for termination (2,3). This effect reflects the co-transcriptional nature of poly(A) site enhancement by splice signals. Analysis of the human b-globin gene terminator sequence, located 1 kb downstream of the poly(A) signal, revealed that transcripts of it are the target of a co- transcriptional RNA cleavage activity termed CoTC ( CoTranscription- al Cleavage (Dye and Proudfoot 2001; Cell 105: 669–681). Subsequent studies have shown that CoTC followed by transcript degradation is essential for Pol II transcriptional termination (4,5) . In further studies we have shown that induction of co-transcriptional cleavage in a nas- cent intron transcript has no effect on splicing or subsequent mRNA accumulation (6). These observations have developed into a model in which we envisage that CoTC and degradation of non-coding sections of pre-mRNAs might aid Pol II transcription in the complex and cramped nuclear environment (7). We are currently investigating how the human b-globin gene terminator sequence works in terms of its properties as a target or mediator of, co-transcriptional pre-mRNA cleavage, pre-mRNA degradation and release of Pol II from the DNA template. References: 1. Whitelaw and Proudfoot, 1986 EMBO J 5, 2915–2922 2. Baure ´ n, et al. Genes. Dev. 1998; 12: 2759–2769 3. Dye and Proudfoot. Mol. Cell. 1999; 3: 371–378 4. West et al. Nature 2004; 432: 522–525 5. West et al. Mol. Cell. 2008; 29 6. Dye et al. Mol. Cell 2006; 21: 849–859 7. Dye et al. Cold Spring Harbor Symp. Quant. Biol. 2006; 71: 275–284 IL2A-2 At the heart of the spliceosome R. Luehrmann Max Planck Institute for Biophysical Chemistry, Goettingen, GERMANY Formation of catalytically active RNA structures within the spliceo- some requires the assistance of proteins. However, little is known about the number and nature of proteins needed to establish and main- tain the spliceosome’s active site. Here we affinity-purified human spliceosomal C complexes and show that they catalyse exon ligation in the absence of added factors. Comparisons of the composition of the pre-catalytic versus the catalytic spliceosome revealed a marked exchange of proteins during the transition from B to the C complex, with apparent stabilization of Prp19-CDC5 complex proteins and destabilization of SF3a/b proteins. Disruption of purified C complexes led to the isolation of a salt-stable ribonucleoprotein (RNP) core that contained both splicing intermediates and U2, U5 and U6 small nuclear RNA plus predominantly U5 and human Prp19-CDC5 pro- teins and Prp19-related factors. Our data provide insights into the spliceosome’s catalytic RNP domain and indicate a central role for the aforementioned proteins in sustaining its catalytically active structure. We have also carried out similar studies with purified yeast spliceo- somes and have analysed the 3D structure of selected human and yeast spliceosomal complexes using electroncryomicroscopy. IL2A-3 The SMN complex: a molecular assembly machine for RNPs G. Dreyfuss Howard Hughes Medical Institute, University of Pennsylvania School of Medicine, Philadelphia, PA, USA Cells contain several different classes of RNAs (snRNAs, mRNAs, etc.) and an enormous assortment of RNA-binding proteins. RNAs exist in cells as RNPs (RNA-protein complexes) and each class is associated with a specific subset of RNA-binding proteins. Based on select in vitro experiments it was widely believed that RNPs in cells form spontaneously, by self-assembly. Our studies on the survival of motor neurons (SMN) complex, however, indicated that this is not the case. We found that the biogenesis of snRNPs, the essential components of the pre-mRNA splicing machinery, and likely other RNPs in eukaryotic cells, is carried out by an assemblyosome, the SMN complex. The SMN complex binds to specific RNA-binding proteins (the Sm snRNP proteins) and to snRNAs which it recog- nizes by specific sequence features (the snRNP code) that distinguish them from other classes of RNAs. The purpose of the SMN com- plex is to confer stringent specificity to otherwise potentially illicit RNA-protein interactions. We have further identified the protein of the SMN complex, Gemin 5, which binds directly the snRNP code, and is critical for snRNP biogenesis. Gemin 5, a novel type of RNA-binding protein, is the snRNA-determining factor in cells. SMN deficiency is the cause of the most common motor neuron degenerative diseased (SMA). The consequences of SMN deficiency, potential therapeutic approaches to SMA and small molecule screens for modulators of SMN complex expression and activity will be described. IL2A-4 Identification of in vivo targets of human splicing factor SF1 A. Kraemer Department of Cell Biology, Faculty of Sciences, University of Geneva, Geneva, SWITZERLAND Splicing factor SF1 cooperates with U2AF in the recognition of the 3’ splice site at the onset of spliceosome assembly. Although SF1 is essential for viability in S. cerevisae, C. elegans and human cells, genetic and biochemical depletion from yeast and mammalian extracts only modestly affected splicing of reporter pre-mRNAs. In line with this observation, RNA interference-mediated knock-down of SF1 from human cells did not compromise the splicing of several pre-mRNAs, suggesting that it plays a kinetic role or is only required for the splicing of a subset of pre-mRNAs. Furthermore, SF1 shuttles between the nucleus and the cytoplasm, a property that is not understood. To confirm a function of SF1 in at least some splicing events or detect a role in other steps of mRNA biogenesis, we have isolated in vivo RNA targets of SF1 from HeLa cells by Cross-Linking and ImmunoPrecipitation (1). A large portion of SF1 binding sites was found in introns and at 3’ splice sites, as expected. Validation experiments, including electrophoretic mob ility shift assays and analysis of splicing of endogenous pre-mRNAs after SF1 deple- tion from HeLa cells, indicated that SF1 is indeed a splicing factor, but plays a role in alternative rather than constitutive splicing. We also found many SF1 target sites in 3’ terminal exons/3’ UTRs, highlighting a possible function of SF1 downstream of splicing, which may be related to its shuttling properties. Reference: 1. Ule et al. Science 2003; 302: 1212–1215 2A.Transcription, RNA Processing and Export Abstracts FEBS Journal 275 (Suppl. 1) 4–64 (2008) ª 2008 The Authors Journal compilation ª 2008 FEBS 13 . INVITED LECTURES 1. Chromosome Architecture and Nuclear Dynamics 1A. Chromosome