FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2078693172> ?p ?o ?g. }

• W2078693172 endingPage "8" @default.
• W2078693172 startingPage "1" @default.
• W2078693172 abstract "EMBO Member's Review4 January 1999free access Dividing the empire: boundary chromatin elements delimit the territory of enhancers Andor Udvardy Corresponding Author Andor Udvardy Institute of Biochemistry, Biological Research Center of the Hungarian Academy of Sciences, H-6701 Szeged, PO Box 521, Hungary Search for more papers by this author Andor Udvardy Corresponding Author Andor Udvardy Institute of Biochemistry, Biological Research Center of the Hungarian Academy of Sciences, H-6701 Szeged, PO Box 521, Hungary Search for more papers by this author Author Information Andor Udvardy 1 1Institute of Biochemistry, Biological Research Center of the Hungarian Academy of Sciences, H-6701 Szeged, PO Box 521, Hungary *Corresponding author. E-mail: [email protected] The EMBO Journal (1999)18:1-8https://doi.org/10.1093/emboj/18.1.1 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Introduction Chromatin is the most complex supramolecular organization of the cell. It has a dual role: to compact the chromosomal DNA and to ensure a highly efficient regulation of gene expression. In consequence of the very compact packaging of the DNA, chromatin is highly repressive for transcription. Enhancers are key regulatory elements which can relieve the chromatin-induced repression. The interplay between the chromatin-mediated repression and the activating function of enhancers ensures that the difference between the induced or repressed level of gene expression in eukaryotic cells is several orders of magnitude higher than that in prokaryotic cells. The ability of enhancers to activate a gene independently of their distance from a promoter poses a topological problem: how is the territory of enhancer action delimited? It is supposed that the higher-order chromatin structure divides the genome into topologically constrained loops which are attached to the nuclear matrix. This model implies that the loops are both structural and functional units of the chromatin in the sense that genes under common transcriptional regulation are located within the same loop and regulated by the same set of enhancers (Forrester et al., 1986; Orkin, 1990; Bonifer et al., 1991). In this attractive model, the monogamy of enhancers is ensured by the attachment of the bases of the chromatin loops to the nuclear matrix, which insulates the promoters located within the loop from the regulatory influence of enhancers present in the neighboring loops. Two distinct approaches have been applied to prove the validity of this model. In the structural approach, the main goals were to demonstrate that chromatin is indeed anchored at defined sites to a nuclear matrix, and to identify and characterize DNA sequences at these attachment sites. Structural studies have led to the discovery of scaffold attachment region (SAR) sequences (Mirkovitch et al., 1984). The role of SAR sequences as domain boundary elements, however, is still controversial (Stief et al., 1989; Phi-Van et al., 1990; McKnight et al., 1992; Poljak et al., 1994; Kalos and Fournier, 1995). The functional approach was based on the observation that induction of gene expression is always accompanied by a structural reorganization of the underlying chromatin. A chromatin domain can therefore be defined as a segment of the genome which exhibits chromatin structural changes upon induction of the gene present in that domain. As the expression of heat shock genes can be induced easily, this approach was first applied to the Drosophila hsp70 heat shock domain (Udvardy et al., 1985). It was shown that both the proximal and distal sides of the hsp70 heat shock locus are flanked by novel chromatin structures called ‘specialized chromatin structures’ (scs and scs′), which are the most remote sequences relative to the hsp70-coding sequences that respond with a gross chromatin structural rearrangement after heat induction. It was presumed that scs and scs′ structures define the proximal and distal boundaries of the heat shock domain and represent boundary elements with insulating properties. The validity of this approach was justified later by a direct in vivo enhancer-blocking assay (Kellum and Schedl, 1992). Since then, a large number of boundary elements have been defined and characterized (Grosveld et al., 1987; Greaves et al., 1989; Bonifer et al., 1990; Holdridge and Dorsett, 1991; Carson and Wiles, 1993; Palmiter et al., 1993; Diaz et al., 1994; Talbot et al., 1994; Jones et al., 1995; May and Enver, 1995; Montoliu et al., 1995; Gdula et al., 1996). On the basis of functional properties, these elements can be divided into two categories: insulators and locus control regions (LCRs). Insulators The scs elements and the suppressor of the Hairy-wing [su(Hw)] protein-binding region of the gypsy retrotransposon are the best-characterized insulators. The scs and scs′ elements are defined by two closely spaced nuclease-hypersensitive (HS) sites arranged around a central nuclease-resistant segment. Characteristic changes occur in the fine structure of this complex HS region in response to heat induction (Udvardy et al., 1985). Additionally, there are localized topoisomerase II sites in the HS complex, and a redistribution of topoisomerase II takes place on heat induction (Udvardy and Schedl, 1993). When DNA fragments carrying either the scs or the scs′ region were placed between the yolk protein-1 (yp-1) enhancer and the hsp70 promoter–LacZ fusion gene, the stage-, sex- and tissue-specific activation of the reporter gene by the yp-1 enhancer was completely blocked. The blocking activity was independent of the orientation of the scs elements. No insulation occurred, however, when the scs elements were inserted upstream of the enhancer (Kellum and Schedl, 1992). Furthermore, the expression of a transgene became copy number-dependent and independent of the site of integration when it was flanked at the 5′ end by an scs element and at the 3′ end by an scs′ element (Kellum and Schedl, 1991). On these criteria, the scs elements are bona fide insulators, because they can fulfill the two most important functions required for a boundary element: insulation of a gene from the influence of an enhancer and protection of a transgene from chromosomal position effect. Insulators, however, do not have enhancer activity. Detailed analysis has revealed that DNA fragments associated with the two HS regions of scs are essential for full enhancer-blocking activity, while the central AT-rich nuclease-resistant region is dispensable (Vasquez and Schedl, 1994). Deletions which shorten the HS region impaired the insulator activity, but this activity could be fully restored by multimerization of DNA fragments from either HS region. The scs insulator is therefore assembled from several independent, functionally redundant DNA sequences, which may be binding sites for sequence-specific DNA-binding proteins. One of these proteins has already been purified, characterized and cloned (Zhao et al., 1995). BEAF-32A recognizes a palindromic sequence which abuts the HS site of the scs′ boundary element. In an in vivo enhancer blocking assay, a 515 bp scs′ fragment carrying one of these palindromic sequences exhibited strong blocking activity, and even a 7mer repeat of a 48 bp oligonucleotide with a BEAF-binding site produced inhibition. Immunoprecipitation of formaldehyde cross-linked chromatin with a specific antibody developed against the recombinant BEAF-32A protein demonstrated that, in vivo also, BEAF-32A binds to scs′ sequences. Significantly, immunolocalization studies revealed that BEAF-32A is present in several hundred interband regions and at many puff borders on the Drosophila polytene chromosome. This indicates that scs′ is not specific to the hsp70 heat shock domain, but is probably a multipurpose boundary element. Recently, a second scs′-binding protein (BEAF-32B) has been identified and cloned (Hart et al., 1997). It differs from BEAF-32A only in its N-terminus. The unique 5′ sequences of these proteins are probably encoded by two independent exons of the same single-copy gene. Three domains have been identified in these proteins. The C-terminal domain is responsible for the observed heterocomplex formation of BEAF-32A and -32B. DNA binding is mediated by the unique N-terminal domains, and this explains the differences observed in DNA-binding specificity detected by DNase I footprinting. The function of the identical middle domain of BEAF-32A and -32B is still not clear. Another sequence-specific DNA-binding protein (scs-binding protein, SBP) which recognizes a 27 bp sequence of scs recently has been identified and cloned (M.Gaszner, J.Vasquez and P.Schedl, personal communication). Multimers of the 27 bp recognition sequence exhibited in vivo enhancer-blocking activity. Mutations in this oligonucleotide disrupted the in vitro binding of the recombinant SBP and eliminated the enhancer-blocking activity in vivo. A chromatin immunoprecipitation assay revealed that SBP binds in vivo to a region of scs which corresponds to the in vitro binding site. The vital role of the boundary function is supported by the observation that the strong loss-of-function mutations of the SBP gene are recessive cell autonomous lethal, while weaker alleles have pleiotropic effects on development. This is consistent with the assumption that the scs, like scs′, is a multipurpose insulator involved in the regulation of several different genes, and SBP has an important role in the maintenance of chromatin architecture. Insertion of the gypsy retrotransposon causes a variety of mutations in Drosophila. Some of these mutations exhibit typical tissue or developmental specificity. Analysis of mutations induced by insertion of the gypsy into different regions of the yellow gene revealed that the phenotype of the mutation depended on the location of the integration site relative to the position of tissue-specific yellow enhancers. Enhancers located upstream of the integration site relative to the position of the promoter were inactivated. A short region of the gypsy transposon carrying a repeated octamer sequence motif flanked by AT-rich sequences is responsible for the mutations (Geyer et al., 1988; Smith and Corces, 1995). Insertion of this short sequence between an enhancer and the promoter blocks the activity of the enhancer (Holdridge and Dorsett, 1991; Jack et al., 1991; Geyer and Corces, 1992). A transgene flanked on both sides by this short sequence is protected from the chromosomal position effect (Roseman et al., 1993). On these criteria, the gypsy is a true insulator. The repeated octamer sequence motif is the binding site of Su(Hw), a Drosophila nuclear protein with 12 zinc fingers in the central domain, with a leucine zipper domain and with two highly acidic domains at the N- and C-terminal segments of the protein (Spana et al., 1988; Mazo et al., 1989; Dorsett, 1990). Mutation analysis revealed that the zinc finger and the leucine zipper motifs are critical for enhancer blocking (Harrison et al., 1993). The modifier of mdg4 [mod(mdg4)] protein interacts with Su(Hw), this interaction being essential for the insulator activity (Gerasimova et al., 1995). The phenotype of gypsy-induced mutations at a variety of loci is modified in a mod(mdg4) background (Georgiev and Gerasimova, 1989). Analysis of the effect of the mod(mdg4) mutation on the gypsy-induced y2 phenotype of the yellow gene led to the postulation that the binding of the Su(Hw)–mod(mdg4) protein complex to the gypsy insulator ensures a unidirectional enhancer-blocking activity, whereas Su(Hw) alone converts the unidirectional insulator to a bi-directional silencer which inactivates all the tissue-specific enhancers of the yellow gene, irrespective of the site of insertion of the gypsy retrotransposon (Gerasimova et al., 1995). This change, however, is not universal; there are genes where the mod(mdg4) mutation increases the gypsy-induced phenotype (Georgiev and Kozycina, 1996). This suggests that mod(mdg4) interacts both with Su(Hw) and with other chromosomal proteins, and these latter interactions are gene specific. Several different models have been proposed with regard to the mode of action of insulators (Figure 1). The domain boundary model attempted to link the physical organization of chromosomes with the most important functional properties of insulators. In this model (Figure 1A), the three-dimensional folding of a chromatin domain, i.e. a DNA segment delimited by insulator elements, would facilitate productive interactions between regulatory elements located within the domain, while it would preclude equivalent interactions between these elements and regulatory elements outside the domain (Udvardy et al., 1985). In the ‘tracking’ model (Figure 1B), transcription factors bound to enhancer elements track along the DNA in search of target promoters. Insulator-bound sequence-specific DNA-binding proteins block transcription factors from reaching the target promoter (Chung et al., 1993). Insulators may also reduce the accessibility of promoters to complexes formed on enhancer elements (Figure 1C; Chung et al., 1993). In the transcriptional decoy model (Figure 1D), the complex formed on the insulator may function as an enhancer sink by attracting the complex assembled on the enhancer into a non-productive interaction (Geyer, 1997). The data available today are insufficient to allow an unambiguous distinction between these models. One important new observation should be borne in mind in connection with all the models proposed for the mode of insulator action: an enhancer prevented from acting on one promoter by the insertion of an insulator remains fully active for another uninsulated promoter (Figure 1E; Cai and Levine, 1995; Scott and Geyer, 1995). This indicates that insulators do not propagate the formation of a repressive chromatin structure. Figure 1.Models of insulation. Different models of insulation, as described in detail in the text, are shown. E, enhancer; Pr, promoter; Ins, insulator. Download figure Download PowerPoint Recent studies suggest that proteins involved in chromatin domain border formation may also be involved in the overall organization of the nucleus. The punctated immunostaining pattern of mod(mdg4) and su(Hw) proteins in the nucleus of Drosophila imaginal disc cells and their co-localization with lamin suggest that many individual binding sites for these proteins come together in the same subnuclear region, and protein components of the insulators might establish this organization by attaching the chromatin fiber to the nuclear matrix (Gerasimova and Corces, 1998). There is some debate as regards the involvement of chromatin structure in the process of insulation. The complex HS structure observed on scs and scs′ sequences suggests that the blocking activity of insulators requires a supporting chromatin architecture. It has been shown recently, however, that scs or scs′ elements can block enhancer-activated transcription on a plasmid DNA microinjected into Xenopus laevis oocytes (Dunaway et al., 1997; Krebs and Dunaway, 1998). Although the enhancer-blocking activity appeared earlier during the microinjection experiment than the completion of the nucleosome assembly, this is not a direct indication that enhancer blocking does not require a chromosomal context. The microinjection-based assay cannot measure separately the transcriptional activity of a fully assembled, partially assembled or unassembled plasmid DNA. Enhancer activity is known to be almost 50-fold higher on chromatin-assembled plasmid DNA as compared with naked DNA (Workman et al., 1991). If a similar dependence of the insulator activity on the structure of template DNA is assumed, a small fraction of nucleosome-assembled plasmid may confuse the results obtained in the microinjection-based assay. Locus control regions The discovery of LCRs was a joint success of researches in the fields of molecular biology and molecular medicine. The initial failure of transgenic experiments to achieve high-level, copy number-dependent transgene expression suggested that, besides the enhancers identified in transient expression experiments as being essential for the appropriate regulation of the transgene, additional, hitherto unidentified regulatory sequences are also required. The identification of these missing sequences was based on the observation that the deletion of DNA sequences far upstream from the β-globin gene in Hispanic (γδβ) thalassemia resulted in a severely reduced β-globin expression. Analysis of the chromatin architecture of the DNA segment deleted in (γδβ) thalassemia revealed the presence of five tissue-specific DNase HS structures (Groudine et al., 1983; Tuan et al., 1985). Transgenes of the β-globin gene flanked by the DNA segment carrying these HS sites resulted in the expected high-level, copy number-dependent expression, which was independent of the site of integration (Grosveld et al., 1987). The 15 kb DNA segment with these HS sites was called the LCR. Several different LCRs subsequently have been identified and characterized. The structures and functions of the individual components of the LCRs have been studied by different techniques. Sequence analysis demonstrated that there are multiple erythroid-specific and ubiquitous transcription factor-binding motifs in the HS sites (Hardison et al., 1997). It is important to note that, although in different numbers and spatial arrangements, each of the HS sites contains the same set of transcription factor-binding sites. Some of these binding sites are also present in the enhancers responsible for the tissue-specific and developmental stage-specific expression of the different globin genes. Transient or stable expression studies and transgenic experiments revealed that HS site 2 of the human globin LCR exerts enhancer activity (Tuan et al., 1987; Collis et al., 1990). HS site 3 will cause high-level expression of the β-globin gene after stable integration, but will not enhance its transient expression (Collis et al., 1990), while HS sites 1 and 4 have no enhancer activity (Pruzina et al., 1991). Deletion of the individual HS sites only mildly reduces the expression of the globin genes, indicating that the HS sites cooperatively regulate the expression of all the genes of the locus. The coordinated enhancer activity of these HS sites is believed to be responsible for the chromatin domain opening activity, which is the most characteristic property of the LCRs. HS site 5 of the human globin LCR (Li and Stramatoyannopoulos, 1994), HS site 4 of the chicken globin LCR (Chung et al., 1993; Pikaart et al., 1998) and element A of the chicken lysozyme LCR (Bonifer et al., 1994) exhibit in vivo enhancer-blocking activity characteristic of insulators. The enhancer-blocking activity of the chicken HS site 4 is retained in Drosophila. These HS sites with insulator activity may have a pivotal role in the position-independent expression of the transgenes, while the domain opening activity may ensure the high-level and copy number-dependent expression. Despite the accumulation of a large body of information on the structures of LCRs, and on their interactions with sequence-specific DNA-binding proteins, the detailed molecular mechanism(s) of the mode of their protective action is far from being resolved. Two different models have been proposed (Martin et al., 1996; Dillon et al., 1997). Although there is a sharp difference in the interpretation of the data which form the basis of these models, the contradictions, as shown below, are not irreconcilable. The competition model is based on a detailed analysis of the in vivo dynamics of transcription complex interactions which occur in the developmentally regulated chromatin architecture of the globin locus. All five human β-globin genes are under the regulatory control of the globin LCR. These genes are arranged in the 70 kb globin locus in the sequence of their developmental expression. It was presumed earlier that the different HS elements of the LCR interact with different genes of the locus, and each HS region is responsible separately for the developmental regulation of a given globin gene (Engel, 1993). However, it is now generally accepted that all the HS elements of the LCR act together, probably by forming a holocomplex. This followed from the observation that each of the genes required all of the HS elements for full activity (Bungert et al., 1995; Fiering et al., 1995; Hug et al., 1996). In situ hybridization with gene-specific hybridization probes on individual embryonic and fetal transgenic erythroid cells can identify those globin genes which are expressed at a given time, at individual alleles, in individual cells. The in situ hybridization experiments detected single gene signals in the majority of the cells. Kinetic analysis based on the rate of primary transcript processing revealed that multigene signals present in a small fraction of cells represented a recent switch of transcription of different globin genes and not concurrent transcription from multiple genes. From the aspect of the chromatin interaction dynamics these data suggested that although the LCR holocomplex interacts with only a single promoter at a time, it can switch back and forth between the different promoters of the globin locus by a flip–flop mechanism, and this switching occurs continuously throughout all the developmental stages of globin gene expression (Wijgerde et al., 1995). The prevailing LCR–promoter interaction in a given period of development is determined by the abundance of gene-specific transcription factor(s) and the distance of the promoter from the LCR (Dillon et al., 1997). The transition from a given developmental phase to the subsequent one is governed by the change of transcription factor abundance, which is accompanied by a gradual transition in the prevailing LCR–promoter interaction. However, this transition does not abolish the potential for occurrence of the LCR–gene interaction which was predominant in the previous developmental phase. It has been calculated that in vivo the LCR–gene interaction is switched every 10–15 min, and this flip–flop does not require the replication of DNA. The continuous availability of all the genes of the globin locus for a single, coordinately acting LCR requires continuous access for the governing LCR toward all the regulatory entities (promoters and enhancers) of all the genes of the globin locus. This means that the globin chromatin domain in erythroid cells should be multihooked, i.e. all the regulatory entities of the 70 kb globin locus should maintain dynamic interactions with the LCR (Gribnau et al., 1998), and the intervening DNA sequences should be looped out. How is this complex chromatin architecture formed? The main problem in the looping model of enhancer action is that it does not take into account the energetic burden of DNA looping. The concentration of DNA in the nucleus of a human cell is ∼7–8 mg/ml. The viscosity of a megabase pair long chromatin fiber in this concentration range is very high. What will provide the energy required to move a segment of the chromatin several thousands of base pairs away? Transcription factors, having neither ATPase nor GTPase domains, cannot provide the energy. It is much more reasonable to suppose that the complex chromatin architecture of the whole globin locus is formed coordinately at the time when the commitment for erythroid-specific differentiation occurs. This coordinated formation of the chromatin architecture is probably initiated by the appearance of high-affinity, sequence-specific DNA-binding protein(s) which specifically recognize the globin LCR and tether the DNA at the LCR to a protein frame (we may call this the ‘nucleoskeleton’ to distinguish it from the nuclear matrix). This process probably requires the disruption of the nucleosomal organization of the chromatin at the site of tethering, and it may therefore occur during the replication of DNA. Alternatively, the appearance of special chromatin remodeling factor(s) may catalyze this rearrangement. Tethering the DNA at the LCR by this high-affinity DNA-binding protein(s) may initiate the looping of the globin locus (Figure 2). If the same or different sequence-specific DNA-binding protein(s) recognize and tether the DNA to the nucleoskeleton at the different intradomain enhancers of the globin locus, the multihooked chromatin architecture can be formed coordinately. GATA- and CACC-binding motifs are present both in the enhancers flanking the globin genes and also in the HS sites of the LCR. The higher concentration of these factor-binding sites in the LCR may be responsible for their predominant role in the formation of the primary contacts at the LCR, while the presence of these sites in the enhancers present in the vicinity of the globin genes may ensure the coordinated formation of the whole-domain, multihooked chromatin architecture which establishes the structural basis of all the dynamic contacts between the LCR and the different globin genes. This newly formed multihooked chromatin architecture should be stabilized subsequently by non-histone chromosomal proteins and heritably maintained until the high-affinity binding protein(s) is present in the cell. This multihooked chromatin organization is a dynamic structure, its formation is dependent on the transcriptional activation of the domain, and it should be clearly distinguished from the chromatin loops organized by the SAR sequences. Figure 2.Model for the establishment of a whole-domain, multihooked chromatin architecture. During the replication of DNA, occurring as DNA is reeled through a nucleoskeleton-bound ‘replication factory’ (Cook, 1991), DNA is tethered to the nucleoskeleton at the LCR and at internal enhancer sequences due to the binding of sequence-specific DNA-binding proteins. The dominant role of LCR in the replication of the globin locus is well documented (Forrester et al., 1990; Aladjem et al., 1995). R.F., replication factory; E1 and E2, internal enhancers of the locus. Download figure Download PowerPoint The role of cellular differentiation in the reorganization of the chromatin architecture and in the transcriptional activation of a domain is well documented in the case of the chicken lysozyme locus (Bonifer et al., 1997). When a transgene carrying an intact LCR integrates close to the centromeric heterochromatin, the LCR completely protects the locus from the repressive effect of heterochromatinization. Pericentromeric integration of a transgene carrying the globin locus with an incomplete LCR results in a severely reduced globin gene expression, this frequently being due to clonally inherited position effect variegation (Milot et al., 1996). This suggests that there is strong competition for the establishment of heritably stabilized higher-order chromatin structures. Heterochromatinization, which is a highly penetrative process, is totally blocked in the presence of a complete LCR, probably because of the very high stability of the presumed complex formed on the LCR. With an incomplete LCR, the formation of this primary complex is slower, and/or the complex is less stable. As it is a stochastic process, the timing of the formation of this primary complex will vary from cell to cell. In those cells where the binding of the presumed LCR-binding protein(s) to its cognate recognition sequence precedes the spread of heterochromatinization, epigenetic mechanisms will stabilize this LCR-specified primary interaction, resulting in the formation of a clonally inherited stable higher order chromatin structure which is permissive for globin expression. In those cells where the spread of heterochromatinization precedes the primary LCR–protein interaction, however, heterochromatin will be stabilized and heritably transmitted, resulting in the appearance of position effect variegation. Similar results were obtained after the deletion of HS site 3 of the human CD2 gene LCR (Festenstein et al., 1996). Transgenes with a deleted HS site 3 exhibited variegated expression when integrated in the centromere. DNase I HS site analysis of the nuclei of fluorescence-activated cell sorting (FACS)-separated CD2+ and CD2− cells of the variegating cell population revealed that HS site 1 was completely missing in CD2− cells, whereas it was fully retained in CD2+ cells. This is a direct indication of the epigenetic stabilization of a clonally inherited stable higher order chromatin structure. Deletion of essential DNA sequences from the LCR is not the only way to influence the competition between the formation of different higher-order chromatin structures. Similar results can be achieved by eliminating a protein component essential for the function of a regulatory chromatin element. Thus, deletion of mod(mdg4) results in a variegated expression pattern of gypsy-induced mutations similar to those chromosomal rearrangements that bring the gene juxtaposed to heterochromatic sequences (Gerasimova et al., 1995). The second model proposed for the mechanism of LCR action was based on the ‘probability’ theory of enhancer activation. When a γ-globin promoter–LacZ-neoR fusion gene with or without a flanking enhancer sequence was stably integrated into K562 erythroleukemia cells, a flanking 5′-HS site 2 from the β-globin LCR, which is a strong erythroid-specific enhancer, produced a 23-fold increase of G418-resistant colonies (Walters et al., 1995). When FACS-separated β-gal-positive cells were assayed by a quantitative β-gal activity test, no differences were observed between the enhancer-containing and en" @default.
• W2078693172 created "2016-06-24" @default.
• W2078693172 creator A5060187574 @default.
• W2078693172 date "1999-01-04" @default.
• W2078693172 modified "2023-10-18" @default.
• W2078693172 title "Dividing the empire: boundary chromatin elements delimit the territory of enhancers" @default.
• W2078693172 cites W139144773 @default.
• W2078693172 cites W1431639342 @default.
• W2078693172 cites W1556415839 @default.
• W2078693172 cites W1594396900 @default.
• W2078693172 cites W1595583901 @default.
• W2078693172 cites W1600994582 @default.
• W2078693172 cites W1681326588 @default.
• W2078693172 cites W1778216591 @default.
• W2078693172 cites W1822511846 @default.
• W2078693172 cites W1822687156 @default.
• W2078693172 cites W1832452463 @default.
• W2078693172 cites W1833324305 @default.
• W2078693172 cites W1850964221 @default.
• W2078693172 cites W1856683646 @default.
• W2078693172 cites W1888261674 @default.
• W2078693172 cites W1893658886 @default.
• W2078693172 cites W1963900436 @default.
• W2078693172 cites W1968473524 @default.
• W2078693172 cites W1978713358 @default.
• W2078693172 cites W1990006714 @default.
• W2078693172 cites W1990967581 @default.
• W2078693172 cites W1995985230 @default.
• W2078693172 cites W1997849302 @default.
• W2078693172 cites W1998729431 @default.
• W2078693172 cites W2001668645 @default.
• W2078693172 cites W2004060771 @default.
• W2078693172 cites W2006247010 @default.
• W2078693172 cites W2013919517 @default.
• W2078693172 cites W2020246296 @default.
• W2078693172 cites W2021598200 @default.
• W2078693172 cites W2027878791 @default.
• W2078693172 cites W2030629446 @default.
• W2078693172 cites W2040029402 @default.
• W2078693172 cites W2042067464 @default.
• W2078693172 cites W2043734426 @default.
• W2078693172 cites W2049226568 @default.
• W2078693172 cites W2050579225 @default.
• W2078693172 cites W2050987515 @default.
• W2078693172 cites W2051403136 @default.
• W2078693172 cites W2055493522 @default.
• W2078693172 cites W2055643269 @default.
• W2078693172 cites W2057321580 @default.
• W2078693172 cites W2057440187 @default.
• W2078693172 cites W2062438194 @default.
• W2078693172 cites W2065867057 @default.
• W2078693172 cites W2070126970 @default.
• W2078693172 cites W2072606138 @default.
• W2078693172 cites W2076395918 @default.
• W2078693172 cites W2077479386 @default.
• W2078693172 cites W2081875871 @default.
• W2078693172 cites W2089437863 @default.
• W2078693172 cites W2090610011 @default.
• W2078693172 cites W2092941260 @default.
• W2078693172 cites W2094596493 @default.
• W2078693172 cites W2094601780 @default.
• W2078693172 cites W2095400582 @default.
• W2078693172 cites W2100132714 @default.
• W2078693172 cites W2106895159 @default.
• W2078693172 cites W2118167494 @default.
• W2078693172 cites W2118299921 @default.
• W2078693172 cites W2120448616 @default.
• W2078693172 cites W2123450144 @default.
• W2078693172 cites W2125064460 @default.
• W2078693172 cites W2133279102 @default.
• W2078693172 cites W2138096293 @default.
• W2078693172 cites W2138315073 @default.
• W2078693172 cites W2139412202 @default.
• W2078693172 cites W2145899011 @default.
• W2078693172 cites W2150630730 @default.
• W2078693172 cites W2151390187 @default.
• W2078693172 cites W2153477671 @default.
• W2078693172 cites W2158053354 @default.
• W2078693172 cites W2158749272 @default.
• W2078693172 cites W2159192458 @default.
• W2078693172 cites W2168639912 @default.
• W2078693172 cites W2178752953 @default.
• W2078693172 cites W2395174933 @default.
• W2078693172 cites W49806502 @default.
• W2078693172 cites W707389 @default.
• W2078693172 doi "https://doi.org/10.1093/emboj/18.1.1" @default.
• W2078693172 hasPubMedCentralId "https://www.ncbi.nlm.nih.gov/pmc/articles/1171096" @default.
• W2078693172 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/9878044" @default.
• W2078693172 hasPublicationYear "1999" @default.
• W2078693172 type Work @default.
• W2078693172 sameAs 2078693172 @default.
• W2078693172 citedByCount "95" @default.
• W2078693172 countsByYear W20786931722013 @default.
• W2078693172 countsByYear W20786931722014 @default.
• W2078693172 countsByYear W20786931722021 @default.
• W2078693172 crossrefType "journal-article" @default.
• W2078693172 hasAuthorship W2078693172A5060187574 @default.
• W2078693172 hasBestOaLocation W20786931722 @default.

• next