Recruitment and loading of the E1 initiator protein: an ATP-dependent process catalysed by a transcription factor
1998; Springer Nature; Volume: 17; Issue: 23 Linguagem: Inglês
10.1093/emboj/17.23.7044
ISSN1460-2075
AutoresCyril M. Sanders, Arne Stenlund,
Tópico(s)RNA Interference and Gene Delivery
ResumoArticle1 December 1998free access Recruitment and loading of the E1 initiator protein: an ATP-dependent process catalysed by a transcription factor Cyril M. Sanders Cyril M. Sanders Cold Spring Harbor Laboratory, P.O. Box 100, Cold Spring Harbor, New York, NY, 11724 USA Search for more papers by this author Arne Stenlund Corresponding Author Arne Stenlund Cold Spring Harbor Laboratory, P.O. Box 100, Cold Spring Harbor, New York, NY, 11724 USA Search for more papers by this author Cyril M. Sanders Cyril M. Sanders Cold Spring Harbor Laboratory, P.O. Box 100, Cold Spring Harbor, New York, NY, 11724 USA Search for more papers by this author Arne Stenlund Corresponding Author Arne Stenlund Cold Spring Harbor Laboratory, P.O. Box 100, Cold Spring Harbor, New York, NY, 11724 USA Search for more papers by this author Author Information Cyril M. Sanders1 and Arne Stenlund 1 1Cold Spring Harbor Laboratory, P.O. Box 100, Cold Spring Harbor, New York, NY, 11724 USA *Corresponding author. E-mail: [email protected] The EMBO Journal (1998)17:7044-7055https://doi.org/10.1093/emboj/17.23.7044 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Initiation of DNA replication critically depends on ori recognition as well as on catalytic activities of the initiator complex. For replication of papillomaviruses the catalytic activities for initiation are provided by the E1 protein. Here, we show that the transcription factor E2 acts to assemble E1 into a complex active for ori distortion in two steps. First, cooperative DNA binding of E1 and E2 generates a sequence-specific ori recognition complex. In the second ATP-dependent step, E2 is displaced and additional E1 molecules are incorporated. The net result is a final complex with low sequence specificity deposited onto a specific sequence in the DNA. This may be a general strategy to accomplish specific positioning of protein complexes with low sequence specificity. Introduction The original replicon hypothesis (Jacob et al., 1963) postulated the existence of defined genetic elements (replicators) that engage protein or protein complexes (initiators) in order to govern the specific and timely onset of DNA replication. Studies in recent years have largely upheld the central tenets of this hypothesis. The initiation of replication is believed to occur in several steps: specific recognition of the origin precedes origin unwinding, helicase delivery and processive DNA synthesis. In some cases, such as initiation of replication at Escherichia coli oriC, these functions are performed by several different proteins. In others, exemplified by the DNA tumour viruses, a single protein is responsible for all these functions. The replicators of simple eukaryotic genomes, such as viral chromosomes, are modular in nature, being comprised of core and auxiliary components (DePamphilis, 1993, 1996). The core components include a binding site for the initiator protein, an A/T-rich region and a DNA-unwinding element. The latter two are regions of low thermodynamic stability, prone to distortion or strand separation. The auxiliary components are generally binding sites for sequence-specific transcription factors. These were first recognized and subsequently shown to act as potentiators of core activity in viral systems (Bergsma et al., 1982; de Villiers et al., 1984; Guo et al., 1989). In addition to DNA replication, multiprotein complexes assembled on DNA are required for other specialized processes such as transcription and recombination. One of the requirements for understanding how these complexes function is to understand their assembly, including the basis for their recognition of specific sequences in DNA. These complexes may either exist as multiprotein complexes in solution or assemble on the DNA from individual components in an ordered fashion. In either case, the sequence complexity of mammalian organisms presents a great challenge to recognition of specific sites in DNA. This consideration also affects the parasites that infect mammalian cells. Papillomaviruses represent an interesting group of viruses in this regard since in their latent state they behave similarly to cellular genes, directing only very low levels of gene expression from a small number of genomes. At early stages after infection, the viral sequences are present at single-copy levels in a vast excess of cellular DNA. Two obvious alternative strategies exist to achieve occupancy of the appropriate sequences in the viral genome by viral proteins. One alternative is the expression of high concentrations of a factor with low sequence specificity. The other possibility is to generate binding with a high degree of sequence specificity. Papillomaviruses provide an interesting example of how sequence specificity can be achieved through cooperative DNA binding of two different viral proteins. The bovine papillomavirus (BPV) origin of DNA replication contains adjacent binding sites for the two viral proteins E1 and E2, both of which are absolutely required for replication in vivo (Ustav and Stenlund, 1991; Ustav et al., 1991, 1993; Yang et al., 1991). The viral initiator E1 binds DNA but with low sequence specificity. In the presence of the viral transcription factor E2, E1 and E2 bind cooperatively to the ori, resulting in a complex with greatly increased sequence specificity and affinity (Mohr et al., 1990; Lusky et al., 1993; Seo et al., 1993a; Sedman and Stenlund, 1995; Sedman et al., 1997). Although many other E2-binding sites are present in the viral genome, replication initiates at this particular site, demonstrating that E1, in spite of its low specificity, is the ultimate determinant for where this complex forms. Here we have asked how the E2 transcription factor functions in the assembly of a DNA replication pre-initiation complex. The surprising result is that E2 functions transiently and 'catalytically', providing sequence specificity for the formation of an E1–ori complex that can distort the ori DNA and therefore is likely to represent an early step in the initiation of DNA replication. This activity is provided in a two-step process where the first step involves the binding of E1 and E2 to ori, forming a highly sequence-specific complex. The second step involves the assembly of additional E1 molecules onto this complex and the displacement of E2 in a process that requires ATP hydrolysis. The net result is that a complex with very limited sequence specificity is deposited onto a specific sequence in the DNA. We suggest that this 'seeding' may be a general method to accomplish specific positioning of complexes with a low degree of sequence specificity. Thus, the results presented here provide an interesting example of how sequence-specific recognition can occur and how this process can be coupled to the assembly of a multiprotein–DNA complex at specific sites. Results Formation of the E1E2–ori and E1–ori complexes Two complexes, E1E2–ori and E1–ori, can form on the BPV minimal origin of replication (Figure 1A) in vitro (Sedman and Stenlund, 1995; Sedman et al., 1997). In the E1E2–ori complex, E1 is bound to DNA as a dimer, and in E1–ori it is likely that four molecules of E1 interact with binding sites in the DNA (Chen and Stenlund, 1998; unpublished observations). The sequence specificity and affinity of E1 and E2 bound cooperatively to DNA (E1E2–ori) are much greater than that of E1 bound alone in the E1–ori complex. Since E1 alone posesses all the activities required for replication in vitro (Yang et al., 1991, 1993; Seo et al., 1993b), but initiates replication with low sequence specificity, we hypothesized that E2 may function to generate ori specificity in vivo. E1E2–ori could be a precursor required to form the E1–ori complex that is in turn active for replication. This would be consistent with the observation that the ability to replicate in vivo correlates with the ability to form both complexes in vitro. If this were the case, several predictions should be met: (i) we would expect E2 to stimulate the formation of the E1–ori complex; (ii) E1E2–ori should be a preferred substrate for E1–ori complex formation compared with naked ori DNA; and (iii) E1 molecules recruited to ori by E2 should be incorporated preferentially into the final E1–ori complex. To address these three questions, we first investigated the effects of the nucleotide cofactor adenosine triphosphate (ATP/Mg2+) on ori complex formation since E1 activities required for replication are ATP dependent. The results are shown in Figure 1B. Figure 1.E1 and E2 binding to the origin of replication.(A)Structure of the BPV-1 minimal origin (replicator). The minimal ori spans 59 bp (BPV-1 genome coordinates 7914–27) and consists of an A/T-rich region and the E1- and E2-binding sites separated by 3 bp. (B) E1–ori and E1E2–ori complex formation: effect of ATP/Mg2+ and complex stability in the presence of ethidium bromide (EtBr). Parallel binding reactions with wild-type origin probe were assembled with and without ATP/Mg2+. After 30 min, reaction products were cross-linked and divided in two. One sample was adjusted to 25 μg/ml EtBr prior to electrophoresis (lower panel). The E1 titration was from 50, 30, 20, 12.5 to 2.5 nM (lanes 1–5 and 11–15). For formation of the E1E2–ori complex (lanes 6–8 and 16–18), E1 was at 2.5 nM and E2 was increased from 0.5 to 2.5 nM. At the highest concentration of E2 alone, no complex formation is observed (lanes 9 and 19). Download figure Download PowerPoint E1E2–ori and E1–ori complexes form in the absence of ATP/Mg2+ and can be observed by agarose gel electrophoresis, after protein cross-linking with glutaraldehyde (Lusky et al., 1993; Sedman and Stenlund, 1995). For E1 alone, binding to DNA is highly cooperative and of low sequence specificity. At low E1 concentration in the presence of E2, where binding of E1 alone is no longer observed, a highly sequence-specific E1E2–ori complex forms (Sedman and Stenlund, 1995, 1996; Sedman et al., 1997). As shown in Figure 1B, formation of the E1–ori complex was not altered significantly in the presence of nucleotide cofactor (Figure 1B, top panel, lanes 1–5 compared with 11–15), but the half-life of the complex increased significantly from ∼18 min to very much greater than 60 min when ATP/Mg2+ was present (data not shown). ATP/Mg2+ may alter the association rate for E1–DNA binding since the extent of complex formation was similar under both conditions. Formation of the E1E2–ori complex in the presence of ATP/Mg2+ was reduced (Figure 1B, lanes 6–8 compared with 16–18), consistent with a 3-fold decrease in the half-life of the complex from ∼95 to 35 min in the presence of ATP/Mg2+ (data not shown). Under these conditions, binding of E2 alone to the low affinity E2-binding site is not observed by gel shift analysis (Figure 1B, lanes 9 and 19) or DNase I protection (see below; Figure 4). DNase I and dimethyl sulfate (DMS) footprinting assays indicated that the nucleotide cofactor had little effect on the structure of the E1E2–ori complex (data not shown). Therefore, the same complexes can form on the minimal BPV-1 origin of replication, in the presence or absence of ATP/Mg2+. Figure 2.DNase I footprint analysis of E1–ori and E1E2–ori complexes. (A) Analytical gel shift of binding reactions treated with nuclease and examined in the left panel below (E1–ori). Reaction 1 (lane 1) contained probe alone and reaction 2 (lane 2) received E2 (2 nM final concentration). Reactions 3–5 (lanes 3–5) all contained E1 at a final concentration of 12.5 nM. E1–ori complex formation in reaction 5 (lane 5) was stimulated ∼5-fold by the presence of E2 (2 nM). All reactions contained ATP/Mg2+ except reaction 3. In lanes 6–8, the products of reactions 3–5 were analysed after addition of excess E2 antiserum. (B) Nuclease footprints derived from the binding reactions (1–5) shown above. In order to obtain footprints at low site occupancy (only 12 and 14% of the probe was bound in reactions 3 and 4, respectively), reaction volumes were large (up to 1 ml). After DNase I treatment and cross-linking, products were concentrated and then resolved on preparative agarose gels. For E1–ori (left), lanes 1 and 10 are sequence ladders, and lanes 2 and 9 are the DNase I cleavage products of free probe, with and without ATP/Mg2+. Lanes 4 and 5 show footprints for the complex formed in the presence of ATP/Mg2+ without E2 (reaction 4), cross-linked before (lane 4) or after (lane 5) digestion with DNase I. Lane 6 is the footprint of the E1–ori complex whose formation was stimulated by E2 (reaction 5); products were cross-linked after DNase I digestion. Lanes 7 and 8 are footprints for the complex formed in the absence of ATP/Mg2+ (reaction 3), cross-linked after (lane 7) or before (lane 8) digestion with DNase I. Boundaries of the core and flanking protections are indicated, as are the positions of the A/T-rich region and E1/E2-binding sites. BPV-1 nucleotide co-ordinates of the minimal ori sequence are given on the far left. For the E1E2–ori footprint (centre), binding reactions contained a low concentration of E1 and E2 which give no detectable binding alone (lanes 13 and 14, respectively). Cleavage products for the total E1–E2 binding reaction are shown in lane 15, and the footprint of the isolated E1E2–ori complex in lane 16. Lanes 11 and 18 are sequence ladders, and lanes12 and 17 are the DNase I cleavage products of free probe. Lane 21 shows E2 binding to the low affinity E2-binding site (lane 20, free probe; lane 19, G ladder). The E2 protection (hatched box) extends over 23 nucleotides from BPV nucleotide +8 to +31. At this high concentration of E2 (500 nM), required to observe protection of the E2 site, there is also evidence of under-cleavage of adjacent sequences, suggesting that more than one E2 dimer may be bound cooperatively to a proportion of the DNA templates. Download figure Download PowerPoint Our previous characterization of E1–ori and E1E2–ori has shown that E1 and E2 bind to one face of the DNA helix in the E1E2–ori complex, but that E1 molecules encircle the DNA in the E1–ori complex (Sedman and Stenlund, 1996). Consistent with these observations, the two complexes showed different sensitivities to the intercalating agent ethidium bromide (EtBr). When EtBr was added to 25 μg/ml to pre-formed, cross-linked complexes, the E1E2–ori complex dissociated from the ori probe (Figure 1B, lanes 6–8 and 16–18, compare the lower panel with the upper panel), while the E1–ori complex was stable (Figure 1B, lanes 1–5 and 11–15, compare lower and upper panels). One explanation is that in the cross-linked E1–ori complex, the E1-binding site is inaccessible to EtBr. Protein–DNA but not protein–protein interactions are disrupted by intercalation of EtBr that affects DNA structure (Schroter et al., 1985; Lai and Herr, 1992). We used this property to recover selectively the E1–ori complex from mixtures of E1E2–ori and E1–ori complexes. E2 stimulates formation of an E1–ori complex only in the presence of ATP/Mg2+ To determine if E2 could stimulate E1–ori formation, we performed binding reactions in the presence and absence of ATP/Mg2+ and increasing E1 concentration (Figure 2). Under both conditions, E1 bound cooperatively to the origin (Figure 2, lanes 2–5 and 15–18). At low E1 concentration in the presence of E2, the E1E2–ori complex formed (Figure 2, lanes 6 and 19). In the absence of ATP/Mg2+, E2 inhibited E1–ori complex formation to a significant extent (Figure 2, compare lanes 2–5 and 6–9). Addition of excess polyclonal anti-E2 antibody to a sample of reactions 4, 6 and 8 (Figure 2, lanes 11–13, respectively) demonstrated that when E2 was present, in the absence of ATP/Mg2+, the majority of the ori complexes contained E2. In clear contrast to the results described above, in the presence of ATP/Mg2+, E2 stimulated the formation of the E1–ori complex (Figure 2, lanes 15–18 compared with 19–22). All the E1E2–ori complex that formed in the presence of ATP/Mg2+ could be supershifted, generating a diffuse band in the agarose gel (Figure 2, lane 25). Suprisingly, only a minor fraction of the E1–ori complex formed in the presence of E2 and ATP/Mg2+ could be supershifted with excess anti-E2 antibodies (Figure 2, compare lane 21 with 26). This indicates that either E2 is hidden from the polyclonal antisera in this complex or, more likely, that E2 is turned over during complex formation. A similar observation has been made by Lusky et al. (1994). Therefore, E2 inhibits E1–ori formation in the absence of ATP/Mg2+ but stimulates E1–ori formation in the presence of ATP/Mg2+. Also, E2 is not stably maintained in the E1–ori complex whose formation it promotes. Figure 3.Effect of E2 on E1–ori complex formation in the absence or presence of ATP/Mg2+. The additions of E1 (increasing from 2, 12, 20 to 30 nM) and E2 (1.5 nM) are indicated above each lane. Selected reactions (4, 6 and 8, and 17, 19 and 21) were analysed following addition of polyclonal anti-E2 antiserum (lanes 11–13, and 24–26, respectively). The polyclonal antibodies generate diffuse supershifted bands. Download figure Download PowerPoint The sequence-specific E1E2–ori complex is a preferred precursor for E1–ori To test directly whether the E1E2–ori complex could act as a precursor for E1–ori formation in vitro, we designed a protocol based on molecule tagging. The experimental design is outlined in Figure 3A. We first assembled a substrate E1E2–ori complex with E1 that had been tagged at the N-terminus with two copies of the hemagglutinin (HA) epitope (HAE1E2–ori). These reactions were performed at high probe concentrations (0.4 nM) in the absence of ATP/Mg2+, with a 4- to 5-fold molar excess of HAE1 and E2 proteins. Under these conditions, the HAE1E2–ori complex is stable and the concentrations of free HAE1 and E2 protein are low. After a 20 min incubation, we added an excess of unlabelled specific competitor E2-binding site oligonucleotide (high affinity E2BS9; Li et al., 1989) and diluted the reaction 15-fold into buffer containing nucleotide cofactor, non-specific competitor DNA and a 100-fold molar excess of untagged E1 (relative to tagged E1)—the 'assembly reaction'. The E2 oligonucleotide competes with further E1E2–ori formation and acts as a sink for any template-bound E2 that dissociates. Products of the assembly reaction were removed at various times (T = 2.5, 5, 10, 17.5 and 25 min), cross-linked with glutaraldehyde, and analysed with or without specific antibodies on agarose gels. Chemical modification of E1 by glutaraldehyde instantly quenches further association of E1 with DNA, and stabilizes the protein–DNA complexes. Figure 4.Conversion of epitope-tagged HAE1E2–ori complex to the multimeric E1–ori complex. (A) Experimental design. The experiment is divided into three steps: (1) formation of epitope-tagged HAE1E2–ori complex; (2) the assembly reaction itself; and (3) analysis of products by gel shift/supershift assay. Using a final ratio of tagged to untagged E1 of 1:100 (activity was determined to be approximately equivalent for the tagged and untagged proteins by titration, data not shown), the following predictions can be made. If the pre-formed HAE1E2–ori complex is unable to act as a substrate for E1–ori formation, the frequency at which epitope-tagged E1–ori complexes appear would be determined by the input ratio of the two proteins and the order of the E1–ori complex. For a tetrameric complex, ∼4% of the products would be tagged and, likewise, for a hexameric complex, ∼6% of the products would be tagged. If HAE1E2–ori is able to act as a substrate for E1–ori formation, with affinity similar to that of the free probe, the proportion of tagged E1–ori complexes would be determined by the ratio of free probe to HAE1E2–ori complex. If HAE1E2–ori is a preferred substrate, the extent of E1–ori complex formation would exceed that which occurs in the absence of a pre-formed HAE1E2–ori complex. Also, the majority of complexes would be tagged. (B) Conversion of a pre-formed HAE1E2–ori complex to the multimeric E1–ori complex. Left: analysis of reaction products of the experiment outlined in (A). Lanes 1–8: products of the pre-incubation reaction in the absence of ATP/Mg2+ and control reactions. Lane 1, free probe; lanes 2–4, E1–ori complex formation in the absence of E2; lanes 5–7, controls for antibody activity with HAE1E2–ori substrate as indicated; lane 8, substrate HAE1E2–ori complex at T = 0, after pre-incubation, before initiation of the assembly reaction. The following three groups of five lanes are time points T = 2.5, 5, 10, 17.5 and 25 min after initiation of the assembly reaction, analysed with or without specific antibodies, added after cross-linking. Right (lanes 24–35): control assembly reaction where E2 was omitted from the pre-incubation without ATP/Mg2+, but was included upon dilution and addition of untagged E1, ATP/Mg2+ and competitor DNAs. Lanes 24 and 25, markers for the E1–ori complex and E1E2–ori complex; lanes 26–30, time points as left, without antibody; lanes 31–35, analysis with anti-HA epitope antibody. Download figure Download PowerPoint Figure 3B shows the time course of assembly of an E1–ori complex from the HAE1E2–ori substrate. Figure 3B, lanes 2–4 show E1–ori complex formation in the absence of E2 at low, intermediate and high E1 concentrations respectively. As shown in lane 8, in the presence of HAE1 and E2, the expected HAE1E2–ori complex formed, and this complex could be supershifted with excess anti-E1, anti-E2 and anti-HA epitope antibody 12CA5 (Figure 3B, lanes 5–7). Addition of the large excess of untagged E1 and nucleotide cofactor resulted in rapid formation of a complex that co-migrated with the slower migrating multimeric E1–ori complex (Figure 3B, lanes 9–13). Concomitant with formation of this complex, we observed a decrease in the amount of detectable E1E2–ori. Addition of anti-HA epitope antibody (Figure 3B, lanes 14–18) demonstrated that >85% of the slow migrating complex contained HA-tagged E1 at early time points (<5 min), whereas the proportion fell to 60–70% at later times (Figure 3B, lane 18). Had the E1–ori complex been formed independently of the tagged HAE1E2–ori complex, only 4% of the complexes (assuming that the E1–ori complex is a tetramer) would have incorporated tagged E1 and would be supershifted by this antibody. Therefore, the vast majority of the E1–ori complexes that formed initially were formed from the HAE1E2–ori complex, indicating that this is a good substrate for E1–ori formation. The appearance of untagged complexes at later times could be the result of de novo E1–ori formation from free probe; alternatively, the tagged E1 molecules (from HAE1E2–ori and E1–ori) could exchange with untagged E1 in solution, or become masked by further addition of E1. These possibilities could also account for the change with time in the relative proportions of the two anti-HA supershift species observed in Figure 3B, lanes 14–18. As expected, when ATP/Mg2+, ATP or Mg2+ alone were omitted from the complete assembly reaction, no E1–ori complex formation was observed (not shown). In a parallel control reaction, E2 was omitted from the pre-incubation of probe with tagged E1, but was added upon dilution and addition of untagged E1, ATP/Mg2+ and competitor DNAs (Figure 3B, lanes 26–35, right). Lanes 24 and 25 are markers for the E1–ori complex and the E1E2–ori complex respectively. The extent of E1 complex formation from free probe was clearly lower (∼3-fold in this case; Figure 3B, lanes 26–30) in this reaction compared with that described above, where a high proportion of substrate was a pre-formed HAE1E2–ori complex. Also, the complex that formed could not be supershifted with anti-HA antibody (Figure 3B, lanes 31–35). Therefore, the large excess of untagged E1 protein (100-fold) results in an undetectable level of tagged E1–ori complex formation in these assays. Together, the results described above demonstrate that a pre-formed HAE1E2–ori complex is a preferred substrate for E1–ori complex formation compared with free probe. Furthermore, the E1 molecules of precursor HAE1E2–ori were incorporated into the resulting E1–ori complex. To determine the fate of E2 in these reactions, we performed supershift assays using antibody directed against E2 (Figure 3B). At the earliest times (<5 min), up to 50% of the slow migrating E1–ori-like complex could be supershifted with anti-E2 antibody (Figure 3B, lane 19). By 25 min, however, little association of E2 with the E1–ori complexes could be detected (Figure 3B, lane 23). This suggests that E2 dissociates from the nascent E1–ori complex. Indeed, nuclease footprinting showed no evidence that E2 was bound to its site in these complexes (data not shown, but see Figure 4B). Like the E1–ori complex that formed at high E1 concentrations in the absence of E2, the E1–ori complex formed via E1E2–ori is stable in the presence of EtBr. Interestingly, the slow migrating E2-containing complexes apparent at early times (Figure 3B, lanes 19–23) also do not dissociate in the presence of EtBr (data not shown). These are most likely intermediates in formation of the multimeric E1–ori complex, where E2 is cross-linked to E1 that it initially recruited to ori. E1–ori complexes formed in the presence or absence of E2 generate identical nuclease footprints To determine if E1 was similarly disposed in E1–ori complexes formed with or without E2, we footprinted complexes with DNase I (Figure 4). We compared E1–ori complexes formed at low E1 concentration in the presence and absence of ATP/Mg2+ with E1–ori complexes formed in the presence of E2. Also, we analysed complexes that were cross-linked before treatment with nuclease. To obtain solution footprints at low site occupancy, we isolated protein–DNA complexes by gel electrophoresis after digestion with DNase I. Cleavage was terminated by adding EDTA and glutaraldehyde. To sequester any E2-containing complexes from the multimeric E1–ori complex, excess anti-E2 antiserum was added prior to electrophoresis. Bound and free probe were then separated by agarose gel electrophoresis and the respective protein–DNA complexes were recovered after blotting to nitrocellulose. Figure 4A shows an analytical gel shift of the binding reactions treated with nuclease. E1 alone, in the absence and presence of ATP/Mg2+, gave rise to minimal complex formation (Figure 4A, lanes 3 and 4). Upon addition of E2, formation of the E1–ori complex was stimulated (Figure 4A, lane 5). In Figure 4A, lanes 6–8, anti-E2 antibody was added to a sample of reactions 3–5. The nuclease protections (bottom strand) derived from the binding reactions in Figure 4A are shown in Figure 4B, left panel (labelled E1–ori). The centre panel shows the footprint of the E1E2–ori complex for comparison (labelled E1E2–ori). The footprint of E2 bound to the low affinity E2-binding site is also shown on the right (labelled E2). Importantly, the footprints for the E1–ori complex formed in the presence of ATP/Mg2+, without and with E2 (Figure 4B, lanes 5 and 6), are indistinguishable. On this strand, E1 protects a 29 nucleotide 'core' sequence (BPV nucleotides 7934 to +15) over the E1 recognition site. In the flanking regions, extending 18 nucleotides upstream and 19 downstream (to nucleotide +34) of the core, some positions are completely protected, while others only partially. There is also evidence of hypersensitivity in these regions (labelled Hyp.). On this strand, the E1E2–ori protection extends from nucleotide 7935 to +31 (Figure 4B, lane 16). The downstream boundary of this protection is the same as when E2 alone is bound to the probe (Figure 4B, lane 21). Importantly, comparing the E1–ori complex protections over the E2 recognition sequence (and beyond) with the corresponding region in the E1E2–ori complex (Figure 4B, centre panel, lane 16, compared with left panel, lane 6) confirms that E2 does not occupy the E2-binding site in the majority of E1–ori complex whose formation it stimulates [Figure 4B, compare the three bracketed regions (1–3) in the left and centre panels]. Similar results were obtained for the other DNA strand (top, data not shown). On this strand, the core E1–ori protection encompasses 28 nucleotides from nucleotide 7938 to +18. The flanking protections extend 19 nucleotides downstream and 21 nucleotides upstream over the A/T-rich region. The E1–ori complex formed in the absence of ATP/Mg2+ (Figure 4B, lane 7) was virtually indistinguishable from the complex formed in the presence of ATP/Mg2+ (Figure 4B, lanes 5 and 6), suggesting that E1 is bound to ori in the same way in all complexes. The only significant change in the E1–ori footprints was observed when complexes were treated with nuclease after cross-linking (Figure 4B, lanes 4 and 8, left). The core protection is preserved; however, for the cross-linked complex formed in the absence of ATP/Mg2+, flanking protections appear to be lost. Similar results were also obtained for the other strand (data not shown). These results indicate that the E1–ori complexes formed with or without E2 are identical, implying that E2 has a purely quantitative effect on E1–ori complex formation. Also, there is good evidence that E2 does not remain associated with its binding site in this complex. Together, these observations are consistent with a catalytic role for E2 in complex formation. In addition, differences between the nuclease footprints of complexes formed in the presence and absence of ATP/Mg2+ become significant only after cros
Referência(s)