Portal de Periódicos da CAPES

Action of site-specific recombinases XerC and XerD on tethered Holliday junctions

1997; Springer Nature; Volume: 16; Issue: 12 Linguagem: Inglês

10.1093/emboj/16.12.3731

1460-2075

Lidia K. Arciszewska,

DNA and Nucleic Acid Chemistry

Article15 June 1997free access Action of site-specific recombinases XerC and XerD on tethered Holliday junctions Lidia K. Arciszewska Lidia K. Arciszewska Microbiology Unit, Department of Biochemistry, University of Oxford, South Parks Road, Oxford, OX1 3QU UKK.Arciszewska and I.Grainge contributed equally to this work Search for more papers by this author Ian Grainge Ian Grainge Microbiology Unit, Department of Biochemistry, University of Oxford, South Parks Road, Oxford, OX1 3QU UKK.Arciszewska and I.Grainge contributed equally to this work Search for more papers by this author David J. Sherratt Corresponding Author David J. Sherratt Microbiology Unit, Department of Biochemistry, University of Oxford, South Parks Road, Oxford, OX1 3QU UK Search for more papers by this author Lidia K. Arciszewska Lidia K. Arciszewska Microbiology Unit, Department of Biochemistry, University of Oxford, South Parks Road, Oxford, OX1 3QU UKK.Arciszewska and I.Grainge contributed equally to this work Search for more papers by this author Ian Grainge Ian Grainge Microbiology Unit, Department of Biochemistry, University of Oxford, South Parks Road, Oxford, OX1 3QU UKK.Arciszewska and I.Grainge contributed equally to this work Search for more papers by this author David J. Sherratt Corresponding Author David J. Sherratt Microbiology Unit, Department of Biochemistry, University of Oxford, South Parks Road, Oxford, OX1 3QU UK Search for more papers by this author Author Information Lidia K. Arciszewska1, Ian Grainge1 and David J. Sherratt 1 1Microbiology Unit, Department of Biochemistry, University of Oxford, South Parks Road, Oxford, OX1 3QU UK *E-mail: [email protected] The EMBO Journal (1997)16:3731-3743https://doi.org/10.1093/emboj/16.12.3731 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info In Xer site-specific recombination, two related recombinases, XerC and XerD, mediate the formation of recombinant products using Holliday junctioncontaining DNA molecules as reaction intermediates. Each recombinase catalyses the exchange of one pair of specific strands. By using synthetic Holliday junction-containing recombination substrates in which two of the four arms are tethered in an antiparallel configuration by a nine thymine oligonucleotide, we show that XerD catalyses efficient strand exchange only when its substrate strands are ‘crossed’. XerC also catalyses very efficient strand exchange when its substrate strands are ‘crossed’, though it also appears to be able to mediate strand exchange when its substrate strands are ‘continuous’. By using chemical probes of Holliday junction structure in the presence and absence of bound recombinases, we show that recombinase binding induces unstacking of the bases in the centre of the recombination site, indicating that the junction branch point is positioned there and is distorted as a consequence of recombinase binding. Introduction Holliday junction-containing DNA molecules (Holliday, 1964) are intermediates in homologous recombination (reviewed in West, 1992; Kowalczykowski et al., 1994; Eggleston and West, 1996) and in site-specific recombination mediated by the λ integrase family of site-specific recombinases (reviewed in Landy 1989, 1993; Stark et al., 1992). In homologous recombination, a range of different proteins recognize Holliday junctions with little or no sequence specificity, and either cleave the junctions to generate ‘recombinant’ or ‘non-recombinant products’, or promote branch migration of the junction (Duckett et al., 1988; Mueller et al., 1988; Lu et al., 1991b; Bennett and West 1995a; West, 1996; White and Lilley, 1996). In contrast, integrase recombinases are sequence-specific DNA-binding proteins that recognize recombination site DNA whether it is in linear duplex or a Holliday junction-containing molecule, and mediate the strand exchange reactions at precisely defined positions. A substantial body of work has led to insight into Holliday junction structure and branch migration (Lilley and Clegg 1993a,b; Panyutin and Hsieh, 1994; Seeman and Kallenbach, 1994; Panyutin et al., 1995). In the absence of metal ions, Holliday junctions adopt an unstacked square configuration that shows 4-fold symmetry and readily allows branch migration. In this configuration, the bases at the branch point of the junction are sensitive to chemical probes, such as osmium tetroxide or potassium permanganate, that react with unpaired or unstacked thymine residues (Duckett et al., 1988; Bennett and West, 1995b). In the presence of metal ions, the junction folds into a 2-fold symmetric antiparallel right-handed X structure in which the helices of the arms are coaxially stacked pairwise onto each other and the helical stacks are rotated at an angle of ∼60° to each other (Figure 1A). Branch migration of these structures is slow. Parallel isoforms of Holliday junctions are also possible, but they are energetically less favourable (von Kitzing et al., 1990; Lu et al., 1991a). Figure 1.(A) The two isoforms of the right-handed antiparallel stacked X Holliday junction. In the isoform on the left, top strands are crossed (dark blue and red), while in the other isoform the bottom strands are crossed (light blue and yellow). The junction crossover is positioned in the middle of the 6 bp central region. The positions of cleavages of XerC and XerD (open and filled circles, respectively) are mapped onto the phosphate backbone. The arms bound by XerC and XerD are indicated. Note the proximity of XerC cleavage sites in the isoform in which top strands are crossed and the identical proximity of XerD cleavage sites in the isoform in which the bottom strands are crossed. The colour code is as follows: strand I, dark blue; II, light blue; III, yellow; and IV, red. (B) A scheme for Xer recombination adapted from the model of Nunes-Düby et al. (1995) for λ Int. Binding of recombinases is followed by synapsis, shown with the dif recombination sites arbitrarily in antiparallel orientation. XerC mediates the first pair of strand exchanges between the top strands (step a, expanded in the box; McCulloch et al., 1994; Colloms et al., 1996). As with other λ family recombinases, XerC catalyses cleavage and rejoining in a two-step transesterification reaction in which the conserved tyrosine acts as the initial nucleophile generating a 3′ phosphotyrosyl protein–DNA complex. Three nucleotides of each of the top strands are exchanged between the partners in a single swap and then the terminal 5′ OH of the partner strand acts as the nucleophile in the second transesterification step. The branch point of the generated Holliday junction is located in the middle of the central region and the top strands are crossed. Resolution of the junction to products by XerD-mediated exchange between the bottom strands (step d) requires isomerization (steps b + c) to adopt a configuration in which bottom strands are crossed. During isomerization, any cross-core XerC–XerD contacts must be broken and contacts with the new partners need to be re-made. The arrows indicate positions of XerC and XerD cleavages and rejoining; Y depicts the tyrosine nucleophile. Download figure Download PowerPoint In the stacked antiparallel right-handed X structure, the two ‘continuous’ strands run in antiparallel orientation to each other along the axes of the stacked X arms (Figure 1A). The other two strands exchange between the two helical stacks at the branch point and have been called ‘discontinuous’, ‘exchanging’ or ‘crossed’. Here, we refer to them as ‘crossed’ strands in order to avoid confusion with strands that contain nicks (discontinuous) or strands that undergo exchange during recombination (exchanging). Depending on the choice of helical partners during stacking, the junction can form two different isoforms in which a specific pair of strands is either continuous or crossed (Figure 1A). Xer site-specific recombination, which functions in the stable inheritance of circular replicons, is an atypical member of the integrase family because, instead of using a single recombinase, it uses two related recombinases, XerC and XerD, each of which catalyses the exchange of one specific pair of strands (Blakely et al., 1993, 1997; Colloms et al., 1996; reviewed in Sherratt, 1993; Sherratt et al., 1995). Xer recombinases act on a variety of natural sites present in plasmids (e.g. psi present in plasmid pSC101; Cornet et al., 1994; Colloms et al., 1996) and at the site, dif, present in the replication terminus region of the Escherichia coli chromosome (Blakely et al., 1991; Kuempel et al., 1991). Our current view of the Xer recombination mechanism is shown schematically in Figure 1B, which shows recombination sites having a 6 bp dif central region. On supercoiled substrates containing two directly repeated copies of psi, strand exchange is sequential, with XerC mediating the first pair of ‘top’ strand exchanges to generate a Holliday junction-containing intermediate. Then XerD catalyses exchanges between the ‘bottom’ strands to generate recombinant products. By convention, we term the pair of strands exchanged by XerC as the ‘top’ strands and the pair of stands exchanged by XerD as the ‘bottom’ strands (McCulloch et al., 1994; Arciszewska and Sherratt, 1995; Blakely and Sherratt, 1996; Colloms et al., 1996). We do not know if there is a preferred order of strand exchange with dif, though, in the cartoon, we have shown the XerC-mediated exchange occurring first. Key issues in understanding the mechanism of strand exchange during integrase family site-specific recombination are the roles of Holliday junction branch migration and isomerization in the recombination process. This and the accompanying paper by Azaro and Landy address the question of how Holliday junction structure and position determine which of the two pairs of strand exchanges is to occur. Earlier work on recombination mediated by the λ integrase enzymes was interpreted as suggesting that after the first pair of strand exchanges the Holliday junction branch point was located where cleavage and rejoining had occurred (reviewed in Stark et al., 1992; Landy, 1993; Nunes-Düby et al., 1995). Branch migration, accompanied by isomerization, would then move the junction 6–8 bp to the position of the second pair of cleavages, thereby allowing recombination to be completed. Recent data from experiments using substrates in which branch migration is constrained led to the idea that the Holliday junction branch point need not be at the position of strand cleavage for efficient strand exchange to occur (Dixon and Sadowski, 1994; Arciszewska et al., 1995; Nunes-Düby et al., 1995) The ‘strand swapping’ model for λ integrase recombination (Nunes-Düby et al., 1995) proposes that the Holliday branch point is located close to the middle of the 7 bp region between the two cleavage points. During Holliday junction formation and resolution, 2 or 3 bp of each partner duplex is swapped in a single step during each pair of strand exchanges, and branch migration is limited to 1–3 bp. This avoids the need for extensive branch migration and the associated helical rotation, though isomerization is still expected since it provides an equivalent stereochemical environment for each pair of strand exchanges (Figure 1A; Stark et al., 1989). In the model, strands that are crossed, or become crossed on Holliday junction formation, are exchanged by strand swapping. Figure 1B shows this model when applied to Xer recombination. In this case, branch migration may not be required since the Holliday junction branch point could be located in the middle of the 6 bp central region after swapping of three nucleotides, though isomerization might still be necessary to allow both pairs of strand exchanges. Here, we use synthetic Holliday junction-containing substrates carrying the recombination site dif, derived from the E.coli chromosome. Two substrates contain a nine thymine oligonucleotide tether to constrain the junction into one or the other of two antiparallel isoforms. Our results suggest that XerD strand exchange occurs only if its substrate strands are initially in the crossed configuration. XerC also efficiently exchanges crossed strands, though apparently it is able to mediate less efficient recombination on strands that were initially continuous. Binding of the recombinases to the Holliday junctions distorts the branch point and localizes it to the centre of the core recombination site. Results Structure of tethered Holliday junction-containing substrates Two antiparallel tethered dif Holliday junction-containing substrates were constructed. In TSX the top strands are crossed, while in BSX the bottom strands are crossed (Figures 1A and 2A and B). A nine thymine oligonucleotide tether ensures that these molecules adopt only one of the two possible antiparallel isoforms and cannot isomerize to a parallel configuration (Kimball et al., 1990). The 28 bp core recombination site of TSX and BSX is identical to that of the untethered dif Holliday junction, and branch migration can occur throughout this region (see below). The heterologous arms outside of the core site are of similar size and sequence in the tethered and untethered substrates. The expected products of the recombinase-mediated strand exchanges in the tethered substrates are shown in Figure 2C. Figure 2.(A) The structure of the untethered Holliday junction-containing substrate. The substrate carries two 28 bp dif core recombination sites. The two XerC- and two XerD-binding sites (each 11 bp, filled and hatched ovals, respectively) flank 6 bp central regions. The lengths of the arms outside the core region are indicated. The sequences of these arms are unrelated to each other, allowing the junction to migrate only within the limits of the core region (Arciszewska et al., 1995). Strands I and IV are designated ‘top’ strands (thick lines), while II and III are ‘bottom’ strands (thin lines); designation of arms follows the numbers of strands which contribute the 5′ end to the arm's terminus, for example, arm I contains the 5′ end of strand I. Exchange of the top (T) strands generates two linear duplex molecules, each of 76 bp; while exchange of bottom strands (B) produces linear duplexes of 84 and 68 bp. XerC cleavage positions are indicated by the solid arrowheads, while expected XerD cleavage positions are shown by open arrowheads. The junction molecule is presented in unstacked square conformation. (B) The structure of tethered Holliday junction substrates, TSX (top strands crossed) and BSX (bottom strands crossed). In TSX, arms III and IV, while in BSX arms I and III, are brought together. To allow assembly of these substrates from synthetic oligonucleotides, each of them contains a nick (indicated by an arrow) in one of the arms located 18 nucleotides away from the oligo(T) tether. It was anticipated that the nick would not have a significant effect on junction structure or recombination, since it is outside of the core recombination site. Filled circles indicate positions of the label in substrates used in recombination experiments presented below. (C) The products of exchanges between a pair of top and a pair of bottom strands in TSX and BSX substrates. The lengths (nucleotides) of recombinant strands are indicated. Download figure Download PowerPoint Tethered molecules are expected to have properties similar to untethered Holliday junctions; for example it has been shown that in the presence of metal ions tethered junctions adopt a conformation close to that of the untethered right-handed stacked X form, since they exibit similar sensitivities to osmium tetroxide and hydroxyl radical attack (Kimball et al., 1990; Bhattacharyya et al., 1991). The interactions of T4 endonuclease VII and RuvC with tethered and untethered Holliday junction substrates were shown to be similar (Bhattacharyya et al., 1991; Bennett and West, 1995a). Below, we extend these observations by demonstrating that XerC and XerD interactions with tethered and untethered Holliday junctions are also very similar. The products of recombinase-mediated catalysis on tethered Holliday junction substrates Incubation of XerC and XerD with substrate TSX gives a single major product of recombination (Figure 3A), whose electrophoretic mobility is consistent with it having resulted from a pair of top strand exchanges (Figure 2C). Reactions using catalytically inactive mutants XerCY275F or XerDY279F show, as expected, that the recombinant product is generated by XerC. The presence of the recombinant strand was confirmed on sequencing gels (data not shown). The equivalent recombinant product of XerC strand exchange was seen on the untethered dif substrate (HJ). All of the detectable product of XerC strand exchange on TSX had retained the tether. If a TSX substrate had become untethered prior to recombination by dissociation of the 3′ end of the strand II tether from strand IV, a linear product containing 54 nucleotides on one strand and 76 nucleotides on the other strand would have been generated. This product has not been observed in our experiments. Therefore, the great majority of TSX molecules remain tethered before and after strand exchange. No recombinant products of the size predicted for XerD-mediated bottom strand exchange were detected in reactions with either TSX or untethered dif in most of the experiments (Figure 3A, and data not shown). In a few experiments, we have detected what may be a very low level of XerD strand exchange product in reactions containing either TSX or an untethered Holliday junction substrate. It is not clear whether this product is derived from the intact substrate or from another DNA molecule present in small amounts in some recombination reactions. Figure 3.(A) Recombination of TSX substrate. Recombinases were XerC and XerD or their catalytically inactive derivatives XerCY275F or XerDY279F. The gel (4% polyacrylamide containing SDS) shows DNA from 20 min reactions at 20°C. The products of XerC recombination run with the mobilities expected from their structure (Figure 2C). The numbers shown in shadowed font represent the average amounts of strand exchange products generated in the 20°C incubations for 20 min in seven experiments [standard error of means (SEM) are: for TSX/XerC, 4.4; HJ/XerC, 4.3] or in 37°C incubations for 40 min in four experiments (SEM; TSX/XerC, 3.6; HJ/XerC, 1.6); other values represent the amount of recombinant products in the gel shown. nd; not detected (⩽0.2% of total DNA). (B) Recombination of BSX substrate. The gel (4% polyacrylamide containing SDS) shows DNA from 30 min reactions at 37°C. The numbers in shadowed font represent average amounts of strange exchange products generated in five experiments in 30 min incubations at 37°C (SEM; BSX/XerC, 6.3; BSX/XerD, 2.8; HJ/XerC, 1.8); other values give the amounts of strand exchange in the gel shown; also shown are the amounts of products generated in reactions incubated at 20°C for 20 min in one experiment. nd; as for (A). (C) Time course of XerC and XerD strand exchange in tethered and untethered dif substrates. Recombination reactions containing ligase-treated TSX, BSX or untethered dif substrates were incubated in the presence of XerC and XerD at 37°C. The plots present the percentage of substrate DNA converted to rejoined recombinant product. Download figure Download PowerPoint Two major recombinant products were produced in similar amounts when XerC and XerD were incubated with the BSX substrate (Figure 3B). The electrophoretic mobility of these products indicates that one of them was generated through exchanges between a pair of top strands and the other between a pair of bottom strands (Figure 2C). Use of the catalytically inactive mutants demonstrated that the larger product was generated by XerD while the smaller was from XerC-mediated strand exchange (Figure 3B). Analysis of recombination reactions with BSX on a sequencing gel confirmed the presence of recombinant top and bottom strands (data not shown). In addition to the strand exchange products, small amounts of covalent complexes of recombinase with DNA were seen running above the TSX and BSX substrates (Figure 3A and B). As expected, they disappeared when samples were treated with proteinase K before electrophoresis (data not shown). The amounts of covalent complexes generated in reactions with TSX and BSX were similar to those observed with the untethered dif Holliday junction, indicating that cleavage and rejoining in all these substrates occur in a similar way. Taken together, these results suggest that XerD can mediate strand exchange on a dif Holliday junction only when its substrate strands are crossed. XerC appears to be able to mediate strand exchange when its substrate strands are either crossed or continuous. In the above experiments, the 150 mM Na+ present in the recombination reaction partially stacks the arms of the substrates (see below). Similar results of recombination were observed in the presence of 1 mM Mg2+, which should fully stack the junctions prior to recombinase binding. It was not possible to analyse recombination in the absence of metal ions because of recombinase insolubility in these conditions. Recombination of substrates containing an 18 thymine oligonucleotide tether generated the same products as TSX and BSX (data not shown). Relative rates of catalysis on tethered Holliday junction substrates In the experiments shown in Figure 3A and B, the levels of strand exchange were dependent on the incubation temperature of the reaction. The levels of XerC and XerD strand exchange on BSX were low at 20°C when compared with XerC strand exchange on the untethered dif Holliday junction (Figure 3B). Shifting the incubation temperature to 37°C resulted in a several-fold increase in both strand exchanges on BSX when compared with exchange on untethered junction. The levels of XerC strand exchange on TSX and dif are similar at 20°C (Figure 3A). At 37°C, less XerC strand exchange occurred on TSX as compared with dif. In order to gain more insight into the strand exchange reactions on TSX and BSX, we carried out kinetics experiments. Before doing this, we wished to confirm that the recombination behaviour of TSX and BSX was not perturbed by the presence of a nick outside of the core recombination site. Therefore, samples of TSX and BSX that had been treated with DNA ligase in order to close the nick were used as recombinant substrates. Typically 50–60% of the nicks could be rejoined. The level and pattern of products of BSX recombination were identical on ligated and unligated substrates. Ligated TSX gave higher levels of XerC recombinant product than unligated TSX (a mean of 48% as compared with 30% in three 30 min, 37°C incubation experiments; data not shown). No evidence for XerD strand exchange in ligated TSX was obtained. Therefore, the nick which is immediately 5′ of the XerD-binding site does not change the nature of strand exchange products, though it does appear to reduce the level of recombination. We show later that this nick is not a ‘sink’ for branch migration, though it may be responsible for some local DNA distortion. The time course experiments were performed on TSX and BSX that had been treated with ligase (Figure 3C). XerC-mediated recombination of untethered dif junction is fast; typically ∼50% of strand exchange product generated in a 30 min reaction at 37°C was produced within the first minute, a result similar to those observed with Holliday junction-containing substrates carrying other core recombination sites (Arciszewska et al., 1995). The initial rate of XerC-mediated strand exchange on TSX was very similar to that in untethered substrate, as was the level of recombinant product at 30 min. On BSX, the initial rate of XerC-mediated strand exchange was at least 3-fold lower than on TSX. The level of product was >2-fold less at 30 min. The initial rate of XerD strand exchange on BSX was lower than that mediated by XerC, though by 30 min the level of XerD product was some 50% higher than that of XerC. A similar pattern of recombination rates was observed when reactions were analysed at 20°C, though in this case the initial rate of XerC strand exchange on BSX was at least eight times lower than on the untethered junction control (data not shown). The higher initial rates of XerC-mediated exchanges on crossed strands of TSX as compared with continuous strands in BSX indicate that XerC has a preference for exchanging crossed strands. Restriction endonuclease analysis of substrates and products Because of the apparent different substrate specificities for XerC and XerD catalysis, we felt it necessary to verify the structure of the recombination products and the integrity of the substrates using BamHI and BglII restriction endonuclease digestion (Figure 4). In particular, we wished to test whether recombination had occurred on tethered molecules, or on a small sub-population of molecules that were untethered prior to recombination. Despite BamHI digestion being inefficient, because its recognition site is located very close to the linear duplex ends, sufficient cleavage was obtained to allow verification of structure. The product of TSX strand exchange is cleaved with both restriction enzymes, giving DNA fragments of the predicted sizes for a linear molecule retaining a tether. Analysis of reacted BSX DNA revealed the expected pattern of BglII cleavage of the XerD-generated strand exchange product. As expected, BglII did not cleave the XerC recombinant product. Figure 4.Restriction endonuclease digestion of tethered junctions and their recombination products. Substrates were incubated with XerC and XerD for 20 min at 20°C, the reactions were stopped with SDS, treated with proteinase K, phenol extracted and ethanol precipitated. The recovered DNA was digested with BamHI (BH) or BglII (Bg) and analysed on a 6% polyacrylamide gel. DNA not treated with either recombinases or restriction enzymes was included as a control. The sites of cleavage by both enzymes are indicated in the structures of substrates and products (only labelled products are shown). Top strands are drawn as thick lines; filled small circles indicate the positions of the label; the position of the nick in each substrate is indicated with an arrow. Download figure Download PowerPoint Cleavage of TSX by BglII and BamHI, and BSX by BglII, should untether the arms of these substrates and therefore retard their electrophoretic mobility (compare the relative mobilities of the tethered and untethered Holliday junctions; Figure 3A and B). The predicted retarded DNA was observed (Figure 4), though digestion was incomplete. Cleavage of TSX with BglII also generated DNA that ran ahead of the undigested substrate; this most likely reflects the dissociation and loss of an oligonucleotide between the 3′ end of strand IV and the BglII cleavage position (8 nucleotides), or between the strand III nick and the BglII cleavage position (14 nucleotides), or both. These results confirm that the products of XerC strand exchange on TSX and XerD strand exchange on BSX retain their tethers and are therefore derived from recombination on substrates that were tethered. However, the product of XerC strand exchange on BSX is identical to that which would have resulted from recombination on an untethered BSX substrate (Figure 2). We believe that this product derives from XerC exchange on tethered BSX for the following reasons. First, we do not observe slower mobility Holliday junction DNA, indicative of untethered substrate, in recombination reactions containing tethered substrates (Figure 3A and B). Second, we did not detect a XerC strand exchange product derived from untethered molecules that might be present in the TSX substrate (see above). Third, XerC-mediated recombination on BSX is strongly inhibited in the presence of one particular XerD mutant; this inhibition is much less pronounced in the untethered dif substrate, indicating that XerC is acting on different molecules in the two cases (B.Hallet, I.Grainge and D.J.Sherratt, unpublished data). Fourth, any untethering of BSX by denaturation of the region from the tether up to the nick (Figure 2B) would have generated single-stranded regions that would have been hypersensitive to osmium tetroxide and KMnO4 modification; these were not observed (see below). Finally, an experiment was carried out in which BSX (treated with and without ligase) was labelled on strand III (Figure 2). This allowed visualization of the other top strand exchange product, which was seen to have the mobility expected of the small gapped circle that should result from continuous strand exchange on the tethered junction (data not shown). Chemical probing of tethered Holliday junction structure Previous studies on Holliday junctions have shown that in the absence of metal ions, the junctions adopt a 4-fold symmetric square configuration and the base pairs located at the branch point are sensitive to osmium tetroxide and potassium permanganate, which react specifically with the C5–C6 double bond of unstacked or unpaired thymine bases (Lilley and Palacek, 1984; Duckett et al., 1988, Sasse-Dwight and Gralla; 1989; Bennett and West, 1995b). Metal ion-induced coaxial stacking of the junction arms in the right-handed stacked X structure, reduces this sensitivity. Therefore, these chemicals provide a useful probe for determining the extent of coaxial stacking and the position of the junction bra