Journal home
Advance online publication
Current issue
Press releases
Guide to authors
Online submissionOnline submission
For referees
Free online issue
Contact the journal
About this site
For librarians
Nature Research Resources
Nature Cell Biology
Nature Reviews Molecular Cell Biology
The EMBO Journal
Nature Reports Avian Flu
NPG Subject areas
Clinical Medicine
Drug Discovery
Earth Sciences
Evolution & Ecology
Materials Science
Medical Research
Molecular Cell Biology
Browse all publications
Nature Structural Biology  8, 126 - 130 (2001)

Structure of free BglII reveals an unprecedented scissor-like motion for opening an endonuclease

Christine M. Lukacs1, 2, Rebecca Kucera3, Ira Schildkraut3 & Aneel K. Aggarwal1

1 Structural Biology Program, Department of Physiology and Biophysics, Mount Sinai School of Medicine, 1425 Madison Avenue, New York, New York 10029, USA.

2 Present address: Hoffman-La Roche Inc., 340 Kingsland St., Nutley, New Jersey 07110, USA.

3 New England Biolabs, 32 Tozer Road, Beverly, Massachusetts 01915, USA.

Correspondence should be addressed to Aneel K. Aggarwal
Restriction endonuclease BglII completely encircles its target DNA, making contacts to both the major and minor grooves. To allow the DNA to enter and leave the binding cleft, the enzyme dimer has to rearrange. To understand how this occurs, we have solved the structure of the free enzyme at 2.3 Å resolution, as a complement to our earlier work on the BglII−DNA complex. Unexpectedly, the enzyme opens by a dramatic `scissor-like' motion, accompanied by a complete rearrangement of the alpha-helices at the dimer interface. Moreover, within each monomer, a set of residues — a 'lever' — lowers or raises to alternately sequester or expose the active site residues. Such an extreme difference in free versus complexed structures has not been reported for other restriction endonucleases. This elegant mechanism for capturing DNA may extend to other enzymes that encircle DNA.

Protein−DNA selectivity is a central event in many biological processes, ranging from transcription and replication to restriction and modification. Type II restriction endonucleases are ideal systems for unraveling the events leading to specific DNA recognition, with structures of both free and DNA bound forms of several enzymes now available1, 2, 3, 4, 5, 6, 7, 8. Most type II restriction enzymes exist as homodimers in solution that recognize short 4−8 base pair palindromic DNA sites, requiring only Mg2+ as a cofactor for DNA hydrolysis9. Despite a lack of sequence homology amongst these endonucleases, the crystal structures reveal a similar alpha/beta core that consists of a central beta-sheet flanked by several alpha-helices10, 11. A striking feature to emerge from these structural studies is the range of conformational transitions that can occur on DNA binding, including the unwinding and bending of DNA1, 12, 13, 14, the global movement of protein domains1, 4, 15, and the folding and unfolding of local protein regions1, 4, 15.

BglII is a 223-amino acid restriction endonuclease, originally isolated from Bacillus globigii16, that recognizes and cleaves the DNA sequence AGATCT. Recently, we reported the 1.5 Å resolution crystal structure of BglII bound to its DNA recognition site14. The overall fold of BglII bears a strong resemblance to that of endonuclease BamHI, which recognizes nearly the same DNA site (GGATCC). This structural similarity lies primarily in the major alpha/beta core domain, which carries the majority of the DNA recognition and all of the catalytic residues. However, BglII contains an extra C-terminal beta-sandwich domain that provides additional DNA specificity14. This domain provides a scaffold for two loops (loops A and D) that reach around to the minor groove side to completely enclose the DNA. The total encirclement of the DNA raises the fundamental question of how the DNA enters the binding cleft for catalysis. To address this question, we have determined the structure of BglII in its unbound state. The structure was solved by a combination of multiwavelength anomalous diffraction (MAD) (Fig. 1a) and molecular replacement methods and refined to 2.3 Å resolution (Table 1). The structure shows that rather than a simple opening of the dimer to accommodate the DNA substrate, the BglII subunits rearrange with respect to each other through a large scissor-like motion. The active site residues are oriented completely differently in the free enzyme compared to the complex, as are the alpha-helices at the dimer interface. These surprising changes in quaternary and tertiary structure provide a basis for understanding the entry and hydrolysis of BglII DNA.

Figure 1. Overall structure.
Figure 1 thumbnail

a, A stereo view of a section of the experimental electron density map from MAD phasing. The map is contoured at 1 sigma (gray) and 5 sigma (magenta), highlighting the positions of two of the selenium sites. The refined BglII structure is superimposed on the map. b, A view of the free BglII dimer with the two-fold axis running vertically. Secondary structural elements, along with the N-terminus and C-terminus, are labeled on one monomer. Loops A−E are labeled on the two-fold related monomer. Loops A and D and a part of loop E are disordered in the free enzyme, and are drawn with dotted lines, corresponding to the conformation seen in the enzyme−DNA complex14. c, Comparison of free and DNA bound dimers. The free enzyme (top) is shown in stereo, with its right subunit in the same orientation as the right subunit in the complex (bottom). The complex is viewed in stereo looking down the DNA axis (bottom). Based on this alignment, the left subunit is offset by as much as approx100° into the plane of the page (or parallel to the DNA axis). Blue and yellow spheres mark the respective positions of Lys 188 and Asn 98 in the free and bound dimers (see text). The figure was produced with Molscript39 and Raster 3D40.

Full FigureFull Figure and legend (119K)
Table 1. Data collection, phasing and refinement statistics
Table 1 thumbnail

Full TableFull Table
Overall fold
The BglII alpha/beta core is similar to that of BamHI, containing a mixed six-stranded beta-sheet (beta1, beta3, beta4, beta5, beta6 and beta7) surrounded by five alpha-helices (alpha1, alpha2, alpha3, alpha4 and alpha5), two of which (alpha4 and alpha5) mediate dimerization (Fig. 1b). The loops preceding these dimerization helices (loops B and C) carry the residues that contact bases in the major groove when BglII binds DNA14. BglII also has a beta-sandwich (beta2, beta8, beta9, beta10 and beta11) domain outside of the alpha/beta core, which extends the size of the enzyme relative to BamHI. In the complex, several loops (loops A, D and E) reach outward from the beta-sandwich domain to grip the DNA14. In particular, loops A and D contain the residues Tyr 190, Arg 189, and Asp 38, whose two-fold symmetric interaction at the minor groove is responsible for the enclosure of the DNA. Strikingly, in the free enzyme, both of these loops as well as loop E (Thr 207−Lys 216) are disordered (Fig. 1b,c). This was not entirely unexpected, however, as loops A, D, and E are positioned by the DNA in the complex and they must move to allow the DNA to bind, and also be flexible enough to reach around to the minor groove side of the DNA.

Scissor-like motion
The structure of BamHI has been solved in the absence and in the presence of DNA3, 4. In the free BamHI dimer, the subunits move outward in relation to the complex, to allow the DNA to enter. In the complex, the subunits clamp onto the DNA by a approx10° rotation around the DNA axis. Because of the similarity of the BglII and BamHI structures, we expected to see a similarly opened BglII enzyme. However, this opening turns out to be in a direction parallel, rather than perpendicular, to the bound DNA axis (Fig. 1c). Thus, rather than a simple unclamping−clamping motion, BglII opens by a scissor-like motion to allow the entry of DNA (Fig. 1c). Each monomer swings by as much as approx50°, like the blades of a pair of scissors, to open and close the binding cleft. The sheer magnitude of this motion is reflected by the dramatic increase in distances across the binding cleft. For instance, the distance between symmetrically related Lys 188 residues at the rim of the cleft changes from approx17 Å in the complex to an extraordinary approx61 Å in the free enzyme (Fig. 1c). Even at the base of the cleft where the effects of the scissor-like motion are expected to be minimal, the distance between symmetry related Asn 98 residues changes from approx9 Å in the complex to approx22 Å in the free enzyme (Fig. 1c).

Pushing the lever
One of the key elements in the change from free to bound forms is a set of residues (Asn 69−Asp 84) that we call the 'lever'. These residues, which make up parts of strands beta3 and beta4 and the loop connecting them, show the largest local conformational changes between the free and DNA bound monomers (Fig. 2a). In the enzyme−DNA complex, this lever is in the 'up' position, forming rather curved strands that add significant twist to the large central beta-sheet (Fig. 2a). With the lever in this up position, the enzyme not only makes extensive dimer contacts but it has a properly shaped DNA binding cleft. Moreover, the active site residues are exposed and ordered for catalysis. In contrast, in the free enzyme the lever is in the 'down' position (Fig. 2a). To achieve this, strands beta3 and beta4, rather than twisting up, continue in a more standard beta-sheet conformation. Strand beta5 becomes extended by several residues in order to contact the more 'regular' strands beta3 and beta4, making the central beta-sheet significantly larger than in the complex. In this down position, the lever residues fill the space where the active site residues contact the DNA in the specific complex.

Figure 2. Comparison of free and DNA bound monomers to show the lever motion.
Figure 2 thumbnail

a, The lever segment (magenta) and helix alpha4 (blue) undergo the largest local conformational change. The lever projects downward in the free enzyme (left) but upward in the complex (right). The length of helix alpha4 also change in going from free to the complexed state. b, A close up of the active site residues. In the free enzyme (left), the catalytic residues Asn 69, Asp 84, Glu 93 and Gln 95 are sequestered by extensive intramolecular hydrogen bonds with residues Asn 105 and Arg 108 from helix alpha4. Shown also is the position of a putative acetate ion. In the complex (right), several of these hydrogen bonds are broken, and residues reorient to form the active site. Shown also as a ball-stick tetrahedron is the position of the scissile phosphate group in the complex.

Full FigureFull Figure and legend (57K)
Concomitant with this switch in the position of the lever is the motion of helix alpha4. In the complex, this is one of two helices (alpha4 and alpha5) involved in forming the four-helix bundle at the dimer interface14, and residues Leu 111 and Ser 115 at the top of this helix contact residues Phe 76 and Leu 79 in the up-positioned lever. The loop connecting helix alpha4 to strand beta6 (residues Lys 114−Glu 122) is also well defined at the top of the helix. In the free enzyme, this area is completely lost; with the lever in the down position, the loop becomes disordered and helix alpha4 shortens by about one turn (Fig. 2a). More importantly, the helix moves by approx3 Å towards the other dimerization helix alpha5 and is rotated by about one-third of a helical turn, effectively reregistering the helix and leading to a reorganization of the dimer interface (described below). The root mean square (r.m.s.) deviation between the Calpha atoms of free and DNA bound monomers is only 1.08 Å without the lever and helix alpha4 (166 atoms), compared to 4.64 Å in the presence of the lever and helix alpha4 (196 atoms). Also, there is little indication of any hinge motion between the alpha/beta core and the beta-sandwich domain in the two structures.

Active site residues
The most important effect of raising the lever is to expose the active site residues Asn 69, Asp 84, Glu 93 and Gln 95 (Fig. 2b). Asn 69 and Asp 84 lie precisely at the pivot points of the lever, and the raising of the lever results in the reorientation of these residues from pointing into the protein core in the free enzyme to being solvent exposed and available for catalysis in the complex (Fig. 2b). For instance, in the free enzyme Asp 84 is co-opted by a salt bridge formed with Arg 108 from helix alpha4, but in the complex this salt bridge is broken and Asp 84 is reoriented into the binding cleft. Similarly, Gln 95 and Glu 93 point into the protein core in the free enzyme and are involved in hydrogen bonds with Asn 105 and Arg 108 at the top of helix alpha4. When the lever flips up, these side chains rotate and together with Asp 84 and Asn 69 coalesce to form the ordered active site (Fig. 2b). Gln 95 is important in positioning the nucleophilic water for attack on the scissile phosphodiester in the complex, while Asp 84 plays a crucial role in coordinating a Mg2+ in the active site14. This unusual rearrangement of the active site residues is coupled closely to the rotation and translation of helix alpha4, which moves in the direction of Gln 95 and Glu 93 to maintain some of the hydrophilic interactions. For instance, the Nalt epsilon of Gln95 remains hydrogen bonded to Asn 105 (which moves almost 4 Å) on helix alpha4, while Glu 93 retains one hydrogen bonding contact with Arg 108 (Fig. 2b).

Structural reorganization of the four-helix bundle
The dimer interface is completely different in the free enzyme. The dimerization helix pairs, alpha4(A)−alpha4(B) and alpha5(A)−alpha5(B), are much more angled, rather than aligned to form the parallel four-helix bundle seen in the complex14. In the complex, the dimer interface extends symmetrically along the entire length of helix alpha4 as well as most of helix alpha5. This interface starts with the Asn 98(A)−Asn 98(B) hydrogen bonds at the base of helix alpha4, and rises through the helix with mostly hydrophobic van der Waals contacts. Several of these hydrophobic residues (for example, Pro 100, Leu 103, and Val 107) contact symmetrically equivalent residues from helix alpha4 of the other monomer. There are also a series of alpha4−alpha5 intermolecular contacts at the top of this interface as well as a few contacts between alpha5 and the raised lever residues around Phe 78. Together, these contacts bury approx1,160 Å2 of surface area from each monomer in the complex.

In contrast, the four-helix bundle is almost nonexistent in the free enzyme (Fig. 1b,c). Only approx640 Å2 of surface area from each monomer is involved in the dimer interface (as calculated by the program GRASP17). The rotation of helix alpha4 contributes to this completely different dimer interface by exposing a different pattern of residues. Correspondingly, there are no direct hydrogen bonds across the interface, and the few hydrophobic contacts are confined mainly between the bottoms of helices alpha4 (for example, residues Tyr 99, Pro 100 and Leu 103) and the tops of helices alpha5 (for example, Ser 154, Leu 155 and Tyr 158). This buried surface area is at the low end of the range reported by Janin et al.18 for homodimers, varying from 670 Å2 for superoxide dismutase to 4,890 Å2 for citrate synthase. A more recent analysis of protein−protein interactions calculates buried areas as low as 368 Å2 for homodimers19, and recent crystal structures of human protein kinase CK2 and human survivin dimers show buried areas of <550 Å2 (refs 20,21). We have determined the oligomeric state of BglII in solution using dynamic light scattering, and found it to be a dimer, in accordance with the crystal structure.

Type II restriction enzymes are characterized by a highly complementary protein−DNA interface, which has made them a paradigm for the study of events leading to protein−DNA recognition. The DNA in almost all cases is accommodated in a tight binding cleft10, 11, which prompts the question of what conformational changes the protein and the DNA undergo upon binding. The free BglII structure reported here shows surprising differences in both tertiary and quaternary structure that provide a basis for understanding how the DNA enters the cleft for hydrolysis.

Within a single BglII monomer, we find a lever-type motion of residues Asn 69−Asp 84. When the lever is in the 'down' position in the free enzyme, the central beta-sheet is larger but less twisted than in the complex. Remarkably, the active site residues are sequestered, and only become exposed for catalysis when the lever is raised to the 'up' position in the complex. This sequestering of active site residues is different from other type II restriction endonucleases (for example, BamHI, EcoRV and PvuII) as well as homing endonucleases (for example, I-CreI and I-PpoI), in which most of the active site residues face the solvent in the free proteins1, 3, 5, 22, 23. The closest analogy in terms of sequestering of active site residues is with the type IIs endonuclease FokI7. FokI is composed of two separate domains, one for DNA recognition and one for nonspecific cleavage. The cleavage domain is sequestered by the recognition domain, but is thought to swing over to the distant cleavage sites on specific DNA binding7. For FokI, it has been suggested that the sequestering of the cleavage domain is a mechanism to avoid hydrolysis at nonspecific DNA sites7. It is tempting to think that the sequestering of active site residues in the free BglII structure fulfills a similar function of preventing accidental cleavage.

At a quaternary level, the BglII monomers rearrange with respect to each other through an unprecedented scissor-like motion. This has the same effect of opening the binding cleft as in free BamHI3, but the motion of the subunits is in a direction parallel rather than perpendicular to the DNA axis. Because of the total encirclement of the DNA in the complex, it was clear from the outset that the BglII monomers would have to undergo a large motion to loosen their grip on the DNA. The 'tongs-like' motion seen in BamHI creates the problem of 'squeezing' the dimerization interface, which the scissor-like motion avoids by essentially allowing the dimerization helices to slide past each other. Interestingly, PvuII also completely encircles its DNA6, but it does not undergo the scissor-like motion seen here. Instead, PvuII opens by the tongs-like motion, which is possible because each monomer is organized into dimerization and DNA binding subdomains5, 6, and the hinge between them allows the DNA binding subdomains to move outward without any squeezing of the dimer interface. This is reminiscent of transcription factor NF-kappaB, which is also divided into dimerization and specificity subdomains24, 25, with the specificity subdomain changing configuration on different DNA sites26. All in all, BglII is the first endonuclease to show alternative modes of dimerization, going from a parallel four-helix bundle in the complex to a more wedge-shaped arrangement of alpha-helices in the free enzyme. Although a parallel four-helix bundle is a common dimerization motif in DNA binding proteins, ranging from endonucleases such as BamHI3, 4 and EcoRI12 to bHLH transcription factors such as Max27 and E47 (refs 28,29), a similar reorganization of the bundle has not been previously described. However, the plasticity we see within the BglII four-helix bundle may extend to other four-helix bundle systems in which the subunits undergo a large conformational rearrangement, such as the binding of apolipoprotein E to phospholipid30 or the fusion of vesicles to target membranes31.

Based on the free and DNA bound structures, we can propose a scenario for the events that might occur when BglII encounters DNA. In the free state, BglII is a loose dimer with a cleft wide enough to accommodate DNA. When we model B-DNA into the cleft between the two monomers (Fig. 3), the distance between the recognition residues (Asn 140, Ser 141, Ser 97 and Asn 98) and the edges of bases in the major groove is >12 Å, with little possibility for making the hydrogen bonds we see in the specific complex14. To 'sense' the DNA, there may be an intermediate configuration in which BglII undergoes a partial scissor-like motion around nonspecific DNA, as it scans the genome for the cognate sequence. An analogy would be the intermediate configuration seen in the BamHI−noncognate DNA structure, in which BamHI adopts a configuration that is on the pathway between free and specific DNA bound forms of the enzyme32. In BglII, a full upward motion of the lever segment may only be realized when the enzyme encounters its cognate sequence, resulting in the rotation of the catalytic residues into the active site, and the repositioning of helix alpha4 for a more hydrophobic dimer interface. As the protein completes the scissor-like motion around the cognate DNA (Fig. 3), base pairs at the center of the recognition sequence would undergo the unwinding (>15°) seen in the BglII−DNA complex14. The free energy lost in breaking some of the intramolecular hydrogen bonds (such as Asp 84−Arg 108 and Glu 93−Arg 108) is probably gained through the formation of new specific bonds with the DNA as well as new dimer contacts. Together, the two BglII structures provide a new basis for understanding how an enzyme can capture and hydrolyze its DNA substrate.

Figure 3. A view of the scissor-like motion.
Figure 3 thumbnail

The complex (right) is shown with the DNA aligned horizontally. The two protein monomers are colored blue and purple, respectively. The free enzyme (left) is shown with a DNA modeled between the two monomers. The two monomers move symmetrically in opposite directions along the DNA axis.

Full FigureFull Figure and legend (25K)
Selenomethionyl (SeMet) BglII was prepared as reported14. The protein was concentrated to 10 mg ml-1 in 10 mM Tris pH 7.4, 100 mM KCl, 10% (v/v) glycerol, and 1 mM dithiothreitol. A 0.75 mul drop of protein was combined with 0.5 mul of mother liquor containing 30% (v/v) isopropanol, 0.2 M sodium acetate, and 100 mM Bis-tris at pH 6.0, and equilibrated at 4 °C against 1 ml of the same solution in a hanging drop plate. Crystals appeared overnight and reached their maximum size after several days. Before data collection, crystals were transferred to a solution of mother liquor containing 15% (v/v) glycerol for flash freezing.

Dynamic light scattering (DLS).
The measurements were performed on a DynaPro 801TC instrument (Protein Solutions, Inc.), using BglII at 1.6 mg ml-1 concentration. The molecular mass was approx51.3 kDa, close to the calculated dimeric size of approx51.5 kDa.

Data collection and structure determination.
An initial data set to 2.6 Å resolution was collected from a SeMet BglII crystal, using a R-AXIS IV image plate detector installed on a rotating anode X-ray generator (Table 1). The data were processed and reduced using DENZO and SCALEPACK33. The space group was determined to be P3(2)21 with cell dimensions a = b = 79.1 Å, c = 66.5 Å, alpha = beta = 90°, gamma = 120° and one monomer in the asymmetric unit. One monomer from the BglII−DNA complex (PDB code 1D2I) minus residues 36−47 (loop A), 185−194 (loop D), and 206−217 (loop E) was used as a starting model for molecular replacement (MR). The MR was carried out with the program AMoRe34 resulting in a top solution with a correlation coefficient of 0.455. This top solution was subjected to rigid body, positional, and B-factor refinement at 3.0 Å resolution, using the program CNS35, which resulted in an R-factor of 0.370 (Rfree 0.497) and gave decent electron density maps. Based on these maps, residues 71−84 and 98−124 were also removed from the model. Several cycles of simulated annealing, positional and B-factor refinement, gradual increases in the resolution to 2.6 Å, and the addition of several more residues and waters using the program O36, resulted in a structure with an R-factor of 0.228 (Rfree 0.282).

At this point, a MAD experiment was performed on the same SeMet BglII crystal, at beamline X25 at the Brookhaven National Synchrotron Light Source (NSLS; Table 1). At this source, the crystal diffracted to 2.3 Å resolution, and an X-ray fluorescence scan was obtained to determine the three wavelengths for use in MAD data collection. About 60° of data were collected at each of the three wavelengths, corresponding to the Se edge, Se peak and a remote point. Three Se sites taken from the refined MR solution, which were verified by anomalous difference Patterson maps, were input into SHARP37, using the data measured at the remote point as a reference. After density modification with SOLOMON38 a new experimentally phased map was calculated. The refined MR model fit the experimental map well. Loops A, D, and E and the loop connecting helix alpha4 and strand beta5 were not visible in the experimental map either, indicative of their disorder. The model was further refined against the remote point data. Simulated annealing, followed by several rounds of fitting and refinement, resulted in the final structure containing residues 1−37, 45−114, 123−189, 192−206, 217−223, 81 waters, and one acetate ion with a final R-factor of 0.214 (Rfree 0.258).

Coordinates have been deposited in the Protein Data Bank (accession number 1ES8).

Received 22 August 2000; Accepted 16 November 2000

  1. Winkler, F.K. et al. EMBO J. 12, 1781−1795 (1993). | PubMed | ChemPort |
  2. Newman, M., Strzelecka, T., Dorner, L.F., Schildkraut, I. & Aggarwal, A.K. Structure 2, 439−452 (1994). | Article | PubMed | ISI | ChemPort |
  3. Newman, M., Strzelecka, T., Dorner, L.F., Schildkraut, I. & Aggarwal, A.K. Nature 368, 660−664 (1994). | Article | PubMed | ISI | ChemPort |
  4. Newman, M., Strzelecka, T., Dorner, L., Schildkraut, I. & Aggarwal, A.K. Science 269, 656−663 (1995). | PubMed | ISI | ChemPort |
  5. Athanasiadis, A. et al. Nature Struct. Biol. 1, 469−475 (1994). | Article | PubMed | ISI | ChemPort |
  6. Cheng, X., Balendiran, K., Schildkraut, I. & Anderson, J.E. EMBO J. 13, 3927−3935 (1994). | PubMed | ISI | ChemPort |
  7. Wah, D.A., Hirsch, J.A., Dorner, L.F., Schildkraut, I. & Aggarwal, A.K. Nature 388, 97−100 (1997). | Article | PubMed | ISI | ChemPort |
  8. Wah, D.A., Bitinaite, J., Schildkraut, I. & Aggarwal, A.K. Proc. Natl. Acad. Sci. USA 95, 10564−10569 (1998). | Article | PubMed | ChemPort |
  9. Roberts, R.J. & Halford, S.E. In Nucleases (eds., Linn, S.M., Lloyd, R.S. & Roberts, R.J.) (Cold Spring Harbor, New York; 1993).
  10. Aggarwal, A.K. Curr. Opin. Struct. Biol. 5, 11−19 (1995). | Article | PubMed | ISI | ChemPort |
  11. Pingoud, A. & Jeltsch, A. Eur. J. Biochem. 246, 1−22 (1997). | Article | PubMed | ISI | ChemPort |
  12. Kim, Y., Grable, J.C., Choi, P.J., Greene, P. & Rosenberg, J.M. Science 249, 1307−1309 (1990). | PubMed | ISI | ChemPort |
  13. Deibert, M., Grazulis, S., Janulaitis, A., Siksnys, V. & Huber, R. EMBO J. 18, 5805−5816 (1999). | Article | PubMed | ChemPort |
  14. Lukacs, C.M., Kucera, R., Schildkraut, I. & Aggarwal, A.K. Nature Struct. Biol. 7, 134−140 (2000). | Article | PubMed | ISI | ChemPort |
  15. Perona, J.J. & Martin, A.M. J. Mol. Biol. 273, 207−225 (1997). | Article | PubMed | ISI | ChemPort |
  16. Anton, B.P. et al. Gene 187, 19−27 (1997). | Article | PubMed | ISI | ChemPort |
  17. Nicholls, A., Sharp, K. & Honig, B. Proteins 11, 281−296 (1991). | PubMed | ISI | ChemPort |
  18. Janin, J., Miller, S. & Chothia, C. J. Mol. Biol. 204, 155−164 (1988). | Article | PubMed | ISI | ChemPort |
  19. Jones, S. & Thornton, J.M. Proc. Natl. Acad. Sci. USA 93, 13−20 (1996). | Article | PubMed | ChemPort |
  20. Chantalat, L. et al. EMBO J. 18, 2930−2940 (1999). | Article | PubMed | ChemPort |
  21. Chantalat, L. et al. Mol. Cell 6, 183−189 (2000). | Article | PubMed | ISI | ChemPort |
  22. Heath, P.J., Stephens, K.M., Monnat, R.J., Jr. & Stoddard, B.L. Nature Struct. Biol. 4, 468−476 (1997). | Article | PubMed | ISI | ChemPort |
  23. Galburt, E.A. et al. J. Mol. Biol. 300, 877−887 (2000). | Article | PubMed | ISI | ChemPort |
  24. Ghosh, G., Van Duyne, G., Ghosh, S. & Sigler, P.B. Nature 373, 303−310 (1995). | Article | PubMed | ISI | ChemPort |
  25. Muller, C.W., Rey, F.A., Sodeoka, M., Verdine, G.L. & Harrison, S.C. Nature 373, 311−317 (1995). | Article | PubMed | ISI | ChemPort |
  26. Muller, C.W., Rey, F.A. & Harrison, S.C. Nature Struct. Biol. 3, 224−227 (1996). | Article | PubMed | ISI | ChemPort |
  27. Ferre-D'Amare, A.R., Prendergast, G.C., Ziff, E.B. & Burley, S.K. Nature 363, 38−45 (1993). | Article | PubMed | ISI | ChemPort |
  28. Ellenberger, T., Fass, D., Arnaud, M. & Harrison, S.C. Genes Dev. 8, 970−980 (1994). | PubMed | ISI | ChemPort |
  29. Fairman, R., Beran-Steed, R.K. & Handel, T.M. Protein Sci. 6, 175−184 (1997). | PubMed | ISI | ChemPort |
  30. Lu, B., Morrow, J. & Weisgraber, K.H. J. Biol. Chem. 275, 20775−20781 (2000). | Article | PubMed | ISI | ChemPort |
  31. Sutton, R.B., Fasshauer, D., Jahn, R. & Brunger, A.T. Nature 395, 347−353 (1998). | Article | PubMed | ISI | ChemPort |
  32. Viadiu, H. & Aggarwal, A.K. Mol. Cell 5, 889−895 (2000). | Article | PubMed | ISI | ChemPort |
  33. Otwinowski, Z. & Minor, W. Methods Enzymol. 276, 307−326 (1997). | Article | PubMed | ISI | ChemPort |
  34. Navaza, J. In Molecular replacement (eds., Dodson, E.J., Gover, S. & Wolf, W.) (Science and Engineering Research Council, Daresbury Laboratory, Warrington, UK; 1992).
  35. Brunger, A.T. et al. Acta Crystallogr. D 54, 905 (1998). | PubMed | ChemPort |
  36. Jones, A.T., Zou, J.Y., Cowan, S.W. & Kjeldgaard, M. Acta Crystallogr. A 47, 110−119 (1991). | Article | PubMed | ISI |
  37. Fortelle, d.L. & Bricogne, G. Methods Enzymol. 276, 472−494 (1997).
  38. Abrahams, J.P. & Leslie, A.G.W. Acta Crystallogr. D 52, 32−42 (1996). | Article |
  39. Kraulis, P. J. Appl. Crystallogr. 24, 946−950 (1991). | Article | ISI |
  40. Merritt, E.A. & Murphy, M.E.P. Acta Crystallogr. D 50, 869−873 (1994). | Article | PubMed | ISI | ChemPort |
We thank L. Berman and H. Lewis for facilitating data collection at NSLS. A.K.A. is supported by a grant from the NIH, and C.M.L. is supported by a Cancer Research Fund of the Damon Runyon-Walter Winchell Foundation Fellowship.

Previous | Next
Table of contents
Download PDFDownload PDF
Send to a friendSend to a friend
Save this linkSave this link


Figures & Tables
Export citation
Export references

Search buyers guide:

Nature Structural & Molecular Biology
ISSN: 1545-9993
EISSN: 1545-9985
Journal home | Advance online publication | Current issue | Archive | Press releases | Supplements | For authors | Online submission | Permissions | For referees | Free online issue | About the journal | Contact the journal | Subscribe | Advertising | work@npg | naturereprints | About this site | For librarians
Nature Publishing Group, publisher of Nature, and other science journals and reference works©2001 Nature Publishing Group | Privacy policy