• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. Sep 27, 2005; 102(39): 13801–13806.
Published online Sep 16, 2005. doi:  10.1073/pnas.0506430102
PMCID: PMC1215310

Out-of-plane motions in open sliding clamps: Molecular dynamics simulations of eukaryotic and archaeal proliferating cell nuclear antigen


Sliding clamps are ring-like multimeric proteins that encircle duplex DNA and serve as mobile DNA-bound platforms that are essential for efficient DNA replication and repair. Sliding clamps are placed on DNA by clamp loader complexes, in which the clamp-interacting elements are organized in a right-handed spiral assembly. To understand how the flat, ring-like clamps might interact with the spiral interaction surface of the clamp loader complex, we have performed molecular dynamics simulations of sliding clamps (proliferating cell nuclear antigen from the budding yeast, humans, and an archaeal species) in which we have removed one of the three subunits so as to release the constraint of ring closure. The simulations reveal significant structural fluctuations corresponding to lateral opening and out-of-plane distortions of the clamp, which result principally from bending and twisting of the β-sheets that span the intermolecular interfaces, with smaller but similar contributions from β-sheets that span the intramolecular interfaces within each subunit. With the integrity of these β-sheets intact, the predominant fluctuations seen in the simulations are oscillations between lateral openings and right-handed spirals. The tendency for clamps to adopt a right-handed spiral conformation implies that once opened, the conformation of the clamp can easily match the spiraling of clamp loader subunits, a feature that is intrinsic to the recognition of DNA and subsequent hydrolysis of ATP by the clamp-bound clamp loader complex.

Keywords: clamp loader, DNA replication, DNA sliding clamps, β-sheet distortion, conformational transitions

Sliding clamps are ring-shaped proteins that can encircle duplex DNA and remain topologically linked to it (15). The sterically loose but topologically tight linkage to DNA enables sliding clamps to provide a mobile platform that allows DNA polymerases to achieve very high processivity and allows various associated enzymes to hop on and off the DNA substrate as required (6, 7). The architecture of sliding clamps is conserved throughout evolution, with each clamp consisting of six similarly folded domains arranged in a flat circle around a large central hole that is lined by 12 α-helices (Fig. 1 and refs. 1 and 812).

Fig. 1.
Overview of PCNA architecture. (A) Crystal structure of PCNA from S. cerevisiae (PDB ID code 1PLQ) (8). The structure consists of three subunits, each with two domains connected by a long linker. In the molecular dynamics simulation, one subunit was removed. ...

In prokaryotes, each clamp subunit (the β-subunit of DNA polymerase III, referred to as β) contains three domains, with a complete six-domain clamp generated by a dimer of β-subunits (1). In eukaryotes and archaea, the clamp is known as the proliferating cell nuclear antigen (PCNA), and, in contrast to dimeric β, PCNA is an assembly of three subunits, each with two domains (8, 9, 11). The efficient placement of sliding clamps on DNA requires the action of five-subunit ATP-dependent complexes known as clamp loaders, which bind to and stabilize the open forms of the sliding clamps in the presence of ATP (for reviews, see refs. 13 and 14). Much of our understanding of the structural basis for clamp opening has come from study of the Escherichia coli system, where it has been shown that the clamp loader loads the clamp by opening one interface in the β-subunit dimer (15).

The crystal structure of a β-subunit monomer bound to an isolated clamp loader subunit (called δ) reveals a more “relaxed” curvature for the clamp monomer (16). Molecular dynamics simulations of the monomeric β-subunit that started with the “closed” conformation revealed a spontaneous and rapid (≈1 ns) lateral opening of the clamp, resulting in a relaxation to a conformation similar to that seen in the crystal structure of the δ:β complex (16). These results suggest that β-subunits in the closed-ring form of the dimer are under tension and that the disruption of one intersubunit interface results in the relaxation of the structure into a more open form that would allow the entry of DNA into the inner region of the sliding clamp (16).

In the crystal structure of the yeast clamp loader complex replication factor C (RFC) bound to PCNA, PCNA is closed and has the same flat, disk-like organization that is observed in the crystal structures of isolated clamps (17). The right-handed spiral organization of the clamp loader subunits suggested a mechanism for DNA recognition, because the pitch of the RFC spiral is approximately congruent with the helical geometry of duplex DNA (17). The general features of this mode of DNA recognition, in which the nucleotide-binding domains of the clamp loader wrap around double-helical DNA that is threaded through the center of the complex, have been confirmed recently by mutagenesis of the E. coli clamp loader complex (18).

That the clamp bound to the clamp loader is in a closed form in the crystal structure of the RFC–PCNA complex came as a surprise, because the engagement of an ATP-bound clamp loader with the clamp is known to result in clamp opening (15, 19, 20). The crystal structure may correspond to a state of the system just before the release of the clamp on DNA. The structure of the E. coli clamp loader determined in the absence of the clamp shows that it also adopts a spiral structure, although it is one in which the nucleotide-binding domains are more spread apart than in the RFC–PCNA complex (17, 21, 22).

Clamp loaders are AAA+ ATPases (23), and the ATP-loaded forms of many other AAA+ ATPases, such as N-ethylamide-sensitive fusion protein (NSF), are flat, circular structures that are similar in overall shape and dimensions to PCNA (24, 25). It is natural to wonder whether the clamp loader complex flattens out into an NSF-like structure to fully engage the PCNA clamp. Simple modeling suggests that if the nucleotide-binding domains were to flatten out, then the integrity of the C-terminal collar of the clamp loader would be compromised. How, then, does the clamp loader engage the clamp before ATP hydrolysis and release on DNA?

We now report the results of a series of molecular dynamics simulations of yeast, human, and archaeal PCNA in fully solvated environments. Our results suggest that when it is not constrained to be a closed ring, the structure of PCNA oscillates spontaneously between planar and nonplanar conformations, with a marked tendency for right-handed spiral distortions. It seems that once an intermolecular interface in PCNA is broken, either spontaneously or by intervention of the clamp loader, the PCNA molecule can adopt a right-handed spiral conformation that would bind with the entire distal surface of the RFC assembly. These results are consistent with recent electron microscopic reconstructions of an archaeal RFC–PCNA complex (26).


Molecular dynamics trajectories were generated by using a version of amber 7.0 that is optimized for the calculation of electrostatics using the particle mesh Ewald summation method (pmemd) (27, 28). The PARM98 force field was used for all calculations (29). The following x-ray crystal structures were used as the initial structures for the simulations: PCNA from budding yeast, Saccharomyces cerevisiae (Protein Data Bank ID code 1PLQ) (8), human PCNA (PDB ID code 1AXC) (9), and PCNA from the archaebacterium Pyrococcus furiosus (PDB ID code 1GE8) (11).

The crystal structure of yeast PCNA contains an intact molecule, which was used in the simulations. The crystal structure of human PCNA was determined in complex with an inhibitor peptide, which was removed before initiating any calculations. Missing residues (at 107–108 and 186–191 in chain C and residues 189–190 in chain E) were modeled based on yeast PCNA. The C-terminal residues (256–261) in both chains of human PCNA were disordered in the crystal structure and were not included in the simulations. Residues 1 and 248–249 (C-terminal segment) are disordered in the archaeal PCNA and not included in the crystal structure model or the molecular dynamics simulations. Residues 118–125 in the long extended linker between domains were also disordered in the crystal structure; this region was modeled based on the structure of yeast PCNA.

Coordinates were prepared for molecular dynamics by deleting one subunit to form PCNA dimers and then building hydrogen atoms and adding counter ions to neutralize the total charge. The protonation state of side chains corresponded to that expected for isolated residues at pH 7.0. The systems were solvated by using a box of TIP3P water molecules (30) such that the waters extended 9–15 Å from the protein molecule, depending on the specific simulation. Some waters were substituted with Na+ and Cl ions to reflect a 0.15 M ionic strength solvent environment (Table 1, which is published as supporting information on the PNAS web site).

Energy calculations used periodic boundary conditions and the calculation of electrostatic energy using particle mesh Ewald summations (31). After 1,500 steps of initial energy minimization, the system was heated to 298 K over a period of 15 ps. Harmonic positional restraints were placed on the protein atoms and reduced gradually over the initial 55 ps of the simulation. The remainder of the molecular dynamics simulation was carried out in the absence of positional restraints, under constant temperature and constant pressure conditions. A 2-fs integration time step was used, with the SHAKE algorithm (32) applied to maintain covalent bonds that include hydrogens. Different random number seeds were used for the initial velocity assignments for each simulation. The archaeal I and human I simulations (Table 1) were generated by using 32 processors on an SP2 computer (IBM, White Plains, NY) at the National Energy Research Scientific Computing Center at Lawrence Berkeley National Laboratory. The other simulations were calculated in-house on two to four processors each of an IBM cluster using 3.06-GHz Xeon processors (Intel, Santa Clara, CA) and a gigabit ethernet switch. On in-house computers, these simulations required ≈12 h of computational time for 100 ps. The results of the simulations were analyzed by using charmm-29 (33), vmd (34), and pymol (35). Principal component (quasiharmonic normal mode) analyses were carried out by using standard procedures in charmm-29 (33). Using the same approach, a 2.75-ns simulation of the intact yeast PCNA trimer was also calculated.

Results and Discussion

General Strategy. To see what happens to the structure of PCNA if it is no longer constrained to be a closed ring, we removed one of the three subunits that constitute the PCNA trimer and then generated molecular dynamics trajectories for the dimer. We have preferred to use this approach over artificially opening the trimeric ring by applying a driving force, because we know little about the molecular steps involved in ring opening and do not have crystal structures of the open PCNA. The T4 bacteriophage clamp has been shown to be an open trimer in solution, indicating that each of the remaining intermolecular interfaces is stable (36). The bacterial clamp has been shown to open at only one interface in the presence of the clamp loader, which again indicates that some stability is retained at each remaining interface when the ring is opened (15).

The intermolecular interfaces in PCNA are held together by the formation of a β-sheet that extends across the molecular boundary (Fig. 1). By monitoring the hydrogen bonds formed by the two β-strands that are at the intermolecular interface, we confirmed that the β-sheet remains intact during the course of the simulations (Fig. 6, which is published as supporting information on the PNAS web site). Another indication that the dimers are stable during the simulations is that over the 10-ns simulation period, there are several oscillations through the conformational changes that are the focus of discussion below.

General Aspects of the Conformational Changes. We focus most of our analysis on a 10.5-ns molecular dynamics trajectory of a yeast PCNA dimer (yeast I). The structure undergoes a large conformational change over the course of the simulation, with the rms deviation in Cα atom positions from the starting structure peaking at >6 Å at ≈7.8 ns (omitting the three flexible C-terminal residues) (Fig. 2A). This large deviation from the starting structure is mainly due to rigid body-like displacements between the individual domains and subunits, because the fluctuations in the structures of the individual domains are much smaller (Fig. 2A). For yeast I, the rms deviation in Cα positions averaged over the whole simulation is 1.4, 1.0, 1.4, and 1.3 Å for domains I, II, III, and IV, respectively (see Fig. 1 for the notation used to identify the domains).

Fig. 2.
Comparisons between simulated structures and the starting structure reveal rigid body motions between domains. (A) rms deviation in Cα positions from the starting structure for the yeast I simulation. The large deviations over the entire dimer ...

The origin of the large change in the overall structure of the molecule during the course of the simulation can be understood by superimposing structures from the trajectory onto the starting structure, using only one domain (domain II) located at the intermolecular interface for the alignment. Once the instantaneous structures are aligned in this way, we calculate the rms deviation from the crystal structure for the two adjacent domains in the simulated structure without further realignment. The rms deviation in the Cα atoms of domain I (located in the same molecule) rise to ≈3 Å during the course of the simulation, with transient displacements of as much as 4 Å (Fig. 2B). This structural change corresponds to a rotation of domain II with respect to domain I but without breakage of the intramolecular β-sheet (Fig. 2C). The rms deviation in the positions of Cα atoms in domain III (located in the next molecule) with respect to the starting structure are about twice as large, rising to as much as 8–9 Å during transient excursions (Fig. 2B). Again, the structural changes correspond to a rotation of the two adjacent domains with respect to each other, without breakage of the intermolecular β-sheet (Fig. 2C). These changes in structure are a consequence of the removal of the ring closure constraint, because they are not evident in the simulation of trimeric PCNA (Fig. 7, which is published as supporting information on the PNAS web site).

Lateral Relaxation of the PCNA Ring. From visual inspection of the molecular dynamics trajectories, it was evident that the PCNA subunits had “sprung open,” as seen earlier in the β-monomer simulations (16). To determine the extent of the lateral opening, a trimer of PCNA was constructed artificially by using two copies of each instantaneous structure (Fig. 3A). The first copy was not manipulated, but the second copy was transformed so that domain II was matched onto domain IV of the first copy, thereby allowing domains III and IV of the transformed structure to occupy the position of the third subunit in a trimeric PCNA, generating domains V and VI artificially. This technique creates an artificial ring in which the fluctuations of domains V and VI are perfectly correlated with those of domains III and IV. In reality, the conformational changes between subunits are unlikely to be correlated. Nevertheless, the simulations do indicate that the observed conformational changes at the interfaces are thermally accessible.

Fig. 3.
Lateral and out-of-plane movements for PCNA simulations. (A) A trimer was constructed artificially from the simulated dimers, and the shortest distances between any atoms in domain I and an artificial domain VI (see Fig. 1) were calculated. (B) Using ...

The shortest distance between any atoms in domain I and the artificial domain VI was calculated for each trajectory. The gap distance between these domains as a function of time for the yeast I, human I, and archaeal I simulations are shown in Fig. 3A (time series for the other six simulations are shown in Fig. 8A, which is published as supporting information on the PNAS web site). In each of the yeast and human simulations, the trimeric structure oscillates through a series of openings and closings, with transient gaps of 20–30 Å between domains I and VI (Fig. 3A). A qualitatively similar pattern was seen in the trajectories for archaeal PCNA, although the gaps are generally smaller. In none of the simulations were the gaps wide enough for passage of duplex DNA. Clamp loaders require single-stranded or nicked DNA to load clamps, and, as has been suggested earlier for the E. coli β-clamp (16), this relatively narrow opening might mean that it is single-stranded DNA that is passed through the gap in PCNA during the first steps of the process of clamp loading. Alternatively, interaction with the clamp loader might further widen the gap in PCNA.

Out-of-Plane Movements of PCNA. We monitored the extent of nonplanarity in the structure by calculating the vertical displacement of the center of mass of domain I with respect to the plane of the PCNA ring in the starting structure. This analysis was done by first aligning the PCNA crystal structure in a principal axis frame such that the z axis is perpendicular to the plane of the trimeric ring. Each instantaneous structure in a trajectory was then aligned onto domain IV of the crystal structure, and the value of the z coordinate of the center of mass of domain I was calculated. Positive and negative values of the z coordinates correspond to right-handed and left-handed out-of-plane deformations, respectively. Time series for the vertical displacement of domain IV that were calculated in this way are shown in Fig. 3B for yeast I, human I, and archaeal I (time series for the other trajectories are shown in Fig. 8B, and results for yeast III are discussed later).

Each of the trajectories shown in Fig. 3B displays a similar behavior. The PCNA dimer rapidly becomes nonplanar and then oscillates back toward planarity, with each oscillation taking ≈0.2–1 ns to complete. The maximum vertical displacements of the center of mass of domain I are in the range of ≈10–15 Å for yeast I, ≈6–9 Å for human I, and ≈8–12 Å for archaeal I (Fig. 3B). For reference, in the structure of the yeast RFC complex, the vertical rise on moving from one subunit to the next is ≈5.5 Å on average, implying that if PCNA were to match the spiral structure of yeast RFC perfectly, then the vertical rise of domain I would be ≈16 Å. There are rare transient fluctuations in each of the simulations that come close to matching this vertical displacement.

β-Sheet Distortions, Particularly at the Intermolecular Interface, Drive the Conformational Changes. A distinctive feature of the structure of PCNA is the set of six antiparallel β-sheets that span the three intermolecular and three intramolecular interfaces (Fig. 1). The overall structures of the two kinds of sheets are very similar, with one difference being that the intramolecular sheet has packed against it a long linker segment that connects domain I to domain II, whereas the exterior face of the intermolecular β-sheet is unhindered.

We have analyzed the structural fluctuations of the β-sheets in the molecular dynamics trajectories of the PCNA dimer by using principal component analysis (quasiharmonic normal modes) (37). As is typically the case, the lowest frequency normal modes account for the bulk of the mean-square displacements of the atoms. The two lowest-frequency modes for the intermolecular β-sheet are illustrated in Fig. 4A. One mode corresponds primarily to a bending of the sheet, whereas the other corresponds to an oscillation in the degree of right-handed twist of the sheet. Visual inspection of molecular dynamics trajectories and comparison with normal modes calculated for the whole structure indicate that the bending mode underlies the lateral displacement of the PCNA molecule, whereas the twisting modes account for the out-of-plane distortions.

Fig. 4.
β-sheet distortions underlying the primary domain displacements of simulated PCNA. (A) The two lowest-frequency normal modes are displayed for the intermolecular β-sheet in the yeast I simulation. The lowest-frequency mode involves a bending ...

The normal modes for the intramolecular β-sheet are qualitatively similar to the modes for the intermolecular sheet, with the two lowest-frequency modes corresponding to bending and twisting of the sheet in both cases (Fig. 9, which is published as supporting information on the PNAS web site). The mean-square displacement of a mode is inversely related to the square of the normal mode frequency (37). The frequencies of the two lowest modes for the intermolecular sheet are 4.82 and 6.78 wavenumbers (cm–1), and the frequencies are 7.51 and 9.33 wavenumbers (cm–1) for the intramolecular sheet. These results show that the distortions of the intermolecular β-sheet make the dominant contribution to the overall displacements of the PCNA molecule, consistent with the analysis of the overall rms deviations of the domains (Fig. 2).

Our analysis of the β-sheet distortions in the molecular dynamics trajectories of PCNA is consistent with the results of principal component analysis of the variation in β-sheet conformation between different protein structures (38). Antiparallel β-sheets, such as the ones in PCNA, show considerably more variation in structure across different proteins, implying a greater degree of flexibility than for parallel β-sheets. The principal modes of deformation of β-sheets are global bending and twisting of the sheets (38). Interestingly, there is a noticeable anticorrelation between the extrema of the lateral and out-of-plane distortions of PCNA in all of the trajectories (a superposition of time series for the two displacements is shown in Fig. 10, which is published as supporting information on the PNAS web site). The maximum extent of lateral opening occurs when the PCNA molecule is more or less planar, and the extreme upward displacement of the domains into a spiral form occurs when the gap between domains I and VI in the artificial trimer is minimal. This phenomenon can be understood as arising from constraints placed on these motions by a presumed necessity to preserve the hydrogen-bonded network in the β-sheets, which would be torn by maximally exciting the bending and twisting modes simultaneously.

Out-of-Plane Distortions of PCNA Tend to Be Right-Handed. It is interesting to note that the simulations for yeast, human, and archaeal PCNA all show a pronounced tendency for right-handed out-of-plane distortions (Figs. (Figs.3B3B and 8B). For seven of the nine trajectories, there is little evidence for left-handed spiraling by PCNA, with the exceptions being yeast III and the final 0.4 ns of the human II trajectories, in which PCNA is mostly in a left-handed spiral conformation. Close examination of the yeast III and human II trajectories is informative in terms of understanding the basis for the preference for right-handed out-of-plane distortions in the other trajectories.

Each of the β-sheets in PCNA has a right-handed twist (39), and analysis of the trajectories reveals that the adoption by PCNA of right-handed spiral structures corresponds to an increase in the degree of right-handed twist in the β-sheets (Fig. 4B). To adopt the left-handed spiral conformation, the central intermolecular β-sheet in the yeast III trajectory has a reduced right-handed twist (i.e., the sheet can be described as undergoing an overall left-handed twist relative to the crystal structure). This reduction in right-handed twist cannot, apparently, be tolerated by the β-sheet, which tears between strands βH1 and βI1 (following the nomenclature of ref. 8; Fig. 4C). The dislocation of strands persists through the simulation, maintaining the left-handed spiral conformation. Interestingly, the same tearing of this β-sheet is observed in the final 0.4 ns of human II. We presume that such distortions have a significant energetic penalty, which is why it is seen in only two of nine simulations. The energetic consequences of β-sheet twisting are complex (4042), and a simple structure–energy correlation that describes the out-of-plane distortions in PCNA is not evident.

Transient Conformations of PCNA Allow Extensive Engagement with a Spiral RFC Complex. In the crystal structure of RFC bound to PCNA, the nucleotide-binding domains of RFC, which contain the clamp-interacting elements, are in a spiral arrangement, whereas the PCNA ring is planar (17). Matching the last subunit of the yeast I PCNA onto the corresponding subunit of PCNA in the RFC–PCNA crystal structure reveals how PCNA may interact with RFC when it is in its open conformation. Interestingly, instantaneous structures from the simulation in which PCNA is highly spiraled track the spiral conformation of RFC. In such structures, PCNA appears well placed for contacting the clamp-interacting elements of all five RFC subunits (Fig. 5).

Fig. 5.
Schematic shows how PCNA may open. In solution, PCNA is normally in a closed-ring form. Either spontaneously or through its initial interaction with RFC, the clamp opens. When domain VI of the simulation is superimposed on the corresponding domain in ...

Concluding Remarks. Crystal structures of PCNA reveal flat trimeric disks, whereas the clamp loader complexes appear to be spiral assemblies. We were therefore interested in understanding the extent to which PCNA is flexible once the ring is opened. There are no crystal structures available for an intact but open clamp, and so we elected to mimic the effects of clamp opening by removing one of the three subunits of the PCNA trimer. In this way, the simulations start with the native interfaces between subunits, and the manner in which the structure relaxes after the constraint of ring closure is released can be monitored.

Our simulations of yeast, human, and archaeal PCNA all show a spontaneous lateral relaxation of the subunits, as seen before for the E. coli β-clamp (16), as well as transient adoption of spiral conformations. The extent of the out-of-plane motions was surprising because such displacements are not apparent in crystal structures of disassembled β- or PCNA monomers (16, 43). In a previous simulation of the E. coli β-monomer, a spiraling of the structure was not observed for the three simulated domains (16). In contrast to the simulations presented here, there is no intermolecular β-sheet in the monomeric clamp. In this regard, it will be interesting to pursue additional simulation studies of the open form of bacterial and bacteriophage sliding clamps. It is natural, from considerations of entropy, for circular structures to go out of plane when the constraint of circularity is removed. A notable aspect of the simulation results is a clear preference to adopt right-handed rather than left-handed spiral conformations. The flexibility of the PCNA structure centers on the subunit interface between the two subunits. The native hydrogen-bonding pattern between the β-strands at the interface is maintained, and it seems that it is the intrinsic constraint on the degree of right-handed twist in antiparallel β-sheets that results in the tendency to assume a right-handed spiral conformation.

Our simulations do not, of course, address the question of how RFC opens PCNA, and this step in the process is poorly understood. In the case of the T4 sliding clamp, the structure appears to be predominantly open in solution (36). Although not likely to be the case for PCNA (44), transient fluctuations that lead to ring opening may still underlie the formation of an initial-encounter complex between RFC and PCNA. The simulations do suggest that the mechanical flexibility of PCNA appears to be more than sufficient for it to readily adapt itself into an extensive interaction with the clamp loader complex once the initial ring cleavage event occurs. Morikawa and coworkers (26) have recently extended the resolution of their negatively stained electron microscopic analysis of archaeal RFC complex bound to PCNA and DNA. This analysis has led to a reinterpretation of earlier results obtained at lower resolution (45) and now reveals a striking consistency with the general features of the x-ray crystallographic analysis of yeast RFC–PCNA (17). An important result is that the clamp is seen to be open in the electron microscopic reconstructions and fully engaged with all five subunits of the RFC complex in a spiral organization.

Supplementary Material

Supporting Information:


We thank Kosuke Morikawa and coworkers for sharing results before publication and Olga Kuchment, Randall McNally, and Eric Goedken for helpful discussions. This work was supported in part by National Institutes of Health Grants GM45547 (to J.K.) and GM38839 (to M.O.) and Department of Energy Grant DOEDEAC03-7600098 (to the Advanced Light Source and the National Energy Research Scientific Computing Center at Lawrence Berkeley National Laboratory for supercomputer work).


Author contributions: S.L.K., Y.Z., G.D.B., M.O., and J.K. designed research; S.L.K., Y.Z., G.D.B., and J.K. performed research; and M.O. analyzed data.

Abbreviations: PCNA, proliferating cell nuclear antigen; RFC, replication factor C.


1. Kong, X. P., Onrust, R., O'Donnell, M. & Kuriyan, J. (1992) Cell 69, 425–437. [PubMed]
2. Stukenberg, P. T., Studwell-Vaughan, P. S. & O'Donnell, M. (1991) J. Biol. Chem. 266, 11328–11334. [PubMed]
3. Stillman, B. (1994) Cell 78, 725–728. [PubMed]
4. Kelman, Z. & O'Donnell, M. (1995) Annu. Rev. Biochem. 64, 171–200. [PubMed]
5. O'Donnell, M. (1992) BioEssays 14, 105–111. [PubMed]
6. Waga, S. & Stillman, B. (1998) Annu. Rev. Biochem. 67, 721–751. [PubMed]
7. Warbrick, E. (2000) BioEssays 22, 997–1006. [PubMed]
8. Krishna, T. S. R., Kong, X.-P., Gary, S., Burgers, P. & Kuriyan, J. (1994) Cell 79, 1233–1243. [PubMed]
9. Gulbis, J. M., Kelman, Z., Hurwitz, J., O'Donnell, M. & Kuriyan, J. (1996) Cell 87, 297–306. [PubMed]
10. Moarefi, I., Jeruzalmi, D., Turner, J., O'Donnell, M. & Kuriyan, J. (2000) J. Mol. Biol. 296, 1215–1223. [PubMed]
11. Matsumiya, S., Ishino, Y. & Morikawa, K. (2001) Protein Sci. 10, 17–23. [PMC free article] [PubMed]
12. Shamoo, Y. & Steitz, T. A. (1999) Cell 99, 155–166. [PubMed]
13. Majka, J. & Burgers, P. M. (2004) Prog. Nucleic Acid Res. Mol. Biol. 78, 227–260. [PubMed]
14. Bowman, G. D., Goedken, E. R., Kazmirski, S. L., O'Donnell, M. & Kuriyan, J. (2005) FEBS Lett. 579, 863–867. [PubMed]
15. Turner, J., Hingorani, M. M., Kelman, Z. & O'Donnell, M. (1999) EMBO J. 18, 771–783. [PMC free article] [PubMed]
16. Jeruzalmi, D., Yurieva, O., Zhao, Y., Young, M., Stewart, J., Hingorani, M., O'Donnell, M. & Kuriyan, J. (2001) Cell 106, 417–428. [PubMed]
17. Bowman, G. D., O'Donnell, M. & Kuriyan, J. (2004) Nature 429, 724–730. [PubMed]
18. Goedken, E. R., Kazmirski, S. L., Bowman, G. D., O'Donnell, M. & Kuriyan, J. (2005) Nat. Struct. Mol. Biol. 12, 183–190. [PubMed]
19. Alley, S. C., Shier, V. K., Abel-Santos, E., Sexton, D. J., Soumillion, P. & Benkovic, S. J. (1999) Biochemistry 38, 7696–7709. [PubMed]
20. Gomes, X. V., Schmidt, S. L. & Burgers, P. M. (2001) J. Biol. Chem. 276, 34776–34783. [PubMed]
21. Jeruzalmi, D., O'Donnell, M. & Kuriyan, J. (2001) Cell 106, 429–441. [PubMed]
22. Kazmirski, S. L., Podobnik, M., Weitze, T. F., O'Donnell, M. & Kuriyan, J. (2004) Proc. Natl. Acad. Sci. USA. 101, 16750–16755. [PMC free article] [PubMed]
23. Neuwald, A. F., Aravind, L., Spouge, J. L. & Koonin, E. V. (1999) Genome Res. 9, 27–43. [PubMed]
24. Yu, R. C., Hanson, P. I., Jahn, R. & Brunger, A. T. (1998) Nat. Struct. Biol. 5, 803–811. [PubMed]
25. Lenzen, C. U., Steinmann, D., Whiteheart, S. W. & Weis, W. I. (1998) Cell 94, 525–536. [PubMed]
26. Miyata, T., Suzuki, H., Oyama, T., Mayanagi, K., Ishino, Y. & Morikawa, K. (2005) Proc. Natl. Acad. Sci. USA 102, 13795–13800. [PMC free article] [PubMed]
27. Duke, R. E. (2003) pmemd (Univ. of North Carolina, Chapel Hill).
28. Case, D. A., Pearlman, D. A., Caldwell, J. W., Cheatham, T. E. I., Wang, J., Ross, W. S., Simmerling, C., Darden, T., Merz, K. M., Stanton, R. V., et al. (2002) amber 7.0 (Univ. of California, San Francisco).
29. Cheatham, T. E. I., Cielplak, P. & Kollman, P. A. (1999) J. Biomol. Struct. 16, 845–862. [PubMed]
30. Jorgensen, W. L. (1981) J. Am. Chem. Soc. 103, 335–340.
31. Darden, T., York, D. & Pedersen, L. (1995) J. Chem. Phys. 98, 10089–10092.
32. Ryckaert, J. P., Ciccotti, G. & Berendsen, H. J. C. (1977) J. Comput. Phys. 23, 327–336.
33. Brooks, B. R., Bruccoleri, R. E., Olafson, B. D., Swaminathan, S. & Karplus, M. (1983) J. Comput. Chem. 4, 187–217.
34. Humphrey, W., Dalke, A. & Schulten, K. (1996) J. Mol. Graphics 14, 33–38. [PubMed]
35. DeLano, W. L. (1998) pymol (DeLano Scientific, San Carlos, CA).
36. Millar, D., Trakselis, M. A. & Benkovic, S. J. (2004) Biochemistry 43, 12723–12727. [PubMed]
37. Hayward, S. & Go, N. (1995) Annu. Rev. Phys. Chem. 46, 223–250. [PubMed]
38. Emberly, E. G., Mukhopadhyay, R., Tang, C. & Wingreen, N. S. (2004) Proteins 55, 91–98. [PubMed]
39. Salemme, F. R. (1983) Prog. Biophys. Mol. Biol. 42, 95–133. [PubMed]
40. Ho, B. K. & Curmi, P. M. (2002) J. Mol. Biol. 317, 291–308. [PubMed]
41. Yang, A. S. & Honig, B. (1995) J. Mol. Biol. 252, 366–376. [PubMed]
42. Wang, L., O'Connell, T., Tropsha, A. & Hermans, J. (1996) J. Mol. Biol. 262, 283–293. [PubMed]
43. Matsumiya, S., Ishino, S., Ishino, Y. & Morikawa, K. (2003) Protein Sci. 12, 823–831. [PMC free article] [PubMed]
44. Yao, N., Turner, J., Kelman, Z., Stukenberg, P. T., Dean, F., Shechter, D., Pan, Z. Q., Hurwitz, J. & O'Donnell, M. (1996) Genes Cells 1, 101–113. [PubMed]
45. Miyata, T., Oyama, T., Mayanagi, K., Ishino, S., Ishino, Y. & Morikawa, K. (2004) Nat. Struct. Mol. Biol. 11, 632–636. [PubMed]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...