Keywords: nucleotide kinase, nucleotide phosphatase, DNA repair, catalytic mechanism

Substrate recognition by the phosphatase domain

Crystals of PNK with a ssDNA bound to the phosphatase domain were obtained by incubating a murine PNK construct lacking the N-terminal FHA domain (PNKΔFHA: amino acids 111-512) with DNA in a 1.5:1 DNA:protein ratio, in the presence of a ~10-fold excess of Mg²⁺-ATP. Crystals appeared after a period of several weeks and were further soaked overnight in the precipitant, augmented with the ssDNA at 0.5mM, before cryoprotecting for data collection. Data were collected from a single soaked crystal, which diffracted to a resolution of 2.0Å (Table 1), and the structure was solved by molecular replacement (see Materials and Methods).

The difference maps from these crystals (Table 1 - PNK-DNA1) revealed the presence of a short stretch of single-stranded DNA bound to a narrow channel traversing the phosphatase domain (Figure 1A,Supplementary Figure S1). The channel also contained a well-ordered magnesium ion coordinated by the side chains of Asp170 and Asp288, along with the backbone carbonyl of Asp172 (Figure 1B). The location and coordination of the magnesium ion identifies it as the essential co-factor for the phosphatase activity based on analogy to other proteins of the haloacid dehalogenase (HAD) enzyme family, to which PNK-phosphatase domain is structurally related (Bernstein et al., 2005). Additionally, Mg²⁺-ADP and a molecule of HEPES buffer are bound to the kinase active site (see below).

The three most 3’ deoxynucleotides (C₃-A₄-C₅) of the single-stranded DNA are clearly visible in the channel. The bases of C₃ and A₄ are stacked against each other in van der Waals contact, and the stacked pair are themselves sandwiched between the side chains of Lys225 and Phe184 (Figure 1C). The interconnecting phosphate (5’P-A₄) is coordinated through a bi-dentate interaction with the guanidine head group of Arg258 and the backbone nitrogen of Met219 (Figure 1D). The terminal base (C₅) is not stacked contiguously with C₃ and A₄, but is instead rotated perpendicularly to the other bases to sit between the hydrophobic side-chains of Phe184 and Phe305 (Figure 1E). An oxygen atom of the phosphate bridging to A₄ makes a ligand interaction with the catalytic magnesium bound by Asp170, Asp172 and Thr216.

Based on studies of other phosphoester hydrolase enzymes in the haloalkane dehydrogenase family, PNK phosphatase is believed to operate via an SN2 reaction mechanism (Bernstein et al., 2005), in which the Mg²⁺ ion activates the terminal 3’-phosphate for nucleophilic attack, by Asp170 in the case of PNK, generating a carboxyl–phosphate ester intermediate, that is hydrolysed in the next stage of the mechanism. Comparison with the structure of a phospho-aspartate mimetic reaction intermediate of the HAD-family enzyme phosphoserine phosphatase, (Cho et al., 2001) (PDB: 1j97), suggests that the phosphatase active site is fully assembled and poised for catalysis. However the reaction cannot proceed as the bridging phosphate bound to the magnesium ion is not positioned correctly for hydrolysis due to the conformational restrictions imposed by its connection to the C₅ base stacked between Phe184 and Phe305 (Figure 1F). Abona fide 3’-terminal phosphate similarly located would be unencumbered and able to interact productively with the active site.

Rather than leading into a blind-ended pocket that would provide explicit selective binding of a 3’-phosphorylated terminus, the phosphatase channel is open-ended and could provide favourable backbone interactions with segments of ssDNA 3’ to the active site. Thus, although its catalytic activity is restricted to terminal 3’-phosphates, PNK appears to be fully capable of binding non-terminal segments of ssDNA. The combination of the phosphate backbone interactions made by Arg258, the phosphate contact made by the catalytic magnesium ion, and the base stacking interaction provided by Phe184 and Phe305, would allow isoenergetic sliding of ssDNA segments through the channel. This would facilitate location of 3’-phosphate ends on extended 3’-overhanging substrates such as those generated in double-strand break resection by DNA2 or EXO1 in collaboration with BLM, RPA and the MRN complex, as part of the homologous recombination pathway of DNA-repair (Gravel et al., 2008;Mimitou and Symington, 2008;Nimonkar et al., 2011;Zhu et al., 2008).

While the DNA backbone and base-stacking interactions are generic and impose no significant DNA sequence specificity on the phosphatase activity, they are totally incompatible with the 3’-phosphorylated substrate DNA strand remaining base paired. The requirement for ssDNA in the phosphatase channel does imply that PNK must be able to facilitate a substantial local disruption of DNA base-pairing when the substrate 3’-phosphorylated end is presented in the context of a gapped or nicked DNA duplex (see below).

As there has been no previous experimental observation of DNA bound to a polynucleotide 3’-phosphatase, we were concerned to confirm the role of the amino acids implicated in recognition and hydrolysis by our crystal structure.

We therefore generated a series of mutants within the phosphatase domain of PNKΔFHA, and determined their ability to bind DNA using an electrophoretic mobility shift assay (EMSA), and their catalytic activity using a fluorescent gel-based assay (Dobson and Allinson, 2006) (Figure 2A,B).

As expected, incubation of the single-stranded oligonucleotide FAM1 (see Materials and Methods) with increasing concentrations of wild-type PNKΔFHA resulted in rapid removal of the 3’ phosphate, visualised by the decrease in electrophoretic mobility of the fluorescent oligonucleotide on denaturing polyacrylamide gels. A charge-reversal mutation of Arg258 to a glutamic acid (R285E) resulted in a substantial decrease in phosphatase activity, consistent with a key role in binding the phosphodiester backbone of the 5’-GTCAC-3’ oligonucleotide substrate, as indicated by the crystal structure (Figure 1D). An F184G mutation, which removes the favourable stacking interaction between A₄ and the side-chain of Phe184 had minimal effect in isolation, but combined with the R258E mutation, substantially reduced phosphatase activity at all but the highest enzyme concentrations.

Studies of the mono-functionalS.cerevisiae DNA 3’-phosphatase Tpp1, to which the 3’-phosphatase domain of mammalian PNK is homologous, have implicated a key aspartic acid residue (Asp 35) as both a ligand for the catalytically essential Mg²⁺ co-factor, and as the phosphate acceptor in the covalent mechanism (Deshpande and Wilson, 2004). Consistent with this, mutation of the equivalent residue in human PNK (Asp 171) to alanine resulted in complete loss of phosphatase activity (Dobson and Allinson, 2006). We made a D170N mutation in murine PNK and assayed its DNA binding and phosphatase activity as above, as expected, PNK-D170N was virtually catalytically inactive, however unlike the F184A/R258E mutant, its ability to bind ssDNA was essentially wild-type (Figure 2C).

Substrate recognition by the kinase domain

Despite extensive trials, we were unable to obtain PNK crystals with DNA bound at the kinase active site using the wild-type protein and conventional nicked dsDNA substrates generated by annealing three separate oligonucleotides. We therefore tried a number of alternative strategies culminating in incubation of phosphatase-dead PNK-D170N with a 10-fold excess of ssDNA oligonucleotide and ATP, in the presence of Mn²⁺ rather than Mg²⁺. In this way we obtained a new crystal form diffracting to 2.2Å, in which DNA was evidently bound in the kinase domain (Figure 3A,Supplementary Figure S2).

These crystals (PNK-DNA2) contain a single copy of PNK-D170N in the asymmetric unit, associated with five copies of the ssDNA 5’-GTCAC-3’. The oligonucleotides are annealed to each other to form a self-complementary blunt-ended duplex with a central C:C mismatch (DNA-A* / DNA-B*) and a nicked-duplex structure in which one 5’-GTCAC-3’ oligonucleotide (DNA-A) is base-paired to a second oligonucleotide (DNA-B) via the 5’-GTC sequence, and to a third (DNA-C) via the AC-3’ sequence. The exposed base-pairs at the end of the duplex segments stack on each other and on symmetry related duplex ends to form pseudo-continuous DNA duplexes that builds the crystal lattice (Supplementary Figure S3).

The 5’-nucleotide (G*) of DNA-B lies adjacent to a pocket in the kinase domain of PNK flanked by α-helices 8 and 12 (using the nomenclature and numbering described in (Bernstein et al., 2005)). In the DNA complex the pocket is occupied by an ADP molecule and the 5’-OH of G* (and all other visible 5’-ends) is phosphorylated, suggesting that the enzyme has turned over during crystallogenesis and that the crystals contain a product complex. Fluorescence-based kinase assays confirmed that the PNK-D170N enzyme has essentially the same activity when either magnesium or manganese is provided as a co-factor (seeSupplementary Figure S4)

Within this pocket N1 and N6 of the adenine ring are hydrogen-bonded to the backbone carbonyl and peptide NH of Arg503 and Gln505 respectively (Figure 3B), in a similar manner to the ‘hinge’ binding observed in protein kinases (Johnson et al., 1996;Noble et al., 2004). Hydrophobic residues Ile345, Phe462, Phe502, Leu504, Leu508 and Tyr516 form the walls of the adenine pocket. The Oγ of Thr466, coordinates the O3’ position of the ribose moiety through a water-mediated interaction (Figure 3C). Numerous polar interactions occur between ADP phosphate oxygens and the protein (Figure 3D), including extensive backbone contacts with NH groups of the Walker A / P-loop motif 374-GAGKST-379. Additional interactions are provided by the side-chains and peptide NH groups of Lys377, Ser378 and Thr379, and the guanidinium head group of Arg463.

A Mn²⁺-ion and associated solvent bridges the side chain of Ser378 and the β-phosphate of the ADP to the 5’-phosphate of the DNA-B strand and the carboxylate side chain of Asp396 (Figure 3E). Arg395, which hydrogen bonds to the phosphate group on the 3’-side of the 5’-guanine, also interacts with Asp396, orientating its carboxylate group towards what would have been the terminal 5’-OH of the DNA strand. The 5’-phosphate group receives a hydrogen bond from the imidazole side chain of His470. Based on their position and interactions in the product complex Asp396 is well placed to act as a general base in the phosphoryl transfer reaction, abstracting a proton from the 5’-OH of the DNA, consistent with the significant loss of activity observed when this residue is mutated to asparagine (Bernstein et al., 2009).

Recognition of Strand Discontinuity by PNK

The 5’-phosphorylated guanine bound in the active site, as well as the two bases 3’ to that in the DNA-B molecule, are fully base paired to the DNA-A molecule. The conformation of the base-paired segment is essentially canonical B-form DNA (very slightly perturbed by the central T:T mismatch) and could continue unimpeded as linear duplex DNA beyond that (Figure 4A). The phosphate backbone of the continuous strand makes no contact with the protein, but the backbone of the discontinuous strand lies in a shallow basic channel with the phosphates for the three nucleotides upstream of the 5’-nucleotide interacting with side chains of Arg395, Arg432 and Trp401. The exposed face of the base-pair formed by the 5’-phosphorylated guanine and its cytosine partner, packs against an extended hydrophobic platform, generated by the side chains of Pro373 and Val473, and by Met476 and Val477 at the N-terminal end of helix α13, whose axis runs perpendicular to the axis of this segment of duplex DNA.

The insertion of helix α13 into the path of the duplex DNA, completely blocks any possibility of maintaining base stacking on either strand, across the site of the nick. Instead, the backbone of the continuous strand undergoes a substantial change in direction, folding around the surface of α13, with the 3’-AC bases resuming base-pairing with the 5’-end of a third oligonucleotide (DNA-C). As with the last base-pair on the 5’-side of the nick, the exposed face of the first base-pair on the 3’-side packs against a hydrophobic surface formed by Met476, Phe479 and Ser480 from α-helix 13, with Arg482 and Lys483 providing polar interactions for the phosphate backbone as it twists around the surface of α13 (Figure 4B).

To determine the significance of the interactions provided by α13 we mutated Val477, to disrupt the hydrophobic interface with the exposed base-pair surface, and Arg482 and Lys 483, to disrupt the phosphate backbone interactions (in the context of the phosphatase-dead PNK-D170N protein). Compared to the wild-type amino-acids, V477R and R482E, and K483E mutations significantly diminished the ability of the enzyme to phosphorylate a 5’-OH in the context of a nicked duplex (Figure 5A,B). The substantial reduction in kinase activity observed for these mutants confirms the biological significance of the interactions observed in the crystal structure, and defines a key role for the α13-helix in exposing and positioning the 5’-substrate in the kinase active site.

Although only two base-pairs of duplex are formed in the crystals beyond α13, the conformation of these is almost perfect B-form, and there is no obstacle to the DNA continuing as a linear duplex beyond this (Figure 6A,B), potentially making favourable electrostatic interactions with a number of extended basic loops (involving lysines 182, 225, 301, 302 and 303) on the surface of the phosphatase domain. Thus the phosphatase domain may play a role in assisting the kinase domain in its activity towards nicked and gapped substrates.

The 3’-nucleotides on DNA-C 3’ not involved in base-pairing with DNA-A, curl into the mouth of the ssDNA-binding channel we identify in the phosphatase domain (see above), with the bases of the two most 3’ nucleotides occupying the base-stacking sites between Lys225 and Phe184 and between Phe184 and Phe305 (Figure 3A,Figure 7A-B). As in the complex with ssDNA, Arg258 interacts with the phosphate backbone, but in this case with the phosphate 5’ of the 3’-terminal nucleotide. While no catalytic metal is present due to the D170N mutation, and the terminal nucleotide carries a 3’-OH, the position of the terminal base suggests that were a 3’ phosphate present, it would be able to interact productively with the active site. Thus, the two active sites in PNK appear to be simultaneously able to make productive contact with their cognate end-chemistries within the same DNA substrate.

Induced-Fit in the PNK kinase domain

Comparison of the PNK-kinase DNA complex and an apo-structure of PNK determined at 1.6Å resolution (see Materials and Methods) shows a number of structural rearrangements in the kinase domain that accompany binding to DNA and ADP (Supplementary Figure S5). Strand β16, helices α9 and α10, and their connecting loops, swing towards the centre of the protein, hinging at Gly389 and Gly415. This movement establishes a cluster of direct and water-bridged interactions with the phosphate backbone of the nicked strand, involving Trp401, Gln402 and Arg395, and moves the catalytic Asp396 ~ 5Å from its position in the apo-structure to lie above the 5’-oxygen of the terminal nucleotide (FIGURE 7C). A second more subtle rearrangement occurs in the loop carrying the amino acid His470, and the hinge segment at residues 503-508, which open out to accommodate and interact with the ADP/ATP (FIGURE 7D).

The observed differences between the apo and substrate — bound kinase domain structures are compatible with solution circular dichroism and UV-difference spectroscopy studies (Mani et al., 2001) that indicated a conformational change on binding ATP to human PNK. Thus the DNA 5’-kinase domain of PNK appears to display active and inactive conformations, which are reminiscent of the active and inactive conformations that mediate allosteric regulation of protein kinases (Johnson et al., 1996). Recent observations suggest that the DNA 5’-kinase activity of PNK may also be subject to allosteric activation via direct interaction with the XRCC1 scaffold protein (Lu et al., 2010), but whether this is mediated by the conformational flexibility of the β16-α9-α10 segment, is currently unknown.

Gap Recognition and DNA Distortion

The crystal structures presented here provide a definitive view of the interaction of PNK with both of its DNA substrates, and fully explain the distinctive and individual specificities of the phosphatase and kinase functions. The phosphatase active site lies within a narrow channel that precludes binding of duplex DNA. This explains the observation thatin vitro PNK preferentially dephosphorylates single-stranded DNA or duplex DNA substrates with at least a three nucleotide 3’-overhang (Bernstein et al., 2005). Nonetheless, PNKin vivo must also act on 3’-phosphates in the context of single-strand gaps and double-strand breaks, so that it must be able to stabilise unstacking of the bases and local unwinding of the DNA in the vicinity of a strand discontinuity in order to access a 3’-end.

By contrast, the DNA-binding site in the kinase domain readily accommodates undistorted duplex DNA upstream of the site of the 5’-terminus bound in the active site. However, the chemistry of phospho-transfer from ATP bound in the co-factor pocket of the kinase domain, requires all-round access to the 5’-terminal hydroxyl group that is not possible in the context of a fully stacked DNA duplex. Thus, as for the 3’-phosphatase function, the 5’-kinase function requires PNK to stabilise unstacking of the duplex DNA for substrates in which the 5-OH is presented in the context of a nick or small gap. For both enzymatic functions, the necessary DNA distortion is provided by helix α13, which inserts into the DNA providing a two-sided hydrophobic ‘wedge’ to support the exposed base-pair surfaces from both phosphatase and kinase substrates, and charged interactions to stabilise the distorted path of the sugar-phosphate backbone of the continuous strand that connects the two. The ~70° bend observed in the path of the DNA as it traverses the adjacent active sites, was not predicted by previous low-resolution analysis and the orientation of the duplex DNA as it engages with the kinase active site, is significantly different to models for DNA interaction by PNK suggested on the basis of those studies (Bernstein et al., 2009).

Functional coupling of 3’-phosphatase and 5’-kinase domains

The presence of two distinct enzymatic activities in PNK has always raised the tantalising possibility that their functions might in some way be coupled. Certainly many DNA damage events generating single-strand gaps or double-strand breaks can produce both 3’-phospho and 5’-hydoxyl ends, requiring both of PNK’s catalytic activities for repair. The DNA complex presented here provides a composite view of the binding modes of substrate DNA ends for both enzymatic functions within a single complex. The duplex segments define a clear trajectory for DNA substrates with nicks and short gaps, which are by far the most common substrates for PNKin vivo. A single common binding mode for nicked/gapped DNAs that facilitates access to either active site, fully explains the intriguing observation that binding of 3’-phosphorylated substrates to a D171A phosphatase-dead human PNK mutant, is inhibitory of the 5’-kinase activity towards gapped duplex DNA (Dobson and Allinson, 2006). This common binding mode would allow either active site to readily access a range of DNA double-strand break substrates, in whose repair by the non-homologous end-joining system PNK also play an important role.

The structures presented here explain how the composite domain structure of PNK utilises a single DNA binding mode to achieve two different enzymatic reactions required to facilitate single-strand gap break repair. PNK is however only one of at least four enzymes known to be associated with the XRCC1-scaffolded short-patch repair complex, all of which require access to DNA ends for their function. How these competing requirements for end-access by APTX, Polβ, Lig3β and PNK are coordinated and regulated, remains one of the major challenges in understanding the structural basis for single-strand gap repair.

Cloning, expression and purification of PNKΔFHA

AnE.coli codon-optimised sequence for full-length murine PNK was generated by gene synthesis (Eurofins Medigenomix, Martinsried, Germany). PCR primers were then used to amplify the region encoding the phosphatase and kinase domains (PNKΔFHA; amino acids 111-522) with additional flanking restriction sites (NdeI andEcoRI) to allow cloning into the vector pTWO-E; an in-house modified pET-17b vector (Novagen, Nottingham, UK) engineered to encode an N-terminal, 3C-Protease cleavable, His₆ affinity tag.

The recombinant plasmid was transformed into theE.coli strain Rosetta2(DE3) pLysS (Merck Chemicals, Nottingham, UK). A 250 ml flask containing 100 ml of L-broth (1% w/v tryptone, 0.5% w/v NaCl, 0.5% w/v yeast extract), supplemented with 100 μg/ml ampicillin and 34 μg/ml chloramphenicol, was inoculated with a single transformed bacterial colony. This was grown, overnight, in an orbital incubator set at 37 °C and 225 rpm. From this ‘starter culture’, 10 ml was then used to inoculate a 2 l flask containing 1 l of L-broth supplemented, as before, with antibiotics. The culture was grown in an orbital incubator set at 37 °C and 225 rpm. Once the A600 of the cell culture had reached 0.6, the incubator temperature was reduced to 20 °C and expression of PNKΔFHA induced by the addition of IPTG to a final concentration of 0.5 mM. The culture was grown for a further 16-18 hours at 20 °C, after which the cells were harvested by centrifugation (3100 g, 10 minutes, 10 °C), and the pellet stored at −80 °C until required.

The pellet arising from 10 l of cell culture was resuspended in 150 ml of buffer A: 50 mM HEPES-HCl pH 7.4, 1000 mM NaCl, 0.5 mM TCEP supplemented with protease inhibitor tablets (Roche, Burgess Hill, UK). The cells were lysed by sonication, after which cell debris and insoluble material were removed by centrifugation at 40 000 ×g for 30 min at 4 °C. The supernatant arising from this step was applied to a batch/gravity column containing 10 ml of Talon affinity resin (TaKaRa Bio, Saint-Germain-en-Laye, France) equilibrated in Buffer A. The column containing the cell extract and resin was rotated/rolled at 4 °C for a period of 1 hour to facilitate protein binding. The resin was allowed to pack under gravity, and then washed with successive applications of Buffer A (approximately 250 ml in total). Any retained protein was eluted from the column with the application of Buffer B: 50 mM HEPES-HCl pH 7.4, 1000 mM NaCl, 0.5 mM TCEP, 300 mM imidazole. Fractions containing PNKΔFHA were pooled, then concentrated to a final volume of 10 ml using Vivaspin 20 (5000 MWCO) centrifugal concentrators (Sartorius Stedim Biotech, Goettingen, Germany) before loading onto a HiLoad Superdex 75 size exclusion column (GE Healthcare, Little Chalfont, UK) equilibrated with Buffer C: 50 mM HEPES-HCl pH 7.4, 500 mM NaCl, 0.5 mM TCEP. Fractions containing PNKΔFHA were again identified by SDS-PAGE, then pooled and concentrated as before. The affinity tag was cleaved from the protein by incubation with rhinovirus 3C-protease overnight at 4°C (PreScission protease, GE Healthcare), before being passed back through a Talon batch/gravity column (to remove any uncleaved protein and the residual tag), then reapplied to a S75 size exclusion column, equilibrated this time in Buffer D: 10 mM HEPES-HCl pH 7.4, 100 mM NaCl, 0.5 mM TCEP. Fractions containing the purified protein were identified, and then concentrated to 20 mg/ml, where it was either stored at 4°C for immediate use, or flash-frozen on dry-ice then stored at −80°C until required.

Crystallisation, Data Collection and Phasing

1. Abramoff MD, Magelhaes PJ, Ram SJ. Image Processing with ImageJ. Biophotonics International. 2004;11:36–42. [Google Scholar]
2. Adams PD, Grosse-Kunstleve RW, Hung LW, Ioerger TR, McCoy AJ, Moriarty NW, Read RJ, Sacchettini JC, Sauter NK, Terwilliger TC. PHENIX: building new software for automated crystallographic structure determination. Acta Crystallogr D Biol Crystallogr. 2002;58:1948–1954. doi: 10.1107/s0907444902016657. [DOI] [PubMed] [Google Scholar]
3. Ali AA, Jukes RM, Pearl LH, Oliver AW. Specific recognition of a multiply phosphorylated motif in the DNA repair scaffold XRCC1 by the FHA domain of human PNK. Nucleic Acids Res. 2009;37:1701–1712. doi: 10.1093/nar/gkn1086. [DOI] [PMC free article] [PubMed] [Google Scholar]
4. Bernstein NK, Hammel M, Mani RS, Weinfeld M, Pelikan M, Tainer JA, Glover JN. Mechanism of DNA substrate recognition by the mammalian DNA repair enzyme, Polynucleotide Kinase. Nucleic Acids Res. 2009;37:6161–6173. doi: 10.1093/nar/gkp597. [DOI] [PMC free article] [PubMed] [Google Scholar]
5. Bernstein NK, Williams RS, Rakovszky ML, Cui D, Green R, Karimi-Busheri F, Mani RS, Galicia S, Koch CA, Cass CE, et al. The molecular architecture of the mammalian DNA repair enzyme, polynucleotide kinase. Mol Cell. 2005;17:657–670. doi: 10.1016/j.molcel.2005.02.012. [DOI] [PubMed] [Google Scholar]
6. Caldecott KW, Aoufouchi S, Johnson P, Shall S. XRCC1 polypeptide interacts with DNA polymerase beta and possibly poly (ADP-ribose) polymerase, and DNA ligase III is a novel molecular ‘nick-sensor’ in vitro. Nucleic Acids Res. 1996;24:4387–4394. doi: 10.1093/nar/24.22.4387. [DOI] [PMC free article] [PubMed] [Google Scholar]
7. Cappelli E, Taylor R, Cevasco M, Abbondandolo A, Caldecott K, Frosina G. Involvement of XRCC1 and DNA ligase III gene products in DNA base excision repair. J Biol Chem. 1997;272:23970–23975. doi: 10.1074/jbc.272.38.23970. [DOI] [PubMed] [Google Scholar]
8. CCP4 Programs for protein crystallography. Acta Crystallogr. 1994;D50:760–763. doi: 10.1107/S0907444994003112. [DOI] [PubMed] [Google Scholar]
9. Cho H, Wang W, Kim R, Yokota H, Damo S, Kim SH, Wemmer D, Kustu S, Yan D. BeF(3)(−) acts as a phosphate analog in proteins phosphorylated on aspartate: structure of a BeF(3)(−) complex with phosphoserine phosphatase. Proc Natl Acad Sci U S A. 2001;98:8525–8530. doi: 10.1073/pnas.131213698. [DOI] [PMC free article] [PubMed] [Google Scholar]
10. Delano WL. The PyMOL Molecular Graphics System.http://www.pymol.org.
11. Deshpande RA, Wilson TE. Identification of DNA 3′-phosphatase active site residues and their differential role in DNA binding, Mg2+ coordination, and catalysis. Biochemistry. 2004;43:8579–8589. doi: 10.1021/bi049434n. [DOI] [PubMed] [Google Scholar]
12. Dobson CJ, Allinson SL. The phosphatase activity of mammalian polynucleotide kinase takes precedence over its kinase activity in repair of single strand breaks. Nucleic Acids Res. 2006;34:2230–2237. doi: 10.1093/nar/gkl275. [DOI] [PMC free article] [PubMed] [Google Scholar]
13. Emsley P, Cowtan K. Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr. 2004;60:2126–2132. doi: 10.1107/S0907444904019158. [DOI] [PubMed] [Google Scholar]
14. Gravel S, Chapman JR, Magill C, Jackson SP. DNA helicases Sgs1 and BLM promote DNA double-strand break resection. Genes Dev. 2008;22:2767–2772. doi: 10.1101/gad.503108. [DOI] [PMC free article] [PubMed] [Google Scholar]
15. Habraken Y, Verly WG. The DNA 3′-phosphatase and 5′-hydroxyl kinase of rat liver chromatin. FEBS Lett. 1983;160:46–50. doi: 10.1016/0014-5793(83)80933-8. [DOI] [PubMed] [Google Scholar]
16. Habraken Y, Verly WG. Chromatin 3′-phosphatase/5′-OH kinase cannot transfer phosphate from 3′ to 5′ across a strand nick in DNA. Nucleic Acids Res. 1986;14:8103–8110. doi: 10.1093/nar/14.20.8103. [DOI] [PMC free article] [PubMed] [Google Scholar]
17. Jilani A, Ramotar D, Slack C, Ong C, Yang XM, Scherer SW, Lasko DD. Molecular cloning of the human gene, PNKP, encoding a polynucleotide kinase 3′-phosphatase and evidence for its role in repair of DNA strand breaks caused by oxidative damage. J Biol Chem. 1999;274:24176–24186. doi: 10.1074/jbc.274.34.24176. [DOI] [PubMed] [Google Scholar]
18. Johnson LN, Noble ME, Owen DJ. Active and inactive protein kinases: structural basis for regulation. Cell. 1996;85:149–158. doi: 10.1016/s0092-8674(00)81092-2. [DOI] [PubMed] [Google Scholar]
19. Kocsis E, Trus BL, Steer CJ, Bisher ME, Steven AC. Image averaging of flexible fibrous macromolecules: the clathrin triskelion has an elastic proximal segment. J Struct Biol. 1991;107:6–14. doi: 10.1016/1047-8477(91)90025-r. [DOI] [PubMed] [Google Scholar]
20. Kubota Y, Nash RA, Klungland A, Schar P, Barnes DE, Lindahl T. Reconstitution of DNA base excision-repair with purified human proteins: interaction between DNA polymerase beta and the XRCC1 protein. EMBO J. 1996;15:6662–6670. [PMC free article] [PubMed] [Google Scholar]
21. Leslie AGW. MOSFLM Users Guide. MRC Laboratory of Molecular Biology; Cambridge, UK: 1995. [Google Scholar]
22. Loizou JI, El-Khamisy SF, Zlatanou A, Moore DJ, Chan DW, Qin J, Sarno S, Meggio F, Pinna LA, Caldecott KW. The protein kinase CK2 facilitates repair of chromosomal DNA single-strand breaks. Cell. 2004;117:17–28. doi: 10.1016/s0092-8674(04)00206-5. [DOI] [PubMed] [Google Scholar]
23. Lu M, Mani RS, Karimi-Busheri F, Fanta M, Wang H, Litchfeld DW, Weinfeld M. Independent mechanisms of stimulation of polynucleotide kinase/phosphatase by phosphorylated and non-phosphorylated XRCC1. Nucleic Acids Res. 2010;38:510–521. doi: 10.1093/nar/gkp1023. [DOI] [PMC free article] [PubMed] [Google Scholar]
24. Mani RS, Karimi-Busheri F, Cass CE, Weinfeld M. Physical properties of human polynucleotide kinase: hydrodynamic and spectroscopic studies. Biochemistry. 2001;40:12967–12973. doi: 10.1021/bi011383w. [DOI] [PubMed] [Google Scholar]
25. McCoy AJ. Solving structures of protein complexes by molecular replacement with Phaser. Acta Crystallogr D Biol Crystallogr. 2007;63:32–41. doi: 10.1107/S0907444906045975. [DOI] [PMC free article] [PubMed] [Google Scholar]
26. Mimitou EP, Symington LS. Sae2, Exo1 and Sgs1 collaborate in DNA double-strand break processing. Nature. 2008;455:770–774. doi: 10.1038/nature07312. [DOI] [PMC free article] [PubMed] [Google Scholar]
27. Nash RA, Caldecott KW, Barnes DE, Lindahl T. XRCC1 protein interacts with one of two distinct forms of DNA ligase III. Biochemistry. 1997;36:5207–5211. doi: 10.1021/bi962281m. [DOI] [PubMed] [Google Scholar]
28. Nimonkar AV, Genschel J, Kinoshita E, Polaczek P, Campbell JL, Wyman C, Modrich P, Kowalczykowski SC. BLM-DNA2-RPA-MRN and EXO1-BLM-RPA-MRN constitute two DNA end resection machineries for human DNA break repair. Genes Dev. 2011;25:350–362. doi: 10.1101/gad.2003811. [DOI] [PMC free article] [PubMed] [Google Scholar]
29. Noble ME, Endicott JA, Johnson LN. Protein kinase inhibitors: insights into drug design from structure. Science. 2004;303:1800–1805. doi: 10.1126/science.1095920. [DOI] [PubMed] [Google Scholar]
30. Potterton L, McNicholas S, Krissinel E, Gruber J, Cowtan K, Emsley P, Murshudov GN, Cohen S, Perrakis A, Noble M. Developments in the CCP4 molecular-graphics project. Acta Crystallogr D Biol Crystallogr. 2004;60:2288–2294. doi: 10.1107/S0907444904023716. [DOI] [PubMed] [Google Scholar]
31. Rass U, Ahel I, West SC. Defective DNA repair and neurodegenerative disease. Cell. 2007;130:991–1004. doi: 10.1016/j.cell.2007.08.043. [DOI] [PubMed] [Google Scholar]
32. Sheldrick GM. A short history of SHELX. Acta Crystallogr A. 2008;64:112–122. doi: 10.1107/S0108767307043930. [DOI] [PubMed] [Google Scholar]
33. Weinfeld M, Mani RS, Abdou I, Aceytuno RD, Glover JN. Tidying up loose ends: the role of polynucleotide kinase/phosphatase in DNA strand break repair. Trends Biochem Sci. 2011 doi: 10.1016/j.tibs.2011.01.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
34. Whitehouse CJ, Taylor RM, Thistlethwaite A, Zhang H, Karimi-Busheri F, Lasko DD, Weinfeld M, Caldecott KW. XRCC1 stimulates human polynucleotide kinase activity at damaged DNA termini and accelerates DNA single-strand break repair. Cell. 2001;104:107–117. doi: 10.1016/s0092-8674(01)00195-7. [DOI] [PubMed] [Google Scholar]
35. Zhu Z, Chung WH, Shim EY, Lee SE, Ira G. Sgs1 helicase and two nucleases Dna2 and Exo1 resect DNA double-strand break ends. Cell. 2008;134:981–994. doi: 10.1016/j.cell.2008.08.037. [DOI] [PMC free article] [PubMed] [Google Scholar]

Supplementary Materials