3¢-to5¢ DNA unwinding by TIP49b proteins Christophe Papin 1,2, * , , Odile Humbert 1,2, *, Anna Kalashnikova 1,2 , Kelvin Eckert 3 , Solange Morera 3 , Emmanuel Ka ¨ s 1,2 and Mikhail Grigoriev 1,2 1 Laboratoire de Biologie Mole ´ culaire Eucaryote, Universite ´ de Toulouse, France 2 CNRS, LBME, Toulouse, France 3 Laboratoire d’Enzymologie et Biochimie Structurales, Gif-sur-Yvette, France Introduction The highly conserved TIP49a and TIP49b proteins (also called pontin and reptin; a complete list of names is provided elsewhere [1]) belong to the AAA + super- family of P-loop ATPases (i.e. ATPases associated with diverse cellular activities), which includes enzymes involved in cellular housekeeping, cell division and dif- ferentiation. This superfamily is composed of a broad variety of enzymes, which appear to have a common property: energy-dependent remodeling of proteins and ⁄ or nucleic acids that results in the unfolding and disassembly of target macromolecules. The AAA + proteins contain conserved ATP-binding and hydroly- sis domains, which are activated by the formation of oligomeric assemblies. ATP binding and hydrolysis induce the conformational changes in the protein that are required for the translocation or remodeling activi- ties of target substrates [2–4]. TIP49a and TIP49b are implicated in a variety of cel- lular processes, such as chromatin remodeling during double strand-break repair, transcriptional regulation, genome stability, small nucleolar RNA biogenesis, telo- merase holoenzyme assembly and cellular division dur- ing mitosis [1,5,6]. Being essential proteins in yeast and embryonic-lethal in higher eukaryotes, they show a com- plex network of protein–protein interactions where, in some cases, they play antagonistic roles during the Keywords AAA+ ATPases; DNA binding; reptin; Rvb2p; TIP49b Correspondence E. Ka ¨ s or M. Grigoriev, IBCG, 118 route de Narbonne, 31062 Toulouse Cedex 9, France Fax: +33 (0)5 61 33 58 86 Tel: +33 (0)5 61 33 59 19; +33 (0)5 61 33 59 59 E-mail: grigor@ibcg.biotoul.fr; kas@ibcg.biotoul.fr *These authors contributed equally to this work Present address IGBMC UMR 7104, Illkirch, France (Received 4 January 2010, revised 9 April 2010, accepted 14 April 2010) doi:10.1111/j.1742-4658.2010.07687.x TIP49b (reptin) is an essential eukaryotic AAA+ ATPase involved in a variety of cellular processes, such as chromatin remodeling during double- strand break repair, transcriptional regulation, control of cell proliferation and small nucleolar RNA biogenesis. How it acts at the molecular level remains largely unknown. In the present study, we show that both human TIP49b and its yeast orthologue, Rvb2p, cooperatively bind single-stranded DNA as monomers. Binding stimulates a slow ATPase activity and sup- ports a 3¢-to5¢ DNA unwinding activity that requires a 3¢-protruding tail ‡ 30 nucleotides. The data obtained indicate that DNA unwinding of 3¢-to 5¢ junctions is also constrained by the length of flanking duplex DNA. By contrast, TIP49b hexamers were found to be inactive for ATP hydrolysis and DNA unwinding, suggesting that, in cells, protein factors that remain unknown might be required to recycle these into an active form. Structured digital abstract l MINT-7804328: TIP49b (uniprotkb:Q4QQS4)andTIP49b (uniprotkb:Q4QQS4) bind (MI:0407) by electron microscopy ( MI:0040) l MINT-7804638: tip49b (uniprotkb:Q4QQS4) and tip49b (uniprotkb:Q4QQS4) bind (MI:0407) by molecular sieving ( MI:0071) Abbreviation ssDNA, single-stranded DNA. FEBS Journal 277 (2010) 2705–2714 ª 2010 The Authors Journal compilation ª 2010 FEBS 2705 transcriptional regulation of gene expression and embry- onic development. However, their molecular mecha- nisms of action remain to be elucidated, particularly in view of such diverse functions. Although ATP hydrolysis is most likely essential for the functions of these proteins [7,8], most biochemical analyses of TIP49 proteins have failed to detect significant ATPase activity of the purified recombinant proteins [9–12]. Some studies have indicated helicase activity of TIP49 proteins, although opposite direction- alities of DNA unwinding have been reported [13–15]. However, as in the case of ATP hydrolysis, little pro- gress has been made with respect to further elucidating the DNA processing activities of TIP49 proteins or establishing whether their wide range of cellular func- tions is related, if at all, to transactions based on DNA unwinding. Because TIP49 proteins are found in chromatin remodeling complexes such as TIP60, Ino80 and SWR1 and are necessary for their activity, it is plausi- ble that they might partake in protein ⁄ DNA interac- tions that play a role in functions such as DNA repair or transcriptional control, in addition to the scaffold- ing role attributed to them in the formation of multi- protein complexes [1]. In the present study, we demonstrate that TIP49b and its yeast orthologue Rvb2p are indeed DNA-binding proteins. Cooperative binding to single-stranded DNA (ssDNA) stimulates a weak ATPase activity, which in turn leads to subse- quent DNA unwinding off 3¢-protruding tails ‡ 30 nucleotides in length. We show that these properties, (i.e. ssDNA-dependent weak ATPase and slow 3¢-to 5¢ unwinding) have been conserved between the mam- malian and yeast TIP49b proteins. Results To examine the biochemical properties of human and yeast reptin (TIP49b and Rvb2p, respectively), we purified them from Escherichia coli as N-terminal FLAG ⁄ His 6 - and C-terminal His 6 -tagged forms, respectively. The gel filtration profile on Superdex S-200 revealed that TIP49b is eluted in three well-sepa- rated fractions (Fig. 1A): the high molecular-weight peak, containing TIP49b aggregates (not used in the present study) and two other peaks eluting in the ranges 440–158 and 67–43 kDa, respectively. Native protein gel electrophoresis (the inset in Fig. 1A shows a silver-stained gel) and electron microscopy analysis (Fig. 1B) of these fractions confirmed that they differ in their oligomeric state: fraction 1 was found to be mostly hexamers (Fig. 1B, left), whereas fraction 2 mostly contained TIP49b monomers (Fig. 1B, right). SDS ⁄ PAGE analysis of aliquots of the protein prepa- rations followed by silver staining and western blotting using aFLAG or aTIP49b sera revealed the presence MonomerHexamer TIP49b: 020 100 nm 100 nm 40 60 80 100 mL mAU 20 A B CD 15 10 5 0 2000 158 67 43 kDa 440 TIP49b RecA –+ + + 123 ssM13 antiFLAG antiTIP49b kDa 12 55 36 31 21 66 97 Western blottingSilver staining Hexamer Monomer Fig. 1. TIP49b proteins used in the present study. (A) Gel filtration profile of TIP49b on a Superdex S-200 column. The fractions further analyzed are indicated by the horizontal bars. Silver-stained aliquots of each fraction analyzed by native PAGE are shown in the inset. (B) Electron microscopy images of TIP49b purified as hexamers (left) and monomers (right) used in the present study. Magnifica- tion: · 220 000; scale bar = 100 nm. (C) Purified proteins (500 ng) were analyzed by SDS ⁄ PAGE on 4–20% Tris–glycine SDS gels followed by silver staining and western blotting using aFLAG (Sigma) or aTIP49b (BD Biosciences) sera as indicated. Black arrowheads indicate TIP49b doublets; the grey arrowhead shows the minor GLMS_ECOLI contaminant. (D) UV cross-linking in the presence of radiolabeled [cP 32 ]-ATP followed by SDS ⁄ PAGE per- formed with RecA (lane 1) or with TIP49b in the absence (lane 2) or in the presence (lane 3) of ssM13. Black arrowheads show the TIP49b doublets (lanes 2 and 3); grey arrowheads indicate minor species that most likely correspond to UV-crosslinked RecA (lane 1) or TIP49b (lanes 2 and 3) intramolecular species. DNA unwinding by TIP49b C. Papin et al. 2706 FEBS Journal 277 (2010) 2705–2714 ª 2010 The Authors Journal compilation ª 2010 FEBS of purified TIP49b monomers (Fig. 1C), which appear as doublets (black arrowheads) suitable for further bio- chemical analysis. The major contamination detected by LC-MS ⁄ MS was a 67 kDa GLMS_ECOLI protein (Fig. 1C, grey arrowhead). We did not detect contami- nation by bacterial ATPases and helicases in our TIP49b and Rvb2p protein preparations. UV cross-linking in the presence of radiolabeled ATP was performed next (Fig. 1D). RecA (33 kDa) (5 lm), used as a positive control (lane 1), or TIP49b (lanes 2 and 3) were incubated in the presence of [c-P 32 ]-ATP and ssM13 at 50 ngÆlL )1 , UV-crosslinked and analyzed by SDS ⁄ PAGE on Tris–glycine gels. Although TIP49b (black arrowheads) showed a lower efficiency of ATP binding compared to that of RecA, regardless of the presence of ssDNA, these results sug- gest that the monomeric TIP49b preparations used throughout the present study bind ATP and are not detectably contaminated by other ATP-binding pro- teins (Fig. 1D), except for very weak bands visible in both RecA and TIP49b preparations that most likely correspond to UV-crosslinked intramolecular species (grey arrowheads). ATPase activity of TIP49b and DNA binding properties ATP hydrolysis to ADP and inorganic phosphate by TIP49b was assayed by incubating 1 lm of the protein with 200 lm ATP containing [a-P 32 ]-ATP tracer in the presence of different nucleic acids (10 ngÆlL )1 ) and sep- aration of reaction products by TLC (Fig. 2A). A small accumulation of ADP above background was detected in the presence of the protein alone, indicating intrinsic ATPase activity (Fig. 2A, lane 2). As expected [13], the ATPase activity was stimulated by single-stranded M13 DNA (ssM13, lane 4), but not by total RNA or circular supercoiled pBR322 plasmid DNA (Fig. 2A, lanes 3 and 5). The steady-state ATPase activity of human TIP49b and yeast Rvb2p (not shown) in the presence of ssM13 was also measured in parallel: both proteins obeyed Michaelis–Menten kinetics within the range of 0–2 mm ATP concentrations and showed similar kinetic parameters, classifying both proteins as weak ATPases. The hexameric fraction of TIP49b was also assayed and found to be inactive for ATP hydrolysis under the same experimental conditions (Fig. 2B). This suggests that oligomerization of the protein might influence its enzy- matic activity and DNA-binding properties (see below). We next tested the effect of single-stranded oligonu- cleotides on ATP hydrolysis by TIP49b compared to that in the presence of ssM13 DNA (Fig. 2C). These experiments were performed with 1 lm TIP49b, 200 lm ATP and saturating concentrations of synthetic oligonucleotides of different lengths (21, 50, 80 and 115 mers; 10 ngÆlL )1 ). Stimulation of ATP hydrolysis by TIP49b was found to be strongly dependent on oli- gonucleotide length, with a plateau of approximately 75% and 45% for ssM13 and ss115, respectively, indi- cating a correlation between ATPase activity and ssDNA-binding properties. Accordingly, we measured these as a function of protein concentration in gel-shift experiments using synthetic oligonucleotides as probes (Fig. 2D). In the titration shown, TIP49b retarded the migration of ss115 (Fig. 2E, lanes 1–4) in a concentra- tion-dependent manner but did not show significant binding to ss21, whereas binding to ss85 was detected at intermediate protein concentrations. The binding curve of yeast Rvb2p to ss115 was comparable to that of TIP49b, indicating the similar affinities of these con- served proteins for ssDNA (data not shown). It is significant to note that the length dependence of ATP hydrolysis (Fig. 2C) parallels that seen for DNA binding (Fig. 2D). Taken together, these results suggest that physical interaction with ssDNA might be directly involved in regulating the ATPase activity of TIP49b. We also tested binding to duplex DNA under the same conditions, demonstrating detectable yet weaker binding with a similar concentration response (Fig. 2E, lanes 5–8). Note that these experiments detect distinct strong (black arrowheads) and weak (grey arrowheads) complexes. At present, the nature of these different complexes remains unknown, although we speculate that they might correspond to species of dif- ferent conformations preserved by native gel electro- phoresis and ⁄ or to different stoichiometries of TIP49b proteins bound to ss- or ds115 substrates. In the latter case, the requirement for excess protein over DNA in electrophoretic mobility-shift assays renders any esti- mate of the number of TIP49b molecules bound to ss- or ds115 necessarily speculative at best. DNA unwinding by TIP49b and Rvb2p The putative helicase activity of TIP49b is controverted in the literature [9,13,15,16]. We addressed this issue by measuring the ability of TIP49b monomers to unwind short DNA duplexes in our experimental system (Fig. 3). Using ss115-mer as a template, we prepared 5¢- and 3¢-ss ⁄ ds junctions (ss94ds21, containing a 21 bp duplex region and a 94-nucleotide single-stranded tail; Fig. 3A) to detect DNA unwinding activity, as well as to investigate the direction of unwinding and the influ- ence of different adenosine phosphate cofactors on this process. The substrates were pre-incubated on ice with protein for 15 min before the addition of cofactors. C. Papin et al. DNA unwinding by TIP49b FEBS Journal 277 (2010) 2705–2714 ª 2010 The Authors Journal compilation ª 2010 FEBS 2707 Reactions were allowed to proceed at 37 °C for 30 min in the presence of a ten-fold excess of a trap oligonucleotide, complementary to the unwound radio- labeled probe, and then stopped by deproteinization in a solution containing a 100-fold trap excess before analysis by native gel electrophoresis. The results presented in Fig. 3 demonstrate that TIP49b indeed possesses a DNA-unwinding activity. However, under our experimental conditions, displacement of the labeled oligonucleotide is seen only in the presence of ATP and has a strict specificity for 3¢- rather than for 5¢-ss ⁄ ds DNA junctions (Fig. 3A, left, lanes 1–9). Yeast Rvb2p showed identical proper- ties (Fig. 3A, center panel, lanes 10–18). Significantly, as was the case for ATP hydrolysis, purified TIP49b hexamers were found to be unable to unwind DNA with a 30-nucleotide 3¢ extension (Fig. 3A, right, compare lanes 20 and 22) or a 5¢ extension (data not shown). Comparison of the unwinding time courses obtained with TIP49b and Rvb2p (Fig. 3B) demonstrates that they display similar apparent first- order kinetics of unwinding under our experimental AB C E D Fig. 2. ATPase and DNA binding activities of TIP49b and Rvb2p. (A) Nucleic acid requirement for ATP hydrolysis. TIP49b was incubated with the nucleic acids shown (10 ngÆlL )1 ) in the presence of radiolabeled ATP (200 l M final) for 45 min at 37 °C before analysis by TLC (lanes 1– 5). The small accumulation of ADP above background detected in the presence of the protein alone indicates intrinsic ATPase activity (lane 2). (B) ATP hydrolysis by TIP49b monomers and hexamers in the presence of ssM13 DNA using purified TIP49b monomers (filled circles) or hexamers (open circles). The inset shows the corresponding TLC for monomers (mono) and hexamers (hex) as a function of time (t). (C) ATP hydrolysis as a function of ssDNA length: ssM13 (filled circles), ss115 (open circles), ss80 (filled triangles), ss50 (open triangles), ss21 (filled squares) or in the absence of ssDNA (open squares). Reaction mixtures contained a fixed concentration of protein (1 l M) and 200 lM ATP. (D) Length-dependent ssDNA binding by TIP49b. Binding reactions were performed with 21, 85 or 115 mer oligonucleotides (0.1 nM) and increasing concentrations of TIP49b monomers. Binding dependence on ssDNA length follows that seen for ATP hydrolysis. (E) Compar- ison between ssDNA and duplex DNA binding by TIP49b. Major and minor complexes are indicated by black and grey arrowheads, respec- tively. The probable nature of the different complexes thus detected by electrophoresis on native 5% acrylamide gels is discussed in the main text. Note that ds115 shows a higher mobility than ss115 on these gels. The use of 8% acrylamide gels restores the expected relative mobilities of single-stranded and duplex DNA. DNA unwinding by TIP49b C. Papin et al. 2708 FEBS Journal 277 (2010) 2705–2714 ª 2010 The Authors Journal compilation ª 2010 FEBS conditions (k app = 0.24 ± 0.02 min )1 for Rvb2p and 0.16 ± 0.02 min )1 for TIP49b with 1 nm DNA sub- strate, 1 lm protein, 1 mm ATP and 10 nm reanneal- ing trap). We also measured unwinding activity as a function of protein concentration, obtaining identical curves, with a midpoint at 80.6 ± 3.6 nm for TIP49b and 85.3 ± 4.7 nm for Rvb2p (Fig. 3C). It is impor- tant to note that this similarity in terms of DNA unwinding parallels the ATPase and DNA-binding activities of both proteins, highlighting the remarkably strong conservation of these three activities. ssDNA-tail length requirement for DNA unwinding by TIP49b and Rvb2p The results reported above demonstrate that the DNA-binding and ATPase activities of TIP49b are sensitive to ssDNA length. This property might in turn affect the efficiency of DNA unwinding. To test this, we next constructed a set of 3¢-ss ⁄ ds junctions contain- ing a 21 bp duplex region and 3¢ single-stranded exten- sions of 0, 4, 19, 30, 39, 59 or 94 nucleotides (Fig. 4A), and measured DNA unwinding activity under our standard experimental conditions. The results obtained in these experiments (Fig. 4B) reveal that the efficiency of DNA unwinding by TIP49b and Rvb2p depends on the length of the 3¢ single-stranded tail. A sharp transition is observed between 19 and 39 nucleotides, with a midpoint at approximately 30 nucleotides of 3¢ single-stranded DNA (26 ± 2 nucleotides and 32 ± 4 nucleotides for TIP49b and Rvb2p, respectively, as estimated by the sigmoid fit of the data). Hence, short 3¢-protruding single-stranded tails are not sufficient to trigger unwinding. However, increasing the length of 3¢-protruding tails results in a sharp activation above a threshold length of approxi- A BC Fig. 3. DNA unwinding by TIP49b is preferentially exerted on 3¢-to5¢-ss ⁄ ds junctions. (A) Unwinding by 1 lM TIP49b (left, lanes 1–9) or Rvb2p (center, lanes 10–18) was tested on DNA (1 n M) in the absence of co-factors (lane 4) or in the presence of 1 mM AMP (lane 5), ADP (lane 6), ATP (lane 7), AMP-PNP (lane 8) or ATP-c S (lane 9) for 30 min at 37 °C. Asterisks denote the 5¢ [P 32 ]-label. 5¢-to3¢ and 3¢-to5¢ sub- strates used to assay unwinding activities are shown to the left of the gels. The right panel (lanes 19–22) shows a comparison of 3¢-to5¢ DNA unwinding by purified TIP49b monomers (lanes 19–20) or hexamers (lanes 21–22). Lanes 19 and 21 are no-ATP controls. (B) Time course of 3¢-to5¢ unwinding reactions performed with TIP49b (open and filled triangles show data from two independent experiments) or Rvb2p (open circles). The lines are the best fit to a single exponential with an apparent first-order rate constant of 0.24 ± 0.02 min )1 for Rvb2p and 0.16 ± 0.02 min )1 for TIP49b (deviations are from the fit). (C) Unwinding measurements performed at varying TIP49b (triangles) or Rvb2p (open circles). Insets in (B) and (C) show representative experiments used for quantifications. C. Papin et al. DNA unwinding by TIP49b FEBS Journal 277 (2010) 2705–2714 ª 2010 The Authors Journal compilation ª 2010 FEBS 2709 mately 30 nucleotides. Additionally, we note that both proteins were unable to unwind either blunt-ended short DNA substrates, such as three- and four-way junctions or nick-containing three-way junctions (data not shown), consistent with the results obtained in previously studies [16]. Duplex length affects the efficiency of DNA unwinding by TIP49b Although circular duplex DNA does not stimulate the ATPase activity of TIP49b (Fig. 2A), we show next that the protein is capable of binding to linear duplex oligonucleotides. This property led us to test possible effects of duplex DNA length on the unwinding effi- ciency of TIP49b. We next constructed a set of 3¢-ss ⁄ ds junctions containing a 30-nucleotide 3¢-tail and progressively longer duplex regions of 21, 45, 55, 65 and 85 bp (ss30ds21, ss30ds45, ss30ds55, ss30ds65 and ss30ds85; Fig. 4C). Reactions were performed with 1nm of DNA substrates, 100 nm TIP49b and 1 mm ATP. Increasing the duplex length decreased the total extent of unwinding and the overall reaction rate, with a sharp drop between 45 and 65 bp (Fig. 4D), suggest- ing that DNA unwinding occurs as a result of a short- range activity of the protein on duplex DNA. The observed drop in the efficiency of DNA unwinding could be a result of reannealing of DNA strands during or after the unwinding reaction. Accordingly, the final amplitude of the reaction would be underestimated. Indeed, the annealing rates of 21- or 25-nucleotide oligonucleotides and ss115 at 37 °C are in the range of 0.25 ± 0.03 to 0.41 ± 0.02 min )1 , respectively, at a nanomolar DNA concentration (data AB C E D Fig. 4. DNA length requirement for DNA unwinding by TIP49b. (A) Schematic presen- tation of the DNA substrates containing 3¢ ssDNA tails of different lengths used in the present study. Oligonucleotides used con- tain a 21 bp duplex region and a 3¢ single- stranded extension of 0, 4, 19, 30, 39, 59 or 94 nucleotides. Asterisks denote the 5¢ [P 32 ]-label. (B) The DNA unwinding activity of TIP49b or Rvb2p depends on the length of the 3¢ single-stranded tail. Reactions were performed for 30 min at 37 °C. Solid circles: TIP49b; open circles: Rvb2p. Data obtained from two independent experiments for each protein were fitted with a sigmoid curve. The inset shows representative bind- ing experiments used for quantifications. (C) 3¢-to5¢-ss ⁄ ds junctions containing a 30-nucleotide ssDNA tail and duplex regions of different lengths. Oligonucleotides a–d were used as traps, as described in the main text, and are complementary to the unlabeled top strand. (D) Time courses of DNA unwinding of the substrates containing duplex regions of different lengths. Filled circles, ss94ds21; open circles, ss30ds21; filled triangles, ss30ds45; open triangles, ss30ds55; filled diamonds, ss30ds65; open diamonds, ss30ds85. (E) Unwinding ampli- tude versus duplex length in the absence of in the presence of reannealing traps whose permutations correspond in each case to the extent of each duplex region tested (filled circles and open circles, respectively). DNA unwinding by TIP49b C. Papin et al. 2710 FEBS Journal 277 (2010) 2705–2714 ª 2010 The Authors Journal compilation ª 2010 FEBS not shown), or approximately two- to three-fold faster than the overall unwinding rate of TIP49b. To rule out the possibility that the effects of duplex length we observed are a result of reannealing occurring faster than unwinding, yielding a drop in the overall mea- sured unwinding efficiency, we repeated these experi- ments in the presence of reannealing traps consisting of short 20-nucleotide oligonucleotides complementary to the unlabeled top strand (10 nm; Fig. 4E), with this length being chosen because it does not support bind- ing by TIP49b (Fig. 2D). The traps (a–d; Fig. 4C, bot- tom) hybridize to the different duplex regions tested and were added alone or in combination to match each duplex. Because the addition of these permuta- tions of the reannealing traps did not change the final amplitude of DNA unwinding by TIP49b (Fig. 4E, compare filled and open circles), we conclude that reannealing of DNA during the reaction is not respon- sible for the sharp decrease in the total extent of unwinding seen for duplexes of 65 and 85 bp. As dis- cussed below, these results suggest that the unwinding properties of TIP49b documented in the present study differ from those of processive hexameric-ring helicases. Discussion We have studied the biochemical properties of mam- malian TIP49b and yeast Rvb2p purified as monomers. We found that the ATPase activity of TIP49b depends on ssDNA in a length-dependent manner and can be correlated with its ssDNA-binding properties. ATP hydrolysis as a result of a contaminant can be ruled out based on routine MS analysis of purified proteins and a lack of other detectable ATP-binding activities (Fig. 1D). In addition, the hexameric fraction of TIP49b did not show detectable ATPase activity, nor did it support DNA unwinding in our experimental system. The availability of TIP49b ATPase mutants would serve as a useful control as well as a powerful tool to dissect the mechanisms that account for DNA binding and unwinding. However, we and others have failed to generate such mutants, and such a puzzling failure has been discussed in a recent review [1]. Para- doxically, in principle, the presence in TIP49b of pre- sumptive essential motifs, such as Walker A and B boxes, R-finger, and Sensor 1 and 2 motifs, should allow for rational mutagenesis, although all efforts have remained unsuccessful. We speculate that the controlled and ssDNA-dependent ATPase activity of TIP49b is unusually sensitive to protein conforma- tional changes that might be induced by ATP binding. In this case, attempts to mutate ATP binding pockets might lead to more active rather than inactive mutants. We demonstrate that both TIP49b and yeast Rvb2p possess a slow ATP-dependent strand separation activ- ity, which is exerted on 3¢-ss ⁄ ds junctions containing a protruding single-stranded tail ‡ 30 nucleotides. The results obtained suggest that, on these substrates, DNA unwinding by TIP49b occurs as a result of pro- tein action over short distances along duplex DNA, whose length affects the efficiency of strand separation. Given a 3¢-protruding tail of fixed length, the extent of strand separation is inversely correlated with the length of duplex DNA lying beyond the junction (Fig. 4D, E), indicating a mechanism of action strictly conserved between mammalian and yeast TIP49b proteins, but differing from that of processive hexameric-ring heli- cases. Because TIP49b also binds duplex DNA, a sim- ple explanation would be that distinct interactions of monomers bound to ssDNA or adjacent duplex DNA ‡ 45 bp intrinsically limit the extent of the unwinding reaction. The molecular basis for such a limitation remains unknown at present, although it could implicate ATP-induced conformational changes undergone by TIP49b at ss- ⁄ dsDNA junctions. Further analysis of the nature of the complexes formed by TIP49b with ssDNA or dsDNA will be required to address this question in more detail. The strict 3¢-to5¢ directionality of DNA unwinding by TIP49b and Rvb2p documented in the present study on well defined substrates is consistent with that of the TIP60 and Ino80 chromatin-remodeling com- plexes containing TIP49a and TIP49b. However, it dif- fers from the initial report on the properties of purified rat TIP49b on much longer ssDNA substrates [13]. Using similar ssM13-helicase substrates and high pro- tein concentrations, we also detected both 3¢-to5¢ and 5¢-to3¢ unwinding by TIP49b and Rvb2p (data not shown). Although it is difficult to compare these out- comes given the significant differences in experimental conditions, our observation that short DNA substrates and low protein concentrations support a 3¢-to5¢ polarity of unwinding by TIP49b and Rvb2p could indicate that the ratio of ss- to dsDNA in these differ- ent substrates influences the polarity of DNA strand separation. We also note that a previous study of Rvb2p did not reveal a detectable unwinding activity using 3¢-to5¢ or 5¢-to3¢-ss ⁄ ds junctions (composed of a 15-nucleotide ss-tail and 28 bp duplex), even at high (3 lm) protein concentrations [15]. However, this apparent discrepancy is consistent with our results showing the strong dependence of unwinding on the length of the ssDNA tail, suggesting that 15-nucleotide tails are not long enough to support efficient DNA C. Papin et al. DNA unwinding by TIP49b FEBS Journal 277 (2010) 2705–2714 ª 2010 The Authors Journal compilation ª 2010 FEBS 2711 unwinding by TIP49b and Rvb2p (Fig. 4B). Finally, we note the possibility that TIP49b ⁄ TIP49a hetero- hexamers might exhibit characteristic properties. This has been previously investigated [15], but using protein concentrations in the range 5–40 lm. The use of such concentrations for TIP49b alone leads, in our hands, to protein aggregation and we are currently attempting to develop new approaches conducive to controlled hexamerization, which might also be extended to the study of TIP49a ⁄ TIP49b interactions. Because the protein purification protocol used in the present study yields both monomeric and hexameric fractions of TIP49b (Fig. 1), we also analyzed hexa- mers side-by-side with monomers. TIP49b hexamers showed some residual DNA binding activity in gel-shift assays but supported neither ATP hydrolysis (Fig. 2B), nor DNA unwinding (Fig. 3A). Hence, hexameric TIP49b assemblies appear to be inactive. Because hexamers would ultimately need to be con- verted back into an active form of the protein, these findings, if biologically relevant, suggest that yet unknown protein cofactor(s) might be required to recycle TIP49b hexamers in cells. Materials and methods Protein purification A pET3a vector containing the coding sequence of human TIP49b was a kind gift from V. Ogryzko (IGR, Villejuif, France) [9]. The protein was expressed in the BL21(DE3) pLysS E. coli strain. A 2 L culture was grown in LB med- ium at 37 °C until D 600 of 0.5 was reached before induction with 1 mm isopropyl thio-b-d-glactoside for 6 h at 25 °C. Cells were lysed in 100 mL of a buffer containing 20 mm Tris–HCl (pH 8.0), 500 mm NaCl, 10% glycerol, 1 mm dithiothreitol and 10 mm imidazole for 30 min on ice in the presence of lysozyme at 1 mgÆmL )1 and protease inhibitor cocktail tablets (Roche Diagnostics, Basel, Switzerland) and sonicated on ice for 3 · 10 s. The clarified supernatant was applied to a Ni 2+ -sepharose column (HisTrapÔ FF, 5 mL; GE Healthcare, Milwaukee, WI, USA) equilibrated with the same buffer, then washed in the presence of 10 mm imidazole and subjected to two step elutions with 100 and 500 mm imidazole using a buffer containing 20 mm Tris– HCl (pH 8.0), 100 mm KCl, 10% glycerol and 1 mm dith- iothreitol. Two TIP49b-containing fractions were detected. Fraction ‘low imidazole’ was eluted from the column dur- ing the 100 mm imidazole step, whereas the rest of the pro- tein eluted at 500 mm (‘high imidazole’). Gel filtration on a HiLoad 16 ⁄ 60 Superdex S-200 PG column (GE Healthcare) followed by SDS ⁄ PAGE revealed that the ‘low imidazole’ fraction contained monomers used throughout the study after gel filtration (fraction 2), hexamers (fraction 1) and high molecular-weight aggregates. Where indicated, TIP49b hexamers were from fraction 1. TIP49b and Rvb2p prepa- rations were routinely controlled by LC-MS ⁄ MS for the absence of contamination by bacterial ATPases and heli- cases. In all cases, pooled purified monomer and hexamer fractions were quantified using the Bradford assay and aliquots were used directly without additional concentra- tion. The pET-9aSN1 vector (a kind gift from S. Cheruel, IBBMC, University Paris-Sud, Orsay, France) containing the coding sequence of the yeast Rvb2 protein was used to transform E. coli BL21(DE3) STAR. Transformed cells were grown in 2TY medium at 37 °C until D 600 of 0.8 was reached. Expression of C-terminal His-tagged Rvb2p was induced with 0.5 mm isopropyl thio-b-d-glactoside for 3 h. Cells were lysed in 50 mm Tris–HCl (pH 7.5), 300 mm NaCl and 5 mm b-mercaptoethanol. After sonication and centrifugation for 30 min at 25 000 g, the clarified lysate was applied to a Fast-Flow Ni-NTA agarose column (Qiagen, Valencia, CA, USA). Rvb2p was eluted with a 10–300 mm imidazole gradient in lysis buffer. Pooled peak fractions were diluted ten-fold in buffer containing 25 mm Bis-Tris propane (pH 6.5), 5 mm b-mercaptoethanol and loaded onto a MonoQ 5 ⁄ 50 GL column (GE Healthcare). Elution was performed with a 0–1 m NaCl gradient. Frac- tions containing pure Rvb2p were collected and dialyzed against 20 mm Tris–HCl (pH 8.0), 100 mm KCl, 10% glyc- erol and 1 mm dithithreitol. Rvb2p preparations were found to be a mixture of dimers and monomers, as judged by gel filtration. TIP49b and aFLAG antibodies were purchased from BD Biosciences (Franklin Lakes, NJ, USA, USA) and from Sigma (St Louis, MO, USA), respectively. RecA protein was purchased from New England Biolabs (New England Biolabs, Beverly, MA, USA). DNA substrates The DNA substrates for gel-shift experiments and helicase assays were prepared by annealing of equimolar amounts of the corresponding synthetic oligonucleotides in a buffer containing 10 mm Tris–HCl (pH 7.5), 1 mm EDTA and 100 mm NaCl and analyzed by native gel electrophoresis. The specificity of DNA unwinding is defined here by the protruding ss-tail. The oligonucleotides used are shown in Table S1. ATPase assays The ATPase activity of the proteins (1 lm) was measured in a 5 lL reaction volume containing 25 mm Hepes-KOH (pH 8.0), 2.5 mm Mg(CH 3 COO) 2 , 100 mm KCl, 0.2 mm dithiothreitol, 100 lgÆmL )1 BSA (Sigma) and 0.6 lCiÆlL )1 [a-P 32 ] ATP, and unlabeled ATP up to 2 mm. This reaction buffer was used throughout the study. The reaction DNA unwinding by TIP49b C. Papin et al. 2712 FEBS Journal 277 (2010) 2705–2714 ª 2010 The Authors Journal compilation ª 2010 FEBS mixtures were incubated at 37 °C for 0, 2, 5, 10 and 30 min, stopped on ice and analyzed by TLC on PEI-Cellu- lose plates (Merck, Darmstadt, Germany) using 0.75 m KH 2 PO 4 as a migration buffer; plates were then dried and quantified. DNA binding TIP49b or Rvb2p recombinant proteins at 0–1 lm concen- trations were incubated in a 10 lL reaction volume con- taining 25 mm Hepes-KOH (pH 8.0), 2.5 m m Mg(CH 3 COO) 2 , 100 mm KCl, 0.2 mm dithiothreitol and 100 lgÆmL )1 BSA (Sigma). DNA substrates at the indicated concentrations were incubated for 45 min and electrophore- sed on native 8% polyacrylamide gels (19 : 1) using 1 · TBE as a running buffer. Gels were dried and quanti- fied on a Fuji BAS 3000 phosphorimager (Fuji Life Sciences, Tokyo, Japan). Helicase assays 5¢-to3¢ or 3¢-to5¢ helicase substrate (1 nm) was preincu- bated with 0.1 or 1 lm TIP49b or Rvb2p as indicated for 15 min on ice. The reaction mixtures were complemented or not with 1 mm nucleotide cofactor, as indicated, and incubated at 37 °C. A ten-fold excess of a trap oligonucleo- tide (complementary to the unwound radio-labeled probe) was added to the reaction mixture to prevent reannealing. The reactions were stopped by the addition of 2 lLofa solution containing 1 mgÆmL )1 proteinase K, 1.25% SDS, 10 mm Tris–HCl, 0.06% bromophenol blue, 0.06% xylene cyanol and 30% glycerol in the presence of 100-fold excess of the trap oligonucleotide. The samples were analyzed by native electrophoresis on 8% polyacrylamide gels using 1 · TBE as running buffer. Gels were dried and quantified on a Fuji BAS 3000 phosphorimager. Electron microscopy TIP49b or Rvb2p (2 lm) were spread on carbon-coated copper grids (100 or 400 mesh). After 30–60 s, the drops were blotted dry and the grids were stained with 1% uranyl acetate for 1 min, blotted again, and washed or not once in water for 30 s. Grids were dried and examined with a trans- mission electron microscope (ME Hitachi 200 kV; Hitachi, Tokyo, Japan). Acknowledgements We thank Simon Lebaron for RNA samples; Ross Tamaino for mass spectroscopy; Ste ´ phanie Balor and Nathalie Laviolette for technical help; Patrick Schultz for help with electron microscopy; Dave Lane for comments on the manuscript; and Martine Obadia for helpful discussions. This work was supported by grants from the Agence Nationale pour la Recherche (ANR grant ‘DNAMOTORS’ No. 143704) and the Association pour la Recherche sur le Cancer (ARC) to M.G. and E.K., Universite ´ Paul Sabatier and the Centre National de la Recherche Scientifique (CNRS). K.E. was supported by a grant from the Agence Nationale pour la Recherche (ANR-05-JCJC No. 015101) to S.M. C.P. and A.K. were recipients of PhD fellowships from the French Ministry for Research. C.P. received additional fellowship support from the ARC. References 1 Jha S & Dutta A (2009) RVB1 ⁄ RVB2: running rings around molecular biology. Mol Cell 34, 521–533. 2 Ammelburg M, Frickey T & Lupas AN (2006) Classifi- cation of AAA+ proteins. J Struct Biol 156, 2–11. 3 Erzberger JP & Berger JM (2006) Evolutionary relation- ships and structural mechanisms of AAA+ proteins. Annu Rev Biophys Biomol Struct 35, 93–114. 4 Iyer LM, Leipe DD, Koonin EV & Aravind L (2004) Evolutionary history and higher order classification of AAA+ ATPases. J Struct Biol 146, 11–31. 5 Baek SH (2008) When ATPases pontin and reptin met telomerase. Dev Cell 14, 459–461. 6 Gallant P (2007) Control of transcription by pontin and reptin. Trends Cell Biol 17, 187–192. 7 Jonsson ZO, Dhar SK, Narlikar GJ, Auty R, Wagle N, Pellman D, Pratt RE, Kingston R & Dutta A (2001) Rvb1p and Rvb2p are essential components of a chro- matin remodeling complex that regulates transcription of over 5% of yeast genes. J Biol Chem 276, 16279–16288. 8 King TH, Decatur WA, Bertrand E, Maxwell ES & Fournier MJ (2001) A well-connected and conserved nucleoplasmic helicase is required for production of box C ⁄ D and H ⁄ ACA snoRNAs and localization of snoRNP proteins. Mol Cell Biol 21, 7731–7746. 9 Ikura T, Ogryzko VV, Grigoriev M, Groisman R, Wang J, Horikoshi M, Scully R, Qin J & Nakatani Y (2000) Involvement of the TIP60 histone acetylase complex in DNA repair and apoptosis. Cell 102, 463– 473. 10 Jonsson ZO, Jha S, Wohlschlegel JA & Dutta A (2004) Rvb1p ⁄ Rvb2p recruit Arp5p and assemble a functional Ino80 chromatin remodeling complex. Mol Cell 16, 465–477. 11 Matias PM, Gorynia S, Donner P & Carrondo MA (2006) Crystal structure of the human AAA+ protein RuvBL1. J Biol Chem 281, 38918–38929. 12 Qiu XB, Lin YL, Thome KC, Pian P, Schlegel BP, Weremowicz S, Parvin JD & Dutta A (1998) An C. Papin et al. DNA unwinding by TIP49b FEBS Journal 277 (2010) 2705–2714 ª 2010 The Authors Journal compilation ª 2010 FEBS 2713 eukaryotic RuvB-like protein (RUVBL1) essential for growth. J Biol Chem 273, 27786–27793. 13 Kanemaki M, Kurokawa Y, Matsu-ura T, Makino Y, Masani A, Okazaki K, Morishita T & Tamura TA (1999) TIP49b, a new RuvB-like DNA helicase, is included in a complex together with another RuvB-like DNA helicase, TIP49a. J Biol Chem 274, 22437–22444. 14 Makino Y, Kanemaki M, Kurokawa Y, Koji T & Tam- ura T (1999) A rat RuvB-like protein, TIP49a, is a germ cell-enriched novel DNA helicase. J Biol Chem 274, 15329–15335. 15 Gribun A, Cheung KL, Huen J, Ortega J & Houry WA (2008) Yeast Rvb1 and Rvb2 are ATP-dependent DNA helicases that form a heterohexameric complex. J Mol Biol 376, 1320–1333. 16 Puri T, Wendler P, Sigala B, Saibil H & Tsaneva IR (2007) Dodecameric structure and ATPase activity of the human TIP48 ⁄ TIP49 complex. J Mol Biol 366, 179– 192. Supporting information The following supplementary material is available: Table S1. Oligonucleotides used in this study. This supplementary material can be found in the online version of this article. Please note: As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors. DNA unwinding by TIP49b C. Papin et al. 2714 FEBS Journal 277 (2010) 2705–2714 ª 2010 The Authors Journal compilation ª 2010 FEBS . sup- ports a 3¢ -to5 ¢ DNA unwinding activity that requires a 3¢-protruding tail ‡ 30 nucleotides. The data obtained indicate that DNA unwinding of 3¢ -to 5¢ junctions. of approxi- A BC Fig. 3. DNA unwinding by TIP49b is preferentially exerted on 3¢ -to5 ¢-ss ⁄ ds junctions. (A) Unwinding by 1 lM TIP49b (left, lanes 1–9)