J. Phys. Chem. B 2008, 112, 5997-6007
5997
New Insights into the Transition Pathway from Nonspecific to Specific Complex of DNA with Escherichia Coli Integration Host Factor† Paula Vivas,£ Serguei V. Kuznetsov,£ and Anjum Ansari*,£,‡ Department of Physics (M/C 273), and Department of Bioengineering (M/C 063), UniVersity of Illinois at Chicago, 845 West Taylor Street, Chicago, Illinois 60607 ReceiVed: July 30, 2007; In Final Form: January 25, 2008
To elucidate the nature of the transition-state ensemble along the reaction pathway from a nonspecific proteinDNA complex to the specific complex, we have carried out measurements of DNA bending/unbending dynamics on a cognate DNA substrate in complex with integration host factor (IHF), an architectural protein from E. coli that bends its cognate site by ∼180°. We use a laser temperature jump to perturb the IHF-DNA complex and monitor the relaxation kinetics with time-resolved FRET measurements on DNA substrates end-labeled with a FRET pair. Previously, we showed that spontaneous bending/kinking of DNA, from thermal disruption of base-pairing/-stacking interactions, may be the rate-limiting step in the formation of the specific complex (Kuznetsov, S. V.; Sugimura, S.; Vivas, P.; Crothers, D. M.; Ansari, A. Proc. Natl. Acad. Sci. USA 2006, 103, 18515). Here, we probe the effect of varying [KCl], which affects the stability of the complex, on this rate-limiting step. We find that below ∼250 mM KCl, the observed relaxation kinetics are from the unimolecular bending/unbending of DNA, and the relaxation rate kr is independent of [KCl]. Above ∼300 mM KCl, dissociation of the IHF-DNA complex becomes significant, and the observed relaxation process includes contributions from the association/dissociation step, with kr decreasing with increasing [KCl]. The DNA bending step occurs with a positive activation enthalpy, despite the large negative enthalpy change reported for the specific IHF-DNA complex (Holbrook, J. A.; Tsodikov, O. V.; Saecker, R. M.; Record, M. T., Jr. J. Mol. Biol. 2001, 310, 379). Our conclusion from these studies is that in the uphill climb to the transition state, the DNA is kinked, but with no release of ions, as indicated by the salt-independent behavior of kr at low [KCl]. Any release of ions in the unimolecular process, together with conformational changes in the protein-DNA complex that facilitate favorable interactions and that contribute to the negative enthalpy change, must occur as the system leaves the transition state, downhill to the final complex.
Introduction Gene regulation frequently involves formation of proteinDNA complexes that require special proteins that kink, bend, or curve DNA.1 Sharply bent DNA is also critical for the packaging of DNA inside the cell.2 Specific binding of proteins to DNA also induces extensive conformational rearrangements in the protein to facilitate favorable interactions with DNA.3,4 Many DNA binding proteins find their target sites by “indirect readout” of the DNA sequence. Rather than direct interactions between the protein and specific DNA base pairs, this mechanism relies on the sequence dependence of the DNA’s conformation and deformability. Indirect readout is particularly important for proteins that bend DNA, including many involved in DNA packaging, transcription, and repair, and the conformational flexibility of both the protein and the DNA is believed to play an important role in the ability of proteins to recognize their DNA binding sites. In this study, we focus on the dynamics of binding and bending of a cognate DNA substrate, the H′ binding site from λ-phage DNA, to integration host factor (IHF), an architectural protein from E. coli. The eubacterial IHF/HU family of DNA†
Part of the “Attila Szabo Festschrift”. * To whom correspondence should be addressed. E-mail:
[email protected]. Phone: (312) 996-8735. Fax: (312) 996-9016. £ Department of Physics. ‡ Department of Bioengineering.
bending proteins are ubiquitous in prokaryotes and aid in chromosomal compaction as well as in the assembly of higherorder nucleoprotein complexes necessary for replication initiation and site-specific recombination.5 The cocrystal structure of the IHF-H′ complex shows that the DNA is bent by nearly 180° over ∼35 base pairs (Figure 1).6 A comparison of the structures of the proteins from this family, with and without bound DNA, reveals that flexible β-ribbon arms of the protein, which are disordered in the structures without bound DNA, are wrapped around the DNA substrate and lie in the minor groove of the DNA (Figure 2). Two sharp kinks in the DNA, separated by about nine base pairs, are stabilized by the intercalation of highly conserved proline residues located in the β-arms. This family of proteins interacts exclusively with the minor groove of the DNA, making very few contacts with the bases themselves.7 Thus, these proteins have become paradigms for investigating the indirect readout mechanism. Thermodynamic studies on a wide range of protein-DNA complexes have provided a wealth of information on the various contributions to their stability and specificity. However, a detailed picture of the dynamics of conformational rearrangements in proteins and DNA, which are key to the recognition of DNA binding sites, is lacking. Fundamental questions that remain to be resolved are: What is the sequence of molecular rearrangements in the protein-DNA complex that lead to the transition from the nonspecific to the specific complex; and what
10.1021/jp076042s CCC: $40.75 © 2008 American Chemical Society Published on Web 05/08/2008
5998 J. Phys. Chem. B, Vol. 112, No. 19, 2008
Vivas et al.
Figure 3. Kinetic scheme for binding of IHF to DNA, with a bimolecular association/dissociation step with corresponding rates of kns and k-ns to form the nonspecific complex and a unimolecular DNA bending/unbending step with corresponding rates kbend and kunbend to form the specific complex.
Figure 1. (a) Structure of IHF bound to the H′ site from bacteriophage λDNA, reproduced with permission from Swinger et al.41 (Copyright 2003, Nature Publishing Group). (b) The H′ binding site used in this study, with fluorescein and TAMRA attached to the thymine at the 5′ end of the top and bottom strand, respectively. The position of the kinks in the crystal structure is indicated by the two arrows. The crystal structure was obtained with a nicked H′ substrate, with a nick in the sugar phosphate backbone located one position to the right of the left arrow.6
Figure 2. The structure of (a) the IHF-H′ complex (PDB code 1IHF) and (b) E. coli HU (PDB code 1HUE). The R- and β-chains are shown in blue and red, respectively. A portion of the bound H′ substrate is shown in yellow in panel (a). Comparison of the structures in (a) and (b) show the large conformational changes in the β-ribbon arms of the protein when the DNA substrate is bound. The proline residues that intercalate between the kinks in the DNA are indicated in green.
role does the DNA flexibility/“bendability” play in the recognition mechanism? To elucidate the energetics and mechanism underlying site recognition requires kinetics measurements that can probe the conformational dynamics in the protein-DNA complex. The kinetics of protein-DNA interactions for several systems in which the DNA is bent in the complex have been investigated primarily with stopped-flow techniques.8-12 Until recently, the limited time resolution of stopped-flow measurements failed to resolve the dynamics of DNA bending from the overall bimolecular association step.8-11 We have started a detailed investigation of the DNA bending/unbending dynamics in IHFDNA complexes, using laser temperature jump (T jump) to perturb the complex. The enhanced time resolution of ∼200 ns of our laser T jump spectrometer provides a powerful complement to stopped-flow and single-molecule techniques to probe
the dynamics in protein-DNA complexes on the submicro-tomillisecond time scales, which is precisely the time scale on which large-scale conformational fluctuations are known to occur at the protein-DNA interface in the nonspecific complex, prior to forming the specific complex.13 T jump measurements in our laboratory, in combination with stopped-flow measurements carried out in the laboratory of Professor Donald Crothers at Yale, provided the first direct observation of the bending/unbending kinetics of a 35 base pair duplex DNA oligomer containing the H′ binding site (Figure 1b).12,14 These measurements resolved a long-standing question in this field as to whether binding and bending of DNA by DNA-bending proteins occurs in a concerted manner, or sequentially,15 and demonstrated distinct binding and bending steps, as illustrated in the kinetic scheme of Figure 3. More importantly, these measurements revealed that the rates and activation energy of DNA bending, in complex with IHF, are similar to those of a single A-T base pair opening inside of duplex DNA, indicating that transient thermal disruption in baseparing and/or -stacking interactions could lead to spontaneous bending/kinking of the DNA, which may be sufficient to overcome the free-energy barrier separating the nonspecific from the specific complex.14 Spontaneous bending/kinking of DNA, in the absence of a bound protein, has not been directly observed, although it has been implicated in recent studies,16-20 raising the question as to whether the DNA is capable of undergoing thermal fluctuations in which distorted conformations are energetically accessible.14,21-25 Here, we report the effect of changes in the ionic strength on the relaxation kinetics of the IHF-DNA complex in response to a T jump perturbation. The binding of proteins to DNA depends strongly on the ionic environment in solution due to the polyelectrolyte nature of DNA, which results in the condensation of high concentrations of cations in the vicinity of DNA.26-28 The uptake and release of ions by the protein, which is believed to be accompanied by protein conformational changes, also plays a significant role in the observed salt dependence.29 These ion effects must be reflected in the salt dependence of each of the steps indicated in the kinetic scheme of Figure 3. The observed effect of changes in the ionic strength on the unimolecular bending/unbending process reveals new insights into the nature of the transition-state ensemble and provides a glimpse into the sequence of molecular rearrangements along the transition pathway that includes DNA bending, ion release, and coupled protein conformational changes. Experimental Methods Materials. All DNA oligonucleotides were synthesized by the Keck Foundation (Yale University). The oligonucleotides were labeled with fluorescein (Fl) and TAMRA (Rh) by
Transition Pathway from Nonspecific to Specific Complex incorporating Fluorescein-dT and TAMRA-dT (Glen Research, Sterling, VA) at the 5′ ends of the top and bottom strands, respectively, as indicated in Figure 1b. Concentrations of the single-stranded oligomers were determined by measuring the absorbance at 260 nm, using the extinction coefficients for the bases from Borer,30 and extinction coefficients of 38 800 M-1 cm-1 for Fl-dT and 32 300 M-1 cm-1 for Rh-dT, obtained from Glen Research. The extinction coefficient was calculated to be 4.0 × 105 M-1 cm-1 for the Fl-labeled strand (top strand in Figure 1b) and 4.15 × 105 M-1 cm-1 for the Rh-labeled strand (bottom strand in Figure 1b). To verify the extent of labeling, Fl and Rh concentrations were also determined in the labeled samples by measuring the absorbance of Fl-labeled strands at 494 nm, using 75 000 M-1 cm-1 for Fl-dT, and Rhlabeled strands at 556 nm, using 89 000 M-1 cm-1 for Rh-dT, and comparing them with the concentrations of the oligomers obtained from measurements at 260 nm. To make duplex DNA, equal molar concentrations of the complementary strands were mixed in the appropriate buffer conditions, and the sample was heated to 90 °C followed by slow cooling to room temperature to allow for complete annealing. The IHF protein was a generous gift from the laboratory of Professor Phoebe Rice at the University of Chicago. IHF concentrations were obtained from absorbance measurements at 276 nm, with an extinction coefficient of 5800 M-1 cm-1.31 All measurements were carried out in 20 mM Tris-HCl and 1 mM EDTA, pH 8, with KCl concentrations ranging from 100 to 500 mM. Equilibrium FRET Measurements. The steady-state fluorescence emission spectra were measured on a FluoroMax2 spectrofluorimeter (Jobin Yvon, Inc., NJ). The equilibrium FRET measurements were made by measuring the fluorescence emission spectra of double-labeled, Fl-H′-Rh, substrates, with and without bound IHF, in the wavelength range of 500-600 nm, with excitation at 485 nm. The emission spectra were obtained as a function of temperature, with the cuvette temperature controlled by a circulating water bath. The sample temperature was measured using a thermistor (YSI 44008; YSI Inc., Yellow Springs, OH) in direct contact with the sample cell. The efficiency of energy transfer (E) between the donor and acceptor is given by E ) 1 - IDA/ID, where ID is the donor fluorescence intensity of the single-labeled (Fl-DNA) substrate and IDA is the donor fluorescence intensity of the double-labeled (Fl-DNA-Rh) substrate under identical conditions (e.g., with bound protein).32 Measurements on Fl-DNA and Fl-DNARh substrates, with and without bound IHF, showed that the fluorescence emission spectra of Fl-DNA are unperturbed in the presence of IHF and that the donor intensities of Fl-DNA and Fl-DNA-Rh are identical, in the absence of IHF.33 Thus, for this system, the FRET efficiency characteristic of the complex can be obtained from E ) 1 - Icomplex/Ifree, where Icomplex is the donor florescence intensity of the Fl-DNA-Rh in complex with IHF and Ifree is the corresponding donor intensity in the absence of IHF. The error in our FRET efficiency measurements is about (0.05. The primary source of error is from variations in the IHF and DNA concentrations from one sample to another. Measurements of the Binding Affinity from Equilibrium FRET Titration. For the equilibrium titration measurements, about 20-25 samples were prepared containing final concentrations of 0.1, 1, or 5 nM labeled DNA for titration measurements at low [KCl] and 50 nM labeled DNA for titration measurements at high [KCl] (see Table 1). The titration experiments were carried out at 100, 200, 300, 350, and 400 mM [KCl]. For each
J. Phys. Chem. B, Vol. 112, No. 19, 2008 5999 [KCl], fluorescence emission spectra were collected for samples with varying concentrations of IHF. Each sample was thermally equilibrated prior to the acquisition of the emission spectra. The FRET values were calculated as described in the previous section. The resulting binding isotherm at each [KCl] concentration, obtained from plotting the FRET efficiency E versus [IHF], was fitted to the following equation to obtain the dissociation constant
E)
Elim [(Ptotal + Dtotal + KD) 2Dtotal
x(Ptotal + Dtotal + KD)2 - 4(PtotalDtotal)]
(1)
Here, Elim is the limiting FRET efficiency expected when all of the DNA is in complex, Ptotal and Dtotal are the total IHF and H′ concentrations, respectively, and KD is the dissociation constant. The two parameters that were varied to obtain the best fit are Elim and KD. An independent estimate of the dissociation constant was obtained by describing the FRET efficiency as a function of [KCl], obtained at 25 and 40 °C, for two sets of IHF-H′ concentrations, [7.5 µM: 5 µM] and [18 µM: 18 µM]. The FRET efficiency was described in terms of eq 1, and the observed dependence on [KCl] was parametrized by assuming a linear dependence of log KD versus log[KCl] with slope SKD, as follows
KD ) KD0
( ) [KCl] [KCl]0
SKD
(2)
Here, KD0 is the dissociation constant at a reference salt concentration [KCl]0, which was chosen to be 350 mM. The FRET efficiency versus [KCl] data for the two sets of IHF-H′ concentrations can thus be simultaneously described in terms of three parameters, Elim, KD0, and SKD, which were varied to obtain the best fit. Laser Temperature Jump Spectrometer. Our T jump spectrometer consists of a Q-switched Nd:YAG laser (Continuum Surelite II, fwhm ) 6 ns, 300 mJ/pulse at 1.06 µm) that is used to pump a 1 m long Raman cell consisting of highpressure methane gas to obtain ∼60 mJ pulses at 1.54 µm. The 1.54 µm beam, focused to ∼1 mm (fwhm) on the sample cell, is used to heat the sample. The probe source was a 20 mW cw diode laser at 488 nm (Newport PC13589), which was focused to a spot size of ∼300 µm at the center of the heated volume of the sample. The fluorescence emission was collected perpendicular to the excitation direction using a combination of a long-pass filter (>490 nm) and a short-pass filter (
