Letter pubs.acs.org/JPCL
Probe Position-Dependent Counterion Dynamics in DNA: Comparison of Time-Resolved Stokes Shift of Groove-Bound to BaseStacked Probes in the Presence of Different Monovalent Counterions Sachin Dev Verma, Nibedita Pal, Moirangthem Kiran Singh, and Sobhan Sen* Spectroscopy Laboratory, School of Physical Sciences, Jawaharlal Nehru University, New Delhi 110067, India S Supporting Information *
ABSTRACT: Time-resolved fluorescence Stokes shifts (TRFSS) of 4′,6-diamidino-2phenylindole (DAPI) inside the minor groove of DNA are measured in the presence of three different monovalent counterions: sodium (Na+), rubidium (Rb+), and tetrabutylammonium (TBA+). Fluorescence up-conversion and time-correlated single photon counting are combined to obtain the time-resolved emission spectra (TRES) of DAPI in DNA from 100 fs to 10 ns. Time-resolved Stokes shift data suggest that groove-bound DAPI can not sense the counterion dynamics because the ions are displaced by DAPI far from the probe-site. However, when these results are compared to the earlier base-stacked coumarin data, the same ions are found to affect the nanosecond dynamics significantly. This suggests that the ions come close to the probe-site, such that they can affect the dynamics when measured by basestacked coumarin. These results support previous molecular dynamics (MD) simulation data of groove-bound and base-stacked probes inside DNA. SECTION: Biophysical Chemistry and Biomolecules
W
inside DNA can sense the ion motions as this probe does not perturb the ion distributions around DNA.16 One way to observe DNA−ion interaction could be through the direct measurement of ion dynamics around DNA or by the measurement of ions’ effect on the overall DNA motion. Timeresolved fluorescence Stokes shift (TRFSS) and related experiments have the capability to measure the ion, hydration, and biomolecular dynamics directly on the time-scales of their motions.15−29 In general, TRFSS experiments follow the solvation dynamics of a system by recording the change in electrostatic interaction energy of a fluorescent probe with its surrounding environment.29 Upon optical excitation, the dipole moment of probe changes due to the perturbation in charge distribution in the probe. In order to stabilize the excited probe dipole, surrounding charged and dipolar molecules (in this case water, ions, and DNA) reorient themselves by exerting reaction electric field on the probe. This phenomenon decreases probe’s interaction energy with its surroundings that leads to the shifts in the fluorescence spectra of probe toward lower frequencies. Hence, the solvation dynamics is directly reflected in probe’s time-dependent fluorescence Stokes shifts.29 Even though DNA has much less structural diversity than proteins, recent TRFSS experiments show that the dynamics in DNA depend strongly on the position of probes inside DNA,15 effects similar to those seen in proteins.25−27,37 Molecular dynamics (MD) simulations, on the other hand, provided the
ater and ions play an important role in stabilizing the DNA double helical structure.1,2 Because DNA is a negatively charged polyion, counterions interact strongly with DNA.2 It is found that almost 75% of DNA charge can become neutralized by the mobile counterions present in the “condensation” layer surrounding the DNA within radius of ∼9 Å.3−5 Counterions can bind inside the grooves of DNA,4−7 sometimes inducing the groove-width8 and conformation of DNA.9 Binding of proteins to DNA is also found to critically depend on the specific positioning of ions inside DNA−protein complexes as well as the displacement of ions by protein residues.10−14 Building the molecular picture of DNA−ion interaction is essential to understand how the dynamic ions around DNA stabilize its structure and control its interaction with proteins and small molecules. Efforts have been made to understand the direct and/or indirect role of ions on the DNA structure,4−9 DNA−protein complexation,10−14 and in the interactions of DNA with small molecules.15,20,24,31,33 However, it is not known whether the ion dynamics around DNA can change when the DNA interacts with proteins and small molecules. The question remains whether the ion-dynamics depend on the ion-distributions around DNA when it forms complexes with proteins and small molecules. This Letter tackles these questions in the context of DNA interaction with small molecules and shows that a groovebound (probe) molecule can not sense the ion-dynamics as the ions are displaced by the groove-bound molecule far from the probe-site. However, a covalently attached base-stacked probe © 2012 American Chemical Society
Received: July 12, 2012 Accepted: August 31, 2012 Published: August 31, 2012 2621
dx.doi.org/10.1021/jz300934x | J. Phys. Chem. Lett. 2012, 3, 2621−2626
The Journal of Physical Chemistry Letters
Letter
Figure 1 compares the excitation and emission spectra of groove-bound DAPI to base-stacked coumarin13 in the
details of contribution of water, ion, and DNA motions that couple together to control the total dynamics in DNA.30−32 However, disagreements are persistent about the relative contributions of components to the total dynamics.32,34−36 Nonetheless, these studies imply that the observed coupled motion of water, ions, and biomolecule should depend on the position of probe inside the biomolecule where the dynamics is being measured. Furthermore, the total dynamics should depend on the relative motions of water, ions, and biomolecule that the probe can sense. Thus, changing the probe-position inside DNA would result in observable change in the ion dynamics, which would depend on whether the probe can sense the ion motions through modulation of ions’ electric-field at the probe-site. Long-range electrostatic coupling among the components, however, expects that the probe must sense similar ion dynamics, along with similar water and DNA dynamics, even if the position of probe is changed inside DNA. This paper will show that the former concept holds; that is, a groove-bound probe can not sense the ion motions as the probe displaces the ions far from the grooves, whereas the same ions can affect the dynamics when measured by a base-stacked probe that does not perturb the ion-distribution around DNA.16 This suggests that ions can access the base-stacked probe-site inside DNA. TRFSS experiments are extensively used to study dynamics in proteins.25−28 In DNA, the slow dynamics was first observed by following the TRFSS of a covalently attached base-stacked coumarin, opposite an abasic-site, inside DNA.17 Subsequently, dynamics in DNA were measured using other base-stacked and groove-bound probes.15,18−24 MD simulations were also performed to gain molecular details of the dynamics and compare MD results with TRFSS experiments.4−6,30−35 However, all these studies focused on DNA dynamics with only Na+ as counterion. Sen et al. reported the first elaborate nanosecond TRFSS study on counterion dynamics in DNA by replacing Na+ with eight other monovalent counterions.16 This study found that in the presence of small alkali ions (except Rb+), the coumarin can sense similar ion dynamics that follow a power-law, but, for larger ions, the dynamics beyond ∼1 ns depend on the hydrodynamic size of the ions.16 Here, we measure the TRFSS of a minor groove binder, 4′,6diamidino-2-phenylindole (DAPI), in 14-mer DNA, d(5′CGCGCAATTGCGCG-3′)2, in the presence of three different monovalent counterions of different sizes: sodium (Na+), rubidium (Rb+) and tetrabutylammonium (N(C4H9)4+ or TBA+) (see Supporting Information for Materials and Methods). Time-resolved emission spectra (TRES) of DAPI in DNA are measured from 100 fs to 10 ns independently in the presence of the three ions. Results show identical Stokes shifts of DAPI in the presence of all three counterions. These results are then compared to the ion-dependent dynamics measured earlier using a covalently attached base-stacked coumarin, which replaces a base-pair inside DNA, in the presence of the same three ions (and also other ions).16 The dynamics measured by coumarin show significant difference in the nanosecond time-range among the three ions.16 We choose only three ions for this study based on the fact that Sen et al. found maximum ion dependence for these three ions with basestacked coumarin, where Na+ shows a power-law dynamics, Rb+ shows anomalous behavior, and TBA+ affects slow dynamics the maximum.16 Thus, it is expected DAPI would show maximum ion dependence, if there is any, in the presence of these three counterions.
Figure 1. Normalized fluorescence emission and excitation spectra of DAPI bound to minor groove of DNA (solid lines), and base-stacked coumarin (dashed lines) in the presence of Na+ (blue), Rb+ (red), and TBA+ (N(C4H9)4+) (green). Data show substantial ion effects on coumarin spectra.
presence of the three counterions. DAPI spectra are identical, which imply that there are no ion effects. However, appreciable change is found in the spectral shapes and peak positions of coumarin in the presence of the same ions. The spectra in the presence of Rb+ are broader compared to other ions, whereas, TBA+ spectra are blue-shifted compared to Na+.16 Although the dynamics of water is complete within a few picoseconds,29 in DNA, the dynamics is found to spread over femtoseconds to nanoseconds.15−19,24 To capture this broad dynamic range (from 100 fs to 10 ns), we combine the TRES of DAPI-DNA, constructed independently from the decays measured in fluorescence up-conversion (UPC, TRES range: 100 fs to 1 ns) and time-correlated single photon counting (TCSPC, TRES range: ∼30 ps to 10 ns) (Supporting Information). Figure 2 shows three raw decays of DAPI (see
Figure 2. Time-resolved fluorescence decays (with fits) of DAPI bound to the minor groove of DNA in the presence of Na+ (blue), Rb+ (red), and TBA+ (N(C4H9)4+) (green). (A) Data collected in the UPC technique (up to 1.5 ns), and (B) data collected in TCSPC. Decays at three wavelengths are compared here to minimize clumsiness. See Supporting Information for other decays. 2622
dx.doi.org/10.1021/jz300934x | J. Phys. Chem. Lett. 2012, 3, 2621−2626
The Journal of Physical Chemistry Letters
Letter
Supporting Information for all decays). The decays are almost identical, which suggest that the relative ion-dependence is indistinguishable among the ions. The sums of 3−4 exponentials are used to fit the decays. TRES were constructed using the fitted parameters and steady-state fluorescence, such that the TRES at common time points in UPC and TCSPC match (Supporting Information). Figure 3 plots the TRES of
Figure 3. TRES constructed from UPC (circles) and TCSPC (stars) decays in the presence of Rb+. Matching of TRES from UPC and TCSPC at a common time point (100 ps) is shown. Dashed or solid lines through points denote log-normal fits. Red solid curve is timezero glass spectrum measured in a dry ice/acetone mixture at −78 °C.
DAPI in the presence of Rb+. Similar data are obtained for Na+ and TBA+ (Supporting Information). Log-normal fits to TRES are used to calculate the time-dependent Stokes shifts. The shifts are reported in terms of mean (1st moment) frequencies (Supporting Information). Figure 3 includes the time-zero spectrum of DAPI-DNA with Rb+ in frozen glass, measured in dry ice/acetone mixture at −78 °C. Glass spectrum represents time-zero position, because in glass all diffusive dynamics in sample are frozen, although vibrational and phonon-like inertial motions can persist.15−19 The mean frequency of glass spectrum, thus, gives the timezero position from where the diffusive solvation dynamics starts at room temperature. Hence, the dynamics is reported in terms of “absolute” Stokes shifts as S(t) = vg̅ lass − v(t) ̅ (see Figure 4). Figure 4 compares the Stokes shift dynamics of groovebound DAPI and base-stacked coumarin in the presence of Na+, Rb+, and TBA+. The plots immediately reveal that the DNA dynamics measured with DAPI are identical (within error limit) in the presence of all three counterions. Figure 4A compares only the slow part of the dynamics from 40 ps to 10 ns (base-stacked coumarin data in Rb+ and TBA+ did not extend below 40 ps).16 This plot shows substantial ion dependence in the nanosecond dynamics when measured with coumarin,16 but not with DAPI. The absolute Stokes shifts of coumarin at 40 ps can vary within a large range from 1150 cm−1 to 1425 cm−1 among the ions, whereas for DAPI the shifts are similar (∼1530 cm−1). Coumarin TRFSS in the presence of Rb+ and TBA+ did not extend below 40 ps;16 hence, the direct comparison of coumarin and DAPI data in faster time-scales is not possible. Nonetheless, the large difference in the absolute Stokes shifts of coumarin at 40 ps suggests that there are substantial differences in the subpicosecond and femtosecond dynamics among the three counterions. To obtain a quantitative estimation of the slow iondependent dynamics, we analyzed the absolute Stokes shift data of DAPI and coumarin in the long time-range from 40 ps to 10 ns (see Supporting Information). The sum of two
Figure 4. (A) Comparison of absolute Stokes shift in DNA in the time-range of 40 ps to 10 ns probed by groove-bound DAPI (circles), and base-stacked coumarin (stars)16 in the presence of Na+ (blue), Rb+ (red), and TBA+ (green). Lines through points show the fits to data (see Supporting Information and text for the fitted results). (B) Comparison of absolute Stokes shifts from 100 fs to 10 ns of DAPI in DNA in the presence of Na+, Rb+, and TBA+ with that of coumarin in the presence of only Na+. (C) Comparison of the solvation correlation function C(t) of DAPI and coumarin, constructed from the Stokes shifts in panel B.
exponentials is used to fit the DAPI data. All three data are fitted with similar (exponential) time constants (τ1 = 260 ps, τ2 = 5 ns). However, the coumarin data in Na+ could be fitted with a single power-law (of exponent 0.15).16,18,19 TBA+ and Rb+ data could only be fitted with a power-law plus an exponential (Figure 4A).16 Fits to TBA+ data shows an initial power-law dynamics, similar to that in Na+, and switching of this dynamics to exponential relaxation (τ = 3.2 ns) beyond ∼1 ns.16 Dynamics in Rb+, however, is seen to follow a different power-law (of exponent 0.23) initially, and then switch to slower exponential relaxation (τ = 7.3 ns) compared to TBA+ (see Supporting Information). Berg and co-workers extended the TRFSS measurement of base-stacked coumarin in DNA in the presence of Na+ over six decades of time from 40 fs to 40 ns.18,19 They found that the dynamics follow single power-law (with exponent 0.15) over the entire six decades from 40 fs to 40 ns.18,19 However, within the time-range of 40 ps to 10 ns, TBA+ and Rb+ data initially follow power-law, but beyond a certain time the dynamics switch to exponential type. These results suggest that there is substantial ion dependence on the dynamics when they are measured by a probe that is incorporated as part of DNA and does not perturb the ion distributions around DNA.16 Previous X-ray,38 NMR,39−41 and MD simulation4,8 studies found that ions can bind differently to different base-sequences, 2623
dx.doi.org/10.1021/jz300934x | J. Phys. Chem. Lett. 2012, 3, 2621−2626
The Journal of Physical Chemistry Letters
Letter
still an issue.42 Such simulations may have the predictive ability to show that intrinsic ion dynamics is slowed in proportion to the ions’ diffusion constants (i.e., hydrodynamic size), which then reassemble with water and DNA components to make the full signal that is similar to what is found in TRFSS experiments.16 Contrary to coumarin data, the same ions are found not to affect the dynamics when measured by groove-bound DAPI. In a recent report, Pal et al. compared the dynamics of DAPI with coumarin in the presence of only Na+ over the time range of 100 fs to 10 ns.15 Results show that both DAPI and coumarin follow the same power-law from 100 fs to ∼100 ps, but DAPI dynamics deviate from the power-law beyond ∼100 ps and converge rapidly to equilibrium showing exponential type relaxation.15 In the present study, we compare the absolute Stokes shifts (Figure 4B) and solvation correlation functions (C(t) = (S(∞) − S(t))/S(∞)) (Figure 4C) of DAPI in the presence of Na+, Rb+, and TBA+ with that of coumarin in the presence of only Na+ (coumarin data in Rb+ and TBA+ did not extend below 40 ps). As can be seen, the DAPI dynamics in all three counterions merge within the entire time range, showing similar power-law dynamics before ∼100 ps and deviation from this power-law beyond ∼100 ps. Furse and Corcelli simulated DNA with groove-bound Hoechst in the presence of Na+.31,34,35 They decomposed the overall energy correlation function following the linear response approach and found that Na+ motion contributes merely (
