J. Phys. Chem. B 2005, 109, 22623-22628
22623
Diffusion Coefficient and the Secondary Structure of Poly-L-glutamic Acid in Aqueous Solution Keiichi Inoue, Naoki Baden, and Masahide Terazima* Department of Chemistry, Graduate School of Science, Kyoto UniVersity, Kyoto 606-8502, Japan ReceiVed: June 1, 2005; In Final Form: August 10, 2005
The diffusion coefficients (D) of poly-L-glutamic acid (PLG) at various pHs are investigated by the laserinduced transient-grating method with a new photoreactive probe molecule. The pH dependence of D is compared with that of the helical content of PLG measured by circular dichroism. It is found that the pH dependences of both quantities are very similar. Since the frictions of the translational diffusion of charged and protonated carboxyl groups are found to be similar each other, it is concluded that the conformation of the main polymer chain is the main factor in determining the diffusion process; in other words, the R-helix conformation makes the molecular diffusion faster. This result indicates that the conformational change of a protein can be detected by monitoring the diffusion coefficient.
1. Introduction Conformational changes of biological proteins are essential for their functions. By changing the conformation, the intermolecular interactions and the affinity for other molecules are modified to trigger the subsequent reaction. In particular, since hydrogen bonding plays an important role in protein systems, it is essential to probe the dynamics (creation or disruption) of the hydrogen bonding associated with the conformational change. Although vibrational spectroscopy, such as the infrared or the Raman scattering techniques, may be used for detecting the hydrogen bonding,1-4 it is difficult to distinguish the intermolecular from the intramolecular hydrogen bonding. Furthermore, there are various types of hydrogen bonding (e.g., N‚‚‚H or O‚‚‚H of many residues) in most proteins and such various hydrogen bondings are difficult to separate from the intramolecular vibrational modes. Here, we propose that the translational diffusion coefficient (D) is useful for monitoring the extent of the intermolecular interaction, in particular the hydrogen bonding with different conformations. In fact, Nishida et al. have found that the D value of the denatured structure of cytochrome c is much smaller (about half) than that of the native protein, and the protein-folding process was investigated by monitoring the time dependence of D.5 However, the relationship between D and conformations of macromolecules has not been clarified because the interpretation of the diffusion of a biological protein may be complicated because of the existence of the ternary structure. We focus our attention on the difference in D of a macromolecule with R helix and random-coil structures. Translational diffusion is a random motion of a molecule in solution. Einstein derived a relationship between D and the friction a hundred years ago. Combined with the Stokes equation, D is related with the friction f by6,7
D ) kBT/f
(1)
where kB is the Boltzmann constant and T is the temperature. If * To whom correspondence should be addressed. Phone: +81-75-7534026. Fax: +81-75-753-4026. E-mail:
[email protected].
there is no specific interaction between the diffusing molecule and the solvent, the friction of a spherical molecule in continuous medium is given by
f ) aηr where a is a constant reflecting the boundary condition, η is the viscosity, and r is the radius. This Stokes-Einstein relationship was derived for a spherical rigid body without a specific interaction. It is reasonable to consider that this friction increases from this predicted value, if there is additional intermolecular interactions.8 Hence, D should reflect the presence of intermolecular interactions. Although D is an important property for characterizing the state of molecules, investigations of D of macromolecules, in particular the relationship to the conformation, have not been well accumulated because of the difficulties of experimental methods. For measuring D, various experimental techniques, such as the capillary method, the sedimentation method, the Taylor dispersion method, NMR with a gradient field, dynamic light scattering, and fluorescence correlation spectroscopy, have been developed.6,7,9-12 However, each technique has characteristic advantages and disadvantages in terms of the sensitivity, temporal response, or mathematical treatment for data analysis.13,14 The development of experimental techniques has not progressed much recently. However, more recently, Baden and Terazima proposed a new method for the diffusion measurement using the laser-induced transient-grating (TG) technique.15 Although the TG method possesses unique merits (e.g., fast response, high sensitivity, and simple data analysis), there was a limitation for the general application to many molecular systems: the refractive index modulation should be induced by photochemical reactions. In other words, a photochemical reaction is necessary for monitoring the molecular diffusion process. To overcome this difficulty, the spatial modulation of the refractive index change was induced by a photoreactive cross-linkage reagent, N-hydroxysulfosuccinimidyl-4-azidobenzoate (HSAB). They demonstrated successful D measurements of many photochemically nonreactive proteins in solution.15 Here, we extend this research
10.1021/jp052897y CCC: $30.25 © 2005 American Chemical Society Published on Web 11/03/2005
22624 J. Phys. Chem. B, Vol. 109, No. 47, 2005 by using a new photoreactive probe molecule to study the relationship between the diffusion and the secondary structure, in particular the R-helix content of a polymer. We used a homopolymer, poly-L-glutamic acid (PLG), for this purpose. PLG is known to change its secondary structure from random coil to the R helix when the pH is decreased from neutral to less than 5.16-19 It has been a model system for a long time to elucidate basic information of the secondary structure of proteins. Therefore, the measurement of D for poly-L-glutamic acid at various pHs allows us to investigate the relationship between D and the conformation throughout the entire helixcoil transition. 2. Experimental Section The experimental setup for the TG measurement was similar to that reported previously.20-27 Briefly, a laser beam from an excimer laser (308 nm, XeCl; Lamda Physik Compex 102xc) with a 10 ns pulse was used as an excitation beam. The excitation beam was split into two by a beam splitter, and they were crossed inside a sample cell. The sample is photoexcited by the created interference pattern to induce the refractive index modulation in the sample. Continuous wave (cw) laser light from a diode laser (λ ) 780 nm) was used for the probe light. A part of the probe beam was diffracted by the refractive index modulation (TG signal). The signal was isolated from the excitation laser beam with a glass filter and a pinhole. The signal light was detected by a photomultiplier tube, and the temporal profile was recorded by a digital oscilloscope (Tektronix TDS520). The sample solution was stirred after each irradiation of the pump light to avoid sample depletion. Usually, 10 signals at each condition were averaged to improve the signal-to-noise ratio. For a typical TG measurement, the concentration of PLG was adjusted to 50 µM (3.9 mg/mL), and it was dissolved in a 25 mM NaCl aqueous solution. The sample pH was adjusted by HCl. A probe molecule, 4-[p-azidosalicylamido]butylamine (ASBA), was dissolved in dimethyl sulfoxide (DMSO), and 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (EDC) was added to the sample solution just before the measurement. Concentrations of ASBA and EDC were 50 µM (1:1 ratio to polymer concentration), and the concentration of DMSO was 2 mM. The sample solution was filtered with a cellulose acetate membrane filter (DISMIC 3 (pore size ) 0.45 µm); ADVANTEC) to remove dust, and the solution was put into a quartz cell (optical path length ) 2 mm). Any scattering probe light from the sample solution was carefully removed to avoid any heterodyne contribution to the TG signal. The helical content of PLG at each pH was measured by the CD intensity at far UV. PLG (10 µM (0.78 mg/mL)) was dissolved in a 25 mM NaCl aqueous solution with ASBA (10 µM), EDC (10µM) and DMSO (2mM). The sample solution was contained in a quartz cell (optical path length ) 2 mm) and measured with a spectropolarimater (J720W1, JASCO) with flowing N2 gas. The change of protonation degree (∆χ) of PLG at each pH was determined from the pH change of the solution by titration with HCl as described by Nakamura and Kidokoro.28 The solvent and the PLG solution, which were stirred and bubbled by N2 gas to minimize the effect of CO2, were titrated using HCl, and the pH was measured. The concentration and volume of the titrated PLG solution were 50 µM and 10 mL, respectively, and the reference solvent was a 25 mM NaCl aqueous solution. The pH change of the PLG solution from adding HCl was smaller than that of the reference solvent because of the
Inoue et al. protonation to the carboxylic groups. The amount of H+ bound to PLG was calculated from the difference in pH between the polymer solution and the reference solvent. The values from three independent measurements were averaged. PLG (typical molecular weight ) 7.5 × 104) and poly(4styrenesulfonic acid-co-maleic acid) (PSSMA, typical molecular weight ) 2.0 × 104) were purchased from Sigma-Aldrich Co., and ASBA and EDC were purchased from Pierce Co. These were used without further purification. 3. Principles and Analysis 3.1. TG Method. The principle of the TG measurement has been reported previously.20,21,23,26,29-31 In the TG experiment, a photoinduced reaction is initiated by the spatially modulated light intensity that is produced by the interference of two excitation light waves. The sinusoidal modulations of the concentrations of the reactant and the product lead to the modulation with the same pattern in the refractive index (δn) at the probe wavelength. This modulation can be monitored by the diffraction efficiency of a probe beam (TG signal). The intensity of the TG signal is proportional to δn2. The refractive index change after photoexcitation mainly comes from the thermal energy released (thermal grating) and the chemical species created (or depleted) by the photoreaction (species grating). The signal intensity becomes weaker as the spatial modulations of the refractive index become uniform, which is accomplished by thermal diffusion or translational mass diffusion. Solving the diffusion equation, one may find that the decay rate constant of the thermal grating signal should be Dthq2 (where Dth is thermal diffusivity of the solution and q is grating wavenumber).20-22 Similarly, the species grating signal decays with a rate constant of Dq2, where D is the diffusion coefficient of the chemical species. For a photochemical reaction system, the time development of the refractive index change (δn) can be expressed by a sum of exponential functions
δn(t) ) δnth exp(-Dthq2t) -
∑r δnr exp(-Drq2t) + ∑p δnp exp(-Dpq2t)
(2)
where δnr and δnp are the refractive index changes caused by the presence of the reactant and product, respectively. The sign of δnr is negative because the phase of the spatial concentration modulation is 180° shifted from that of the product. 3.2. Photolabeling by ASBA. The reaction scheme of ASBA with PLG is depicted in Scheme 1.32 First, EDC in solution binds to the carboxyl group of PLG. After the addition of ASBA, EDC bound to PLG is replaced by ASBA. Upon the irradiation with UV light, the azide group of the ASBA-PLG complex releases N2 and generates the nitrene. Because the refractive index of solution is changed by the photodissociation reaction, the TG signals corresponding to the parent molecule and product molecule are created. 4. Results and Discussion 4.1. pH Dependence of the Secondary Structure. Before the diffusion changes of PLG are shown, the helical contents at various pHs are described. Figure 1 depicts the CD spectra and intensity at 222 nm for PLG in our experimental solution at various pHs. The CD spectra are similar to those reported previously, which ensures that the solvent used in the present study did not affect the secondary structure of PLG.18,33 The
D and Secondary Structure of PLG in Solution
J. Phys. Chem. B, Vol. 109, No. 47, 2005 22625
Figure 2. TG signals of the ASBA solution (dashed line) and the solution containing PLG (50 µM) (solid line). The best-fitted curve by the biexponential function for the ASBA solution is also shown by the solid line overlapped with the ASBA signal.
to fit the CD intensity change between pH 3.3 and 7.1 with an empirical function
[θ]222 ) a + bpH +
Figure 1. (a) CD spectra of the PLG solution at various pHs (pH 7.13.3). Dashed line shows spectrum at an extremely low pH (pH 3.2). (b) The ellipsity of the PLG solution at 222 nm, which reflects the helical content of PLG (closed circles). The best-fitted curve using eq 3 is shown as the solid line in panel b.
SCHEME 1. Reaction Scheme of ASBA with PLG
CD intensity at 222 nm is a good indicator of the presence of the R helix.34,35 The weak CD intensity at 6 < pH < 7 shows that the polymer forms a random-coil structure.34,36 The CD intensity increases with decreasing pH from around a pH of 5.3 indicating the formation of a R helix.34,35 This change has been reasonably explained in terms of the electrostatic repulsion from the charges at the side chains (-COO-) in a high pH range for forming the random-coil conformation and the reduction of the repulsive destabilization in a lower pH range (-COOH) inducing R-helix formation.17,19,37 Taking into account the gradual changes in the pH < 4.5 and pH > 6 regions, we tried
a′ + b′pH pHhalf - pH 1 + exp e
(
)
(3)
where a, b, a′, b′, and e are fitting parameters and pHhalf corresponds to the pH at which half of this polypeptide forms an R helix. The fitting with this equation is very good, and the value of pHhalf was determined to be 5.28 ( 0.01. In the very low-pH (
