Low-frequency Raman spectra of lysozyme

VOL. 15, 219-225 (1976)

BIOPOLY MERS

Low-Frequency Raman Spectra of Lysozyme L. GENZEL, F. KEILMANN, T. P. MARTIN, G. WINTERLING, and Y. YACOBY,* Max-Planck Institut fur Festkorperforschung, Stuttgart, Federal Republic of Germany; H. FROHLICH, Department of Electrical Engineering, University of Salford, Salford, England; and MARVIN W. MAKINEN? Laboratory of Molecular Biophysics, Department of Zoology, Oxford University, Oxford, England

Synopsis New techniques in laser Raman spectroscopy are used to obtain spectra of aqueous solutions of lysozyme for frequency shifts as small as 5 ern-'. In addition, Raman measurements are made on two crystalline forms of hen egg white lysozyme. The spectra obtained from the solution and from the crystal are found to be similar for frequencies above 100 cm-'. However, a low-frequency band at 25 cm-' observed in crystalline lysozyme is not found in the solution, indicating that this band cannot be attributed to an internal molecular vibration.

INTRODUCTION It has been suggested that long-range forces are of importance in biological systems1.2 and that such forces may be created by dipoles induced by very low-frequency .vibrational modes.3 Recently, Raman bands a t frequencies smaller than 100 cm-l have been observed in solid proteins, such as films of native and chemically modified a-chymotrypsin.4 The observation was made possible by discriminating against the strong elastic Rayleigh scattering. It was hypothesized that the Raman bands arose from low-frequency internal vibrations of the protein mole~ u l e . In ~ the solid state, low-frequency vibrations can also arise from the linking forces between neighboring macromolecules. Thus, it was desirable to search for the existence of these vibrational modes in aqueous solution in order to establish whether the vibrations arise from internal motions within the macromolecule or from collective motions of many macromolecules. In aqueous solution, contributions to the inelastic light scattering can arise from the reorientation of molecules with anisotropic polarizability. The spectral width of this anisotropy scattering depends on the magni* Permanent address: Racah Institute of Physics, Hebrew University, Jerusalem, Israel. + Present address: Department of Biophysics and Theoretical Biology, University of Chicago, Chicago, Ill. 60637. 219 0 1976 by John Wiley & Sons, Inc.

220

GENZEL ET AL.

tude of the rotational diffusion constant; for lysozyme the spectral width was found5 to be as small as cm-l. Water molecules, however, reorient much faster due to their smaller dimensions. Their anisotropy scattering extends up to frequency shifts of -50 cm-*, making it difficult to observe the low-frequency Raman scattering of the lysozyme molecule. Iodine filter techniques or the use of a triple grating spectrometer4 cannot improve the experimental situation in this case. In this paper, we report the application of new methods to overcome the above difficulty and we present Raman spectra of aqueous solutions of lysozyme. These spectra are compared with results obtained for the orthorhombic and triclinic crystal forms of lysozyme.

EXPERIMENTAL METHODS Raman Techniques Raman spectra were recorded with a continuous-wave argon laser and a double monochromator. The signal from the photomultiplier was processed using standard pulse shaping and photon counting techniques. The spectra of aqueous solutions were, in general, not readily reproducible using the conventional Raman techniques alone. Slow fluctuations of the scattered intensity were observed, which could mask the presence of weak Raman bands or could give rise to misinterpretations. These fluctuations seem to arise from macroscopic optical inhomogeneities driven by thermal gradients due to laser beam heating. To analyze the influence of the fluctuations on the recorded spectra, we denote'the unperturbed Raman intensity by Z R ( U ) , where u is the Raman frequency shift. Introducing f ( t ) to represent the intensity variations arising from macroscopic fluctuations, we get for the intensity actually measured Z ( t ) = [l f ( t ) ] z R ( U ) . The spectrum is recorded by changing u a t scanning speed us, thus giving:

+

z(t) =

[I + f(t)]zR(ust)

(1)

The scattered intensity Z R ( u ) can be divided into two components:

IR(v) = ZRB(V)

+ IRS(U)

(2)

where ZRB(U) is a background intensity and ZRS(U) represents the lines varying relatively fast with frequency u. Z ( t ) will then have three components:

+

Z ( t ) = I R ( U , ~ ) ~ ( ~ ) Z R B ( U ,+ ~ )f ( t ) Z ~ s ( ~ , t )

(3)

The second and third terms represent the noise arising from macroscopic fluctuations, which is usually small since If(t)l