Journal of Colloid and Interface Science 253, 265–272 (2002) doi:10.1006/jcis.2002.8563 Magneto-optics of Ferritin A. Dobek,1 M. Pankowska, and J. Gapiński Faculty of Physics, A. Mickiewicz University, Umultowska 85, 61-614 Poznań, Poland Received October 30, 2001; accepted June 27, 2002; published online August 27, 2002 Measurements of Rayleigh light scattering, nonlinear light scattering in DC magnetic fields, and the Cotton–Mouton effect were carried out for 15 mM NaCl and water solutions of ferritin at room temperature. The spherical geometry of the molecule implies that it is optically isotropic. Such a macromolecule should not manifest magnetic anisotropy; however, in solution it shows induced magnetic birefringence (Cotton–Mouton effect) and changes in the intensity of the scattered light components. The analysis of the obtained results indicates the deformation of linear optical polarizability induced in the ferritin by a magnetic field as the main source of the magneto-optical phenomena observed. Light scattering and the CM effects theoretically depend on the linear magneto-optical polarizability, χ , and the nonlinear magneto-optical polarizability, η. Using the theory describing the phenomena as well as the experimental data, the values of the anisotropy of linear magneto-optical polarizability components, χ − χ⊥ = −(1.3 ± 0.7) × 10−22 [cm3 ] (in SI units χ − χ⊥ = −(2.0 ± 1.2) × 10−33 [m3 ]), the linear optical polarizability, α = (α + 2α⊥ )/3 = (3.9 ± 1.0) × 10−20 [cm3 ] (in SI units α = (3.52 ± 0.09) × 10−4 [Cm2 V−1 ]), and its anisotropy, κα = (α − α⊥ )/3α = −(0.06 ± 0.03), nonlinear magneto-optical polarizability, η = (η + 2η⊥ )/3 = −(4.7 ± 0.9) × 10−30 [cm3 Oe−2 ] (in SI units η = −(6.7 ± 1.3) × 10−18 [Cm4 V−1 A−2 ]) and its anisotropy, κη = (η − η⊥ )/3η = −(0.15 ± 0.10), were deduced. Here α , η , α⊥ , η⊥ are the optical and magneto-optical polarizability components along the parallel and the perpendicular axes of the axially symmetric molecule, respectively. C 2002 Elsevier Science (USA) Key Words: ferritin; light scattering; Cotton–Mouton effect; linear and nonlinear polarizability; optical anisotropy; magneto-optical anisotropy. INTRODUCTION Light scattering measurements are of general use in biomacromolecule research, in particular protein research (1, 2). They have also been utilized for conformational studies of different protein molecules (3, 4) and for the determination of their size and molecular mass distribution (5). Much fewer studies are performed in solution of proteins oriented by a D.C. magnetic field. Most of them are diamagnetic, and the effect of a magnetic field is difficult to detect. Contrary to the majority of proteins, ferritins present in plant and animal cells belong to a group of 1 To whom correspondence should be addressed. E-mail: dobek@amu.edu.pl. metalloproteides. The micellar tertiary structure of these proteins allows iron accumulation in the form of hydrated oxides and phosphates of this metal. Thus ferritin is a large spherical macromolecular protein with iron compounds situated in the cavity created by a peptide shell (6, 7). If needed for the organism, the iron is released by iron reductase enzyme participation. In the human body, ferritin macromolecules are accumulated in liver, spleen, and medulla cells. Recently Petsev et al. (8) observed aggregation of apoferritin molecules depending on CdSO4 concentration in solution. The authors have indicated the effective diffusion coefficients and hydrodynamic radius for monomers, dimers, and trimers from dynamic light scattering. They also studied an increase of crystal surface in situ by atomic force microscopy. The translational diffusion was also measured in living skeletal muscle fibers by fluorescence recovery after photobleaching and NMR techniques (9). The neutron scattering measurements described by Kilcoyne et al. (10) allowed estimation of the packing factor, the effective rigid sphere radius, and the internal and external radius of the ferritin and apoferritin sphere. The authors found an effective radius smaller than the external radius of the macromolecule and suggested ordering of the macromolecules with increasing electrostatic forces in solution. Haüssler et al. have studied the structure and dynamics of appoferitin solutions over a wide range of protein concentrations and ionic strengths (11). Small angle X-ray scattering and dynamic light scattering were used to determine the form factor and the hydrodynamic radius. From the X-ray data and the apoferritin form factor the structure factors of ordered solutions, and finally of polyelectrolyte ordering within the paracrystalline domains in apoferritin solutions were obtained. Yang et al. (12) have studied the kinetics of 3-(aminomethyl)-proxil migration through molecular canals into and outside the ferritin sphere by EPR techniques. The ferritin molecule is used as a paramagnetic label. Dominguez-Vera et al. have reported the crystallization of ferritin and the process of iron removal from the crystals using chelating agents of high iron (III) affinity (13). The authors used UV–visible spectroscopy, video-microscopy, and X-ray diffraction to detect of iron (III) complexes arising from the iron cores and ferritin– apoferritin conversion. In (14, 15) the separation of latex spheres, polyacrylamide beads, and erythrocyte ghosts by high-gradient magnetic separation (HGMS) is described. Considering the influence of different physical factors on a living organism, the 265 0021-9797/02 $35.00 C 2002 Elsevier Science (USA) All rights reserved. 266 DOBEK, PANKOWSKA, AND GAPIŃSKI effect of magnetic fields cannot be neglected as, among others, it has recently been applied in many different medical diagnostic methods. Therefore, determination of the magnetic field influence on ferritin has become a problem of increasing importance for analysis of its physiological functions. Ferritin solutions have been studied by the Cotton–Mouton effect and nonlinear light scattering in D.C. magnetic fields. The analysis of the obtained results indicates the deformation of linear optical polarizability induced in the biomacromolecule by a magnetic field as the main source of the magneto-optical phenomena observed. THEORY symmetry axis. For its mean linear optical polarizability one has Ru = Vv + 3Hv . 2 Ris = α= n o Mw dn · , 2π N A dc κα2 =  α  − α⊥ 3α 2 ,  4 2π 13 N α 2 κα2 , = 10 λ [9] [10] or from the experimentally determined value of Du as 5Du . 6 − 7Du [2] [11] Nonlinear Light Scattering in D.C. Magnetic Fields [3] where Ris and Ranis are the isotropic and anisotropic components of the Rayleigh coefficient, Ru = Ris + Ranis . Using the measured value of Ru one can calculate Ris and Ranis after the Cabannes relations (19): Let us consider an axially symmetric macromolecule, and let χ , χ ⊥ , stand for the linear magneto-optical polarizability components of the macromolecule in the direction of and perpendicular to its symmetry axis. For its mean polarizability χ one has χ= 6 − 7Du Ris = Ru , 6 + 6Du [4] 13Du . 6 + 6Du [5] Ranis = Ru [8] can be determined from the measured value of Ranis as κα2 = 2Dv Hu = , Vu 1 + Dv [7] where n o is a refractive index of the solvent, Mw is the weight average molecular weight of the macromolecule, c is a weight concentration of the macromolecules, and N A is Avogadro’s number. The squared anisotropy of linear optical polarizability components, defined as Ranis Du =  4 2π 1 N α2, 2 λ where N is the number of molecules per unit volume. One can also find this value from the measured increment of the refractive index dn/dc at λ, which is a wavelength of scattered light, using the formula [1] For the determination of the optical parameters of the molecules from light scattering measurements it is convenient to know the depolarization ratios, defined as follows, Hv Dv = , Vv [6] Experimentally the α value may be obtained using the measured value of Ris and the formula Static Light Scattering We shall consider a Cartesian space in which incident light of intensity Iv is polarized perpendicular to the plane of observation xy (Fig. 1) (16, 17). Small Hv and large Vv intensity components of the scattered light (with the electric vector polarized parallel and perpendicular to the plane of observation xy, respectively) are measured at 90◦ along the x axis. Ru , Dv , and Du stand for the total Rayleigh coefficient and the depolarization ratios for vertically polarized and unpolarized incident light, respectively. For molecules which are small compared to the light wavelength the total Rayleigh coefficient is defined as follows (18): α + 2α⊥ . 3 α= Let us consider an axially symmetric macromolecule, and let α , α ⊥ stand for the linear optical polarizability components of the macromolecule in the direction and perpendicular to its χ + 2χ⊥ . 3 [12] The anisotropy κχ of linear magnetic-optical polarizability components is defined as κχ = χ − χ⊥ . 3χ [13] In the following one considers an interaction of the ferritin molecule with an external D.C. magnetic field H applied in the 267 MAGNETO-OPTICS OF FERRITIN equations describing changes of small-HvH and large-VvH components as dependent on the magnetic field (20), HvH VvH  2 (χ − χ⊥ ) ηκη = H 2, + 21 2kT 2ακα   η 4 (χ − χ⊥ ) κα + H 2, = 15 2kT α  [16] [17] where η= η + 2η⊥ 3 [18] is the nonlinear magneto-optical polarizability and FIG. 1. κη = Geometry of light scattering measurements in a D.C. magnetic field. z direction, Figs. 1, 2. The distinguished symmetry axis (axis ) of the molecule makes an angle ϑ with H . We shell refer to HvH and the VvH as the small and large components of scattered light intensity, respectively, when a magnetic field H is applied, and to Hv0 and Vv0 in the absence of this field. Then one can define the measured quantities of components’ relative changes as HvH = HvH − Hv0 , Hv0 [14] VvH = VvH − Vv0 . Vv0 [15] For small values of H (when U < kT , where U is the potential energy of the molecule in the presence of H , k is Boltzmann’s constant, and T is absolute temperature in K) one can write PULSE COUNTER AMPLIFIER DISCRIMINATOR PM η − η⊥ 3η is the anisotropy of the nonlinear magneto-optical polarizability components, with η , η⊥ being the nonlinear magneto-optic polarizabilities along the  axis and the ⊥ axis of the symmetry axis of the macromolecule, respectively. The nonlinear magnetooptical polarizability η describes changes of the electric dipole moment induced by the electric field of the incident light beam, due to the external magnetic field H . Both equations, [16] and [17], are composed of two parts. The first part contains linear magneto-optical polarizability χ and its anisotropy κχ and describes magnetic molecular orientation. The second part contains nonlinear magneto-optical polarizability η and describes changes of the linear optical polarizabilty α due to the molecular deformations induced by the magnetic field. Cotton–Mouton Effect Let us consider a Cartesian space in which the polarization plane of the light incident onto a medium makes an angle of 45◦ with the z direction of an external magnetic field H (21, 22), Fig. 3. As in the case of light scattering, in the following one ANALYZER SAMPLE X Z POLARIZER LASER 0 Y ELECTROMAGNET FIG. 2. Apparatus for static light-scattering and light-scattering measurements in a magnetic field, schematically. [19] FIG. 3. Geometry of Cotton–Mouton measurements. 268 DOBEK, PANKOWSKA, AND GAPIŃSKI considers an interaction of ferritin (axis ) with an external D.C. magnetic field H applied in z direction. The electric field vector E of the incident light intensity I is composed of two components, parallel E  and perpendicular E ⊥ to H . The difference in the propagation velocities of these two components implies the appearance of the phase difference ϕ over a path L, which can be expressed as ϕ = 2π n − n⊥ , λ [20] where n  and n ⊥ , are the light refraction indices of the light polarized parallel and perpendicular to H , respectively, and λ is the light wavelength. According to the Cotton–Mouton equation, ϕ = 2π LC H 2 , [21] where C is the Cotton–Mouton constant. The molar Cotton– Mouton constant is given by the expression C M = C∗ 6n 2 M , (n 2 + 2)2 ρ [22] where C ∗ is defined as C ∗ = C λn , n is the light refraction index for the medium, M is the molecular mass of the medium, and ρ is its density. Taking into account the nonlinear effects related to deformation of the linear optical polarizability of a diamagnetic/paramagnetic macromolecule appearing under the influence of a magnetic field, we get for noninteracting macromolecules (21) M Cgas   4π N A 3 = ακα (χ − χ⊥ ) + 5η . 45 kT [23] As in the case of the changes in the scattered light components, M the molar Cotton–Mouton constant Cgas is composed of two parts, one describing the contribution of molecular orientation, and the other the molecular deformation caused by the magnetic field. MATERIALS AND METHODS Repeatedly distilled organic liquids (benzene, nitrobenzene, cyclohexane, and carbon tetrachloride) were used to check the measuring device. The measurements were performed at room temperature 23◦ C. Apparatus and Measurements Light Scattering Measurements of Rayleigh light scattering, light scattering, and dynamic light scattering in a D.C. magnetic field were carried out using the apparatus schematically shown in Fig. 2. The light source consisted of an He–Ne laser, λ = 632.8 [nm], or an Ar+ ion laser λ = 514 [nm] (Carl Zeiss Jena, Germany). A Glan polarizer ensured that the laser light polarization was perpendicular to the plane of observation xy, Figs. 1, 2. The sample was contained in a rectangular plane-parallel glass cell (1 cm × 1 cm × 3 cm) placed between pole pieces of an electromagnet. The maximum magnetic field strength available in the z direction was 20 kOe. The scattered light emerging from the cell in the x direction at 90◦ to the y direction of the incident beam traversed the analyzer and a system of two very small diaphragms (“pin holes”) to finally reach the detector and photomultiplier (THORN EMI Electron Tubes). The electric signal created was amplified, discriminated, digitized, and sent to a pulse counter. The apparatus allowed for scattered light intensity of horizontally and vertically polarized light measurements by changing positions of the analyzer at different magnetic field strengths. The positions of the analyzer were fixed with an accuracy of 0.01◦ . The device described was checked by light scattering intensity and correlation function measurements in the absence and presence of magnetic field strengths. The static light scattering data, i.e., depolarization ratios and absolute Rayleigh constants, obtained using some simple organic liquids (listed under Materials) were in good agreement to those in the literature (23–26). Benzene was used as a working standard for this checking. When calculating the absolute Rayleigh constants, we applied the absolute constant for benzene at λ = 632.8 [nm] R B = 8.765 × 10−6 [cm−1 ] (23). The intensities of light scattered by the ferritin molecules in solution were calculated as a difference in light intensities scattered by the solution and solvent. Since only relative changes in the small, Hv , and large, Vv , intensity component of scattered light were measured, no geometrical corrections had to be taken into account. Preparation of Solution Measurements were carried out for solutions of ferritin from Boehringer Mannheim GmbH . The compound is an extract obtained from horse spleen. Studied samples were prepared by dissolving the basic solution (at a concentration of 36.5 g/l) in water and in 15 mM NaCl to the final concentrations 0.075, 0.15, 0.3, 0.8, 1.56, 1.75, 2.08, 3.12, 6.25, and 7 g/l. In order to eliminate mechanical impurities, the water as well as all solutions was passed through a Millipore cellulose filter (pore diameter 0.22 µm) prior to use. The air bubbles were eliminated by centrifugation of the solution at 8000 rpm for 30 min. Cotton–Mouton Effect Measurements of the Cotton–Mouton effect were carried out using the apparatus schematically shown in Fig. 4. The light source consisted of an He–Ne laser λ = 632, 8 [nm] or an Ar+ ion laser λ = 514 [nm] (Carl Zeiss Jena, Germany). A neutral density filter was used to attenuate the intensity of incident light. A Glan–Thomson polarizer ensured that the laser light polarization was adjusted at 45◦ to the z axis, to which the magnetic field H was aligned, Fig. 3. The electromagnet and the holder for the same sample cell as used in the light scattering 269 MAGNETO-OPTICS OF FERRITIN 3 Concentration, c [kg/m ] 0 1 0 1 2 3 4 2 3 4 1, 3346 Refractive index, n 1, 3344 1, 3342 1, 3340 1, 3338 1, 3336 1, 3334 Concentration, c [g/L ] FIG. 4. Apparatus for Cotton–Moutan measurements, schematically. studies were applied. After passing the solution, a light beam traversed an quarter plate for the wavelength used, crossed the analyzer, and reached the detector and photomultiplier (THORN EMI Electron Tubes). The electric signal created was amplified, discriminated, digitized, and counted by a pulse counter. When the magnetic field was on, the ferritin solution became birefringent. The linearly polarized beam became polarized elliptically, which the detector observed as an increase of light intensity. The light polarized elliptically was changed by a properly adjusted quarter plate to be polarized linearly at an angle ϕ with respect to incident light polarization (see Fig. 3). The set-up allowed birefringence measurements at different magnetic field strengths with 0.01◦ accuracy of ϕ angle. The device described was checked by Cotton–Mouton constant measurement for nitrobenzene, which was found in good agreement with those in the literature (27). When calculating the absolute Cotton–Mouton constants, we applied the absolute constant for nitrobenzene at λ = 632.8 [nm] C = 2.31 × 10−12 [Oe−2 cm−1 ] (28). FIG. 5. Refractive index n as a function of the ferritin concentration c for water solutions at λ = 632.8 nm. Fig. 5. From the graphs, a refractive index increment dn/dc = 2.49 × 10−4 [L/g], was calculated (c is the concentration of ferritin in g L−1 ). Small Hv and large Vv intensity components of scattered light were measured for series of ferritin solutions at different concentrations using the geometry and setup shown in Figs. 1 and 2, respectively. In Fig. 6 the depolarization ratio Du (Eqs. [2], [3]) is plotted as a function of ferritin concentration. The dependence of Du on the concentration seen in the figure indicates the presence of weak interactions between protein molecules in the solution. The partial Rayleigh constants Hv and Vv , obtained as described in Apparatus and Measurements allowed for calculation, using Eq. [1], of the absolute total Rayleigh constant Ru . Figure 7 shows the values of the linear optical polarizability α of ferritin macromolecule, calculated using Eqs. [7], [8], as a 3 Concentration, c [kg/m ] 0,04 Light refractive indices of the ferritin solutions were measured at 23◦ C for the light wavelength of 588 [nm] by means of the Pulfrich refractometer IRF-23. The measured values were converted to the increment at λ = 632.8 [nm] using the following formula (29):  dn dc  λ =  dn dc    2 · 104 . · 0.940 + λ2 588 [24] RESULTS AND DISCUSSION Light refractive indices of the studied ferritin solutions, for the light wavelength of 632.8 [nm], are shown graphically in Depolarization ratio, Du Light Refractive Indices 0 2 4 6 8 0 2 4 6 8 0,02 0,00 -0,02 Concentration, c [g/L] FIG. 6. Depolarization ratio Du as a function of ferritin concentration c. 270 DOBEK, PANKOWSKA, AND GAPIŃSKI 3 2 Concentration, c [kg/m ] 2 3 4 5 6 7 20 3 3 2 2 1 Linear optical polarizability, 4 2 -1 α x 10 [Cm V ] 4 3 10 [cm ] 4 α x Linear optical polarizability, 5 1 0 0 0 1 2 3 4 5 6 Square of magnetic field, H x 10 v 1 7 Change of the large component , ∆Vv 0 0,0 0,5 1,0 1,5 function of its concentration. Observed slightly different values of linear optical polarizability α for different c supports previous conclusions concerning existence of weak interactions between macromolecules. The same conclusions can be drawn from the dependence of the anisotropies of ferritin protein κα , calculated using Eqs. [10], [11]. For different concentrations c, κα is shown in Fig. 8. The very small anisotropies obtained are characteristic for the spherical shape of the studied macromolecules. Below, for further calculations, the values of linear optical polarizability α = (3.9 ± 0.1) × 10−20 [cm3 ] (in SI units α = (3.52 ± 0.09) × 10−4 [Cm2 V−1 ]) and its anisotropy, κα = −0.06 ± 0.03, as obtained from extrapolation of measured values to zero concentration were used. The relative changes of the polarized VvH and depolarized HvH components of scattered light (defined in Eqs. [14], [15]) -0,05 1 2 3 as a function of the square of an external D.C. magnetic field are presented in Figs. 9 and 10. In both cases these changes are negative and depend linearly on the square of the magnetic field. Light scattering in a magnetic field is theoretically described by Eqs. [16], [17], in which the linear and nonlinear magnetooptical polarizabilities are involved. Identical magneto-optical polarizabilities describe magnetic birefringence induced in a macromolecular solution, see Eq. [23]. To find molecular values of 3χ κχ = χ − χ⊥ , η, and κη of ferritin macromolecule, the Cotton–Mouton effect for different solute concentrations was measured using the geometry and setup shown in Figs. 3 and 4. The results of magnetic birefringence studies are shown in Figs. 11–13. Figure 11 presents the phase difference ϕ [deg] induced by a magnetic field in solutions of different ferritin Square of magnetic field, H x 10 v 5 6 7 α 0,5 κ Anisotropy of linear polarizability, 4 0,0 -0,5 2 FIG. 9. Relative changes of the polarized component of scattered light VvH as a function of the square of a D.C. magnetic field, H 2 . Change of the small component, ∆ Hv Concentration, c [kg/m3] -8 Square of magnetic field, H x 10 [Oe ] 2 3 2,0 0,00 0 FIG. 7. Linear optical polarizability α (in CGS ES and SI units) as a function of ferritin concentration c. 2 -2 0,05 2 1 2 [A m ] -0,10 Concentration, c [g/L] 0 -12 0, 0 0,5 1,0 -12 2 -2 [A m ] 1,5 2,0 0,1 0,0 -0,1 -0,2 0 1 2 3 4 5 6 7 Concentration, c [g/L] FIG. 8. Anisotropy of linear polarizability κα as a function of ferritin concentration c. 0 1 2 3 2 -8 2 Square of magnetic field, H x 10 [Oe ] FIG. 10. Relative changes of the depolarized component of scattered light HvH , as a function of the square of a D.C. magnetic field, H 2 . 271 MAGNETO-OPTICS OF FERRITIN 0,5 1,0 1,5 -12 2,0 2 3 -2 Concentration, c [kg/m ] [A m ] 0 2,5 0,5 2 3 4 5 6 7 5 Molar Cotton-Mouton constant, 8 3 -2 -1 M Cgas x10 [cm Oe M ] 6,25 g/l Phase difference, ϕ [deg] 1 3,12 g/l 0,4 2,08 g/l 1,56 g/l 0,3 0,2 0,1 8 4 6 3 4 2 2 1 0 0,0 0 1 2 3 -8 2 0 0 4 Molar Cotton-Mouton constant, 18 5 -2 -1 M Cgas x 10 [m A M ] 2 Square of magnetic field, H x 10 0,0 1 2 3 4 5 6 7 Concentration, c [g/L] 2 Square of magnetic field, H x 10 [O e ] FIG. 11. Difference of phase ϕ measured in a solution of ferritin by the Cotton–Mouton effect, as a function of the square of a D.C. magnetic field, H 2 (in CGS EM and SI units). concentration as a function of the square of the magnetic field; see Eq. [21]. For all four concentrations studied the dependence of ϕ = f (H 2 ) is linear, suggesting that no interaction between the macromolecules occurs due to the induced magnetic dipoles. Using the data from Fig. 11, the Cotton–Mouton constants, C [Oe−2 cm−1 ] (in SI units [A−2 m]), and the molar Cotton– M = (2.5 ± 1.0) × 10−8 [cm3 Oe−2 M−1 ] Mouton constants, Cgas M (in SI units Cgas = (3.9 ± 1.6) × 10−18 [m5 A−2 M−1 ]), for different ferritin concentrations were calculated, Eqs. [21], [22]. The concentration dependence of these constants is shown in Figs. 12 and 13. From Fig. 12 one can see that the C value changes linearly with ferritin concentration. The molar Cotton– Mouton constant is independent of the concentration (in the limit of experimental error). Using Eqs. [16], [17] one can estimate 3 Concentration, c [kg/m ] 0 1 2 3 4 5 6 7 10 8 5 4 6 3 4 2 2 Cotton-Mouton constant, 14 -2 C x 10 [mA ] Cotton-Mouton constant, 12 -2 -1 C x 10 [Oe cm ] 6 1 0 0 0 1 2 3 4 5 6 7 Concentration, c [g/L] FIG. 12. Concentration dependence of Cotton–Mouton constant C (in CGS EM and SI units) for ferritin solutions. M FIG. 13. Concentration dependence of molar Cotton–Mouton constant Cgas (in CGS EM and SI units)for ferritin in solutions. η from Eq. [23]; with the value of η, the anisotropy of linear magneto-optical polarizability, χ − χ⊥ , can be calculated and finally κη , the anisotropy of nonlinear magneto-optical polarizability, can be obtained. From the experimental data presented in this paper the anisotropy of linear magneto-optical polarizabilities χ − χ⊥ = −(1.4 ± 0.7) × 10−22 [cm3 ] (in SI units χ − χ⊥ = −(2.0 ± 1.2) × 10−33 [m3 ]), the nonlinear magneto-optical polarizability η = −(4.7 ± 0.9) × 10−30 [cm3 Oe−2 ] (in SI units η = −(6.7 ± 1.3) × 10−18 [Cm4 V−1 A−2 ]), and its anisotropy, κη = −(0.15 ± 0.1), were determined. The measurements of dynamic light scattering in a DC magnetic field excluded the dimerization or any higher order aggregation of ferritin monomers in solution induced by a strong magnetic field. Both the Cotton–Mouton effect and light scattering changes observed in ferritin solution under the influence of a magnetic field can be interpreted as due to two phenomena. One is connected with the orientation of the macromolecule induced magnetic dipole moment by the magnetic field. The second one is linked to the rearrangement of the inner iron core with increasing intensity of the magnetic field. Such a displacement is connected with a deformation of the linear optical polarizability and a change of the value and/or sign of its anisotropy. CONCLUSIONS Ferritin is a large spherical macromolecular protein with iron compounds situated in the cavity created by a peptide shell. Very small anisotropy of linear optical polarizability obtained from static light scattering measurements suggests a spherical shape. In 0.015 M NaCl water solution no aggregation of the macromolecules was observed. Also, no aggregates were detected due to induced magnetic dipole moment interactions. 272 DOBEK, PANKOWSKA, AND GAPIŃSKI Optically isotropic spherical macromolecules should not show anisotropy of the magneto-optical polarizabilities. However, in solution they show induced magnetic birefringence (Cotton– Mouton effect) and changes in intensity of the scattered light components induced by a magnetic field. In view of the theory, only nonlinear interaction of ferritin protein with magnetic field can be responsible for the observed phenomena. Such nonlinearity is proposed to be a result of the displacement or deformation induced by strong magnetic field in the iron compound core. These phenomena can influence the transport of hydrogen or metal ions and other chemicals through the canals of ferritin, e.g., during NMR imaging procedures. REFERENCES 1. Chu, B., in “The Application of Laser Light Scattering to the Study of Biological Molecules” (J. C. Earnshaw and M. W. Steer, Eds.), NATO ASI, Series A: Life Sciences, Vol. 59, Plenum, New York, London, 1982. 2. Pecora, R., in “Dynamic Light Scattering.” Plenum, New York, London, 1985. 3. Mischenko, M. J., Hovenier, J. W., and Travis, L. D., Eds., “Light Scattering by Nonspherical Particles—Theory, Measurements and Applications.” Academic Press, San Diego, 2000. 4. Patkowski, A., Seils, J., Hinssen, H., and Dorfmüller, Th., Biopolymers 30, 427 (1990). 5. Brown, W., Ed., “Dynamic Light Scattering—The Method and Some Applications.” Clarendon, Oxford, 1993. 6. Harrison, P. M., and Arosio, P., Biochim. Biophys. Acta 1275, 161 (1996). 7. Frey, R. F., Donlin, M. J., and Bashkin, J. K., J. Chem. Ed. 75, 437 (1998). 8. Petsev, D. N., Thomas, B. R., Yau, S. T., and Vekilov, P. G., Biophys. J. 78, 2060 (2000). 9. Papadopoulos, S., Jürgens, K. D., and Gros, G., Biophys. J. 79, 2084 (2000). 10. Kilcoyne, S. H., Mitchell, G. R., and Cywinski, R., Physica B 180/181, 767 (1992). 11. Haüsler, W., Wilk, A., Gapiński, J., and Patkowski, A., submitted for publication. 12. Yang, X., Arosio, P., and Chasteen, N. D., Biophys. J. 78, 2049 (2000). 13. Dominguez-Vera, J. M., Rondón, D., Moreno, A., and Garcı́a-Ruiz, J. M., J. Cryst. Growth 168, 138 (1996). 14. Owen, C. S., and Lindsay, J. G., Biophys. J. 42, 145 (1983). 15. Odette, L. L., McCloskey, M. A., and Young, S. H., Biophys. J. 45, 1219 (1984). 16. Kielich, S., Acta Phys. Polon. A 37, 447 (1970). 17. Kielich, S., Opt. Commun. 1, 345 (1970). 18. Kielich, S., “Molekularna Optyka Nieliniowa.” PWN, Warsaw, Poznan, 1977. 19. Cabannes, J., “La diffusion moleculaire de la lumière.” Presses Universitaires de France, Paris, 1929. 20. Dobek, A., Patkowski, A., and Labuda, D., J. Polym. Sci. Polym. Sym. 61, 111 (1977). 21. Kielich, S., PTPN Prace Kom. Mat. Przyr. 11, 65 (1962). 22. Kielich, S., and Surma, M., PTPN Prace Kom. Mat. Przyr. 11, 153 (1962). 23. Kaye, W., and Havlik, A. J., Appl. Opt. 3, 541 (1973). 24. Dezelic, G., Pure Appl. Chem. 23, 327 (1970). 25. Lalanne, J. R., and Bothorel, P., Mol. Phys. 19, 227 (1970). 26. Cardamone, J., Hunt, J. L., and Stevens, J. R., Chem. Phys. Lett. 12, 628 (1972). 27. Surma, M., PTPN Prace Kom. Mat. Przyr. 11, 113 (1962). 28. Surma, M., PTPN Prace Kom. Mat. Przyr. 11, 131 (1962). 29. Pearlmann, G. E., and Longsworth, L. G., J. Am. Chem. Soc. 70, 2719 (1948).