ESR Studies on Lipid-protein Interaction in Gluten

Transcription

1 Agric. Biol. Chem., 45 (9), 1953~1958, ESR Studies on Lipid-protein Interaction in Gluten Junko Nishiyama, Toyo Kuninori, Hiroshi Matsumoto and Atsushi Hyono* Department of Natural Science, Osaka Women's University, Daisen-cho, Sakai, Osaka 590, Japan *Biophysics Laboratory, Osaka City University Medical School, Abeno-ku, Osaka 545, Japan Received December 3, 1980 The interaction of protein with lipid in wheat gluten has been studied by electron spin resonance (ESR). The gluten in the flour suspension was spin-labeled with a fatty acid spin label (Noxyl-4,4'-dimethyloxazolidine derivative of 5-ketostearic acid) and washed out from the flour. The ESR spectra of the spin label incorporated in gluten exhibited clearly separated parallel and perpendicular hyperfine splittings. The orientation of the gluten lipid and its fluidity showed temperature dependence. Phase transition was observed at 25 C. Compared with gluten, vesicles of the lipids extracted from flour were found to be in a less oriented, highly fluid state, and with much lower activation energy for rotational viscosity, while the reconstituted gluten, which was prepared by mixing purified gluten protein and the extracted lipids, had a lipid environment similar to that of gluten. The results indicate that the lipid was immobilized in the gluten matrix by strong interaction with protein. The rheological properties of dough is due to the formation of a gluten network and the proper development of this network may deend partially on the balance of intermolecular and intramolecular disulfide bonds in the proteins and on the existence of a lipidprotein complex. It has been suggested that a lipid-protein complex is formed during gluten preparation and dough formation.1~4) The lipid-protein interaction has been studied by using many techniques including X-ray diffraction,5'^ electron microscopy,4'7* nuclear magnetic resonance spectroscopy,8) solvent extraction,9'10* and baking test.1m2) The lipidprotein interactions play important roles in the physical properties of dough. The formation of the lipid-protein structure appears to be of particular importance in the development of a proper gluten extensibility, for example. Grosskreutz6) postulated a gluten model involving the Danielli-type membrane to account for the extensibility of gluten. Wehrli and Pomeranz13) suggested that covalent and ionic bonds primarily increase cohesiveness of doughs, while dipole, hydrogen, and hydrophobic bonds contribute to elasticity and plasticity. In the present study, electron spin resonance (ESR) techniques were employed for the investigation of the relations between lipid and protein in the gluten matrix. ESR spectra of spin labels incorporated into the lipid position provided much information about the structural and dynamic properties of the lipid in the gluten. MATERIALS AND METHODS Materials. A commercially milled unbleached wheat flour with ll.8% protein, and 0.40% ash on a 13.2% moisture basis was used. The spin labels, N-oxyl-4,4'- dimethyloxazolidine derivatives of 5-ketostearic acid, 12- ketostearic acid, and 16-ketostearic acid (5 NS, 12 NS, and 16 NS), were the products of Syva Co. (Palo Alto, CA). The organic solvents of analytical grade were obtained from Wako Pure Chemical Industries, Ltd. Purified wheat gluten protein was isolated from n- butanol-defatted flour by the method of Jones et a/.14) Flour lipid was extracted according to Folch et al.l5) Reconstituted gluten was obtained by mixing the isolated gluten protein with the extracted lipid (5%, w/w) and water (65%), and followed by lyophilization.

2 1954 J. Nishiyama, T. Kuninori, H. Matsumoto and A. Hyono Preparation of spin-labeled sample. Flour or petroleum ether-defatted flour (0.2 g) was incubated with 5 nmol of one of the spin labels in 1 ml of water under mechanical shaking for 30 min at 30 C. After incubation, starch was thoroughly washed out with water. The residual wet luten was put into a quartz tube for ESR measurement. The isolated gluten protein or the reconstituted gluten (0.2g) was incubated with 1 ~5 nmol of the spin label in 0.5ml of water under mechanical shaking for 30 min at 30 C. The mixture was put on a Sephadex G-50 (100~300fj) column (1.4 x 20cm) which had been equilibrated with water for removal of the uncorporated spin label, and the protein fraction eluted with water was centrifuged for 15 min at 10,000 rpm. ESR measurement as carried out with this precipitate. The flour lipid (10mg) was mixed with 100 nmol of spin label in a small volume of chloroform and the solvent was evaporated. The lipid vesicles (liposomes) were obtained by sonication of the flour lipid with 0.5 ml of water for 30 min at 42 C. The mixture was fractionated by the same Sephadex G-50 column, and the liposome fraction was used for the ESR measurement. In all cases, the amount of spin label used was small enough to prevent spin-spin coupling. ESR measurement. ESR measurements were performed on a Hitachi 771 ESR spectrometer operating in X-band at varied temperatures (10 ~ 50 C). The order parameter, the rotational correlation time and the activation energy of rotational viscosity were obtained from the ESR spectra. The order parameter, S, was determined for the observed hyperfine splittings of ESR spectra by the following equation16): where/a is given as: S=fa(A ll -A1)/(Az-Ax) /a=(^x+^y+^z)/(2^±+^, ) A± and A ^ are perpendicular and parallel splittings of the spectra and Ax, Ay, and Az are the principal values of hyperfine tensor (6.0, 6.0, 32.0 G for 5 NS, 5.5, 5.6, 29.9 G for 12 NS and 16 NS, respectively). The rotational correlation time was calculated by16): Tc=3.418 x l(t10-zf(0) / MO) where A(0) is the peak-to-peak line width of the central line in gauss and him) is the peak height of the line belonging to the nuclear spin quantum number m. The activation energy, 2svis, of rotational viscosity as obtained basing on the following equation17): w \n (xj^y) = C+EvjRT wheret, R, and C are absolute temperature, the gas constant, and another constant, respectively. RESULTS Figure 1 illustrates the ESR spectra of spinlabeled gluten and spin-labeled gluten protein. The signal intensity of gluten which contained 11% of lipid on dry basis was extremely strong compared with the purified gluten protein which had*been delipidated. Thus, it can be supposed that the spin label was selectively incorporated into the lipid environment of gluten. The probes, 5 NS, 12 NS and 16 NS perceived the atmospheres at the depth of 5, 12 and 16 carbon atoms of stearic acid, respectively. The ESR spectra of spin label 5 NS had clearly separated parallel and perpendicular hyperfine splittings, while the spectra of 12 NS and 16 NS were imperfectly separated at the experimental temperature. Spin labeled gluten altered the ESR spectra M-i) Mi)i Fig. 1. ESR Spectra of Spin Labels, 5 NS, 12 NS, and 16 NS in Gluten Recorded at 25 C. Spectra a, b, and c show 5 NS, 12 NS, and 16 NS in gluten, respectively. Spectrum d shows 5 NS in purified gluten protein. The signal level of a, b, and c is one-tenth of that of d.

3 ESR Studies on Lipid-protein Interaction in Gluten 1955 Fig. 2. ESR Spectra of Spin Label, 5 NS in Gluten Recorded at Different Temperatures. j{0.4 " Temp ( C) Fig. 3. Changes in Order Parameter (a) and Rotational Correlation Time (b) of Three Spin Labels in Gluten at Different Temperatures. Spinlabel: 5 NS (#), 12NS (A), 16NS (å ). 50 Fig. 4. Andrade's Plot Microviscosity of Gluten. Spinlabel: 5 NS (#), 12NS (A), 16NS ( Activation Energies of according to the temperature (Fig. 2). The peak-to-peak line width of the central line decreased gradually with increasing temperature but no significant change was observed over the temperature range studied. The order parameters and correlation times were calculated from the spectra obtained at different temperatures. The order parameter decreased with an elevation of temperature as shown in Fig. 3a. The order parameter of5 NS in gluten was higher than those of 12 NS and 16 NS, indicating that the lipid orientation is higher and the motion of hydrocarbon chains is anisotropic in the vicinity of the polar head groups, while the lipid orientation is random and the motion of hydrocarbon chains is sotropic in the hydrophobic region. The relation between rotational correlation

4 G 1956 J. Nishiyama, T. Kuninori, H. Matsumoto and A. Hyono Table I. Activation Energies and Transition Temperatures of Gluten, Defatted Gluten, Lipid Vesicles, and Reconstituted Gluten Determined by Rotational Correlation Time of Three Spin Labels Activation energy, 2svis (kcal/mol); transition temperature, Tc ( C); LP, liquid phase; SP, solid phase. Sample Spin label Gluten luten from PEdefatted flour* Extracted lipid Reconstituted gluten Petroleum ether-defatted flour Temp( C) Fig. 5. Changes in Order Parameter (a) and Rotational Correlation Time (b) of Three Spin Labels in Gluten Washed from Petroleum Ether-defatted Flour at Different Temperatures. Spin label: 5 NS (#), 12NS (A), 16 NS (å ). time and temperature is shown in Fig. 3b. A significant decrease in rotational correlation time was observed for 5 NS with an increase in temperature. The logarithmic function is plotted against \\T (Fig. 4). The plot for 5 NS resulted in two straight lines intersecting at 25 C, where the lipid phase at the depth of 5 carbon atoms from the polar head group would change from solid to liquid. The activation energies calculated from the tangents of the plots were 17.0 kcal/mol below 25 C (solid state) and 7.8 kcal/mol above 25 C (liquid state) for the 5 NS label as shown in Table I. On the other hand, the plots for 12 NS and 16 NS had no break point and gave the activation energies of 6~7 kcal/mol, suggesting that at the depth of 12 and 16 carbon atoms the lipid phase would remain liquid over the whole temperature range studied. Figure 5 shows the relation between the temperature and the order parameter or the rotational correlation time of the spin labels in

5 ESR Studies on Lipid-protein Interaction in Gluten 1957 "0.6K- S.0.4J à" à". 1 I 0I å å å I - T å å.å å å Temp( C) Fig. 6. Changes in Order Parameter (a) and Rotational Correlation Time (b) of Three Spin Labels in Flour Lipid and Reconstituted Gluten at Different Temperatures. Sample: flour lipid (solid symbol), reconstituted gluten (open symbol). Spin label: 5 NS (#, O), 12 NS (A), 16 NS(«). gluten which was prepared from petroleum ether-defatted flour and contained 3.5% of lipid on a dry basis. Both the order parameter and correlation time were higher than those values of non-defatted gluten (Fig. 3) in the range of temperature studied. Andrade's plot for 5 NS and 12 NS was composed of two straight lines, having a break point at 31 C for 5 NS and a breakpoint at 22 C for 12NS, but the plot for 16 NS had no break point. The activation energies are compared in Table I. The relation between the temperature and the order parameter or the rotational correlation time of spin label in the vesicles of flour lipid is shown in Fig. 6. The lipid vesicles had a lower orientation, higher fluidity and much lower activation energy than gluten. The phase transition in lipid vesicles was observed at 23 C only with 5 NS (Table I). Reconstituted lipid-gluten system, which was prepared by mixing purified gluten protein with flour lipids, showed higher values in the order parameter, rotational correlation time and activation energy than those of lipid vesicles, as shown in Fig. 6 and Table I, whereas all these values were similar to those of gluten. A phase transition was observed at 26 C with 5 NS. DISCUSSION A part of the free lipids in flour change into bound lipids during gluten preparation. The formation of lipid-protein complex was observed during doughing procedure.1 ~4) The lipid is bound to the glutenin fraction rather than to the gliadin fraction.18) The hydrophobicity of both fractions is high enough to stabilize intraand inter-molecular hydrophobic bonds.13) Hoseney et al}9) assumed that polar lipids are bound to the gliadin proteins hydrophilically and to the glutenin proteins hydrophobically, and that these bonds may cement these proteins. Grosskreutz6) proposed that a lipoprotein structure occupies some 2 to 5% of he elastic gluten. The fatty acid spin label probes used in this study provide information about physical state of the lipid in gluten. The results obtained suggest the presence of a lipid-protein complex in hydrated gluten. It can be concluded that the lipid was immobilized by the proximity of protein. The difference in the parameters between the gluten and the isolated lipid vesicles implies that the lipid in gluten is bound to the protein with Var der Waals forces, hydrogen bonds, and hydrophobic interaction. For example, it is considered that the hydrogen bond affects the interaction in the gluten but it is minor in the lipid vesicles. Since gluten obtained from petroleum etherdefatted flour gave higher values of the order parameter, motional parameter and transition

6 1958 J.Nishiyama, T. Kuninori, H. Matsumoto and A. Hyono temperature than those of native gluten, the lipid remaining in gluten has a high polarity and may be bound more tightly to protein. We feel that the molecular motion of lipid. is restricted until a depth of 12 carbon atoms by the association with the protein. The reason for the high phase transition temperature observed in this partially defatted gluten may reside in the hydrogen bonding ability of the emaining lipid with protein. It is likely that the lipid molecule in gluten forms a bilayer which orients well in the vicinity of polar head group, and that such a lipid structure helps the aggregation of protein through hydrophobic and hydrophilic interactions. Thus the comparatively strong interaction between lipid and protein seems to maintain the rigid structure of the gluten. The ratio of amounts of lipid to protein, used in this study, is too small to form a uniform membrane-like structure. Rather, the lipid would constant such a partial bilayer as described by Grosskreutz,6) some occurring in forms bound tightly to proteins, and the thers in the form of vesicles in gluten. ESR data is consistent with the notion that the flour lipids may contribute to the rhelogical properties in dough. Acknowledgments. Mrs. S. Kuriyama, Osaka City University Medical School, provided advice for ESR measurement. This work was supported in part by a Grant-in Aid from the Ministry of Education, Science and Culture REFERENCES 1) Y. Pomeranz, "Wheat, Chemistry and Technology," ed. by Y. Pomeranz, AACC, St. Paul, Minn., 1971, p ) H. Zentner, Chem. Ind., 128 (1958); ibid., 317 (1960 3) J. D. McCaigandA. G. McCalla, Can. J. Res. C, 19, 163 (1941). 4) D. B. Cumming and M. A. Tung, Can. Inst. FdSci. Technol. J., 8(2), 67 (1975). 5) K. Hess, Qual. Plant. Mater. Veg., 6, 312 (1950). 6) J. C. Grosskreutz, Cereal Chem., 38, 336 (1961). 7) N. Crozet and A. Guilbot, Cereal Chem., 51, 300 (1974). 8) H. P. Wehrli and Y. Pomeranz, Cereal Chem., 47, 160 (1970). 9) Chien-Mei Chiu and Y. Pomeranz, /. FdSci., 31, 753 (1966). 10) M. Wootton, /. Sci. FdAgric, 17, 297 (1966). ll) Chien-Mei Chiu, Y. Pomeranz, M. Shogren and K. F. Finney, Fd Technol., 22, 1157 (1968). 12) R. D. Daftary, Y. Pomeranz, M. Shogren and K. F. Finney, Fd Technol., 22, 327 (1968). 13) H. P. Wehrli and Y. Pomeranz, BakersDig., 43(6), 2 (1969). 14) R. W. Jones, N. W. Taylor and F. R. Senti, Arch. Biochem. Biophys., 84, 363 (1959). 15) J. Folch, M. Lees and G. H. Sloane Stanley, /. Biol. Chem., 226, 497 (1957). 16) M. A. Hemminga, Chem. Phys. Lipids, 14, 151 (1975). 17) A. Hyono, S. Kuriyama, H. Hara, I. Yano and M. Matsui, FEBS Lett., 103(1), 192 (1979). 18) J. G. Ponte, Jr., V. A. DeStefanis and R. H. Cotton, Cereal Chem., 44, 427 (1967). 19) R. C. Hoseney, K. F. Finney and Y. Pomeranz, Cereal Chem., 4,1, 135 (1970).