H Magnetization Transfer in Hydrated Gluten and Flour: Effects of

The interaction of water with flour or gluten in hydrated samples was investigated by proton magnetization transfer measurements. Flour and gluten fro...
0 downloads 0 Views 132KB Size
Download PDF
Biomacromolecules 2004, 5, 1824-1831

1824

1H

Magnetization Transfer in Hydrated Gluten and Flour: Effects of Wheat Aging Lucia Calucci* and Claudia Forte IPCF-CNR, Area della Ricerca di Pisa, via G. Moruzzi 1, 56124, Pisa, Italy

Luciano Galleschi Dipartimento di Scienze Botaniche, Universita` degli Studi di Pisa, via L. Ghini 5, 56126, Pisa, Italy Received March 22, 2004; Revised Manuscript Received May 25, 2004

The interaction of water with flour or gluten in hydrated samples was investigated by proton magnetization transfer measurements. Flour and gluten from both durum and bread wheat seeds, either unaged or artificially aged over different periods of time, were investigated. Measurements were performed at several radio frequency power levels and frequency offsets, and the data were quantitatively modeled by two interacting pools, a liquid (water) and a solid (macromolecules) one. A super-Lorentzian line shape well described the magnetization of the solid pool. Magnetization transfer was found to be more efficient for flour with respect to gluten samples, in agreement with their hydrophilic/hydrophobic behavior. The aging treatment of seeds resulted in a minor degree of interaction between macromolecules and water. Introduction Flour hydration is a complex process of great relevance in food science since it influences the rheological properties of doughs and the quality of baked products. In dough, water, which acts as a plasticizer, is nonuniformly distributed between starch and protein phases, and in addition, it influences the interactions among the various flour components (starch, proteins, lipids, and sugars).1 Therefore, flour samples from wheats of different quality (kernel hardness, cultivar, harvest, and storage history) can be expected to interact differently with water. In particular, kernel hardness has been found to influence water absorption and retention capacities2 as well as water partitioning between flour components.1,3 Bad wheat storage conditions have been shown to cause a decrease in starch and protein content in flour,4,5 as well as a deterioration of wheat storage proteins constituting gluten.6,7 In fact, flour is obtained from artificially aged seeds in low yield and with reduced capability of giving gluten upon hydration. In this context, it is therefore of interest to investigate the influence of both wheat hardness and aging on the interaction of flour components with water. The degree of interaction between macromolecular systems and water can be monitored by measuring 1H magnetization transfer which takes place between the “mobile” protons of water and the “immobile” protons in macromolecules, either by chemical exchange and/or cross-relaxation via dipolar interactions. The importance of cross-relaxation in the analysis of relaxation times of water and/or macromolecules in hydrated biological systems is well-known ever since the 1970s.8,9 Magnetization transfer techniques have been extensively used to investigate many biological systems with * To whom correspondence should be addressed. E-mail: lucia@ ipcf.cnr.it. Phone: +39-50-3152517. Fax: +39-50-3152442.

a too low quantity of solid phase to be observable by standard solid-state NMR techniques, as for example agar,10 lipid bilayers,11 and tissues,12 but its application has proved useful also for studying solids in more concentrated systems, such as hydrated starch,13-16 flour, and gluten.15 Magnetization transfer rates can be quantitatively determined using the saturation transfer method.17 Saturation transfer involves the selective steady-state irradiation with a radio frequency of one member of an interacting pair of spin systems and observation of the effects of this irradiation on the nonirradiated spin system. If exchange or crossrelaxation is occurring, the irradiation will cause a decrease in the steady-state magnetization of the observed spin system. In a heterogeneous system containing solidlike macromolecular protons and liquidlike water protons it is thus possible to obtain NMR parameters of the solid component through observation of the liquid component, using standard NMR spectrometers. The simplest model used to interpret magnetization transfer data is the so-called “two-pool” model,18 which describes the heterogeneous water/macromolecule system as two proton pools, one liquid and one solidlike, which can be roughly identified with the water and the macromolecular protons, respectively. The coupling of the two pools takes place either through dipolar interactions or chemical exchange of water protons in the surface layer with the macromolecular matrix, the two phenomena yielding the same effect on the observed magnetization transfer. The surface layer of water communicates with the bulk water pool by diffusion, whereas the effect of the interaction with water is transferred to internal protons of the macromolecular matrix via spin diffusion. A quantitative description of the magnetization transfer as a function of the relative proton populations, the transfer rate, and the relaxation behavior of the

10.1021/bm049829m CCC: $27.50 © 2004 American Chemical Society Published on Web 07/10/2004

1H

Magnetization Transfer in Gluten and Flour

two pools has been given starting from the coupled Bloch equations, assuming a Lorentzian NMR line shape for the water pool and a Lorentzian,10,11,19-21 Gaussian,22 superLorentzian,12 or other flexible line shapes23,24 for the macromolecular proton pool. A more rigorous treatment of the semisolid pool using the Redfield-Provotorov theory25 and a more complicated three-pool model26 have also been proposed but have found limited application due to the high number of parameters introduced and, moreover, have not provided significant improvements in the analysis of the data.27 To the best of our knowledge, no quantitative determination of the magnetization transfer rate and the fraction of interacting protons in hydrated flour and gluten has been reported in the literature. In the present work, magnetization transfer measurements were performed on hydrated flour and gluten samples obtained from durum wheat (Triticum durum cv. Cappelli) and bread wheat (Triticum aestiVum cv. Centauro) seeds, either unaged or artificially aged over different periods of time, to test the applicability of a simple two-pool model to these heterogeneous systems and to gain information on the different degree of interaction between water and flour or gluten depending on both the type of wheat examined and the aging treatment. Measurements were performed on each sample at several radio frequency power levels and frequency offsets, as required for a quantitative interpretation of magnetization transfer.22 Materials and Methods Accelerated Aging. Seeds of durum wheat (Triticum durum cv. Cappelli) were cultivated in the experimental fields of the Department of Botanical Sciences of the University of Pisa (Italy). Seeds of bread wheat (Triticum aestiVum cv. Centauro) were provided by the Societa` Produttori Sementi S.p.a. (Bologna, Italy). Accelerated aging was performed according to Delouche and Baskin28 by incubation of wheat seeds at 40 °C and 100% relative humidity for 4, 6, and 10 days, as reported in refs 6 and 7. Flour and Gluten Preparation. Seeds were ground in a break roller-mill (Labormill 4 RB, Italy) and flour was stored at -20 °C. Gluten was manually extracted from flour according to the following procedure. 3 mL of distilled water was added to 5 g of flour and mixed with a glass rod, and the resulting dough was washed drop by drop with 300 mL of distilled water. Gluten was separated from flour and water using a thin mesh net, frozen, lyophilized, homogenized in dry conditions with a Retsch MM2000 mill, and stored at -20 °C. Flour from wheat seeds aged for 10 days did not give gluten. NMR Sample Preparation. Dry flour and gluten samples were obtained by pumping under vacuum for 6 h and dehydrating further in a vacuum desiccator over P2O5 for at least 5 days. Hydrated gluten samples were prepared by weighing about 40 mg of gluten in 5 mm o.d. NMR tubes and adding appropriate amounts of bidistilled water to wet the powders completely. The samples were gently mixed with a thin glass rod to obtain a cohesive mass, and the excess water was eliminated with a syringe in order to obtain

Biomacromolecules, Vol. 5, No. 5, 2004 1825

samples containing 60 ( 1 wt % water. Hydrated flour samples were prepared by weighing about 40 mg of flour in 5 mm o.d. NMR tubes and adding appropriate amounts of bidistilled water in order to make 60 wt % water suspensions. Flour samples were mixed using a vortex mixer for 10 min. All samples were let to stand for 2 h before performing NMR measurements. NMR Experiments. NMR experiments were run on a Bruker AMX300 spectrometer working at 300.13 MHz for proton. The magnetization transfer pulse sequence was composed of a radio frequency (r.f.) irradiation pulse of amplitude ω1, at an offset frequency ∆ with respect to the water proton resonance, applied for 5 s followed by a 6 µs 90° pulse after a delay time of 200 µs.22 A r.f. pulse on the solid of 5 s ensured that the steady state was reached for all samples. Measurements were performed at 5 r.f. amplitudes (ω1/2π ) 310, 640, 1340, 2825, and 6500 Hz) for each gluten sample and at 3 r.f. amplitudes (ω1/2π ) 640, 1340, and 2825 Hz) for each flour sample, with 37 different offset values distributed along a logarithmic scale between 10 Hz and 320 kHz. The ω1 values were determined from the length of the 180° pulse. Eight scans were acquired for each experiment with a recycle time of 4 s. All experiments were run at room temperature (23.0 ( 0.5 °C). The longitudinal relaxation times (T1) of water protons were determined for all of the samples by inversion recovery (IR) experiments with 18 variable delay values ranging from 100 µs to 3 s. Measurements were performed using a relaxation delay of 4 s and acquiring eight scans for all samples. Transverse relaxation times (T2) of water protons were determined using the Carr-Purcell-Meiboom-Gill (CPMG) spin-echo pulse sequence with an echo delay of 1 ms and a maximum of 100 echoes. Data Analysis. The baseline corrected area of the water peak (MAz ) in each recorded spectrum was measured to build curves of MAz /MA0 vs ∆ at different ω1 values, MA0 being the water peak area obtained without off-resonance preirradiation. Experimental values of MAz /MA0 at all ω1 and ∆ values were fitted to the two-pool model described in the Results and Discussion section using a nonlinear least squares procedure based on the normal equations29 and implemented in a homemade program written in Fortran language (Visual Fortran 5.0, Microsoft). Results and Discussion 1 H NMR spectra of hydrated flour and gluten (60 wt % H2O) are dominated by an intense peak at 4.6 ppm which is mainly due to mobile water protons and therefore, for simplicity, indicated throughout the paper as water peak; very weak peaks around 2 ppm ascribable to lipids are also observed. The rigid protons of flour and gluten are not detectable in the high-resolution 1H spectrum. However, their interaction with the mobile water protons gives rise to a decrease in the water protons signal in magnetization transfer (MT) experiments, where they are selectively irradiated with a r.f. pulse that is off-resonance from the liquid water signal. The results of the magnetization transfer experiments are

1826

Biomacromolecules, Vol. 5, No. 5, 2004

Calucci et al.

Figure 1. MT curves of hydrated bread wheat flour (top left), durum wheat flour (bottom left), bread wheat gluten (top right), and durum wheat gluten (bottom right) from unaged seeds obtained at different r.f. amplitudes. Diamonds: ω1/2π ) 6500 Hz; circles: ω1/2π ) 2825 Hz; squares: ω1/2π ) 1340 Hz; down triangles: ω1/2π ) 640 Hz; up triangles: ω1/2π ) 310 Hz. Lines are drawn to guide the eyes.

usually reported by plotting the ratio of the intensity of the bulk water signal with (MAz ) and without (MA0 ) off-resonance r.f. irradiation as a function of the offset frequency, ∆; different curves are obtained at different r.f. amplitudes, ω1. The MT curve reflects the entity of magnetization transfer between water and macromolecule as well as the line shape of the macromolecular protons. In general, protons in more rigid environments give broader NMR resonances, so that possible magnetization transfer effects on the water signal intensity are observed at higher offset values; this is reflected in a shift of the steeply ascending part of the curves to higher offset values. The same effect is observed with increasing r.f. irradiation power. It must be pointed out that, in the absence of magnetization transfer, the MAz /MA0 vs ∆ curve is a sigmoidal curve corresponding to the Lorentzian line shape of the water peak arising from the direct effect of the r.f. irradiation on the water signal.22 The area of the region between this sigmoidal curve and the observed curve is directly related to the efficiency of magnetization transfer; a larger area is therefore indicative of either a faster crossrelaxation and/or a higher proportion of solid protons interacting with water. Representative examples of MT curves obtained for hydrated gluten and flour samples with offset frequencies

ranging from 10 Hz to 320 kHz at different r.f. powers are shown in Figure 1. The MAz /MA0 ratio is close to zero for offset values smaller than 1000 Hz and increases at higher offsets reaching 1 for offset values greater than 40-100 kHz, depending on the ω1 value and on the sample. At low offset frequencies ( 60 kHz, our model could not successfully account for the attenuated signal intensity for the higher r.f. amplitudes. This deviation is ascribable to the choice of the super-Lorentzian curve which does not have a sufficiently long high frequency tail to match the experimental data. Nevertheless, the overall good agreement between experimental and calculated curves confirms the applicability of a two-pool model also for the type of systems here examined and the validity of a super-Lorentzian line shape for the solid proton pool. This is not totally surprising since the solid pool in these systems is quite rigid but characterized by dynamic heterogeneity, also pointed out in

1H

Magnetization Transfer in Gluten and Flour

Biomacromolecules, Vol. 5, No. 5, 2004 1829 Table 2. Longitudinal Relaxation Times of Water Protons Determined from Inversion Recovery Experiments

T1 (ms)

aging time (days) Bread Wheat Flour

840 ( 10 800 ( 10 640 ( 10 580 ( 10

0 4 6 10 Durum Wheat Flour

700 ( 10 630 ( 10 720 ( 10 520 ( 10

0 4 6 10 Bread Wheat Gluten

570 ( 10 450 ( 10 430 ( 10

0 4 6 Durum Wheat Gluten 0 4 6

Figure 3. Examples of experimental points and calculated (lines) MT curves for hydrated flour (top) and gluten (bottom) at different ω1 values. Diamonds: ω1/2π ) 6500 Hz; circles: ω1/2π ) 2825 Hz; squares: ω1/2π ) 1340 Hz; down triangles: ω1/2π ) 640 Hz; up triangles: ω1/2π ) 310 Hz.

previous studies,33,34,36-39 and cannot therefore be suitably described neither by a Lorentzian function nor by a single Gaussian line shape; moreover, the use of a super-Lorentzian function is not totally unfounded since isotropic motional averaging of dipolar interactions is not to be expected for most macromolecular protons in hydrated flour or gluten on the basis of solid-state NMR studies.33,34,37-39 In addition, even though it has been argued that in complex macromolecular systems no a priori choice of line shape should be made, no effects on the fitted parameters have been observed when a super-Lorentzian or a flexible function, obtained from the data through an iterative procedure, have been used,24 therefore justifying the use of the much simpler superLorentzian function. T2B times are quite short for all samples, with flour showing slightly longer values than gluten, indicating that both flour and gluten are quite rigid, in agreement with previous observations.15,33,34,36,39 No significant differences are observed with aging treatment. The best-fit values obtained for 1/RAT2A are not typical of free water but rather indicate water in restricted environments, this phenomenon being more pronounced in flour. The value of the funda-

370 ( 10 340 ( 10 360 ( 10

mental magnetization transfer rate constant R is clearly higher in gluten with respect to flour samples, whereas no differences are observed between the two wheat types nor among samples from differently aged seeds. This behavior suggests stronger dipolar interactions between water and gluten protons, possibly favored by the lower mobility of the solid phase. On the other hand, the lower values of RMB0 /RA observed in all gluten samples indicate that a less efficient magnetization transfer takes place in gluten with respect to flour. The lower values obtained in gluten could be ascribed to a lower MB0 and/or a higher RA value. Higher longitudinal relaxation rates of water protons in gluten were determined from Inversion Recovery experiments, as reported in Table 2. An indication of this difference is also given by the 1/RAT2A values, reported in Table 1, which are markedly higher for flour, also considering that T2 of water protons, determined by CPMG experiments, resulted quite similar for gluten and flour samples (T2 = 20 ms). However, assuming the RA values equal to the relaxation rates obtained from IR experiments and using the best-fit R values, MB0 = 0.15 and 0.05 are estimated for flour and gluten, respectively. Notwithstanding the roughness of this estimate, mainly due to the impossibility of determining the intrinsic RA value for water in these heterogeneous systems, these results indicate that only a fraction of all the flour protons (which, in our samples, would give MB0 = 0.30) contributes to magnetization transfer and, in the case of gluten, an even smaller fraction (all of the gluten protons should give MB0 = 0.40). This confirms a more intimate mixing in the flour/water system which is expected considering the different hydrophilic/hydrophobic properties of starch and gluten. Starch, the major component of flour, has a high affinity with water due to the presence of a high number of OH groups which can give rise to proton exchange with water. On the contrary, gluten is a polymer with quite a high number of nonpolar side chains and few side chains with charged groups;40 gluten is therefore essentially hydrophobic but characterized by a

1830

Biomacromolecules, Vol. 5, No. 5, 2004

small number of sites strongly interacting with water. All this considered, and given the higher mobility of liquid water in hydrated gluten with respect to flour, indicated by the 1/RAT2A values, the gluten/water system can be described as a quite rigid protein network embedding pools of relatively mobile water, whereas in hydrated starch, water is essentially located on the surface of the biomacromolecules. An indication of the presence of relatively mobile water pools in hydrated gluten is given by Near Infrared spectroscopy measurements where, upon hydration, a higher free water signal is observed in starch/gluten mixtures with respect to starch only.41 These observations are also consistent with the different surface wettability of flour and gluten, sensibly higher for the former,42 and with the lower water content in the first hydration layer determined for gluten by means of vapor adsorption studies.43 The differences in the curves observed for bread and durum wheat gluten highlighted in Figure 1 are reflected in the different fitting parameter values obtained for the two sets of samples; in particular, the sensibly lower values of 1/RAT2A in durum wheat gluten samples suggest a more efficient direct relaxation of water protons, which makes the direct effect comparable to the magnetization transfer pathway. This behavior can be ascribed to a higher content of relatively mobile water in durum wheat gluten with respect to the bread wheat one, possibly due to differences in composition and/or structure, which deserve further future investigations. Effects of wheat aging are observed on the parameters RMB0 /RA and 1/RAT2A, which decrease for all samples by prolonging the aging treatment of the seeds. Both effects point to a minor degree of interaction between the macromolecular systems and water. An indication of this effect is also given by the observation of a progressively lower moisture content in bread wheat flour from seeds aged for longer periods.34 Moreover an increasing mobility of the solid pool with aging, highlighted by previous studies on bread wheat flour samples,34 could decrease the water protons/ macromolecule protons dipolar interactions and hence the magnetization transfer. Conclusions Proton magnetization transfer experiments on hydrated gluten and flour (60 wt % water) allowed quantitative information to be acquired on water/biomacromolecules interactions in these systems, as well as on their structural and dynamic properties. Flour and gluten protons interacting with water protons were found to constitute a quite rigid solidlike phase, in which proton-proton dipolar couplings are partially averaged by anisotropic motions. A superLorentzian line shape well described the magnetization of this phase when a theoretical two-pool model based on the modified Bloch equations was employed to analyze MT curves obtained at different r.f. powers and offset values. Magnetization transfer rate showed higher values for gluten than for flour samples, indicating stronger water-protein interactions, probably ascribable to NH, CO, OH, and NH2+ groups. Nevertheless, magnetization transfer was more

Calucci et al.

efficient for flour samples than for gluten samples; this behavior is ascribed to both a higher number of macromolecular protons interacting with water and higher longitudinal relaxation times of water protons in hydrated flour. These findings are in agreement with the high number of OH groups in starch, the main component of flour, which can give both hydrogen bonds and chemical exchange with water protons. Finally, accelerated aging of wheat seeds resulted in a decreased MT efficiency, suggesting that damages to the biomacromolecules (mainly starch and storage proteins) present in seeds upon aging cause a reduction in the interaction of flour and gluten with water. The main macroscopic consequence of this phenomenon is the reduced gluten yield of flour from seeds aged for longer periods.6,7 Acknowledgment. This work was financially supported by CNR-Progetto Giovani Agenzia 2000. We thank Dr. Silvia Ghiringhelli for technical assistance in preparing the samples. References and Notes (1) Matveev, Yu. I.; Greenberg, V. Ya.; Tolstoguzov, V. B. Food Hydrocolloid 2000, 14, 425-437. (2) (a) Merritt, P. P.; Bailey, C. H. Cereal Chem. 1939, 16, 377-383. (b) Preston, K. R.; Tipples, K. H. Cereal Chem. 1978, 55, 96-101. (c) Sollars, W. F. Cereal Chem. 1972, 49, 168-172. (3) Greer, E. N.; Stewart, B. A. J. Sci. Food Agric. 1959, 10, 248-252. (4) (a) McDonald, M. B. Seed Sci. Res. 1998, 8, 265-275. (b) McDonald, M. B. Seed Sci. Technol. 1999, 27, 177-237. (5) Krishnan, P.; Nagarajan, S.; Dadlani, M.; Moharir, A. V. Seed Sci. Technol. 2003, 31, 541-550. (6) Galleschi, L.; Capocchi, A.; Ghiringhelli, S.; Saviozzi, F.; Calucci, L.; Pinzino, C.; Zandomeneghi, M. J. Agric. Food Chem. 2002, 50, 5450-5457. (7) Calucci, L.; Capocchi, A.; Galleschi, L.; Ghiringhelli, S.; Pinzino, C.; Saviozzi, F.; Zandomeneghi, M. J. Agric. Food Chem. 2004, 52, 4274-4281. (8) Edzes, H. T.; Samulski, E. T. J. Magn. Reson. 1978, 31, 207-229. (9) Koenig, S. H.; Bryant, R. G.; Hallenga, K.; Jacob, G. S. Biochemistry 1978, 17, 4348-4358. (10) Eng, J.; Ceckler, T. L.; Balaban, R. S. Magn. Reson. Med. 1991, 17, 304-314. (11) Fralix, T. A.; Ceckler, T. L.; Wolff, S. D.; Simon, S. A.; Balaban, R. S. Magn. Reson. Med. 1991, 18, 214-223. (12) Morrison, C.; Henkelman, R. M. Magn. Reson. Med. 1995, 33, 475482. (13) Wu, J. Y.; Bryant, R. G.; Eads, T. M. J. Agric. Food Chem. 1992, 40, 449-455. (14) Wu, J. Y.; Eads, T. M. Carbohydr. Polym. 1993, 20, 51-60. (15) Vodovotz, Y.; Vittadini, E.; Sachleben, J. R. Carbohydr. Res. 2002, 337, 147-153. (16) Vodovotz, Y.; Dickinson, L. C.; Chinachoti, P. J. Agric. Food Chem. 2000, 48, 4948-4954. (17) Forsen, S.; Hoffman, R. A. J. Chem. Phys. 1963, 39, 2892-2901. (18) Grad, J.; Bryant, R. G. J. Magn. Reson. 1990, 90, 1-8. (19) Caines, G. H.; Schleich, T.; Rydzewski, J. M. J. Magn. Reson. 1991, 95, 558-566. (20) Yeung, H. N.; Swanson, S. D. J. Magn. Reson. 1992, 99, 466-479. (21) Koenig, S. J.; Brown, R. D., III.; Ugolini, R. Magn. Reson. Med. 1993, 29, 311-316. (22) Henkelman, R. M.; Huang, X.; Xiang, Q.; Stanisz, G. J.; Swanson, S. D.; Bronskill, M. J. Magn. Reson. Med. 1993, 29, 759-766. (23) Yeung, H. N.; Tzou, D. L.; Lee, S. M.; Huang, F. Y.; Hur, Y. J. Magn. Reson. B 1996, 113, 167-171. (24) Li, J. G.; Graham, S. J.; Henkelman, R. M. Magn. Reson. Med. 1997, 37, 866-871. (25) Yeung, H. N.; Adler, R. S.; Swanson, S. D. J. Magn. Reson. A 1994, 106, 37-45. (26) Kuwata, K.; Brooks, D.; Yang, H.; Schleich, T. J. Magn. Reson. B 1994, 104, 11-25. (27) Henkelman, R. M.; Stanisz, G. J.; Graham, S. J. NMR Biomed. 2001, 14, 57-64.

1H

Magnetization Transfer in Gluten and Flour

(28) Delouche, J. C.; Baskin, C. C. Seed Sci. Technol. 1973, 1, 427452. (29) Hamilton, W. C. Statistics in Physical Science; The Ronald Press: New York, 1964. (30) Van Duynhoven, J. P. M.; Kulik, A. S.; Jonker, H. R. A.; Haverkamp, J. Carbohydr. Polym. 1999, 40, 211-219. (31) Morrison, C.; Stanisz, G.; Henkelman, R. M. J. Magn. Reson. B 1995, 108, 103-113. (32) Tessier, J. J.; Dillon, N.; Carpenter, T. A.; Hall, L. D. J. Magn. Reson. B 1995, 107, 138-144. (33) Calucci, L.; Forte, C.; Galleschi, L.; Geppi, M.; Ghiringhelli, S. Int. J. Biol. Macromol. 2003, 32, 179-189. (34) Calucci, L.; Galleschi, L.; Geppi, M.; Mollica, G. Biomacromolecules 2004, 5, 1536-1544. (35) Belton, P. S.; Duce, S. L.; Colquhoun, I. J.; Tatham, A. S. Magn. Reson. Chem. 1988, 26, 245-251. (36) Belton, P. S.; Duce, S. L.; Tatham, A. S. J. Cereal Sci. 1988, 7, 113-122.

Biomacromolecules, Vol. 5, No. 5, 2004 1831 (37) Garbow, J. R.; Schaefer, J. J. Agric. Food Chem. 1991, 39, 877-880. (38) Gil, A. M.; Alberti, E.; Santos, D. In AdVances in magnetic resonance in food science - a View to the future; Webb, G., Belton, P. S., Gil, A. M., Delgadillo, I., Eds.; The Royal Society of Chemistry: Cambridge, U.K., 2001; pp 43-53. (39) Ruan, R. R.; Wang, X.; Chen, P. L.; Fulcher, R. G.; Pesheck, P.; Chakrabarti, S. Cereal Chem. 1999, 76, 231-235. (40) Singh, H.; MacRitchie, F. J. Cereal Sci. 2001, 33, 231-243. (41) Wesley, I. J.; Blakeney, A. B. J. Near Infrared Spectrosc. 2001, 9, 211-220. (42) Roman-Gutierrez, A.; Sabathier, J.; Guilbert, S.; Galet, L.; Cuq, B. Powder Technol. 2003, 129, 37-45. (43) Roman-Gutierrez, A. D.; Guilbert, S.; Cuq, B. J. Cereal Sci. 2002, 36, 347-355.

BM049829M