The atomic force experiments showed that gliadin molecules were spherical and tended to aggregate (Plates 3.5, 3.6, 3.7 and 3.8). The diameter was found to be approximately 32nm (Plate 3.10). SANS results gave two different radii of gyration for both gliadin and deamidated gliadin (Table 3.13). If Rg2 is the actual radius of gyration, then Rg1 would have to be an aggregate. AFM results were five times bigger than the SANS results. As the magnification of AFM was pushed to the limit where it was just impossible to immobilize the proteins, a different picture emerged (Plate 3.12). Gliadin was not spherical, which was also mentioned by Krejci and Svedberg (1935). The axial ratio according to Plate 3.12 was approximately 15/4=3.75. Lamm and Poulsen (1936) reported an axial ratio of 8:1 from diffusion measurements whilst Entrikin (1941) reported an axial ratio of 13:1 from dielectric measurements. It could also be argued that the image of gliadin as seen from Plate 3.12 is only part of the gliadin molecule with a hole in the middle as suggested by Grosskreutz (1961). The diameter of the central hole in this study was approximately 5-6nm, whereas, that proposed by Grosskreutz (1961) was 100Å. It is being proposed from this study that molecules of gliadin which are not spherical in shape aggregated to a compact spherical form of diameter 32nm, which can be called a unit. This unit then aggregated with other units to form massive aggregates as shown in Plates 3.5 to 3.8. SAXS and STM measurements made on gamma-gliadin indicated a molecule about 10nm long, with a diameter of 3nm ( Thomson et al., 1992).

Another feature that came to light in the SANS experiment was that alcohol was preferentially adsorbed onto the surface of the gliadin molecules to form an inner solvent layer with water as the external layer (Figure 3.20). This did not occur with deamidated gliadin. The CD results in the present study did not match those found for gamma-gliadin ( Tatham et al., 1990b). It was mentioned in their study that the repetitive domain of gamma-gliadin adopted a beta-reverse turn whilst the non-repetitive domain was alpha-helix. Kasarda et al. (1968) showed gliadin to be 33-34% alpha-helix which matched the results of the computer prediction ( chapter 4). Reverse turns are normally short ( chapter 1). If they are extended as in the gliadins, then there is the possibility that they might form beta-sheets. Changing the polar but non-ionisable glutamine for glutamic acid in deamidated gliadin increased the helix content. Polyglutamic acid is an alpha-helix. Deamidated gliadin was found to be very stable (Figures 3.30 and 3.28). No endotherm was found between 50°C and 94°C from the DSC study and the CD study showed reversibility. The disulphide bonds in both gliadin and deamidated gliadin were found to be in the interior of the protein. If the deamidation process denatured gliadin, then it is likely that some of the disulphide bonds may be reduced and be exposed. In this study it is proposed that, the deamidation process caused some conformational change in the protein without disrupting the disulphide bonds.

It was not possible to find an isolated deamidated gliadin molecule in the AFM study even at the dilution and magnification used (Plates 3.13, 3.14, 3.15, 3.16 and 3.17). This may be due to the fact that the samples were dissolved in distilled water, and surface tension may have pulled the molecules together on drying.

Beta-mercaptoethanol reduced the disulphide bonds and partly unfolded deamidated gliadin which through disulphide interchange with other molecules formed strands as the beta-mercaptoethanol evaporated. This led to the formation of a network (Plate 3.18). When urea was added, the deamidated gliadins formed fibrils of around 343nm diameter (Plate 3.19, 3.20 and 3.21). The difference between the two structures was that although beta-mercaptoethanol destabilized the molecule, water prevented the hydrophobic groups from being exposed to the surface. When urea was added, the hydrophobic groups were exposed and different molecules came together through a combination of hydrophobic interactions and disulphide bond formation. Although deamidated gliadin contains substantial amounts of glutamic acid which would be expected to lead to electrostatic repulsion. The molecules tended to aggregate due to a predominance of hydrophobic groups (Appendix A).

Below 16% (w/v) the viscosity of SWP decreased with increasing temperature (Figure 3.25). Above 16% (w/v), the viscosity increased only when the temperature was above 90°C (Figure 3.26). The exposed hydrophobicity (Figure 3.27) was greater than the surface hydrophobicity. SWP dispersions aggregated at room temperature. On heating, to 60°C the aggregates collapsed (Figure 3.25 and 3.26) leading to the decrease in viscosity. Reduction in the size of the aggregates exposed more hydrophobic groups (Figure 3.27) on the surface of the molecules for both deamidated gliadin and deamidated glutenin.

Staining SWP with haematoxylin and eosin proved that SWP was only partially deamidated. The red colour in Plate 3.24 indicated eosin binding to basic residues in SWP. Most of the basic residues were on the large aggregates which were most likely to be deamidated glutenin since they are known to polymerize into large molecular weight particles. On heating to 60°C the large aggregates broke up (Plate 3.25) and at higher temperatures of 90°C structure formation was observed (Plate 3.26). When an aliquot of a reducing agent, DTT was added to SWP at room temperature the large aggregates broke up into strands (Plate 3.27). Plates 3.28 and 3.29 show the effect of haematoxylin staining on SWP with and without DTT which gave a better definition of the shape and size of the dispersions.

Below 3% (w/v) SWP exhibited Newtonian flow with viscosity independent of shear rate (Figure 3.22). Above 3% (w/v) Newtonian flow was still evident at low shear rates whilst at high shear rates shear thinning occurred. The intrinsic viscosity of SWP was found to be 0.9 dlg-1 (Figure 3.23). Glutenin in pH 3.1, 0.003M buffer had a high intrinsic viscosity of 2.23 (100 mlg-1) that decreased markedly to 0.7 with increase in buffer concentration according to Taylor and Cluskey (1962). They interpreted these findings to indicate that molecules of glutenin are highly asymmetric and fairly loosely organised since they tend to unfold in lower ionic strength solutions. Dissociating agents like urea and guanidine hydrochloride increased the intrinsic viscosity of glutenin due to additional unfolding and gliadin was found to have an intrinsic viscosity of 0.16 (Wu and Dimler, 1964). In the present study the intrinsic viscosity of SWP was probably determined by the deamidated glutenin fraction, because intrinsic viscosity is related to molecular weight by the empirical Mark-Houwink equation. For rigid spheres, Einstein showed that intrinsic viscosity is independent of molecular weight. Thus the intrinsic viscosity of various globular proteins falls within the range 0.031 to 0.043 dl/gm, with no correlation between M and [h] (Yang, 1961). For prolate (cigar-shaped) or oblate (disk-shaped) ellipsoids of revolution the intrinsic viscosity is a function of the axial ratio only and does not depend on the molecular weight of the particle. The factor that determined the intrinsic viscosity of SWP cannot be spherical.