SCANFHNG ELECTRON MICROSCOPY OF ‘ FLOUR-‘WAYER DOUGHS TREATED WITH- . OXIDiZlNG AND REDUCRNG AGENTS _ . Thesis for the Segree of M. S. MECHESM! STAE UE‘EEVERSETY LEAH GM. EVAE‘éS 3.975 IIIIIIIIII IIIIIIIIIIIIIIIIIIIIIIIIIII 3 1293 00651 6458 IIII ABSTRACT SCANNING ELECTRON MICROSCOPY OF FLOUR-WATER DOUGHS TREATED WITH OXIDIZING AND REDUCING AGENTS By Leah Gail Evans The present study was undertaken to investigate the role of oxidizing and reducing agents in flour-water doughs by examining the surface characteristics of stretched and unstretched samples using scanning electron microscopy (SEM). The oxidizing agents used in this study were potassium bromate (75 ppm), potassium iodate (75 ppm) and azodicarbamide (ADA) (H5 ppm). The reducing agents were L-cysteine (75 ppm) and glutathione (75 Ppm). The water absorption and mixing re- quirements of the control and the experimental doughs were determined in the farinograph. Adjusted water absorptions and peak development times were used for each dough so that optimal mixing was defined at a standard peak consistency. The doughs were examined by SEM at four stages of development: (1) when undermixed; (2) at optimal development; (3) after incubation; and (H) after overmixing. SEM revealed that the gluten of the undermixed flour~water dough was not uniformly developed and that the starch granules were clustered in patches on the dough surface. Continued Leah Gail Evans mixing distributed a thin film of gluten throughout the mass. At optimum development a strong starch-protein association was established, which supported the formation of thin fibrils oriented in the direction the dough was stretched. The ability to form gluten fibrils increased after incubation, but de- creased upon overmixing. The gluten of the overmixed dough disintegrated into a spindly web-like structure over the starch granules and failed to form fibrils upon stretching. SEM suggested that the viscoelastic properties of the gluten were lost during overmixing. The reduction in cohesiveness of the gluten may have resulted from forced cleavage of disulfide bonds during excessive manipulation, especially at high mixing speeds. In all doughs, the gluten sheet was pitted with small round pores, probably caused by lipid inclusions in the gluten complex. Continued mixing or stretching enlarged the pores and led to discontinuity of the gluten with ultimate breakdown of the dough. Unstretched samples of the doughs containing oxidizing and reducing agents did not differ significantly from the control. Upon stretching, however, SEM revealed considerable differences in gluten characteristics of the samples. Dough containing potassium bromate was not distinguishable from the control when mixed at 30°C. After incubating the dough bromate strengthened the gluten, as indicated by the formation of thick fibers upon stretching rather than the thin fibrils observed in the control. Leah Gail Evans The optimally mixed and incubated potassium iodate treated doughs formed thin fibrils upon stretching. A reduction in extensibility expected from the oxidizing agent was not ob- served by SEM and suggested that a high level of iodate accom- panied by high speed mixing reversed the improving effect of iodate. Similar to the control, the overmixed iodate treated dough broke down into a discontinuous web of gluten and did not form fibrils upon stretching. ADA treatment restricted the formation of thin rounded fibrils in stretched samples of the dough but instead the gluten tended to divide into wide flat strips. Even after overmixing of the ADA treated dough, the starch-protein bond was maintained and the gluten appeared to resist breakdown. The disulfide reducing agents, cysteine and glutathione, had similar effects on gluten, but differed in degree of reactivity. Cysteine severely damaged the gluten strength and prevented fibril formation at all stages of development. On the other hand, glutathione was detrimental to fibril for- mation only after the dough was overmixed. Generally, breakdown of the gluten sheet was characterized by large irregular holes surrounding the starch granules. The reducing agents promoted thiol-disulfide interchange in the dough allowing extensibility to increase at the expense of elasticity, a change indicated by the extent of fibril for- mation. Upon loss of elasticity, fibrils could no longer form, so the gluten disintegrated and stretched irregularly. Complete destruction of elasticity may be related to the molecular Leah Gail Evans configuration of glutenin, which must be linear in order to impart elastic properties to gluten. Agents reducing the large linear glutenin into smaller globular subunits decrease the molecular interactions required for an elastic network. Cysteine and glutathione appear to progressively reduce glutenin and destroy the visco-elastic properties of the dough. Thus, gluten continuity cannot be maintained under the stresses of stretching and mixing. SCANNING ELECTRON MICROSCOPY OF FLOUR-WATER DOUGHS TREATED WITH OXIDIZING AND REDUCING AGENTS By Leah Gail Evans A THESIS Submitted to ~ Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Food Science and Human Nutrition 1976 ACKNOWLEDGMENTS The author expresses her sincere appreciation to Dr. A. M. Pearson for his support and guidance throughout the preparation of this thesis. Special thanks is given to Dr. G. R. Hooper for his advice in electron microscopy and for serving on the author's committee. Appreciation is also ex- pressed to Dr. M. E. Zabik and Dr. P. Markakis for serving as members on the author's committee. Thanks is given to Dr. T. Volpe of the American Institute of Baking, Chicago, Illinois, for her advice and assistance in designing this project. The author is indebted to the Pillsbury Company, St. Louis, Missouri, for supplying the untreated bread flour and to Penwalt Corporation, Chicago, Illinois, for the sample of azodicarbamide. Finally, the author is grateful to her parents, Woodrow and Vivian, for their encouragement, patience and generosity through twenty-three years of learning. 11 TABLE OF CONTENTS INTRODUCTION. LITERATURE REVIEW . . Methods of Making Dough Principles of Breadmaking . . Formulation . . . . . . . Mixing . . . Fermentation . Baking . . . . . . . Microstructure . Early Studies on Bread Structure Shape and Arrangement of Starch Granules in Dough O O O O O O Theories of Gas Bubble Structure Effects of Mixing . . . . Gluten . . Properties of Glutenin and Gliadin Amino Acid Composition . Thiol- Disulfide Interchange Theory Lipoprotein Model of Gluten . Dough Development . . . . . . . Fermentation . . . . . . Mechanical Development Chemical Dough Development The action of oxidizing and reducing agents . . . . . Reducing agents . Oxidizing agents . MATERIALS AND METHODS . . . . Source and Levels of Ingredients Preparation of Dough . . . . . . Flour-Water Dough . . . Oxidized and Reduced Doughs Scanning Electron Microscopy . . RESULTS AND DISCUSSION . . . . . . Farinograph Data . . . . . . . Flour-Water Dough . . . Oxidized and Reduced Doughs 111 O 5? 0(1) Nudmmzn-ww H (D HHHHHHHHHH \ocnoumv1zmnnndca NH H0 22 Scanning Electron Microscopy (SEM) Starch. . . . . . . . Flour-Water Dough . . . . . . Undermixing . . . . . . Overmixing . . Oxidizing Agents ~. . . . Potassium bromate . . . Potassium iodate . . Azodicarbamide (ADA) Reducing Agents . . . Cysteine . Glutathione SUMMARY . . . . . . . . . . . . . . . . BIBLIOGRAPHY . . . . . . . . . . . . iv Table 1. LIST OF TABLES Dough Properties as Determined in the Farinograph . . . . . . . . . . . . . Figure 10. 11. 12. LIST OF FIGURES SEM of starch granules from wheat flour showing the characteristic small and spherical— or large and lentil-shaped' granules. . . . . . . . . . . . . . . . SEM of nodule attached to the surface of a starch granule . . . . . . . . . . . . . . SEM showing optimally mixed control flour- water dough . . . . . . . . . . . . . . . . SEM of optimally mixed control flour-water dough O O O O O O O O O O O O O O O O O 0 SEM of optimally mixed flour-water dough showing sheet-like gluten . . . . . . . . . SEM of optimally mixed flour-water dough illustrating thin fibrils formed upon stretching . . . . . . . . .-. . . . . SEM of unstretched incubated flour-water dough . . . . . . . . . . . . . . . . . . . SEM of incubated flour-water dough . . . SEM of incubated flour-water dough showing thin thread-like fibrils formed upon stretching. . . . . . . . . . . . . . . . SEM showing small starch granules trapped by thin gluten fibrils in stretched incubated dough . . . .-. . . . . . . . . . SEM showing small round pores pitting the surface of the gluten layer of an untreated flour-water dough o o o o o o o o o o o o 0 SEM showing bumps on the surface of the gluten covering a large starch granule . . vi 33 33 36 36 38 38 Al Al ‘13 “3 A6 A6 Figure l3. 1”. 15. 16. 17. 18. 19. 20. 21. 22. 23. 2M. 25. 26. 27. 28. 29. SEM of undermixed flour-water dough showing lack of uniform development . SEM of undermixed flour-water dough showing lack of development with incomplete sheet formation near center of field . . SEM of stretched undermixed flour-water dough . . . . . . . . . . . . . . . . . SEM of stretched undermixed flour-water dough . . . . . . . . . . . . . . . . SEM of overmixed flour-water dough. . SEM of overmixed flour-water dough showing the weakened gluten . . . . . . . . . . SEM showing the gluten of the overmixed flour-water dough . . . . . . . . SEM of stretched overmixed flour-water dough O 0 O O O O O O O O O O O O I O O O 0 SEM of optimally mixed potassium iodate treated dough O O O O O O I O O O O O 0 SEM showing thin fibrils formed upon stretching optimally mixed bromate treated dough I O O O O O O O O O O O O O O O O 0 SEM showing thick fibers formed upon stretching incubated bromate treated dough . . . . . . . . . . . . . . . . . SEM showing optimally mixed potassium iodate treated dough . . . . . . . . . . . SEM depicting long thin fibrils of stretched optimally mixed dough containing potassium iodate. . . . . . . . . . . . . . SEM showing the thin discontinuous gluten of overmixed iodate treated dough . . . SEM showing optimally mixed ADA treated dough O O O O O O O O O O O O O O 0 O O O 0 SEM of stretched ADA treated dough . . . . SEM of incubated ADA treated dough showing resistance to fibril formation upon stretch- ing . . . . . . . . . . . . . . . . . . . vii ”9 51 51 5A 5A 56 56 59 59 61 61 6H 6h 67 67 69 Figgre 30. 31. 32. 33. 3M. 35. 36. 37. 38. 39. 40. “1. SEM of overmixed ADA treated dough. SEM of optimally mixed cysteine treated dough o o o o o o o o o I o 0' o o o 0 SEM of stretched optimally mixed cysteine treated dough . . . . . . . . . . . . . . SEM of stretched incubated cysteine treated dough . . . . . . . . . . . . . . . . . . SEM of overmixed cysteine treated dough . SEM of overmixed cysteine treated dough . SEM of optimall mixed glutathione treated dough O O O O O O O O O O O O O O O O O 0 SEM of stretched optimally mixed glutathione treated dough showing formation of fibrils upon stretching . . . . . . . . . . . . . SEM of incubated glutathione treated dough showing thin fibrils formed upon stretch- ing 0 o o o o o o o o o o o o o o o o o 0 SEM of stretched incubated glutathione treated dough . . . . . . . . . . . . . SEM of overmixed glutathione treated dough. SEM of stretched overmixed glutathione treated dough showing the disruption of the gluten and resulting irregular holes. viii Page 69 72 72 7A 7A 77 77 80 80 82 82 8A INTRODUCTION Oxidizing and reducing agents are widely used in commer— cial breadmaking. Since they develop a balance between elas- ticity and extensibility in the dough, long periods of bulk fermentation are no longer necessary. Thus, production time can be greatly reduced. Reducing agents accelerate the mixing process and soften the dough, whereas, oxidizing agents are added to stabilize the structure. As these agents have a pronounced effect on the rhe- 01081031 properties of the dough, farinograms and load- extension tests have been useful in understanding their influ- ence. The farinograph is used to determine the water absorption of flour and its response to mixing. Load-extension meters (extensigraphs and alveographs) measure the extensibility and relaxation rate of dough by stretching a sample until it ruptures. A curve is then recorded of load versus elongation (Bloksma, 1971). Even when oxidizing agents have a small effect on the farinogram from a dough, load-extension meters can demonstrate that they decrease extensibility (Locken g§_gl., 1960). Appar- ently, they react with the sulfhydryl groups of the gluten proteins to increase the number of disulfide bonds, and thus give a more tenacious network of molecules (Matz, 1972). In contrast, reducing agents rapidly decrease dough con- sistency in the farinograph (Bloksma, 1971) and increase dough extensibility (Tsen, 1970). Since reducing agents increase the number of sulfhydryl groups in the dough, they promote disulfide interchange reactions, and thereby enhance plastic deformation of the mass. The present study was undertaken in order to supplement existing physical and chemical data on the role of oxidizing and reducing agents in bread dough. Although the effects of oxidizing and reducing agents have been measured rheologically and in baking tests, only a limited amount of information is available as to their effects on microstructure as observed with light microscopy (Moss, 197“). Little or no information is found in the literature on the electron microstructure of chemically treated doughs. Therefore, the objective of this investigation was to evaluate oxidized and reduced doughs by means of scanning electron microscopy (SEM). This instrument resolves the surface characteristics of a sample and is suitable for observing the mechanism and extent of gluten spread over the starch granules. LITERATURE REVIEW Methods of Making Dough According to Ponte (1971) four distinct methods of making dough are used for nearly all bread production in the United States. These include the straight dough method, the sponge dough method, liquid fermentation and continuous breadmaking. The former methods are conventional procedures and have been used since man first produced bread, whereas, the latter two procedures are of recent origin. In the straight dough method all components are mixed into a dough by a single-step procedure (Pyler, 1973). Sponge doughs require an extra step in which water, yeast, yeast food and two-thirds of the total flour are combined and fer- mented three to five hours to form a sponge, which is then mixed with the remaining ingredients (Ponte, 1971). Both conventional methods are followed by two to three hours of bulk fermentation. According to Ponte (1971) the straight dough method requires much less equipment, processing time, labor and energy; but the bread has a characteristic bland flavor, and the dough is less flexible to schedule variation than that of the sponge method (Pomeranz 32 al., 1971). The liquid fermentation method, which is the least popular, begins with a concentrated brew made without flour. The brew is later mixed with the dough ingredients (Ponte, 1971). Continuous breadmaking blends a ferment, which may be made with or without flour, with the remaining ingredients into a homogenous mass (Ponte, 1971). The dough is pumped to a developer for intense mixing, and then is extruded into baking pans. Compared to the conventional methods, continuous breadmaking reduces processing time, saves labor and floor space (Ponte, 1971). Because the resulting bread is "cake- like", lacking strength and resilience, it is generally con- sidered inferior in taste and structure (Magoffin, 197”) in comparison to bread produced by the straight or sponge dough procedures (Ponte, 1971). Principles of Breadmaking Regardless of the dough making procedure, the same prin- ciples generally apply to the production of bread. The straight dough method will be used for discussing the basic steps in- volved (Pyler, 1973). Formulation Ingredient proportions are based on the weight of the flour, but the amount of water is varied to give a dough of optimum consistency (Bloksma, 1971). Water absorption, deter- mined by farinograph methods (Locken g§_al., 1960) generally increases with the protein content of the flour (Pratt, 1971). A typical formulation for commercial bread includes flour, water, yeast, yeast food, salt, sweetener, fat, dough condi- tioners and improvers, spoilage retardants, and nutrients for enrichment (Ponte, 1971). Two of these components, flour and water, are essential for forming a glutinous dough. Adding yeast, which is the source of leavening gas (002) during fermentation, transforms the solid mass into a stable foam. Other ingredients contribute to the flavor, texture, volume of the loaf, mechinability during the automated processes and the shelf-life of the finished bread (Ponte, 1971). Mixing Mixing dough ingredients distributes them evenly and pro- motes the optimum development of gluten structure (Bloksma, 1971). It completely disaggregates the flour particles (Tsen, 1970) yielding an apparently homogeneous mass of starch, fat and yeast in the glutinous medium constituting the dough (Burhans and Clapp, 19A2). At first the dough is wet and lumpy, having little coherence, but with continued stretching and folding it develops cohesive elastic properties and begins to pull away from the sides of the container. Eventually, the woolly appearance changes to a satiny sheen, a change called "clearing" (Marston, 1971). Production of these de- sirable properties is called dough development (Bloksma, 1971). When mixed beyond optimum development, the dough breaks down becoming highly extensible and sticky, approaching flu- idity (M088, 1972; Marston, 1971). If it is correctly done, the mixing process develops the gas retaining properties of gluten, a complex of proteins from the endosperm of wheat. In order to retain gas, the gluten must maintain its integ- rity as a film on the inner surface of the expanding gas bubbles produced by yeast fermentation (Baker, 19A1). Fermentation After the dough is mixed to optimum consistency, it is fermented in a warm humid environment. Gas production leavens the dough, flavors are produced and the dough is further devel- oped or ripened to maturity (Magoffin, l97h). It is punched intermittently whenever the dough doubles in bulk, thus collaps- ing the gas cells and redistributing the yeast cells and their nutritive supply. The inner surfaces of the collapsed gas bubbles draw together to produce glutinous nuclei throughout the dough mass, providing for better bubble structure and expansion during the next phase of fermentation (Baker, 19u1). The benefits of rising and subsequent punching are im- proved loaf volume, a fine crumb grain and thin cell walls in the finished bread (Baker, l9fll). Punching the dough revives elasticity, making it less pliable and resistant to deformation and shaping into loaves, so it must rest several minutes before moulding (Pyler, 1973). The loaves are moulded, panned and allowed to rise once more (the final proofing period) until doubled in bulk (Pomeranz 2£.§l-: 1971). Baking Baking fixes the foam-like structure of the fermented dough into bread. In the initial baking stage glutinous cell walls must support the rapidly expanding gas bubbles and a sudden increase in volume (Marek and Bushuk, 1967). As the temperature increases, the proteins are denatured and become rigid as the water is absorbed by the starch. At the same time the gelatinizing starch granules become flexible and undergo pronounced distortion, but they do not disinte- grate (Sandstedt gt gl.,'l95fl). The heat sets the interior of the bread in a vesicular structure. In addition, non- enzymatic Maillard-type browning reactions take place between reducing sugars and free amino groups during baking, thus, browning the crust (Ponte, 1971). Microstructure Early Studies on Bread Structure Verschaffelt and Van Teutem (1915) made camera lucida drawings of sectioned bread, which visually described the phys- ical relationships between the components of breadstuffs. They showed starch and yeast cells embedded in a gluten struc- ture and elongated starch grains exhibiting distinct lines of flow adjacent to air cells. Scheffer (1916) confirmed the structure of bread using photomicrographs. Katz and Van Teutem (1917) stated that starch cells form a regular pattern in bread from correctly fermented dough, the large cells being interspersed with the small ones. They further stated that starch cells tend to separate in bread from overripe doughs, and that the cells are closer and overlap in underripe doughs, which results in no regular organization of the starch (Butterworth and Colbeck, 1938). Shape and Arrangement of Starch Granules in Dough From photomicrographs, Burhans and Clapp (19h2) iden- tified starch granules of various sizes and shapes, which included small spherical ones, intermediate ovoid ones and large elliptical or roughly ovoid ones. Testing at varying fermentation periods, Butterworth and Colbeck (1938) found no characteristic changes in the arrangement of starch cells. Regular and irregular starch formations were observed in stretched films of underfermented, overfermented, and correctly fermented dough. These studies provided the earliest photomicrographs of dough and support the view that gluten is devoid of organized structure at the microscopic level. They proposed that the gas produced in fermentation causes a mechanical squeezing out of starch cells from the gluten. The cells may then adhere to the surface of the gluten mesh. Thus, the quantity of starch embedded in the gluten decreases during fermentation. The mechanical action segregates the starch granules into groups of 10 to 15 in each according to Burhans and Clapp (19h2). On the other hand, Sanstedt gt El- (l95h) concluded that starch and protein do not segregate to isolated positions in dough films. In general, starch granules are separated frcm each other by the continuous protein phase of the film. Theories of Gas Bubble Structure Swanson (1925) theorized that a glutinous network be- tween the starch granules improved the gas retaining abilities of the dough by increasing the capillarity and surface prop- erties of the water, which lies in and between these materials. On the other hand, Baker (19A1) claimed that gluten must be a continuous phase of protein around starch; thus, its continuity and properties determine the bubble characteristics. Baker (l9hl) also stated that the walls of dough bubbles have a matrix of gluten and starch, which impart strength to the bubbles due to their mutual adhesive properties. Working with bread films simulating bubble walls, Baker (l9hl) tested them for surface starch with an iodine stain. Since the unruptured bread films were not colored by iodine, he concluded that the films had walls of impervious glutinous material com— pletely covering the starch granules, and more or less segre— gating the granules to the interior of the film. Baker (l9hl) developed a theory on the properties of gas bubbles, which explains their expansion. According to this theory gluten must have sufficient integrity (a high viscosity compared to its modulus of elasticity) to prevent breaking. The expanding bubbles must draw more glutinous material from the dough mass into their expanding surfaces and produce gas tight films on the inner surfaces. Baker 10 (l9hl) applied this theory to explain the benefits of punching dough during fermentation. Burhans and Clapp (19A2), following progressive changes in breadmaking, noted that tension in the matrix peripheral to the gas pockets produces concentric axial alignment of the long axis of the starch granules. On moulding, much of the gas is forced out, which further arranges the starch longitudinally in the dough sheet. Conversely, the remaining protein matrix coalesces into stratified sheets. Better gas retention ensues, the gas pockets increase in number and can enlarge far beyond previous limits. The final proof yields a porous honeycomb structure where gas bubbles enlarge by expanding instead of rupturing (Burhans and Clapp, 19A2). Effects of Mixing Further work was not published on bread dough micro- structure until Moss (1972) followed dough development through the early stages of mixing to breakdown caused by overmixing. She showed dough development after 1, 3, 7, and 11 minutes of mixing. She noted that at 1 minute most endosperm cells were broken apart, the proteins were hydrated and large masses of gluten developed which bound the starch granules in the mass. At 3 minutes, no diffuse undeveloped protein remained. Discrete starch-protein massestere stretched into sheets, appearing fibrillar in cross-section. Moss (1972) stated that only small granules were contained by the fibrils, although 11 her published micrographs failed to support her viewpoint. The fi- brils were linked in a continuous network with large spaces con- taining starch between adjacent fibrils. As mixing proceeded, the many discontinuities in the network and the degree of spac- ing between fibrils diminished. The lipids were completely dis- tributed through the dough after 3 minutes. The droplet diameter was generally less than 10 um and most were less than 2 um. After 7 minutes of mixing, the protein fibrils were stretched farther, which resulted in thinning and increased the amount of cross-linking between adjacent fibrils (Moss, 1972). The extent of the starch-protein association also in- creased. According to Moss (1972), fibrillar protein was converted to an enveloping mantle Which flowed around the granules. After 11 minutes, gross overmixing had occurred, which caused a very extensible and very sticky dough (Moss, 1972). No fibril- lar protein network remained, and the gluten became a veil-like mantle surrounding the starch granules in the dough--a condition' associatedeith poor gas retaining properties (Moss, 1972). Moss (197“) showed the light microstructure of chemically treated dough mixed at 37 rpm for 30 minutes. She concluded that oxidized dough was not uniformly developed and that reduced dough was overdeveloped. Gluten Although proteins, starch, lipids and other minor components contribute to the unique properties of dough, it is generally 12 agreed that the gluten proteins are of fundamental importance, since they lend cohesiveness and elasticity to the mass (Bloksma, 1971). According to Bloksma (1971), the amino acid composition of the gluten proteins determines the distinctive chemical and phys- ical properties of molecular shape and types of bonding (hydrogen bonds, disulfide and electrostatic bonds, and Vanckanaals forces). Properties of Glutenin and Gliadin Gluten can be separated into two crude fractions, gliadin and glutenin, which are soluble or insoluble in 70% aqueous ethanol, respectively (Traub 22 21., 1957). According to Huebner (1968), gliadin and glutenin have molecular weights of roughly 26,000 and 250,000, respectively. Even though their amino acid contents are similar, these two protein fractions differ widely in size, shape and dimensional stability (Wall and Beckwith, 1969). When hydrated, glutenin forms a very tough, elastic and co- hesive mass (Wall and Beckwith, 1969). Glutenin consists of many large molecules with random coils offering numerous opportunities for the molecular associations promoting cohesion and elasticity (Wall and Beckwith, 1969). According to Wall and Beckwith (1969), disulfide linkages in glutenin are intermolecular. When gliadin is hydrated, it forms a viscous fluid mass. Gliadin is composed of small, uniformly compact molecules offering little surface area for contact with other molecules (Wall and Beckwith, 1969). Disulfide linkages in gliadin are primarily intramolecular (Wall and Beckwith, 1969). Mixtures 13 of gliadin and glutenin have properties intermediate between those of the separate proteins according to Wall and Beck- with (1969). Amino Acid Composition Bloksma (1971) stated that the polypeptide chains of gluten proteins have a predominantly random configuration. A high proline content (1A1) restricts highly organized heli- cal arrays and favors intermolecular over intramolecular bonding (Bloksma, 1971). Glutamine, the most abundant amino acid residue (37%). participates in hydrogen-bonding. Its terminal amide group is free to interact with other amides or with groups that only accept or donate hydrogen bonds, such as hydroxyls or carbonyls (Wall and Beckwith, 1969). Since the low bond energy of hydrogen bonds is compensated for by their abundance, hydrogen-bonding may extensively effect gluten properties (Bloksma, 1971). Krull and Inglett (1971) pointed out that very few ionizable amino acid residues (2%) are present in gluten proteins, a deficiency accounting for gluten's low solubility in aqueous medium. Likewise, amino acids with nonpolar side chains (17%), such as leucine, interact in hydrophobic bonding and contri- bute to gluten's insolubility (Krull and Inglett, 1971). Although cystine residues (2.1%) are limited in number, their intermolecular and intramolecular disulfide bonds are very important to gluten structure (Bloksma, 1971). 1“ Thiol-Disulfide Interchange Theory Goldstein (1957) first suggested that thiol-disulfide interchange reactions explain the occurrence of permanent deformations in a network with disulfide cross—links. Gold- stein (1957) explained viscous flow as a result of thiol and disulfide interchange by the following series of reactions: RlSSR2 + XSH —4> RlSH + RZSSX Brownian motion R2$SX + R3SH g> R288R3 + XSH RlSSR2 + R3SH ~> RISH + RZSSR3 Where R = a large protein molecule and the subscript designates different protein molecules. XSH is the thiol containing compounds which initiates the 8-8 exchange from RISSR2 to RZSSR3, and again becomes available for another cycle of ex- change reactions. In the relaxed condition, Brownian motion allows reactive groups to approximate each other's vicinity, but under stress the direction of motion is biased yielding measurable deformation. To demonstrate actual interchange reactions in dough, McDermott and Pace (1961) added thiolated gelatin to dough. They isolated the gluten and found hydroxyproline in the hydrolysate. This amino acid is normally found only in gelatin. According to the SH - S—S interchange theory, the pro- teins of separate flour particles become cross-linked to form a continuous protein network in dough (Bloksma, 1971). The theory suggeststhatthe elastic deformation of gluten and dough is 15 restricted by the number of disulfide bonds present, whereas, the thiol groups are required for viscous deformation. It predicts the increased elasticity or resistance of gluten to stretching with increased S-S content and the decreased vis- cosity with increased -SH content. Hence, oxidation of thiol groups, yielding disulfide bonds, should increase both viscosity and elasticity (Bloksma, 1971). The effect of oxidation was demonstrated qualitatively by adding potassium iodate to the dough. Remarkably, the modulus of elasticity increased at least 60%, while the disulfide content increased by only AS (Bloksma, 1968). Lipoprotein Model of Gluten Using X—ray scattering and electron microscopy tech— niques, Grosskreutz (1960) demonstrated that wheat protein is arranged in flat platelets approximately 70 A thick. Under hydration these platelets bond together into sheets capable of large plastic deformation. An applied stress can orient the sheets parallel to one another and to the plane of stretch. Removing flour lipids destroys the ability to bond into co- herent sheets, but it does not effect the basic platelet (Grosskreutz, 1960). Hess (195“) pictured lipids as forming layers between protein fibrils and entering strongly into the hydration reaction. Upon analyzing X-ray patterns, Traub g£_§1. (1957) indi- cated that layers of phospholipid hold protein fibers together. Within these layers, the fat molecules were roughly l6 perpendicular to the protein fibers. Evidence for this phos— pholipid structure favors the presence of well oriented bi- molecular leaflets, as found in myelin figures (Traub §£_al., 1957; Grosskreutz, 1961). However, Grosskreutz (1961) was unable to detect the biomolecular leaflets using transmission electron microscopy. Grosskreutz (1961) postulated a lipoprotein model for wheat gluten structure. It explains the continuity of the glu- ten sheet as a series of protein platelets linked to one another through an aqueous phase, probably by hydrogen bonding. Five percent of the elastic sheet, however, consists of a lipo- protein complex which can form hydrogen or salt-like bridges to the protein platelets. Because of the different relative bond strengths involved, an applied stress will initiate a "slip plane" along the phospholipid interfaces of the lipo- protein complex before reaching the rupture strength of the interprotein bonds. The Grosskreutz (1961) model explains the plasticity of gluten, which is necessary for optimum baking characteristics. Dough Development Fermentation Fermentation in breadmaking was reviewed by Magoffin and Hoseney (197“), who concluded that it had three primary functions: (1) leavening action (002 production of yeast); (2) flavor development (alcohols, acids, esters, and other 17 flavor precursors); and (3) dough development (the result of total fermentation). The manner in which fermentation affects dough develop- ment after mixing is largely unknown (Magoffin and Hoseney, 197A). Nevertheless, it has pronounced effects on the gluten complex, balancing between thin extensible film formation and the rheological properties that allow for maximum gas retention (Pyler, 1973). A combination of factors (proteolysis, fermentation by-product, and hydrogen—ion concentration) alter the behavior of the gluten and may account for the im- proving effects of fermentation (Magoffin and Hoseney, 197“). Working (1928) described two distinct changes occurring in the dough development or ripening during fermentation, both of which are required to balance tenacity and ductility in the dough. Acids produced in fermentation increase the ability of the gluten strands to absorb water; thus, they swell, reducing tensile strength. The phosphatides lubricate the gluten strands allowing for easier slippage over each other, and thus, increase ductility. The reduced pH from C0 and lactic acid production, 2 and the uptake of assimilated ammonia not only affects the hydration and swelling properties of gluten, but also influ- ences the reaction rate of enzymes, oxidation and reduction processes, and the various chemical reactions involving organic salts (Brown and Thomas, 19H5). Regardless of the mechanism of action, bulk fermentation has long been considered necessary to mellow dough, rectifying structural defects and rendering 18 it sufficiently extensible to hold gas (Moss, 1972). Re- cently, however, mechanical dough development and chemical dough development methods have largely replaced bulk fer- mentation (Tipples, 1967). Mechanical Development Swanson and Working (1926) showed that intense mechanical energy put into dough could to a large degree replace tradi- tional bulk fermentation, conserving at least two-and-one- half hours of production time. In the 1950's this concept was implemented in two successful processes, the Do—Maker process (Baker, 1954) and the Amflow process (Anon., 1958). They involve the principles of continuous dough production by mechanical development under pressure of a continuous flow of ingredients, and feature the extrusion of the developed dough directly into baking pans (Ponte, 1971). Mechanical development requires high speed mixers and intense energy to disaggregate the flour proteins and make sulfhydryl groups available for interchange reactions (John- ston and Mauseth, 1972). It also requires a controlled critical rate of work input and the addition of oxidants to set the protein network (Tipples, 1967). High speed mixers impart sufficient energy to the dough in a short time to yield a structure with the desired prop- erties of elasticity and extensibility without extensive bulk fermentation (Pyler, 1973). While developing the gluten network in dough, the extended protein molecules, aided by 19 thiol—disulfide interchange reactions, tend to retract to their original convoluted shape. High speed mixers used in mechanical dough development can overcome this obstacle. They disperse protein more rapidly than conventional mixers. The gluten network builds up faster than the protein molecules can recoil themselves (Johnston and Mauseth, 1972). To prevent deterioration of the structure after mixing, the number of free sulfhydryl groups in the dough must be reduced by adding oxidants. Oxidizing the sulfhydryl groups to disulfide bonds stabilizies the structure (Johnston and Mauseth, 1972). Chemical Dough Development Because mechanically developed dough demands a tremendous energy input, chemical means of development now have economic and practical significance (Marston, 1971). The subject of chemical dough development and the interrelationship of oxidants and reductants has been reviewed by several researchers (Tsen, 1970, 1973; Marston, 1971; Johnston and Mauseth, 1972; Ponte, 1971). A reducing agent chemically develops the dough and its action is followed by that of an oxidizing agent, which sta- bilizes the structure (Tsen, 1970). This produces a well developed and mature dough within a few minutes, which is suitable for breadmaking without conventional bulk fermenta- tion or intensive mechanical action (Tsen, 1970). The action of oxidizinggand reducing agents. Oxidizing compounds have been used by millers and bakers since the 20 early 1900's to mature flour and improve its baking qualities without extensive aging periods (Tsen, 1970). However, re- ducing agents were generally considered to be detrimental to flour performance, and have only been used since the advent of chemical dough development (Sullivan and Howe, 1936). Wheat germ has long been known to hinder the baking qual- ity of flour (Geddes, 1930). Sullivan gt g1, (1936) identified glutathione as the deleterious agent. Their farinograph data showed that glutathione reduced dough development time and stability, but these effects were reversed by using an oxidant (potassium bromate). An indirect mechanism of action was proposed by Jorgensen (1936) and by Balls and Hale (1936). They attributed the weakening effect of glutathione to the activation of proteolytic enzymes, and the effect of bromate to the inactivation of the proteolytic enzymes. However, Howe (19H6) showed that the proteolytic theory was invalid. Meanwhile, Sullivan (19u0) showed that the oxidizing and reducing agents effect sulfur linkages in dough. It has since been concluded that oxidizing and reducing agents react dir— ectly through the sulfur linkages on the flour proteins (Tsen, 1970). Frater 23 El: (1960) stated that at any given protein content, the rheological properties of the dough appear to be directly related to the number of intermolecular disulfide bonds and the rate at which they can interchange with thiol groups. They went on to relate this concept to the action of oxidizing and reducing agents on the dough. They 21 demonstrated that oxidizing agents strengthen the dough by inhibiting disulfide exchange reactions, and possibly by forming new intermolecular disulfide bonds. In contrast, they showed that reducing agents lower the number of intermolecular disulfide bonds and/or increase the rate of -SH - S-S exchange. Thus, the reactions of sulfhydryl groups and disulfide bonds in various dough improvers have been well established (Hird and Yates, 1961; Bushuk and Hlynka, 1962). Reducing agents. Reducing agents act immediately in dough to accelerate development (Tsen, 1973). Those with free sulfhydryl groups promote sulfhydryl-disulfide inter- change (Johnston and Mauseth, 1972), thereby releasing the stresses of the mixing action on the dough (Frater g§_§l., 1960). Cysteine and glutathione reduce the interdisulfide bonds of the flour protein aggregates, splitting them into small extractable units (Tsen, 1969). Cysteine, the -SH containing amino acid, when used in increasing concentrations (below 63 ppm),decreases dough development time logarithmically (Henika and Rodgers, 1965). Depending upon the flour strength, cysteine can reduce peak mix time for unfermented doughs by 30 to 65 percent (Henika and Rodgers, 1965; Finney g£.§l., 1971). It also reduces the energy level required to achieve peak dough development and reduces the critical mixing speed necessary to produce bread of high volume (Kilborn and Tipples, 1973). Cysteine used alone in breadmaking significantly decreases bread quality. Therefore, it must be used in combination with 22 an oxidant, such as bromate, to produce an acceptable loaf (Henika and Rodgers, 1965). Cysteine reduced doughs are very weak and stretch irregularly (Jelaca and Dodds, 1969). When tested on the extensigraph, resistance to stretch de- creases and extensibility increases with cysteine concentra- tion (Frater gt gl., 1960). Upon noting reduced dough strength with increases in cysteine, Frater gt gt. (1960) stated that the reductant destroyed intermolecular disulfide bonds and increased their rate of interchange. Glutathione is a tripeptide containing cysteine. It rapidly decreases dough consistency in the farinograph (Bloks- ma, 1971) and has a pronounced effect on dough character, increasing extensibility with increasing concentrations (Ville- gas gt gt., 1963). Upon using more than 0.6 umoles of glu- tathione per gram of flour, the dough becomes too fluid to be measured by the extensigraph (Tsen, 1970). The effects of glutathione, like cysteine, can be explained by increases in the ratecfi'thiol-disulfide interchange reactions (Tsen, 1970). Oxidizing agents. Once the dough has been developed by mechanical mixing and the use of chemical reducing agents, it is necessary to set the protein network by oxidation of sulfhydryl groups (Johnston and Mauseth, 1972). Oxidants stabilize the structure and improve gas retention in bread (Marek and Bushuk, 1967). Suitable oxidants can be divided into three categories: slow, intermediate, and fast acting (Johnston and Mauseth, 1972). 23 Slow acting oxidants include potassium and calcium bromate (Frater gt gt., 1960; Johnston and Mauseth, 1972). During mixing at 30°C, bromate oxidizes -SH groups more slowly than the surrounding air (Tsen and Bushuk, 1963). It reacts too slowly to sufficiently stiffen the dough (Bloksma, 1971). Its effect is delayed until subjected to high dough tempera- tures (AD-50°C), which occur during the later phase of proving or the early part of baking (Bushuk and Hylnka, 1960; Tsen, 1968; Jalaca and Dodds, 1969). Structural relaxation experi- ments show the bromate reaction is also time-dependent and has no immediate effect on the dough properties (Dempster gt a_1_., 1956). Dehydroascorbic acid is an intermediate reaction rate oxidant that is commonly used in continuous bread making. Ascorbic acid, a reducing agent, is rapidly converted to dehydroascorbic acid, an oxidant, when dough is mixed in air (Johnston and Mauseth, 1972). Apparently it has a dual function (Johnston and Mauseth, 1972), but theories as to its oxidant effect are not conclusive. Zentner (1968) showed that ascorbic acid does not reduce disulfide bonds or block -SH groups. He suggested that ascorbic acid in dough effects hydrogen bonding. Fast acting oxidants, such as potassium iodate and azo- dicarbamide (Frater gt_gl., 1960; Johnston and Mauseth, 1972), decrease extensibility and increase resistance to extension (Bushuk and Hlynka, 1962; Bloksma, 1971). Iodate acts very rapidly having an appreciable immediate effect on the dough 24 (Dempster gt gt., 1956). In a farinograph, potassium iodate alters dough properties by first increasing resistance to mixing and then later by lowering it (Frater gt gt., 1960). Frater gt gt. (1960) related this effect to the disappearance of thiol groups. In come cases, such as high dosages and long mixing, there is a reversal of the effect so that breakdown occurs and leads to increased extensibility (Bloksma, 1971). Tsen and Bushuk (1963) showed that iodate reacts stoichiometri- cally with -SH groups in dough, i.e., the iodate consumed is parallel to the -SH groups oxidized. The same workers showed that breakdown of the iodate-treated dough from prolonged mixing coincides with a reduction of 8-3 bonds. Azodicarbamide (ADA) is a rapid-acting reagent that oxidizes the sulfhydryl groups of flour proteins (Bloksma, 1971). It removes hydrogen atoms from these groups and in- corporates them into its own structure; ADA is thus converted into biurea (Pyler, 1973). ADA lessens the mixing requirements of a dough (Pyler, 1973) and increases the resistance to ex- tension (Bloksma, 1971). Bromate complements the improving action of ADA, when they are used in combination, and ADA reduces the over-all oxidant requirement of a dough (Pyler, 1973). MATERIALS AND METHODS Source and Levels of Ingredients Untreated white bread flour was obtained from the Pills- bury Company in St. Louis, Missouri. The flour contained 11.3% protein, 1A.0% moisture, and 0.A6% ash. The following oxidizing and reducing agents were utilized in this study: potassium bromate and potassium iodate (Mallin- ckrodt Chemical Works), azodicarbamide (Penwalt Corp.), L- Cysteine (Sigma Chemical Co.), and glutathione (Nutritional Biochemicals Corp.). Oxidizing agents were added at the maximum level allowed under current standards of identity (Food and Drug Administra- tion, 1976). Based on the weight of flour, the levels added were 75 ppm for potassium bromate and potassium iodate and AS ppm for azodicarbamide. The reducing agents, cysteine and glutathione, were added to the flour at 75 ppm. Preparation of Dough The farinograph was used to determine the water absorption of the flour (Table I) and to mix each sample. The procedure for a constant dough weight was followed according to the 25 26 standard farinograph methods as outlined by the American Asso- ciation of Cereal Chemists (1962). The water absorption was adjusted so that each dough was mixed to a standard peak consis— tency (500 B.U. on the farinogram). Standard peak consistency was taken to be the point at which the dough offered maximum resistance against the mixing blades of the farinograph. At this stage the dough was considered to be optimally mixed. Flour-Water Dough The flour-water dough prepared without oxidizing or re- ducing agents was mixed to one of four stages of development: (1) undermixed (2.5 minutes); (2) optimally mixed (7.0 min- utes); (3) optimally mixed and incubated for 90 minutes in a proving cabinet at “0°C; and (A) overmixed (11.0 minutes). Oxidized and Reduced Doughs The oxidizing and reducing agents were mixed dry with the flour for 9 minutes in the farinograph bowl. The mixture was then hydrated by adding the required amount of water, and was then mixed to one of three stages of development: (1) optimally mixed (the time being dependent on the particular additive as shown in Table I); (2) optimally mixed and incubated 90 minutes in a proving cabinet at "0°C; and (3) overmixed (11.0 minutes). 27 Scanning Electron Microscopy Immediately after each dough was mixed, two small pieces were removed for freeze-drying. Each sample was approximately 2 mm in diameter and 10—15 mm long. The shape of one sample was maintained so as not to disturb the surface structure of the dough. The second sample was stretched lengthwise to a point just short of breaking, with the amount of stretching being dependent on the extensibility of the dough. Both samples were frozen rapidly by placing in contact with aluminum foil-coated dry ice. They were immediately trans- ferred to previously cooled (-70°C) sample vials and stored in an ethanol and dry ice slush until lyophilization. The samples were lyophilized in a Virtis freeze dryer for at least twelve hours. The freeze-dried dough strips were divided into small segments suitable for mounting on aluminum stubs with a conduc- tive adhesive (Television Tube Koat). The specimens were coated with gold in a sputter coating apparatus and examined 3 with an ISI—Super-Mini scanning electron microscope at 10 KV. Duplicate samples of each dough were prepared and repre- sentative micrographs were taken using Polaroid-105 P/N film. RESULTS AND DISCUSSION Farinograph-Data Flour-Water Dough Although many ingredients are used in breadmaking, flour and water form the glutinous dough. In order to simplify and facilitate interpretation of the results, a flour-water system rather than a complete bread dough formulation was used in this study. The water absorption of the flour was 61.6%. The mixing time required to reach peak consistency was 7.0 minutes (peak time). The dough withstood the mixing force for 10.0 minutes (stability) before losing maximum consistency (Table I). Oxidized and Reduced Doughs The addition of oxidizing and reducing agents to the doughs altered the mixing properties. Both oxidizing and reducing agents, except for potassium bromate, increased the water absorp- tion of the flour, reduced mixing requirements and the stability of the dough (Table I). Apparently, the chemicals enhanced the disaggregation of hydrated protein particles during the mixing process. Disaggregation exposed hydrophilic protein centers so that more water was absorbed as shown by farino- graph measurements (Table I). 28 29 Table I. Dough Properties as Determined in the Farinograph Additive “8°???” P852513“ “IAIN” None 61.6 7.0 10.0 KBrO3 61.6 7.0 10.0 K103 62.5 6.5 “.O ADA(a 62.6 6.0 3.5 L-Cysteine 62.5 2.5 2.5 Glutathione 62.5 3.0 “.0 a) azodicarbamide 30 According to Tsen (1969), reducing agents expedite the disaggregation process by the scission of disulfide bonds. On the other hand, the action of oxidizing agents is more com- plex, resulting from sulfhydryl-disulfide interchange (Tsen, 1969). Once the SH groups are oxidized, the interchange mechanism in the dough becomes much slower. Since there is less disulfide interchange to release the stress, the dough becomes stiffer and more resistant to deformation. Under these conditions, and presumably with the continued shearing and tearing of mix- ing, more protein bonds are forceably cleaved so as to disag- gregate“theprotein particles (Tsen, 1969). Potassium bromate had no effect on the farinogram from dough at 30°C. This agrees with the work of Jelaca and Dodds (1969), who demonstrated that bromate lacks reactivity below “0-50°C. Thus, bromate effects the dough during the later stages of proving and in early baking. Because cysteine and bromate differ in reaction time, they are the reductant and oxidant, respectively, most commonly combined in chemical dough development (Tsen, 1969). The action of bromate is sufficiently delayed to separate its influence from the rapid action of cysteine and to permit sequential development and maturation of the dough." or the oxidants tested in this investigation, ADA had the largest effect on water absorption and the mixing charac- teristics of the dough, even though it was used at the lowest level (“5 ppm versus 75 ppm). Cysteine was a more reactive reducing agent than glutathione according to farinograph peak 31 time and stability measurements (Table I). The greater influ- ence of cysteine on the dough supports the report of Tsen (1969), who demonstrated that cysteine is more effective than glutathione in extracting gluten proteins from flour-water suspensions. ScanninggElectron Microscgpy,(SEM) Starch The chemical treatments did not effect the starch granules as observed by SEM. The starch granules were of different shapes and sizes, which confirms previous observa- tions by light microscopy (Burhans and Clapp, 19“2) and by scanning electron microscopy (Aranyi and Hawrylewicz, 1968). SEM showed that the large granules tended to be lentil-shaped while the smaller granules were generally spherical, as shown in Figure 1. Although most of the granules were smooth, some exhibited surface irregularities. It is possible that enzymatic attack initiated the shallow depressions observed on some of the granules, as shown by the arrows in Figure 1. If enzymes, such as amylases, were responsible for the depressions, their occurrance may be expected to be increased with the age of the flour, had it not been refrigerated. Frequently, nodules were attached to the surface of the large starch granules (Figure 2). The nodules were smaller than most of the small starch granules and seemed to be 32 Figure 1. SEM of starch granules from wheat flour showing the characteristic small and spherical- or large and lentil-shaped granules. Irregularities on the granule surfaces may be the site of en- zyme attack (arrow). 900 X. Figure 2. SEM of nodule attached to the surface of a starch granule. 2500 X. 33 3“ attached tc the large granules by an adhesive material. Aranyi and Hawryewicz (1968) suggested that structural formations, similar to the nodules, may be the adhering protein referred to by Hess (1955). The adhering protein that Hess (1955) described has an average thickness of 0.2 um, whereas, the nodules are greater than 1.0 pm in diameter. Due to their size, it is probable that the nodules are a variety of starch granule unrecognized in earlier studies. The starch granules appeared to be distributed randomly throughout the dough without segregating into groups or accord- ing to size (Figures 1 and 3). This agrees with the conclusions of Sandstedt gt_gl. (195“), who observed the starch granules to be located at random throughout the dough. Flour-Water Dough Figure 3 shows an optimally mixed flour—water dough, which appears as a pebbly mass of starch granules bound by a continuous gluten matrix. In some areas the gluten formed a thin film, which flowed over and between the granules to con- form with their underlying shapes (Figure “). In other areas, the gluten had a sheet-like character (Figure 5), in which the starch granules were covered by a thick gluten sheet ob- scuring their shapes. The gluten was closely associated with the starch, so there was no visible space between the gluten and starch granules. Figure 6 shows that the gluten and starch granules adhered strongly to each other, even under the severe stress imposed 35 Figure 3. SEM showing optimally mixed control flour-water dough. 800 X. Figure “. SEM of optimally mixed control flour-water dough. Note the pores and depressions in the gluten sheet. 3700 X. 36 37 Figure 5. SEM of optimally mixed flour—water dough showing sheet-like gluten. 1600 X. Figure 6. SEM of optimally-mixed flour-water dough illus- trating thin fibrils formed upon stretching. 500 X. 38 39 by stretching. The stretched gluten formed a multitude of long, thin fibrils extending over and between the large starch granules. The fibrils were oriented in the direction of stress and were continuous beyond the trapped starch granules. During the 90 minute incubation period the dough relaxed becoming less elastic and more extensible. Relative changes between the control (Figure 3) and the unstretched incubated sample (Figure 7) were not discernible by SEM. Although SEM revealed that fibril formation occurred in both the newly mixed dough (Figure 6) and the incubated dough (Figure 9), there was more abundant fibril formation in the incubated sample, probably reflecting its greater extensibility. The stretched samples of both the newly mixed (Figure 6) and incubated doughs (Figure 10) displayed small starch gran- ules, which were trapped in the fibrillar gluten or adhered to it strongly. The thinnest fibrils were separated from the greater dough mass and did not support any starch granules, as shown in Figures 6 and 9. At high magnification, the gluten surface was pitted with small round pores, which ranged in size from less than 0.2 um to about 1.5 um. The pores are clearly shown in Figures “ and 11. They are comparable in size to the lipid-rich inclu- sions in the protein matrix of wheat endosperm and dough, which were reported in earlier work with transmission electron microscopy (Seckinger and Wolf; 1967; Simmonds, 1972; Khoo gt g;., 1975). Seckinger and Wolf (1967) demonstrated that free or nonpolar lipids were rather uniformly distributed “0 Figure 7. SEM of unstretched incubated flour-water dough. 1200 X. Figure 8. SEM of incubated flour-water dough. Note the long thin fibrils reflect the extensibility of the sample. 750 X. “l “2 Figure 9. SEM of incubated flour-water dough showing thin thread-like fibrils formed upon stretching. 700 X. Figure 10. SEM showing small starch granules trapped by thin gluten fibrils in stretched incubated dough. 1500 X. “3 ““ throughout the protein matrix, whereas, bound or polar lipids were found in small osmiophilic inclusions. Simmonds (1972) showed that the lipids originated from remnants of the cytoplas- mic organelles surviving grain maturation. It seems probable that the gluten pores observed in the present investigation are formed upon freeze—drying, air-drying, or critical-point drying of the lipid-rich inclusions for SEM. As the sample is dehydrated, some of the hydrophobic forces on the lipid inclusions would be eliminated thus allowing the original lipid material to be absorbed by the dehydrated dough leaving an empty space. This process may be aided by the vacuum used in drying or in the SEM column. Another possible explanation for the pores could be that the lipid is partially 6r completely volatilized by the vacuum. This suggests that the pores are artifacts of preparation for SEM and probably represent lipid inclusions in the original dough. In addition to the pores, small depressions and bumps were observed in the gluten film (Figures “ and 12). The depres- sions may actually be due to pores which have a slightly dif- ferent appearance from a close underlying layer of gluten. The bumps occur in the gluten, especially where it coats the large starch granules. The origin of the bumps is unknown, but it may be a form of protein, such as the "adhering" protein of wheat flour as described by Hess (1955). The bumps were rarely noticed on the gluten covering the small starch granules and were absent in highly stretched gluten fibrils (Figures 10 and 12). “5 Figure 11. SEM showing small round pores pitting the sur- face of the gluten layer of an untreated flour- water dough. 9000 X. Figure 12. SEM showing bumps on the surface of the gluten covering a large starch granule. 3800 X. “6 1:7 In general, the optimally mixed flour-water dough was characterized by a continuous and strongly adhering gluten layer covering the starch granules. Upon stretching, the gluten formed thread-like fibrils which occurred even more abundantly after the dough relaxed during the incubation period. Undermixing. In flour-water doughs mixed for 2.5 minutes the gluten was not uniformly developed. For the most part, the starch granules tended to be clustered together, crowded and overlapped (Figure 13). In completely undeveloped areas of the dough, the protein and starch were not closely associated and amyloplast membrane residues capped the starch granules (Figure 1“). In some areas the gluten began to develop cohesive properties, and it flowed over and between the granules (Fig- ure 13). With gluten development, the long axis of the starch granules became oriented with the dough surface. Stretched samples of the undermixed dough demonstrated that the starch-protein association was weak. Elongated de- pressions in the gluten surface appeared to be the result of starch granule imprints, which were deformed as the dough was stretched (Figure 15). Some small granules remained in the depressions, but most of them were empty since the protein could not sufficiently bind and retain the starch granules. In addition, the undermixed dough (Figure 16) did not stretch into thin fibrils like the control (Figure 6). This may imply that the starch granules in the control acted as anchor points and facilitated the extension of the gluten into the fibrils. In contrast, the starch-protein association in the undermixed “8 Figure 13. SEM of undermixed flour-water dough showing lack of uniform development. In the center starch granules are clustered together with no apparent matrix. In other areas, the gluten formed a coherent sheet flowing over and be- tween the granules (lower right). 600 X. Figure 1“. SEM of undermixed flour-water dough showing lack of development with incomplete sheet for- mation near center of field. Note the membranous residues capping the starch granules (arrow). 1300 X. “9 50 Figure 15. SEM of stretched undermixed flour-water dough. Note the empty elongated depressions marking the gluten. Arrow points to a granule sitting in a depression in the stretched gluten. l“00X. Figure 16. SEM of stretched undermixed flour-water dough. 500 X. 51 52 dough was not fully developed, so that the granules could not be anchor points for the fibrils and were merely obstacles to the continuity of the gluten. Overmixing: Mixing the flour-water dough beyond peak consistency injured the quality of the gluten. As shown in Figure 17, the gluten sheet became unusually thin and lost the sleek appearance of the control (Figure 3). The thin sheet sagged loosely over the contours of the starch granules, and random patches across the dough surface began to disintegrate. In the disrupted areas, the gluten pulled away from the starch granules and collapsed in a spindly web, surrendering its integrity as a cohesive unit (Fig- ures 18 and 19). The breakdown of the sheet into a fibril- lar web and the loss of integrity were characteristic of overmixing. These findings are in agreement with the ob- servations of Moss (1972). Fewer pores were observed in the gluten of the over- mixed dough (Figure 19) as compared to the control (Figure “). Larger holes were predominant in the overmixed dough 'and were observed at the initiation points of breakdown as observed in Figure 18. It is probable that excessive mix- ing enlarged and distorted the pores, resulting in the large irregular holes associated with breakdown of gluten. Fibrils were absent in the stretched gluten of overmixed dough in areas where breakdown was apparent. As shown in Figure 20, thick strips of gluten were formed with crosslinking by fibrillar webs. The starch-protein association 53 Figure 17. SEM of overmixed flour-water dough. Note the large holes disrupting the continuity of the gluten film. 900 X. Figure 18. SEM of overmixed flour-water dough showing the weakened gluten. 'Note the gluten has contracted away from the starch granules and collapsed into a spindly web. 1“00 X. 55 Figure 19. SEM showing the gluten of the overmixed flour- water dough. Note the wrinkled surface and large holes. 3600 X. Figure 20. SEM of stretched overmixed flour-water dough. Note the thick strips of gluten are crosslinked by fibrillar webs of disintegrated protein (arrow). 900 X. 56 57 appeared to be weakened with overmixing, since the gluten had contracted away from many of the starch granules and fibril formation was absent. Oxidizing Agents Potassium Bromate. Potassium bromate had no measur- able effect on the water absorption or mixing characteris- tics of the dough. The bromate treated dough behaved like the control in the farinograph.. With SEM, the bromate treated dough also appeared to be identical to the optimally mixed (Figure 3) and overmixed (Figure 17) controls. However, stretched samples of the incubated dough containing bromate (Figure 23) formed thick fibers compared to the thinner fibrils observed in the newly mixed sample (Figure 22). The gluten stretched into wide strips or thick fibers (Figure 23) but resisted extension into thin thread-like fibrils. Changes in the bromated dough at elevated temperatures support rheological data obtained by Dempster gt gt. (1956) and Jelaca and Dodds (1969). They found that doughs con- taining potassium bromate behaved normally in relaxation tests from 25 to 30°C, but were less relaxed than controls above 35°C. The stiffer dough at “0°C can be explained by the increased reactivity of bromate, which presumably restricts the SH reactivity of the gluten (Jelaca and Dodds, 1969). Potassium Iodate. Optimally mixed unstretched dough treated with potassium iodate (Figure 2“) showed no appreciable 58 Figure 21. SEM of optimally mixed potassium iodate treated dough. 700 X. Figure 22. SEM showing thin fibrils formed upon stretching optimally mixed bromate treated dough. 900 X. 59 60 Figure 23. SEM showing thick fibers formed upon stretching incubated bromate treated dough. 900 X. Figure 2“. SEM showing optimally mixed potassium iodate treated dough. 1600 X. 61 62 difference from the control (Figure 5). Upon stretching, the iodate treated dough formed fibrils (Figure 25), but their formation over the dough surface was not as widespread as for the unoxidized control (Figure 6). Wherever the fibrils formed (Figure 25), they were long and thread-like, indicating a more extensible dough. The unexpected extensibility of the dough suggests that the iodate did not fulfill its intended purpose, probably as a result of the high level added. Support for the adverse effects of high levels of iodate is found in a report by Bloksma (1971), who concluded that a high level of the oxidiz- ing agent reverses improvement of the dough. It is suggested that the reversal due to high levels of the oxidant may be the result of the sudden reduction of SH groups, which restricts thiol-disulfide interchange in the gluten. Consequently, high mixing speeds cause a high degree of stress and lead to forceful breakage of the disulfide crosslinks. After incubating, the iodate treated dough was unchanged from the newly mixed samples (Figures 2“ and 25) as observed by SEM. Continued mixing of the iodate treated dough led to break- down of the gluten, which reflected a decrease in dough stability (Table I). Observations suggest that excessive manipulation spread the gluten progressively thinner over the granular mass until the prominant contours were obscured (Figure 26). This is in contrast to the optimally mixed dough in which the gluten was draped almost completely around the granules, distinctly revealing their shapes (Figure 2“). Overmixing the iodated dough severely damaged the integrity of the gluten as shown in Figure 26. It appeared very 63 Figure 25. SEM depicting long thin fibrils of stretched optimally mixed dough containing potassium iodate. 500 X. Figure 26. SEM showing the thin discontinuous gluten of overmixed iodate treated dough. Note the obscured shapes of the starch granules. l“00 X. 6“ 65 thin and failed to maintain its continuity over the dough surface. Many granules were left coated with a web-like film where the gluten had apparently collapsed and shriveled. The contracted protein left large gaps around the granules and irregular holes in the sheet. The general effect was similar to that seen in the unoxidized, overmixed dough (Figure 17). Azodicarbamide (ADA). The effect of ADA was not ob- served by SEM in the unstretched samples of optimally mixed or incubated doughs (Figure 27). Upon stretching of the op- timally mixed sample, however, it became apparent that the char- acter of the gluten had changed. Rather than forming fibrils as in the control (Figure 6), the ADA oxidized dough (Figure 28) divided into thick strips of gluten that were oriented in the direction of stress. The tendency to stretch into flat strips instead of rounded fibrils was also noted in the incubated dough, which is shown in Figure 29. With ADA in the dough, deterioration of the gluten struc- ture was not characteristic of the overmixed samples. In Figure 30, a stretched portion of the dough shows that elongated holes seemed to initiate the division of the gluten into flat strips. In contrast to the iodate treated dough (Figure 26), the gluten of the overmixed ADA treated sample (Figure 30) draped snugly over the starch granules revealing their prominant contours. The apparent stability of the gluten in the ADA treated dough after overmixing does not reflect the lack of stability against mixing which was measured in the farinograph (Table I). 66 Figure 27. SEM showing optimally mixed ADA treated dough. 500 X. Figure 28. SEM of stretched ADA treated dough. Note that wide flat strips of gluten have replaced the characteristic rounded fibrils. 550 X. 67 68 Figure 29. SEM of incubated ADA treated dough showing resistance to fibril formation upon stretching. Note that the holes in the gluten elongate into slits dividing the gluten into flat strips. 2700 X. Figure 30. SEM of overmixed ADA treated dough. Note the gohesiveness of the gluten upon stretching. 50 X. 69 70 It is possible that an accumulation of biurea (the reaction product of ADA) during extended mixing reverses the stiffening effect of the oxidant on the dough. The biurea may have reacted to relieve intermolecular stress in the gluten and allowed the gluten to extend by thinning of the sheet rather than by forced cleavage of 8-8 bonds as discussed earlier for the iodate treated sample. Reducing Agents Cysteine. Immediately after mixing, cysteine treated dough was unusually extensible and even more so after incu- hating or overmixing. The reduced dough stretched irregularly and was not elastic. The unstretched samples of the cysteine treated dough (Figure 31) appeared similar to the control (Figure 3). On the other hand, stretched samples of the dough were characterized by severely damaged areas as shown in Figure 32. The gluten sheet was completely disrupted around the granules although the granules were still capped with a smooth protein film. In general, the cysteine treated dough did not form fibrils upon stretching but merely disintegrated. Incubating the cysteine treated dough gave results similar to those of the optimally mixed sample. The fragile sheet was battered with irregular holes and did not stretch into fibrillar extensions (Figure 33). The overmixed cysteine treated dough revealed scattered patches of breakdown as shown in Figure 3“. Some areas ap- peared coherent while others were completely disrupted. The' 71 Figure 31. SEM of optimally mixed cysteine treated dough. 800 X. Figure 32. SEM of stretched optimally mixed cysteine treated dough. Note the severe damage to the gluten sheet. 500 X. 72 73 Figure 33. SEM of stretched incubated cysteine treated dough. Note large holes disrupting the continuity of the gluten. 750 X. Figure 3“. SEM of overmixed cysteine treated dough. Note the loss of integrity surrounding the starch granules but not on their surfaces. 850 X. 7“ 75 gluten maintained an adhesion to the surface of the starch granules but lost its integrity in the areas surrounding the granules (Figure 35). The gluten did not shrivel and disin- tegrate over the surface of the granules as noted in the control (Figure 17). The inability of the cysteine treated dough to form fibrils upon stretching may be caused by decreasing the glutenin fraction of gluten through the scission of the disulfide bonds. Tsen (1969) proposed that cysteine reduced some of the glutenin fraction to a molecular size corresponding to that of the glia- din fraction, which is thought to have a spherical configura- tion. On the other hand, glutenin is said to be a random linear polymer of polypeptide chains (Ewart, 1972). Ewart (1972) reasoned that if highly assymmetric (linear) molecules were dominant in gluten, the orientation caused by stretching would improve intermolecular adhesion and tensile strength in the direction of stress. In contrast, if the molecule tended to be symmetrical, the intermolecular forces per unit volume would not increase appreciably on stretching and the gluten would lack visco-elasticity. In the linear model proposed by Ewart (1972), elasticity arises from the tendency of extended or unfolded polypeptide chains to return to their contracted conformations of lowest free energy. Therefore, the increased extensibility and the fragility of the cysteine reduced gluten may be explained as being due to the reduction of the linear molecules and an increase in the globular conformation. Glutathione. The newly mixed dough containing glutathione 76 Figure 35. SEM of overmixed cysteine treated dough. Observe the expanded pores in gluten creating large holes associated with breakdown during excessive manipulation of the dough. 6000 X. Figure 36. SEM of optimally mixed glutathione treated dough. Note the gluten form a thin film over the starch granules. 1200 X. 77 78 was quite slack. After incubating or after overmixing, the dough lost its elasticity and was very extensible, much like the cysteine treated dough. When the optimally mixed glu- tathione treated dough was examined by SEM, the gluten appeared as a very thin film over the starch granules (Figure 36). The gluten layer conformed readily to the shape of the starch gran- ules and appeared to strongly adhere to the starch. Upon stretching, the newly mixed dough (Figure 37) tended to form fibrils, but their occurrence was not widespread in comparison to the control (Figure 6). After incubation, the stretched sample of the glutathione treated dough (Figure 38) exhibited greater fibril formation across the surface of the dough than the same newly mixed sample (Figure 37). Although the dough had relaxed, the starch- protein association remained strong and cohesive (Figure 39). Unstretched samples of the overmixed glutathione treated dough did not reflect breakdown (Figure “0), which was supported by farinograph readings (Table I). The gluten sheet appeared flat and partially obscured the starch granules. In the stretched overmixed glutathione treated samples, severely disrupted areas were apparent where large irregular holes destroyed the continuity of the sheet (Figure “1). The gluten disintegrated around the starch granules, but maintained a thin film attached to their tops. These results suggest that glutathione con- tributed to destruction of gluten strength during overmixing, while the starch-protein association was at least partially maintained. 79 Figure 37. SEM of stretched optimally mixed glutathione treated dough showing formation of fibrils upon stretching (unidentified debris in center). “50 X. Figure 38. SEM of incubated glutathione treated dough showing thin fibrils formed upon stretching. 500 X. 80 81 Figure 39. SEM of stretched incubated glutathione treated dough. 1300 X. Figure “0. SEM of overmixed glutathione treated dough. Note that the granules lie flat on the dough surface. 1100 X. 83 Figure “1. SEM of stretched overmixed glutathione treated dough showing the disruption of the gluten and resulting irregular holes. 600 X. 8“ 85 In contrast to cysteine, glutathione was not detrimental to gluten continuity until the dough was overmixed. The slower reactivity of glutathione may be due to the lower SH content per gram than for cysteine. In addition, the effect of gluta- thione was more evenly distributed throughout the dough than cysteine, suggesting that a high concentration of cysteine resulted in more pronounced breakdown in localized areas. The discontinuity of the glutathione treated dough after overmixing may be explained by a decrease in the glutenin frac- tion of the gluten as discussed earlier herein for the cysteine treated dough. Tsen (1969) also demonstrated an increase in the gliadin-like fraction of glutathione treated gluten, but the increase was much less than that caused by cysteine. SUMMARY Flour-water doughs containing oxidizing and reducing agents were examined by SEM to study their effects on the surface characteristics of dough. Farinograph readings showed that oxidants and reductants decreased the mixing requirement and increased the water absorption of the flour. In order to define the characteristics of gluten develop- ment and breakdown in a flour-water dough, stretched and unstretched samples were examined as dough development pro- gressed. A continuous gluten sheet began to form early during mixing, and developed a strong starch-protein association at Optimal mixing. SEM suggested that the starch-protein asso- ciation supported the formation of thread-like gluten fibrils in the direction of stress. Fibril formation increased upon incubation of the dough. Overmixing caused the gluten sheet to disintegrate, lose its affinity for starch granules and fail to form gluten fibrils upon stretching. SEM revealed the presence of small round pores in the gluten sheet, which was common to all of the doughs. It was hypothesized that the pores resulted from lipid inclusions in gluten and may be the sites for the initiation of larger holes formed upon breakdown of the dough. The gluten surface 86 87 was also marked by small depressions and bumps. The surface structure of the doughs containing oxidizing or reducing agents was generally like the unstretched, optimally mixed and incubated controls. Changes in the gluten strength of the chemically treated doughs were more clearly observed upon stretching. The effect of potassium bromate on the dough was not apparent until after it was incubated. A stiffening of the gluten was indicated by the formation of thick fibers upon stretching in contrast to the thin fibrils of the control. SEM showed that the optimally mixed and incubated iodate treated samples formed thin fibrils upon stretching. The unexpected sign of extensibility may indicate that a reversal of the oxidant effect occurred due to the high level of iodate. Overmixing of the iodate treated dough caused breakdown of the gluten, which was similar to that of the overmixed control. SEM revealed that all samples of the ADA treated doughs resisted fibril formation upon stretching. The gluten divided into wide flat strips instead of rounded fibrils, emphasizing its tendency to maintain a Sheet-like structure. Unlike the other doughs examined in this investigation, the ADA treated dough showed no appreciable loss of integrity upon overmixing. The addition of reducing agents to the dough tended to destroy its ability to form fibrils upon stretching. How- ever, optimally mixed and incubated glutathione treated doughs showed extensive fibril formation in the stretched samples. 88 Large irregular holes destroyed the gluten continuity of the stretched cysteine treated doughs, but only in the overmixed samples of the glutathione treated dough. BIBLIOGRAPHY BIBLIOGRAPHY Am. Assoc. Cereal Chem. 1962. Approved Methods, 7th ed. The Am. Assoc. Cereal Chem., St. Paul, Minn. 55-21. Anonymous. 1958. The newest of the continuous dough making systems. Baker's Dig. 32(6):“9. Aranyi, C. and Hawrylewicz, E. J. 1968. A note on scanning filectron microscopy of flours and doughs. Cereal Chem. 5:500. Baker, J. C. 19“l. The structure of the gas cell in bread dough. Cereal Chem. 18:3“. Baker, J. C. 195“. Continuous processing of bread. Proc. Am. Soc. Bakery Engin. p. 65. Balls, A. K. and Hale, W. S. 1936. Further studies on the activity of proteinase in flour. Cereal Chem. 13:656. Bloksma, A. H. 1971. Rheology and chemistry of dough. In: Wheat Chemistry and Technologz. Ed. Y. Pomeranz. Am. Assoc. Cereal Chem., Inc., St. Paul, Minn. p. 523. Brown, E. B. and Thomas, J. M. 19“5. Baker's Dig. 18:1. As cited by C. D. Magoffin and R. C. Hoseney, 197“. Burhans, M. E. and Clapp, J. l9“2. A microscopic study of bread dough. Cereal Chem. 19:196. Bushuk, W. and Hlynka, I. 1960. The bromate reaction in dough. II. Inhibition and activation studies. Cereal Chem. 37:3“3. Bushuk, W. and Hlynka, I. 1962. The effect of iodate and N-ethylamaleimide on extensigraph properties of dough. Cereal Chem. 39:189. Butterworth, S. W. and Colbeck, W. J. 1938. Some photomicro- graphic studies of dough and bread structure. Cereal Chem. 15:“75. 89 90 Dempster, C. J., Cunningham, C. K., Fischer, M. H., Hlynka, I. and Anderson, J. A. 1956. Comparative study of the improving action of bromate and iodate by baking data, rheological measurements, and chemical analysis. Cereal Chem. 33:221. Ewart, J. A. D. 1972. Recent research and dough visco- elasticity. Baker's Dig. “6(“):22. Finney, K. F., Tsen, C. C. and Shogren, M. D. 1971. Cys- teine's effect on mixing time, water absorption 'oxida- tion requirement and loaf volume of Red River 68. Cereal Chem. “8:5“0. Food and Drug Administration. 1976. Bakery products revision of standards of identity. Federal Register “1:62“2. Frater, R., Hird, F. J. R., Moss, H. J. and Yates, J. R. 1960. A role for thiol and disulfide groups in deter- mining the rheological properties of dough made from wheaten flour. Nature 186:“51. Geddes, W. F. 1936. Chemical and physico-chemical changes in wheat and wheat products induced by elevated tempera- tures. III. The influence of germ constituents on baking quality and their relation to improvement in flour induced by heat and chemical improvements. Can. J. Research 2:195. Goldstein, S. 1957. Sulfhydryl- and disulfidgruppen der klebereiweisse und ihre beziehung zur backfahigkeit der brotmehle. Mitt. Gebiete Lebensm. Hyg. (Bern) “8:87. Grosskreutz, J. C. 1960. The physical structure of wheat protein. Biochim. Biophys. Acta 38:“00. Grosskreutz, J. C. 1961. A lipoprotein model of wheat gluten structure. Cereal Chem. 38:336. Henika, R. G. and Rodgers, N. E. 1965. Reactions of cysteine, bromate, and whey in a rapid breadmaking process. Cereal Chem. “2:397. Hess, K. 195“. Protein, kleber und lipoide in weizenkorn und mehl. Kolloid Z. 136:8“. Hird, F. J. R. and Yates, J. R. 1961. The oxidation of cys- teine, glutathione and thioglycollate by iodate, bromate, persulphate and air. J. Sci. Food Agr. 12:89. Howe, M. 19“6. Further Studies on the mechanism of the action of oxidation and reduction on flour. Cereal Chem. 23:8“. 91 Huebner, F. R. 1968. Comparative studies on glutenins from different classes of wheat. Am. Assoc. Cereal Chem. — Am. 011 Chem. Soc. meeting, Washington, D. 0., March - April. As cited by S. J. Wall and A. C. Beckwith, 1969. Jelaca, S. and Dodds, N. J. H. 1969. Studies of some improver effefits at high dough temperatures. J. Sci. Food Agr. 20:5 0. Johnston, W. R. and Mauseth, R. E. 1972. The interrelations of oxidants and reductants in dough development. Baker's Dig. “6(2):20. J¢rgensen, H. 1936. On the existence of powerful but latent proteolytic enzymes in wheat flour. Cereal Chem.13: 3“6. Katz, J. R. and VanTeutem, E. 1917. Onder zoekingen naar het oudbakken worden van brood en de middelen om dit te voorkomen. (Published by Dutch government) As cited by W. Butterworth and W. J. Colbeck, 1938. Khoo, U., Christianson, D. D. and Inglett, G. E. 1975. Scan- ning and transmission electron microscopy of dough and bread. Baker's Dig. “9(“):2“. Kilborn, R. H. and Tipples, K. 1973. Factors affecting mech- anical dough development. IV. Effect of cysteine. Cereal Chem.50:70. Krull, L. H. and Inglett, G. E. 1971. Industrial uses of gluten. Cereal Sci. Today 16:232. Locken, L., Loska, S. and Shuey, W. 1960. The Farinograph Handbook. Am. Assoc. Cereal Chemists, St. Paul, Minn. Magoffin, C. D. and Hoseney R. C. 197“. A review of fermen- tation. Baker's Dig. “8(6):22. Marek, C. J. and Bushuk, W. 1967. Study of gas production and retention in doughs with a modified Brabender oven-rise recorder. Cereal Chem. ““:300. Marston, R. E. 1971. Chemical activation of dough development under slow mixing conditions. Baker's Dig. “5(6):16. Matz, S. A. 1972. Bakin Technolo and En ineerin . 2nd ed. AVI Publishing Co., nc., Westport, Conn. p. 158. ‘_. McDermott, E. E. and Pace, J. 1961. Modification of the Eggpggties of flour protein by thiolated gelatin. Nature 92 Moss, R. 1972. A study of the microstructure of bread doughs. CSIRO Food Res. Quarterly 32:50. Moss, R. 197“. Dough microstructure as affected by the addition of cysteine, bromate and ascorbic acid. Cereal Sci. Today 19:557. Pomeranz, Y. and Shellenberger, J. A. 1971.- Bread Science and Technology. AVI Publishing Co., Inc., Westport, Conn. p. ““. Ponte, J. G., Jr. 1971. Bread. In: Wheat Chemistry and Technology. Ed. Y. Pomeranz. Am. Assoc.ICereal Chem., Inc., St. Paul, Minn. p. 675. Pratt, D. B., Jr. 1971. Criteria of flour quality. In: Wheat Chemistry and Technology. Ed. Y. Pomeranz. Am. Assoc. Cereal Chem., Inc., St. Paul, Minn. p. 201. Pyler, E. J. 1972. Baking Science and Technology Vol. I. p. 582. Siebel Publishing Co., Chicago, Ill. Sandstedt, R. M., Schaumburg, and Fleming, J. 195“. The microstructure of bread and dough. Cereal Chem. 31:“3. Scheffer, W. 1916. Mikroskopisch Dunnschnitte durch Gebacke Z. ges. Getreide-Muhlen. u Backerview 8:6. Seckinger, H. L. and Wolf, M. J. 1967. Lipid distribution in the protein matrix of wheat endosperm as observed by electron microscopy. Cereal Chem. ““:669. Simmonds, D. H. 1972. The ultrastructure in the protein ma- trix of wheat endosperm as observed by electron micro- scopy. Cereal Chem. ““:669. Sullivan, B. and Howe, M. 1936. The isolation of glutathione from wheat germ. J. Am. Chem. Soc. 59:27“2. Sullivan, B., Howe, M. and Schmalz, F. D. 1936. On the pres- ence of glutathione in wheat germ. Cereal Chem. 13:665. Swanson, C. O. 1925. A theory of colloid behavior in dough. Cereal Chem. 2:265. Swanson, C. O. and Working, E. B. 1926. Mechanical modification of dough to make it possible to bake bread with only the fermentation in the pan. Cereal Chem. 3:65. Tipples, K. H. 1967. Recent advances in baking technology. Baker's Dig. “1(3):18. 93 Tsen, C. C. 1968. Oxidation of sulfhydryl groups of flour by bromate under various conditions and during the bread- making process. Cereal Chem. “5:531. Tsen, C. C. 1969. Effects of oxidizing and reducing agents on changes of flour proteins during dough mixing. Cereal Chem. “6:“35. Tsen,ufi.uc. 81970. Chemical dough development. Baker's Dig. ( ):2 . Tsen, C. C. 1973. Chemical dough development. Baker's Dig. “7(5):““. Tsen, C. C. and Bushuk, W. 1963. Changes in sulfhydryl and disulfide contents of doughs durin mixing under various conditions. Cereal Chem. 0:399. Traub, W., Hutchinson, J. B. and Daniels, D. G. H. 1957. X-ray6studies of the wheat protein complex. Nature 179:7 9. Verschaffelt, E. and van Teutem, F. E. 1915. Die anderung der mikroskopiskhen. Strukter des brostes beim alback— enwerden. Z. Physiol. Chem. 95:130. Villegas, E., Pomeranz, Y. and Shellenberger, J. A. 1963. The effects of thiolated gelatins and glutathione on Eheglggical properties of wheat doughs. Cereal Chem. 0: 9 . Wall, 3. J. and Beckwith, A. C. 1969. Relationship between structure and rheological properties of gluten pro- teins. Cereal Sci. Today 1“:16. Working, E. B. 1928. The action of phosphatides in bread dough. Cereal Chem. 5:223. Zentner, H. 1968. Effect of ascorbic acid on wheat gluten. J. Sci. Food Agr. 19:“6“. ICHIGRN STRTE UNIV. LIBRRRIES 31293006516458