Microstructures of Bread Dough and the Effects of Shortening on Frozen Dough

Shigeo AIBARA,1;y Noriko OGAWA,2 and Masaaki HIROSE1

1Division of Applied Life Science, Graduate School of Agriculture, Kyoto University, Gokasho Uji,

Kyoto 611-0011, Japan2Faculty of Home Economics, Gifu Women’s University, 80 Taromaru, Gifu 501-2529, Japan

Received October 7, 2004; Accepted November 26, 2004

Three types of straight doughs different in combina-

tion of yeast and shortenings (RLS20, FTS20, and

FTS80) were prepared, and the structure of the frozen

doughs was examined under a microscope after staining

protein or lipid droplets. Even after 2 months of frozen

storage, distinct changes were not found in the gluten

network of FTS80, although significant damages in the

dough structures of FTS20 and RLS20 appeared after

only one month of frozen storage. These results suggest

that the gluten networks loosen and decrease in the

water retention ability, and it may be concluded that the

lipid is removed from the gluten protein due to the

decrease in water in the continuous protein phase. The

resulting product from the damage to the gluten matrix

gave rise to fusion of lipid droplets and an increase in

their size. Because of the difference in fatty acid

composition, the lipids of shortening S80 are presumed

to interact more strongly with gluten proteins and to

keep the gluten matrix from damage in comparison with

the lipids of shortening S20.

Key words: frozen dough; shortening; lipid droplet; loaf

volume; microstructures of bread dough

Lowering of yeast fermentation and gas retention ofbread dough have been pointed out as the two factorsaffecting loaf volume of frozen dough bread.1) Even ifone of these two abilities decreases, a sufficient loafvolume of the bread is fails to be secured. Particularlythe ability of gas retention is associated with breaddough structures (gluten network); in other words, theexpansibility of bread dough is depend on the relevantbalance of the network structure of gluten proteins andlipids as a plasticizer. Proofing time of the regular yeastdough prolonged gradually depending on the period offrozen storage. However, using freeze-tolerant yeastimproved retardation of proofing time and maintainedalmost the same level of proofing time as one weekfrozen storage-dough even after three or four months offrozen storage.2) On the other hand, the frozen doughbread is not able to recover the loaf volume completely

just by using frozen-tolerant yeast. These facts clearlyindicated that the dough structures rather than yeastfermentability closely participated in the loaf volume ofthe frozen dough bread.Gluten protein and starch granules are badly damaged

by long term frozen storage. Varriano-Marston et al.3)

have reported that water plays a role in maintainingbread dough structures, and Tipples4) has reported thatdamaged starch granules which have separated act like asponge in the presence of water, and possibly drawwater away from the gluten network with more damage.The loss of ability of gluten proteins to retain waterresults in the separation of starch granules and bringsabout the deterioration of dough structures. From theviewpoint of the microstructure of bread dough, starchgranules are embedded firmly in the intact glutennetwork, but once the gluten network has been damaged,starch granules exhibit a reticular pattern5,6) and separatefrom the gluten network.Lipids derived from added shortening are classified

into two groups from an aspect of the microstructure ofbread dough: one is the lipids taken into the dispersephase as lipid droplets during mixing, and the other isthe lipids interacting with gluten proteins,7) as indicatedin Fig. 1. The added shortening has a role as a plasticizerand closely relates to the expansion of the bread doughduring the second fermentation and baking processes.8)

Many investigators have reported abundant data on theeffects of shortenings on the bread-making of frozen andnon-frozen doughs.9–17) We noted in the previous paperthat the fatty acid composition of the shortening isimportant for the expansibility of the frozen doughbread, and the size of lipid droplets in the disperse phaseof the bread dough grew large when thawing frozendough.2) However, the details of deterioration of doughstructures are not yet clear, and it is ambiguous how theshortening relates closely to the frozen dough expansion,or if some other factors are involved.In this paper, we observed microscopically the gluten

matrix and lipid droplets of bread dough after long-termfrozen storage for elucidation of the relationship

y To whom correspondence should be addressed. Fax: +81-774-33-3004; E-mail: aibara@kais.kyoto-u.ac.jp

Abbreviations: RLS, regular yeast and shortening; FTS, freeze-tolerant yeast and shortening

Biosci. Biotechnol. Biochem., 69 (2), 397–402, 2005

between the microstructures of the bread dough and adecrease in loaf volume of frozen dough bread.

Materials and Methods

Flour. The wheat flour used in his study was acommercial hard wheat flour (‘‘Superking’’ from NisshinFlour Milling, Tokyo, Japan).

Yeast. Two kinds of compressed, raw yeast, regular(RL) and freeze-tolerant (FT) yeast (Oriental Yeast,Tokyo, Japan) were bought and consumed in 2 weeks,and they were stored in a refrigerator at 4 �C before use.

Dough formulation. On a flour weight basis, wateraddition was 61%, compressed yeast was at 5%, sugarwas added at 5%, salt at 2%, and plastic shortening at5%. Water addition was a little less than the optimumamount due to preparation of the frozen straight dough.Two types of plastic shortening different in fatty acid

composition were used, and they were named S20 andS80 according to degree of saturation of total fattyacids.2) S20, containing about 75% of unsaturated fattyacid, oleic acid, was a soft type of shortening, but S80was a hard type and contained considerable amounts(over 60%) of lauric and myristic acids. S80 wasextremely hard and was pre-heated to 25 �C for 10minand smashed to a paste before use.

Straight dough method. Straight dough was preparedusing a mixing machine (Kanto mixer, HMS-30, KantoMixer Industry) according to the procedures of theprevious study.2) Three types of straight doughs differentin the combination of yeast and shortenings (RLS20:regular yeast and shortening S20; FTS20: freeze-tolerantyeast and shortening S20; FTS80: freeze-tolerant yeastand shortening S80) were prepared.

Freeze-and-thaw treatment. After divided the straightdough into 320 g portions, the pieces were sheet-molded(to a diameter of approximately 18 cm and 1.5 cm thick),and subjected to freeze treatment. The doughs, whichhad been formed into a disk slab shape, were frozen to�10 �C on aluminum vats in an upright superfreezer(Super Freezer MDF-U281: Sanyo Electric, Japan) at�85 �C for 40min. Then the frozen dough was put into asealed polyethylene bag and stored in a freezer (MedicalFreezer MDF-U332, Sanyo) at �20 �C.

After a given period of frozen storage, each fourpieces of the frozen dough was subjected to a two-stepthawing in the sealed polyethylene bags. After the doughwas thawed and re-rounded, the average temperature ofthe thawed dough before proofing was lower than that ofthe non-frozen dough before proofing.

The total period of the freeze and thaw treatment wasrepresented by 1M for 1 month of storage, 2M for 2months of storage and so on, and the number of times offreeze-and-thaw treatment during storage was added tothe period of the storage: 1M1, 2M1, 1M4, 2M4, and soon. 1M1 or 2M1 means that the dough was first thawedat the end of the frozen storage. In the case of 1M4, thefreeze and thaw treatment was carried out every week, in2M4 the treatment was applied every 2 weeks, and every2 weeks of the last two months for 3 month frozenstorage.

Measurement of dough temperature. Measurement ofdough temperature was carried out with a chromel-alumel thermocouple equipped with a digital multimeter(Model 3457A, Hewlett-Packard). A sensor was insertedthrough the side to the center of the dough. Temper-atures of the doughs in each step are indicated inTable 1.

Measurement of loaf volume. Measurement of loafvolume was done after cooling the bread to room

Fig. 1. Dough Composition.

Classification of dough components according to Bloksma (1990).

398 S. AIBARA et al.

temperature for 120min. The rapeseed method wasused.

Light microscopy. Small blocks were cut from thebread dough after bench resting and frozen in liquidnitrogen, and sectioned to 6–8 mm thick with a cryostatand then fixed on a slide glasses at room temperaturewith a gentle ventilation. Lipid in the bread dough wasstained with Sudan IV and protein by the PAS method.

Analysis of size distribution of lipid droplets. The sizedistribution of lipid droplets was analyzed using theprogram WinROOF. Lipid droplets were distinguishedby color, and the minimum size was set to 0.370 mm.

Results

Distribution and size of lipid droplets in bread doughLipid droplets, which look like small orange-colored

particles under the optical microscope, as shown inFig. 2, were distributed across almost the whole area inthe gluten matrix. The area ratio of the lipid dropletsoccupied 3.7–8.7% (Table 2). Lipid particles grew large,to 10 to 20 mm, and the number of them increased in theregular yeast dough (RLS20) after 1 month of frozenstorage (Fig. 2F), and the loaf volume of the breaddecreased a little (Table 3). This suggests that thestructure of the gluten network was altered by thefreeze-and-thaw treatment and that the fusion of lipidparticles in the frozen dough propagated by coming intofrequent contact with them. Using freeze-tolerant yeast,the lipid droplets remained relatively small (Figs. 2Band 2E) and the loaf volume maintained the level of thefresh, non-frozen dough, but there was a tendency toincrease in the frozen dough FTS20 (Fig. 2I) after 4cycles of freeze-and-thaw treatment. The frozen doughof RLS20 treated with 4 cycles of freeze-and-thawprocess did not have enough volume to bake the breadduring the second fermentation (Fig. 2G and 2L,Table 3). It is suggested that the yeast could no longershow the fermentation ability after 4 cycles of thefreeze-and-thaw treatment since the dough structure did

not suffer from such profound damages. Nevertheless,the ratio of lipid droplets area clearly increased(Table 2). In the case of using shortening S80 withfreeze-tolerant yeast (FTS80), the size of the lipiddroplets was not influenced even if repeating the freeze-and-thaw treatment 4 times (Fig. 2C), and almost nochanges in their distribution were found as compared tothe fresh, non-frozen dough.In contrast, after 2 months of frozen storage, the area

ratio of lipid droplets in the frozen doughs (RLS20 andFTS20) significantly increased, to 8% or more (Fig. 2Kand 2M, Table 2), and the loaf volume decreased, eventhough the lipid droplets of the frozen dough (FTS80)remained small and the loaf volume had the normalvalue without repeating the freeze-and-thaw treatment(Fig. 2D). This indicated the protecting effects of

Table 1. Changes in Temperature during the Process of Freezing

Dough

Flour- and water-temperatures should be kept at 20 and 8.5 �C

respectively, in order to adjust the dough temperature after full mixing

to 23 �C.

ProcessTemperature

(�C)

After preliminary mixing for 5min 12:0� 0:6

After full mixing for 10min 22:8� 0:8

Frozen �20:0

Thawed in refrigerator 4:1� 1:1Thawed in electric

fermentation cabinet 13:3� 0:8

After bench resting 16:5� 0:5After second fermentation 27:8� 0:4

Table 2. Statistics on the Size and Number of Lipid Droplets in

Frozen Dough

Mark Number Area AverageSt. dev.

Max.

Name in the of ratio of diameter(mm)

diameter

figure particles (%) (mm) (mm)

FTS80 A 1992 3.7 1.457 1.061 7.693

FTS201M1 E 2788 4.8 1.368 1.080 7.701

FTS201M4 I 3948 7.3 1.400 1.123 14.171

FTS801M1 B 4225 5.1 1.223 0.930 7.981

FTS801M4 C 2953 6.2 1.245 0.964 7.912

RLS201M1 F 3434 5.7 1.296 1.103 21.012

RLS201M4 G 3815 8.0 1.438 1.262 13.221

FTS202M1 J 5241 5.7 0.998 0.958 11.443

FTS202M4 M 9340 8.3 0.953 0.799 9.370

FTS802M1 D 5683 5.7 1.145 0.676 7.374

FTS802M4 H 8267 6.7 0.899 0.777 9.979

RLS202M1 K 8570 8.7 0.961 0.920 10.041

RLS202M4 L 2996 8.5 1.545 1.601 14.267

FTS803M1 N 5869 6.3 1.093 0.827 10.183

Average 4937 6.5 1.215 1.006 10.881

Table 3. Specific Volume of Frozen Dough Bread with Various

Shortenings

Dough weight of one loaf was 320 g. The specific volume is

expressed as the loaf volume divided by the dough weight. �Times is

the number of freeze-and-thaw processes. The dough was treated at the

end of the frozen storage, but in the case of plural times, the dough was

treated every week for 1 month, every two weeks for 2 months, and

every two weeks of the last 2 months for 3 months frozen storage. ‘‘—’’

means no experiment, and ‘‘nd’’ not possible to bake due to too long a

second fermentation.

Frozen Period RLS20 FTS20 FTS80

Week Times� ml/g ml/g ml/g

0 0 5:273� 0:023 5:317� 0:051 5:267� 0:019

1 1 — 5:331� 0:031 —

2 1 5:286� 0:060 5:297� 0:092 —

3 1 — 5:328� 0:078 —

4 1 5:193� 0:036 5:297� 0:045 5:254� 0:096

4 4 nd 4:581� 0:061 5:163� 0:022

8 1 5:052� 0:077 5:164� 0:096 5:250� 0:1008 4 nd 4:501� 0:053 5:032� 0:035

12 1 4:819� 0:095 5:047� 0:076 5:123� 0:057

12 4 nd 4:478� 0:126 4:931� 0:043

16 1 4:758� 0:061 4:969� 0:047 5:081� 0:079

Microstructures of Frozen Bread Dough 399

shortening S80 in comparison with shortening S20.After 2 months of frozen storage with 4 cycles of freeze-and-thaw treatment, the lipid droplets of frozen doughFTS80 became large (Fig. 2H). After 3 months of frozen

storage, the size of the lipid droplets resulted in almostthe same level as the other frozen doughs (Fig. 2G and2J), as shown in Fig. 2N.

Fig. 2. Changes in Lipid Droplets in Frozen Dough Structures.

A, FTS80-Non; B, FTS80-1M1; C, FTS80-1M4; D, FTS80-2M1; E, FTS20-1M1; F, RLS20-1M1; G, RLS20-1M4; H, FTS80-2M4; I, FTS20-

1M4; J, FTS20-2M1; K, RLS20-2M1; L, RLS20-2M4; M, FTS20-2M4; N, FTS80-3M1.

Fig. 3. Changes in Gluten Matrix in Frozen Dough Structures.

A, FTS80-Non; B, FTS80-1M1; C, FTS80-1M4; D, FTS80-2M1; E, FTS20-1M1; F, RLS20-1M1; G, RLS20-1M4; H, FTS80-2M4;

I, FTS20-1M4; J, FTS20-2M1; K, RLS20-2M1; L, RLS20-2M4; M, FTS20-2M4; N, FTS80-3M1.

400 S. AIBARA et al.

Gluten matrix in the bread doughGluten proteins in the dough were stained to a purple

red color by the PAS method, as shown in Fig. 3. In thefresh, non-frozen dough, in which the gluten networkextended well and stained to a deep purple red, starchgranules lay embedded in the gluten matrix of thecontinuous protein phase (Table 1 and Fig. 3A). Thestructures of the gluten matrix did not change signifi-cantly after one month of frozen storage (Fig. 3B, 3E,and 3F). Therefore, the decrease in the loaf volume offrozen dough RLS20 is considered to result from adecrease in the fermentability of yeast. An indication ofslight damage to the dough structures might haveoccurred in the size of the lipid droplets (Fig. 2F).Repeating the freeze-and-thaw treatment, the damage tothe dough structures was slight in the frozen dough ofFTS20, as shown in Fig. 3I. However, the damageclearly appeared in the regular yeast dough (Fig. 3G),and the color of the dough became thin and open seamswere noticeable.

After two months of frozen storage, starch granules ofthe frozen doughs (RLS20 and FTS20) became separatefrom the gluten matrix (Fig. 3J and 3K, respectively).Further starch granules were dissembled and the closeconnection of the dough structures disappeared byrepeating the freeze-and-thaw treatment (Fig. 3L). Incontrast, frozen dough FTS80 maintained a thick glutenmatrix and kept a deep color (Fig. 3B, 3C, and 3D) andno indication of separation was recognized. Repeatingthe freeze-and-thaw treatment, however, starch granulesof the frozen dough of FTS80 became separate after twomonths of frozen storage (Fig. 3H), and those of frozendough FTS20 were almost completely destroyed(Fig. 3M). It is clear that the difference in the degreeof damage to the dough structures was caused by thedifference in the type of shortening. Thus, we consideredthat the effect of the shortening was attributable to thefatty acid composition of shortenings S20 and S80.2) Thereason baking was possible irrespective of the damage tothe doughs after more than two months of frozen storage(Fig. 3J, 3K, 3M, and 3N) is the freeze-tolerant yeast. Inthe case of regular yeast, repeating the freeze-and-thawtreatment did take very long to obtain enough volume ofthe second fermentation for baking, and dough at onemonth (Fig. 3G) as well as two months of frozen storage(Fig. 3L) was not capable being baked.

Discussion

It was after 2 months of frozen storage of doughsRLS20 (Fig. 3K) and FTS20 (Fig. 3J) that significantdamage to the gluten network appeared and the loafvolume of the frozen dough bread decreased. On theother hand, the size of the lipid droplets of RLS20(Fig. 2F) was found to grow large after 1 month offrozen storage, and in the case of repeating the freeze-and-thaw treatment, enlargement of lipid dropletsoccurred also in FTS20 (Fig. 2I) even after 1 month of

frozen storage, and RLS20 (Fig. 2G and 3G) could notbe baked due to too long a second fermentation.However, no difference in the length of the secondfermentation was found between one month- and twomonth-frozen storage doughs of FTS20 and FTS80.These facts suggest that the condition of the glutennetwork was closely related to the oven spring of thefrozen dough as well as the freeze-tolerant yeast.Furthermore, the chemical property of the lipids inter-acting with gluten proteins as a plasticizer played a keyrole in protecting the gluten network from disintegra-tion, taking into account that the effects of changes inthe particle size of lipid droplets varied with the type ofshortening. As mentioned with reference to the glutennetwork, that is, the interaction of gluten proteins withthe lipids derived from the shortening on mixing, glutenprotein is of high hydrophobicity and poor in solubilitybecause of the high content of proline and glutamine inthe amino acid composition. Thus, the interaction withthe lipids is presumed to occur partly by sticking thefatty acid moiety of triacylglycerol to gluten proteinsand partly by adhering to the surface of gluten proteins.If so, the middle chain saturated fatty acids are capableof forming a more stable interaction with gluten proteinsthan the long chain unsaturated fatty acid does. There-fore, the chemical structure of the fatty acid moiety ofthe lipids as the plasticizer is important to avoid theirseparation from the gluten proteins before disintegrationof the dough structures. Further, we emphasize that theshortening possessing middle chain saturated fatty acidshad good effects on expansibility of the frozen doughbread.Furthermore, the lack of water in the gluten matrix

may be a trigger, taking into consideration that both therepetition of freeze-and-thaw treatment and the increasein the size of the lipid droplets were controlling factorsinvolved in disruption of the gluten network. Water inthe continuous protein phase of the frozen dough repeatschanges in the states, ice and water, through the freeze-and-thaw treatment, and then water is removed fromgluten matrix every thawing time, and migrates to thedamaged starch granules in the disperse phase. The lackof water in the gluten matrix makes it easy to remove thelipids interacting with the gluten proteins due to leadingthe hydrophobic environment in the gluten matrix, andlipid droplets re-arrange and fuse since the gluten matrixgives rise to a condition of less water content. Then thestarch granules embedded in the gluten network arereleased from the gluten matrix, and the dough struc-tures results in the fatal damage. Thus the lipid dropletsgrow large before the damage to the gluten networkextends significantly. This is a possible reason why thesize of the lipid droplets increases.Hydrophobicity in the gluten matrix increases due to a

decrease in water with prolonging of the duration offrozen storage or repeating the freeze-and-thaw cycle,and it causes a fusion of the lipid droplets in the dough.Further, depletion of triacylglycerol from the gluten

Microstructures of Frozen Bread Dough 401

protein by itself results in destruction of the glutenmatrix. In conclusion, shortening S80 was able to leadthe dough structures to a stable gluten network under theconditions mentioned above by making firmer interac-tion of the lipids with gluten protein than shortening S20did, and by retaining water in the gluten matrix, sincethe former frozen dough FTS80 was rich in the middlechain saturated fatty acids.

Acknowledgment

We would like to thank Nisshin Flour Milling Co.,Ltd., for the donation of bread flour ‘‘Superking’’,Oriental Yeast Co., Ltd., for the donation of yeast, andFuji Oil Co., Ltd., for shortening S80. We would alsolike to thank Mr. Seiji Fukusawa of the Mitani Corpo-ration for the analysis of the size distribution of lipiddroplets.

References

1) Rasanen, J., Harkonen, H., and Autio, K., Freeze-thawstability of prefermented frozen lean wheat doughs:effect of flour quality and fermentation time. CerealChem., 72, 637–642 (1995).

2) Aibara, S., Nishimura, K., and Esaki, K., Effects ofshortening on the loaf volume of frozen dough bread.Food Sci. Biotechnol., 10, 521–528 (2001).

3) Varriano-Marston, E., Hsu, K. H., and Mahdi, J.,Rheological and structural changes in frozen dough.Bakers’ Dig., 54, 32 (1980).

4) Tipples, K. H., The relation of starch damage to thebaking performance of flour. Bakers’ Dig., 43, 28(1969).

5) Berglund, P. T., Shelton, D. R., and Freeman, T. P.,Comparison of two sample preparaion procedures forlow-temperature scanning electron microscopy forfrozen bread dough. Cereal Chem., 67, 139–140 (1990).

6) Berglund, P. T., Shelton, D. R., and Freeman, T. P.,

Frozen bread dough ultrastructure as affected by dura-tion of frozen storage and freeze-thaw cycles. CerealChem., 68, 105–107 (1991).

7) Bloksma, A. H., Dough structure, dough rheology, andbaking quality. Cereal Foods World, 35, 237–244(1990).

8) Fu, J., Mulvaney, S. J., and Cohen, C., Effect of addedfat on the rheological properties of wheat flour doughs.Cereal Chem., 74, 304–311 (1997).

9) Kline, L., and Sugihara, T. F., Factors affecting thestability of frozen bread doughs. I. Prepared by thestraight dough method. Bakers’ Dig., 42, 44 (1968).

10) Bell, B. M., and Fisher, N., The binding of modelshortenings during mixing of mechanically developedbread doughs from fresh and stored flours. J. Am. OilChem. Soc., 54, 479–483 (1977).

11) Bell, B. M., Daniels, D. G. H., and Fisher, N., Theeffects of pure saturated and unsaturated fatty acids onbread making and on lipid binding, using Chorleywoodbread process doughs containing a ‘Model Fat’. J. Sci.Food Agric., 30, 1123–1130 (1979).

12) Bell, B. M., Daniels, D. G. H., and Fisher, N., Vacuumexpansion of mechanically developed doughs at prooftemperature. Effect of shortening. Cereal Chem., 58,182–186 (1981).

13) Krog, N., Theoretical aspects of surfactants in relation totheir use in breadmaking. Cereal Chem., 58, 158–164(1981).

14) Addo, A., and Pomeranz, Y., Lipid binding and fattyacid distribution in flour, dough, and baked and steamedbread. Cereal Chem., 68, 570–572 (1991).

15) Hosomi, K., Nishio, K., and Matsumoto, H., Studies onfrozen dough baking. I. Effects of egg yolk and sugarester. Cereal Chem., 69, 89–92 (1992).

16) Inoue, Y., Sapirstein, H. D., and Bushuk, W., Studies onfrozen doughs. IV. Effect of shortening systems onbaking and rheological properties. Cereal Chem., 72,221–226 (1995).

17) Silva, R. F., Uses of alpha-crystalline emulsifiers in thesweet goods industry. Cereal Foods World, 45, 405–411(2000).

402 S. AIBARA et al.