ORIGINAL PAPER

Texturized Pinto Bean Protein Fortification in Straight DoughBread Formulation

Courtney W. Simons & Emily Hunt-Schmidt &Senay Simsek & Clifford Hall & Atanu Biswas

# Springer Science+Business Media New York 2014

Abstract Pinto beans were milled and then air-classified toobtain a raw high protein fraction (RHPF) followed by extru-sion to texturize the protein fraction. The texturized highprotein fraction (THPF) was then milled to obtain flour, andcombined with wheat flour at 5, 10, and 15 % levels to makebread. The air-classification process produced flour with highconcentration of lipids and phytic acid in the protein-richfraction. However, extrusion significantly reduced hexaneextractable lipid and phytic acid. However, the reductionobserved may simply indicate a reduction in recovery due tobind with other components. Total protein and lysine contentsin composite flours increased significantly as THPF levelsincreased in composite flour. Bread made with 5 % THPFhad 48 % more lysine than the 100 % wheat flour (control).The THPF helped to maintain dough strength by reducingmixing tolerance index (MTI), maintaining dough stabilityand increasing departure time on Farinograph. Bread loafvolume was significantly reduced above 5 % THPF addition.THPF increased water absorption causing an increase in breadweights by up to 6 %. Overall, loaf quality deteriorated at 10and 15 % THPF levels while bread with 5 % THPF was notsignificantly different from the control. These results supportthe addition of 5 % THPF as a means to enhance lysinecontent of white pan bread.

Keywords Texturized bean protein . Lysine . Bread . Dough,amino acid

AbbreviationsAAA Amino acid analyzerRHPF Raw high protein fractionTHPF Texturized high protein fractionWAI Water absorption indexWSI Water solubility indexMTI Mixing tolerance index

Introduction

Pulses and cereal grains complement each other nutritionallysince pulses are high in lysine and low in sulfur-containingamino acids such as methionine, cysteine and tryptophanwhile cereals are low in lysine but have abundant amountsof sulfur-containing amino acids [1]. Pulse protein fractionshave been obtained using both wet fractionation and air-classification methods [2–5]. Wet fractionation can producehigher protein extraction levels. However, the process is moretime consuming, complex, expensive and may increase therisk of contamination due to use of water. In contrast, airclassification produces a lower protein yield but is faster andless expensive. Finer, less dense materials are collected as ahigh protein fraction while coarser and more dense materialsare collected as a high starch fraction [5].

Both whole pulse flour and protein fractions have beenadded to bread to improve nutritional quality. Despite nutri-tional improvements, overall physical characteristics of breadare typically compromised due to gluten dilution [2, 5–9].This problem can be addressed by texturization of pulseproteins using extrusion. Texturization imparts structuralchanges in the protein matrix; resulting in strong protein-

C. W. SimonsWright State Unversity, Lake Campus, 234 Dwyer Hall, 7600 LakeCampus Drive, Celina, OH 45822, USA

E. Hunt-Schmidt : S. Simsek :C. Hall (*)Department of Plant Science, North Dakota State University, NDSUDept. 7670, P.O. Box 6050, Fargo, ND 58108-6050, USAe-mail: clifford.hall@ndsu.edu

A. BiswasAgricultural Research Services, USDA, Peoria, IL, USA

Plant Foods Hum NutrDOI 10.1007/s11130-014-0421-1

protein association [10, 11]. These associations deliver a fi-brous texture to pulse proteins. The structural changes couldproduce unique functionality in bread that is not typicallyobserved when pulses are not texturized. So far, no work hasbeen reported on the effects of texturized high-protein pulseflours in bread. Therefore, the objectives of this research wereto enhance protein quality of white pan bread using texturizedpinto bean flour and determine the effects of this proteinsource on bread quality.

Materials and Methods

Fractionation and Extrusion Processing

Dried pinto beans were purchased from Kelley Bean Co.(Hatton ND) and pin-milled and air-classified at Particle Con-trol Inc. (Albertville, MN). After milling, the flour consisted of88 % 44 μm particles. The high protein stream was collectedand stored at room temperature until ready for extrusion.Period between air-classification and extrusion was approxi-mately 3 weeks.

Extrusion was completed on a co-rotating twin screw ex-truder (Wenger TX-52, Sabetha, KS) as reported previously[12]. The onlymodifications included an extruder screw speedof 250 rpm and a square shaped die opening (13 mm×13 mm). The moisture of the raw flour was raised from 12to 40 % by addition of water and steam during cooking.Specific mechanical energy, i.e., a measure of work done onthe feed during extrusion, of 0.06 kWh/kg was calculated asreported previously [12].

Extrudates were dried in an impingement oven (Lincoln,Middlesex, Great Britain) at 150 °C for 3 min and then placedon racks overnight to cool. Final moisture of extrudates was7.4 %. Samples were collected in plastic bags and stored in arefrigerator at 4 °C until ready for further analysis and breadmaking. For characterization and bread making, driedextrudates were milled using a centrifugal mill with a0.5 mm sieve (Retsch, Haan, Germany). Three 1 kg batchesof composite flours were made consisting of white flour fromwheat (Dakota Miller’s Choice, ND State Mill, Grand ForksND) and 5, 10 and 15 % texturized high protein fraction(THPF), respectively. Bread made from 100 % wheat flourwas used as the control.

Chemical Properties

Protein was determined in triplicate using a nitrogen combus-tion analyzer (Leco FP-528, St. Joseph, MI) according toAACC Method 46–30.01 [13] using a nitrogen conversionfactor of 6.25. Total starch was determined in triplicate usingMegazyme total starch analysis procedure (Megazyme Inter-national, Bray, Ireland). Amino acid contents were determined

on composite flours [14–16] using an automatic amino acidanalyzer (AAA) (HITACHI L-8800, Japan) withtransgenomic ion-exchange column. An amino acid standardsolution for protein hydrolysate (Sigma-Aldrich A-9906, St.Louis Missouri) was used to determine response factors and tocalibrate the AAA for all amino acids before analysis. Tryp-tophan, cysteine, and methionine were not reported sincethese amino acids are destroyed during acid hydrolysis.

Total hexane extractable lipids was determined in triplicatebased on AOCS [17] methods Af 3–53, Am 2–93, and Aa 4–38. Ash and phytic acid were determined in triplicate followingAACCMethod 08–01.01 [13] and the phytate/total phosphorusanalysis procedure (Megazyme International, Bray, Ireland),respectively.Water absorption index (WAI) andwater solubilityindex (WSI) were determined as previously described [12, 18].

Dough, Baking and Bread Evaluation

Prior to bread baking, rheological dough properties of thecomposite flours were determined using Farinograph method54–21.02 and Extensograph method 54–10.01 of AACC In-ternational [13]. Farinograph measurements were determinedin triplicate and extensograph measurements in quadruplet.

Breads were baked on two separate days. On each of thesedays, one control and three replicates of bread at each treatmentlevel (5, 10 and 15%) were made. Baking experiments follow-ed AACC International Method 10–09.01 [13] and wateraddition adjusted according to the Farinograph water absorp-tion values for each flour, i.e., treatment. Mixing time wasdetermined as the time taken for full dough development basedon visual observation. Bread volume was determined based onAACC International Method 10–05.01 [13]. Specific volumewas calculated by dividing the volume of individual loaves bytheir weight. Crumb color was determined using a Minoltacolorimeter to determine L, a, and b values on theHunter scale.Color of the crust was based on a subjective color evaluationchart ranging from one to ten; where ten was the darkest color.

Statistical Analysis

Data was analyzed using statistical analysis systems software(SAS Institute, Cary, NC), based on a completely randomizeddesign. An analysis of variance (ANOVA) and means com-parison following LSD procedure was used to establish sig-nificant differences (P≤0.05).

Results and Discussion

Chemical Properties

Total protein decreased significantly from 44.2% in the RHPFto 39.2 % in the THPF (Table 1). WSI was significantly lower

Plant Foods Hum Nutr

for the THPF (30.9 %) compared to RHPF (39.6 %),supporting observations that extrusion processing reducesprotein solubility [19]. However, utilization of THPF to pro-duce composite flours resulted in a significant increase inprotein and all essential amino acids (Table 2). Total proteinincreased from 13.5 % in the control wheat flour to 17.4 % inflours with 15 % THPF. Lysine content increased by 48, 89.8and 139.3 % in flours fortified with 5, 10 and 15 % THPF,respectively.

Total starch contents in RHPF and THPF was 1.3 and2.1 %, respectively (Table 1), and they were not significantlydifferent. Czarnecki et al. [3] also reported low starch content

in protein fraction of air-classified pinto beans (2.5 %). Al-though not significantly different, the lower starch content inRHPF could be underestimated as the high level of phytic acidin the protein fraction may interfere with starch digestion byα-amylase in the starch assay. Yoon et al. [20] reported thatphytic acid could retard α-amylase activity by binding calci-um. Therefore, the slightly higher starch in the THPF could berelated to the lower phytic acid.

Ash content in RHPF and THPF (Table 1) was high.Although the ash contents were significantly different, thedifference is small and may not be of practical difference.The small difference in ash content supports that the mineralcontent (i.e., ash) did not change substantially during theextrusion process. Czarnecki et al. [3] also reported high ashcontent in air-classified pinto bean flour (7.1 %). High ashcontent was likely due to the presence of high phytic acidlevels, which results in mineral binding [21]. Phytic acidcontent in RHPF was 21.8 mg/g and decreased significantlyduring extrusion (i.e. 19.4 mg/g in THPF). However, thereduction may be simply a recovery issue and therefore maynot be of practical significance.

Total hexane extractable lipids in RHPF were significantlyhigher compared to THPF (Table 1). This data supports thatlipids form complexes with other components during extru-sion and thus makes it more difficult to extract the lipids [22,23]. Reduction in total hexane extractable lipids may helplower the rate of rancidity caused by lipid oxidation sincethe lipids would be tied up in the protein network. Starch-lipid complexes have been extensively reported and thus thehigh protein of the THPF likely interacted with the lipid toform a complex as only a small amount of starch was presentin the THPF.

Dough, Baking and Bread Evaluation

Mixing time of dough (Table 3) significantly increased withaddition of THPF. This corresponded to an increase in arrivaltime at higher THPF additions (Table 4). Therefore, as THPFincreased, dough took a longer time to reach a standardviscosity of 500 Brabender Units (BU) on the Farinograph.Loaf weight increased significantly with the addition of THPF,i.e. from 131.0 g in the control to 139.2 g in 15 % THPFtreatment (Table 3). Although WAI was significantly reduced,from 3.6 in RHPF to 3.1 in THPF, increased levels of THPFfortification cause increased water absorption percentage incomposite flours (Table 4). This resulted in higher loaf weightwith THPF addition (Table 3).

Peak time (Table 4) was not significantly different betweencontrol and bread with 5 % THPF. However, the peak timewas longer by approximately 2 min in dough containing 10and 15 % THPF. Lorimer et al. [2] also reported increasedpeak times when wheat flour was replaced with high-proteinlegume flours at 5 and 10 %. Silaula et al. [9] observed an

Table 1 Characteristics of protein fraction before (RHPF) and after(THPF) extrusion

Fraction type1

Property RHPF THPF

Moisture (% db) 11.6b 7.4a

Total protein (% db) 44.2b 39.2a

Total starch (% db) 1.3a 2.1a

Total lipids (% db) 3.2b 1.2a

Ash (% db) 6.5a 6.7b

Phytic acid (mg/g) 21.8b 19.4a

WAI 3.6b 3.1a

WSI (%) 39.6b 30.9a

RHPF raw high protein fraction

THPF texturized high protein fraction1Values with different superscripts in the same row indicates significant(P≤0.05) difference between raw and texturized fractions

Table 2 Effect of THPF addition on total protein and essential aminoacids (EAA) in composite flours

Treatment1,2

Control 5 % 10 % 15 %

Total protein (%) 13.5a 15.0b 16.1c 17.4d

EAA(mg/g)

Threonine 3.89a 4.68b 5.33c 6.17d

Valine 5.91a 6.80b 7.46c 8.12d

Isoleucine 5.09a 5.90b 6.51c 7.21d

Leucine 9.98a 11.44b 12.55c 13.91d

Phenylalanine 7.24a 8.29b 9.14c 10.06d

Histidine 3.14a 3.68a 4.33b 4.58c

Lysine 3.09a 4.58b 5.87c 7.40d

Arginine 5.31a 6.41b 7.38c 8.45d

THPF texturized high protein fraction1Values with different superscripts in the same row indicates significant(P≤0.05) difference between raw and texturized fractions2 Treatment represent the control (no THPF) and flours containing 5, 10,and 15 % THPF

Plant Foods Hum Nutr

increase in peak time required for dough development withaddition of dry-roasted air-classified pinto and navy beans towheat flour at 10, 15 and 20 % levels.

The MTI, i.e., how much the dough will soften after acertain degree of mixing, of the control and 5 % THPFaddition were not significantly different. However, MTI de-creased at 10 and 15 % levels (Table 4). Stability of the doughwas not significantly different between the 5 and 10 % THPFlevels, but was significantly different between the 5 and 15 %THPF levels. Results of stability, MTI and departure timeindicated a general resistance to over-mixing. The resistance

to over-mixing also was supported by Extensograph data(Table 5). The general trend was that with increasing THPFaddition to the dough, a reduction in extensibility occurredwhile an increase in resistance to extension was observed. Thedough extensibility and resistance values with 15 % THPFwere, with a few exceptions, significantly different from theother dough samples. In contrast, a reduction in stability andincreased MTI with addition of 10, 20 and 30 % non-texturized pulse flour to wheat flour previously was reported[8]. In our study, texturized protein was used and may accountfor the resistance to over-mixing due to structural propertyimparted by the texturized protein.

Loaf volume and specific volume bread with 5 % THPFwas not significantly different from the control, but decreasedat 10 and 15 % THPF levels. These results were consistentwith results observed for the addition of grasspea wholemealto traditional white bread formulations [24]. Fenn et al. [4]reported that legume protein addition to wheat flour reducedspecific volume with significant reduction observed at 5 %addition. Fleming and Sosulski [6, 7] reported that supple-mented proteins disrupt the protein-starch complex observedin wheat flour bread. In addition, they observed small pores inthe cell walls of the supplemented bread, which they sug-gested allowed gases to escape from the structure duringbaking. Therefore, the volume loss observed with increasingTHPF levels might be related to thin cell wall structure created

Table 3 Effect of THPF addition to wheat flour on mixing time ofdough, and loaf weight, volume, density and color of bread

Treatment1,2

Control 5 % 10 % 15 %

Mixing time (s) 210a 240b 255c 300d

Loaf weight (g) 131.0a 133.5b 136.7c 139.2d

Loaf volume (cc) 932c 904c 809b 710a

Specific volume (cc/g) 7.1c 6.8c 5.9b 5.1a

Crust color 9a 9a 10b 10b

Crumb color4

L* 79.6b 76.3b 73.2ab 70.9a

a* −0.3a 1.4b 2.8c 3.9d

b* 15.1a 15.3ab 15.8bc 16.3c

THPF texturized high protein fraction1Values with different superscripts in the same row indicates significant(P≤0.05) difference between raw and texturized fractions2 Treatment represent the control (no THPF) and flours containing 5, 10,and 15 % THPF3 L*a*b* = color values based on Hunter scale where L = brightness, a =greenness (−) and redness (+), and b = blueness (−) and yellowness (+)

Table 4 Farinograph properties of dough without (control) or with addedTHPF

Treatment1,2

Dough property Control 5 % 10 % 15 %

Water absorption 67.2a 70.1b 74.2c 76.1d

Arrival time (min) 2.0a 3.1a 8.1b 8.2b

Peak time (min) 10.1a 10.5a 11.8b 11.8b

Stability (min) 16.8ab 17.8b 16.3ab 13.8a

MTI (BU) 40b 36.7b 30a 28.3a

Departure (min) 18a 20.8b 24.4c 21.9b

THPF texturized high protein fraction1Values with different superscripts in the same row indicates significant(P≤0.05) difference between raw and texturized fractions2 Treatment represent the control (no THPF) and flours containing 5, 10,and 15 % THPF

Table 5 Extensograph properties of dough without (control) or withadded THPF

Treatment1,2

Indices Proofingtime (min)

Control 5 % 10 % 15 %

Energy (cm2) 45 137.0b 130.0ab 111.8ab 103.3a

90 186.0c 163.8b 146.0ba 129.0a

135 176.3c 140.8ab 151.5ab 121.3a

Extensibility (mm) 45 160.5b 151.5b 145.8b 123.8a

90 156.8b 149.3b 141.3b 115.8a

135 148.8b 124.0b 135.8b 108.5a

Resistance toextension (BU)

45 423.0a 469.8ab 430.8a 515.3b

90 599.5a 594.8a 590.5a 723.0b

135 644.0a 720.3a 661.0a 762.0a

Ratio number3 45 4.2a 3.1a 3.9a 4.8b

90 6.1a 5.8a 5.6a 7.5a

135 6.3a 7.9b 6.6b 8.1b

THPF texturized high protein fraction1Values with different superscripts in the same row indicates significant(P≤0.05) difference between raw and texturized fractions2 Treatment represent the control (no THPF) and flours containing 5, 10,and 15 % THPF3Ratio number = Resistance/Extensibility

Plant Foods Hum Nutr

by the addition of the THPF and reduced gas retention. Fur-thermore, differences in particles size between THPF flourand wheat flour may have contributed to observed dough andbread characteristics. Other researchers have reported the ef-fect of particle size on dough and bread characteristics[25–27].

Bread loaves with 5 % THPF were similar to the controlexcept for coarser break and shred (Fig. 1). Break and shredrefers to the shredded or comb-like pattern observed on oneside of the bread due to loaf expansion and stretching of glutenfibers. A good break and shred should be uniform and smooth[28]. Coarseness in break and shred at or above 10 % THPFlevel was more accentuated compared to control. This sug-gests that THPF imparts rigidity to the dough system, reduc-ing ability of gluten fibers to expand smoothly and uniformly.Crust L and b values of 5 % THPF addition level were notsignificantly different from control (Table 3). However, allother color values were significantly different from the con-trol. Crumb color gradually lost brightness as the amount of

THPF increased, and had higher red and yellow intensities(Table 3), which resulted in increased browning in the bread.Fenn et al. [4] also reported reduced brightness (L*) as pulsesubstitution increased from 2 to 8 %.

Conclusion

The inclusion of texturized THPF between 5 and 10 % pro-duced the best dough and bread characteristics of the threelevels tested. However, the 5 % THPF product was closest tothe control. At the 5 % THPF level, lysine concentration wasincreased by 48 % without causing any significant negativeeffects on quality parameters. THPF can therefore be used asan ingredient in bread to enhance protein quality. However,levels of 15 % are not recommended due to the negativeimpact on dough and bread characteristics even though thelysine increased by 139 %.

A

B

CFig. 1 Comparison of control bread (far left) and bread with a 5% THPF, b 10% THPF, and c 15% THPF. Triplicate bread runs containing with THPFare to the right of the control bread

Plant Foods Hum Nutr

Acknowledgments We express our thanks to Northarvest BeanGrowers Association for research funding, Kristin Whitney and DeLaneOlsen for providing technical assistance during dough analysis and breadbaking and Rilie Morgan from the Northern Crops Institute for technicalassistance during extrusion.

Conflict of Interest The authors claim no conflict of interest. Thisarticle does not contain any studies with human or animal subjects.

References

1. Duranti M (2006) Grain legume proteins and nutraceutical properties.Fitoterapia 77:67–82

2. Lorimer NL, ZabikME, Harte JB, StachiwNC, UebersaxMA (1991)Effect of navy bean protein flour and navy bean globulin(s) oncomposite flour rheology, chemical bonding, and microstructure.Cereal Chem 68:213–220

3. Czarnecki Z, Khan K, Gujska E (1993) Extrusion of pinto bean highprotein fraction pretreated with papain and cellulase enzymes. J FoodSci 58:395–398

4. FennD, LukowOM,HumphreysG, Fields PG, Boye JI (2010)Wheat-legume composite flour quality. Int J Food Prop 13:381–393

5. Schutyser MAI, Goot AJ (2011) The potential of dry fractionationprocesses for sustainable plant protein production. Trends Food SciTechnol 22:154–164

6. Fleming SE, Sosulski FW (1977) Breadmaking properties of fourconcentrated plant proteins. Cereal Chem 54:1124–1140

7. Fleming SE, Sosulski FW (1978) Microscopic evaluation of breadfortified with concentrated plant proteins. Cereal Chem 55:373–382

8. Deshpande S, Salunkhe D, Sathe S, Rangnekar P (1983) Functionalproperties of wheat-bean composite flours. J Food Sci 48:1659–1662

9. Silaula SM, Lorimer NL, Zabik ME, Uebersax MA (1989)Rheological and sensory characteristics of bread flour and wholewheat flour doughs and breads containing dry-roasted air-classifiedpinto and navy bean high-protein fractions. Cereal Chem 66:486–490

10. KeShun L, Fu-Hung H (2007) Protein-protein interactions in highmoisture extruded meat analogs and heat-induced soy protein gels. JAm Oil Chem Soc 84:741–748

11. KeShun L, Fu-Hung H (2008) Protein-protein interactions duringhigh-moisture extrusion for fibrous meat analogues and comparisonof protein solubility methods using different solvent systems. J AgricFood Chem 56:2681–2687

12. Simons C, Hall C, Tulbek M (2012) Effects of extruder screw speedson physical properties and in vitro starch hydrolysis of precookedpinto, navy, red, and black bean extrudates. Cereal Chem 89(3):176–181

13. AACC International (2011) Approved methods of the AACC, 11thedn. St Paul, MN

14. Moore SD, Spackman H, Stein WH (1958) Chromatography ofamino acids on sulfonate polystyrene resins; and improved system.Anal Chem 30:1185–1190

15. Ozols J (1990) Amino acid analysis. In: Deutscher MP (ed) Methodsin enzymology. Academic, San Diego, pp 587–601

16. Cooper C, Packer N, Williams K (2000) Methods in molecularbiology: amino acid analysis protocols. Humana Press, New Jersey

17. AOCS (1998) Official methods and recommended practices of theAOCS. American Oil Chemists Society Champaign, IL

18. Anderson RA (1982) Water absorption and solubility andamylograph characteristics of roll-cooked small grain products.Cereal Chem 59:265–269

19. Gujska E, Khan K (1991) High temperature extrusion effects onprotein solubility and distribution in navy and pinto beans. J FoodSci 56:1013–1016

20. Yoon JH, Jenkins DJA, Thompson LU (1983) The effect of phyticacid on in vitro rate of starch digestibility and blood glucose response.Am J Clin Nutr 38:835–842

21. JenabM, Thompson LU (1998) The influence of phytic acid in wheatbran on early biomarkers of colon carcinogenesis. Carcinogenesis 19:1087–1092

22. Izzo MT, Ho CT (1989) Protein-lipid interaction during single-screwextrusion of zein and corn oil. Cereal Chem 66:47–51

23. Bhatnagar S, Hanna MA (1994) Extrusion processing conditions foramylose-lipid complexing. Cereal Chem 71:587–593

24. Kasprzak M, Rzedzicki Z (2012) Application of grasspea wholemealin the technology of white bread production. Pol J Food Nutr Sci 62:207–213

25. Cai L, Choi I, Hyun J-N, Jeong Y-K, Baik B-K (2014) Influence ofbran particle size on bread-baking quality of whole grain wheat flourand starch retrogradation. Cereal Chem 91:65–71

26. Zhang D, Moore W (1997) Effect of wheat bran particle size ondough rheological properties. J Sci Food Agric 74:490–496

27. Zhang D, Moore W (1999) Wheat bran particle size effects onbread baking performance and quality. J Sci Food Agric 79:805–809

28. Swanson R (2004) Bakery: yeast-leavened breads. In: Smith SJ, HuiYH (eds) Food processing principles and applications. Wiley-Blackwell, Ames, IA, pp 183–202

Plant Foods Hum Nutr