Fctors Associted with Dough Stickiness s Sensed by Attenuted Totl Reflectnce Infrred Spectroscopy Ewoud J. J. vn Velzen, 1,2 John P. M. vn Duynhoven, 1 Pul Pudney, 3 Peter L. Weegels, 1 nd John H. vn der Ms 4 ABSTRACT Cerel Chem. 80(4):378 382 Attenuted totl reflectnce (ATR) nd Fourier trnsform infrred (FTIR) spectroscopy hve been pplied in the chrcteriztion of sticky dough surfces. The chrcteriztion provides insight in the chemicl distribution of gluten protein, strch, wter, nd ft during dough kneding. ATR is especilly useful for selective smpling of dough surfces becuse the depth of penetrtion of rdition is quite shllow. For dough, it is clculted to be in the order of 0.5 4 µm in the midinfrred, idel for mesurements of stickiness effects, where only the dough surfce is of interest. To investigte the cohesive nd dhesive properties of the individul dough constituents, dough ws peeled from the ATR plte to study the mteril tht dhered to it. The infrred spectr obtined indicte tht ft nd gluten protein pper to be locted t the outer sticky dough surfces, rther thn wter nd strch. In comprison with gluten, the ftty component showed reltively strong dhesive forces to the ATR plte; high residul frction ws mesured fter peeling the dough. Gluten proteins disply different cohesion nd dhesion properties tht re strongly dependent on their hydrtion stte. This indictes tht the degree of hydrtion of gluten proteins contributes to the sticky properties of (overkneded) dough. When nlyzing gluten protein in D 2 O insted of dough mtrix, more or less similr results were obtined. Significnt differences in mide I nd mide II intensities were mesured for kneded nd stretched gluten protein in comprison to untreted, wet gluten. Besides chnges in the vibrtionl properties of the mide groups, conformtionl chnges in the tertiry protein structure lso were observed. It ppers tht kneding nd stretching of dough results in mjor decrese in -helices content, ccompnied by n increse of extended b-sheet conformtions. According to Chen nd Hoseney (1995) nd Hoseney nd Smewing (1999), the rheologist Scott Blir (1936) netly summed up the problem of dough stickiness over 60 yers go when he stted tht stickiness ppers to be the property of dough to which the bker s hnd is most sensitive. Tody, when dough is hndled more by mchine nd less by the bker s hnd, the importnce of sticky dough is even greter. Dough tht is sticky cn cuse immense problems in the mechnicl bredmking process. Dusting flour, up to 5% of the totl flour, is sometimes used to overcome stickiness. However, the consequence of dding flour genertes significntly higher production costs. Although the impct of the efficiency loss is severe, the fundmentl knowledge nd controlling dough stickiness is still limited. To understnd the phenomenon of stickiness better s result of prolonged kneding, spectroscopic study hs been performed. However, mny other processing nd formultion vribles cn contribute to stickiness s well (e.g., kneding temperture, protein composition, pentosn content). These vribles were not evluted in the present study. Even though sticky doughs hve been problem ever since humns first strted mking dough, it received limited scientific study until recently. Previous studies hve linked number of fctors to dough stickiness. These include the mount of wtersoluble pentosns, differences in protein composition, -mylse ctivity, nd proteolytic ctivity (Chen nd Hoseney 1995; Hoseney nd Smewing 1999). However, these studies of fctors ffecting dough stickiness were ll hmpered by lck of n objective test to mesure stickiness. Chen nd Hoseney (1995) nd Hoseney nd Smewing (1999) reported studies focused on fctors ffecting the dhesion properties of dough. Sequentil seprtion, frctiontion, nd reconstitution techniques were used to show tht the entity tht cused the dough to be sticky (dhesive) ws wter-soluble nd of reltively smll size. The compound ws identified s ferulic 1 Unilever Reserch, P.O. Box 114, 3130 AC Vlrdingen, The Netherlnds. 2 Corresponding uthor. E-mil: Ewoud-vn.Velzen@unilever.com. 3 Unilever Reserch, Colworth House Shrnbrook, Bedford MK44 1LQ, United Kingdom. 4 Infrred nd Rmn Group, Utrecht University, Sorbonneln 16, 3584 CA Utrecht, The Netherlnds. Publiction no. C-2003-0614-03R. 2003 Americn Assocition of Cerel Chemists, Inc. cid, esterified to hexose chin. It ws lso shown tht sponifiction of the ester bond destroyed the bility of the entity to cuse sticky dough. Although the exct nture of stickiness is still unknown, one of the most likely hypotheses strts from the ide tht wter undergoes redistribution during kneding, thus ltering the hydrtion stte of the gluten phse. The wter-redistribution phenomenon nd the hydrtion of gluten hve lso been studied by spectroscopic methods. Fourier trnsform infrred (FTIR) spectroscopy hs shown tht the gluten mide bsorptions undergo intensity chnges during kneding tht hve been ttributed to hydrtion. Stickiness is surfce-relted chrcteristic to which humn skin is sensitive. This property cn be defined s the force of dhesion when two surfces re in contct with ech other. In dough mtrices, the dhesion force is combintion of n dhesive force nd cohesive force. It is when the dhesive force is high nd the cohesive force is low tht the dough is perceived s being sticky (Hoseney nd Smewing 1999). Dobrszczyk (1997) sttes tht the cohesive energy term is usully the mjor contribution to the mesured dhesive frcture energy nd is generlly highly dependent on the viscoelstic properties of the dhesive. Therefore, stickiness is not constnt mteril property but is principlly dependent on the viscoelstic rheologicl properties of the dhesive nd, s such, will be dependent on the imposed rte, temperture, nd size of the deformtion. To study rheologicl-chemicl reltionships t sticky dough surfces, it is essentil to use nlyticl techniques with the bility to collect chemicl informtion specificlly from the outer micrometer lyers. Furthermore, crucil limiting fctor is the restricted exposure time with ir to prevent time-relted vribles such s chnges in temperture nd humidity, evportion of wter, nd proofing. In this context, FTIR spectroscopy is probbly one of the most convenient nd strightforwrd nlyticl tools to get the chemicl informtion. In FTIR, the presence of functionl groups (e.g., mide groups in protein) is determined either qulittively or quntittively. This is ccomplished by detecting the bsorption of infrred light in the wvelength rnge between 2.5 nd 25 µm (4,000 400 cm 1 ). The frequency of the light tht is bsorbed is chrcteristic for the presence of specific functionl group or chemicl bond type. The frequency of the light tht specific bond will bsorb depends on the chemicl environment in which the 378 CEREAL CHEMISTRY
functionl group or chemicl bond is contined. For exmple the mide peptide bonds in gluten proteins with n -helix conformtion dsorb light in the frequency of 1,650 1,655 cm 1, wheres mide peptide bonds in b-sheet conformtion dsorb light t 1,620 1,640 cm 1 (Byler et l 1986; Mntsch nd Csl 1986; Whrton 1986). An ttenuted totl reflection (ATR) ccessory ws used for mesuring dough surfces of different sttes of stickiness within the first micrometer lyers s well s the conformtionl behvior of gluten protein s result of kneding, stretching, nd hydrtion. ATR Infrred Spectroscopy In ATR, the smples re put on to n infrred trnsprent crystl of high refrctive index (internl reflectnce element [IRE]) tht permits lmost totl internl reflectnce. ATR hs been pplied for dough nd gluten protein mesurements becuse it offers some importnt benefits. The min chrcteristic of ATR is its surfce sensitivity, which results from the phenomenon of the evnescent infrred wve, estblished between the reflecting ATR crystl nd the smple, penetrting only few micrometers into the smple. The smll penetrtion depth of the evnescent wve requires close contct between the crystl nd the smple to obtin the smple s infrred spectrum. The depth of penetrtion (D p ) of the incident infrred rdition depends on the ngle of incidence (q), the frequency of incident light t ny specific wvelength l, nd the refrctive index difference between ATR element (h 1 ) nd the dough smple (h 2 ) (Hrrick 1979; Wilson 1990; vn de Voort nd Ismil 1991; Dupuy et l 1994; Mrtin 1999). Depth profiling is fesible using different ATR crystls with incresing ngles of incidence (q). In the following eqution, D p is defined s the distnce required for the mplitude of the incident wvelength (l) to fll to n e 1 vlue: Another dvntge of the ATR pproch is tht the smple surfce is protected from the tmosphere. Direct contct between the dough nd the ATR plte implies tht chemicl nd physicl chnges will be restricted during the infrred cquisition time. Furthermore, the bot design of the ATR plte permits the ddition of extr solvent (H 2 O nd D 2 O) to prevent dehydrtion. (1) MATERIALS AND METHODS Smple Preprtion nd Hndling Dough smples were prepred for Bguepi flour (Moulins de Soufflet Pntin, Frnce) with 1.4% (w/w) slt, 4.4% (w/w) yest, 0.2% (w/w) bred improver (Congel, Astr Clve, Frnce), nd 36% (w/w) wter nd kneded in spirl mixer (Diosn Dierks & Söhne GmbH, Germny) for 2 nd 25 min to obtin nonsticky nd sticky dough. During kneding, the temperture ws kept constnt t 20 ± 0.5 C with temperture controlled CO 2 cooling purge. The freshly prepred dough smples (0.5 cm thick) were crefully positioned on horizontl zinc selenide (ZnSe) ATR plte with miniml stress nd sher forces. Immeditely fter cquisition, ech smple ws removed from the plte in one flowing move to hve more or less reproducible residue left on the horizontl ATR plte. After removing the bulk of the dough, second infrred mesurement ws performed on the remining mteril. Ech of the FTIR experiments were performed within 10 min of smple preprtion to prevent moleculr nd structurl chnges s much s possible. All experiments were repeted once to estimte the reproducibility of the obtined results. Gluten protein smples were prepred in D 2 O with 34% dry vitl gluten protein (Avebe, The Netherlnds) nd kneded for 1 nd 13 min using Brbender Frinogrph to investigte conformtionl chnges in the secondry gluten protein structure s result of kneding (Tble I). During kneding, the temperture ws kept t 20 25 C. Like the dough smples, the prepred gluten protein were crefully positioned (without stretching) on ZnSe ATR plte nd covered with D 2 O to prevent dehydrtion. In second experiment, infrred mesurements of the gluten protein smples were collected while pplying sttic stretch. To chieve this, piece of the gluten protein ws stretched out nd fixed to ZnSe ATR plte using n ATR clmp (SpectrTech). Infrred Mesurements ATR FTIR experiments were performed on spectrometer (FTS-6000, Bio-Rd, Cmbridge, MA) equipped with deuterted tri-glycine sulfte (DTGS) detector. For both smple nd bckground, 64 symmetric interferogrms with sptil resolution of 4 cm 1 were co-dded nd Fourier-trnsformed. Bsed on the symmetric pek shpes in the resulting infrred spectr, it ws not necessry to phse-correct the symmetric dt set. WinIR-Pro (v. 2.9, Bio-Rd) nd GRAMS/32 (v. 5.0, Glctic) softwre were used for cquisition nd spectrl processing, respectively. The smples were mesured on n ATR ccessory (SpectrTech) with three zinc selenide (ZnSe) crystls ech possessing different ngle of incidence (q): 40, 45, nd 60. The single-bem spectrum of ech smple ws rtioned ginst the reference single-bem spectrum of the clen ATR crystl. TABLE I Experimentl Conditions of Gluten Protein in D 2 O Experiment Gluten Protein in D 2 O Kneded Stretched 1 Yes No No 2 Yes Yes No 3 Yes No Yes 4 Yes Yes Yes Kneding times: 1 min (underkneded) nd 13 min (overkneded). Fig. 1. Attenuted totl reflectnce (ATR) nd Fourier trnsform infrred (FTIR) spectrum of optimlly developed Bguepi dough (kneded for 13 min) collected on ZnSe ATR crystl with 60 ngle of incidence. nch, liphtic CH stretching of CH 2 nd CH 3 (ft); nco, crbonyl stretching of the ft triglyceride ester linkge; A 1, crbonyl stretching of the gluten protein (mide I) in combintion with the OH deformtion of wter; A 2, NH bending nd CH stretching of gluten protein (mide II); ncoc/nco, C-O-C stretching nd CO (-COH) stretching of strch (Workmn 2001). TABLE II Chrcteristic Frequencies (cm 1 ) nd Assignments of Four Min Amide I Components Wvenumber (cm 1 ) Assignment 1,675 ± 5 b-turn 1,653 ± 4 -Helix 1,633 ± 5 Extended b-sheet 1,618 ± 10 Intermoleculr b-sheet (due to ggregtion) Workmn (2001). Vol. 80, No. 4, 2003 379
Curve Fitting To investigte the spectrl chnges within the complex infrred spectr, the use of curve-fitting techniques (Byler et l 1986) is prcticlly inevitble. In this mthemticl pproch, number of Gussin curves re fitted to set of overlpping peks. For the spectrl region 1,800 1,500 cm 1, it is essentil to resolve three overlpping bnds t 1,745 cm 1 (C=O stretch), 1,640 cm 1 (C=O stretch of mide I nd OH bend), nd 1,550 cm 1 (mide II, NH bend/cn stretch) into seprte peks. These infrred bnds cn be ssocited with ft, gluten protein nd wter, nd gluten protein, respectively (Workmn 2001). Itertive dt processing ws performed with GRAMS/32 softwre nd the Arry-Bsic modules Curve Fitting nd Pek Integrtion. In the itertive curve-fitting procedure, the bndwidths of the infrred bsorptions t 1,745, 1,640, nd 1,550 cm 1 were initilly set to relistic strting vlues (i.e., 22, 76, nd 50 cm 1, respectively). Curve fitting is lso used to investigte conformtionl chnges in the secondry protein structure of gluten under vrying experimentl conditions in D 2 O (Tble II). The use of D 2 O is n importnt prerequisite becuse H 2 O is chrcterized by strong overlpping bsorption with the crbonyl stretching bnd of gluten protein t 1,640 cm 1. Becuse the crbonyl bsorption (C=O stretch, mide I bnd) provides sensitive probe for the secondry protein structure, the bsence of overlpping bnds is evident. To evlute the mjor structure elements in the secondry protein conformtion, curve-fitting techniques re required. According the curvefitting procedure described by Pezolet et l (1992), the mide I bnd cn be resolved into four seprte peks. These seprted peks cn be ssocited with distinct segments of the peptide bckbone tht represent different conformtions: intermoleculr b- sheets (1,612 cm 1 ), extended b-sheets (1,631 cm 1 ), -helixes (1,650 cm 1 ), nd b-turns (1,668 cm 1 ). During itertive curvefitting, the bndwidths of the 1,612, 1,631, 1,650, nd 1,668 cm 1 peks were initilly set to relistic strting vlues (i.e., 33, 40, 27, nd 29 cm 1, respectively). Dividing the sum of the clculted res of ll components ssocited with given conformtion by the totl mide I bnd re gives number tht indictes wht frction of the protein hs this prticulr conformtion (Byler et l 1986; Tthm et l 1990; Pezolet et l 1992; Popineu et l 1994; Mngvel et l 1998; Belton 1999). It should be mentioned tht this experiment is bsed on the ssumption tht the conformtionl behvior of gluten proteins in D 2 O is more or less similr to TABLE III Position, Pek Assignment, nd Quntifiction of Selected Infrred Bnds Representtive of Spectrl Mrkers for Individul Dough Constituents Pek Limits (cm 1 ) Assignment b Clcultion of Pek Are 2,150 ± 300 d def OH (wter) Integrtion 1,745 ± 30 n CO (ester of ft) Integrtion nd curve fitting 1,640 ± 100 n CO (mide I), d bend OH(wter) Integrtion nd curve fitting 1,550 ± 65 d bend NH / n CN (mide II) Integrtion nd curve fitting 1,070 ± 120 n CO / ncoc (strch) Integrtion Workmn (2001). b d: Group vibrtionl bending modes (or deformtions); n: group vibrtionl stretching modes. TABLE IV Penetrtion Depth (D p ) of Infrred Rdition Through Dough Surfce Depending on Angle of Incidence (q: 40, 45, nd 60 ) nd Wvelength Wvelength (cm 1 ) l (µm) D p (µm) 2,150 4.65 2.1, 0.9, nd 0.5 1,745 5.73 2.5, 1.1, nd 0.6 1,640 6.10 2.7, 1.2, nd 0.7 1,550 6.45 2.8, 1.3, nd 0.7 1,070 9.35 4.1, 1.9, nd 1.0 Refrctive index (h 1) for ZnSe ATR element = 2.4. Index for dough smples (h 2) 1.5 on verge (see Eq. 1). tht in H 2 O becuse both solvents possess corresponding liquid properties, volume properties, densities, nd dielectric constnts t room temperture. RESULTS Chemicl Chrcteriztion of Dough Surfces As illustrted in Fig. 1, the infrred spectrum of Bguepi dough developed to optimum mixing time (13 min) shows strong nd distinctive bsorption bnds of the min constituents. Gluten protein, wter, ft, nd strch cn esily be identified due to fvorble infrred ctivities. Wter hs chrcteristic infrred fetures t 3,350 cm 1 (OH stretch), 1,640 cm 1 (OH bend), 2,150 cm 1 (OH def), nd <800 cm 1 (OH def). The infrred bsorption bnds of strch re t 1,020, 1,080, nd 1,150 cm 1, nd cn be ssocited with the coupled C-O nd C-C stretching vibrtions (C-O/C-C stretch) of the polyscchride molecules. Finlly, ft displys some wek bsorptions t 2,958 cm 1 (symmetric stretch of CH 3 ), 2,928 cm 1 (symmetric stretch of CH 2 ), nd 2,854 cm 1 (symmetric stretch of CH 2 ) tht cn be ssigned to the liphtic ftty cid chins. Furthermore, the crbonyl stretching of the ftty ester groups cn be distinguished t 1,743 cm 1. Typicl protein bsorption bnds of mide I nd mide II pper t 1,651 cm 1 (C=O stretch) nd 1,541 cm 1 (NH bend/cn stretch), respectively. The mide I bnd, however, is strongly overlpped by the OH deformtion bnd of wter (OH bend). No wter-subtrcted spectrum is clculted becuse wter is one of the crucil dough ingredients to look for in connection with dough stickiness. Insted of wter subtrction, curve fitting is pplied to resolve the overlpping bnds in the spectrl region 1,800 1,500 cm 1 (Fig. 2). To be ble to compre wter, gluten protein, ft, nd strch distributions t different dough surfces, reltive pek res were clculted from typicl spectrl mrkers, ech representing nother constituent (Tble III). Chemicl effects s n effect of kneding were studied for dough surfces t vrying penetrtion depths (D p ) using four different ZnSe ATR crystls. Depending on the wvelength (l) nd the ngle of incidence (q) of the ATR pltes, depth profiling experiments were performed (Tble IV). A similr pproch ws performed on the remining dough frctions fter removing (peeling) the dough from the ATR crystls. When evluting the collected dt set, significnt effect cn be seen s result of prolonged kneding ction (producing sticky dough). As illustrted in Fig. 3D, the mide II bnd of gluten protein t 1,550 cm 1 increses s n effect of kneding, while the CO/COC stretching bnd of strch t 1,070 cm 1 minly shows decresing trend (Fig. 3E). Fig. 2. Attenuted totl reflectnce (ATR) nd Fourier trnsform infrred (FTIR) spectrum (1,800 1,500 cm 1 ) of optiml developed Bcquepi dough (kneded for 13 min) collected on ZnSe ATR crystl with 60 ngle of incidence. Curve fitting is pplied to resolve infrred bnds t 1,745 cm 1 (CO stretch), 1,640 cm 1 (mide I, OH bend), nd 1,550 cm 1 (mide II). 380 CEREAL CHEMISTRY
Four explntions cn be postulted with respect to these observtions. 1) Hydrtion will led to n incresed trnsition dipole moment of the mide groups, resulting in more intense infrred bsorptions. 2) Hydrtion will led to better physicl contct between the ZnSe ATR substrte nd the (sticky) gluten protein. 3) Hydrtion will promote the gluten protein to migrte to the upper lyers due to n incresing mobility. Consequently, the strch content t the upper lyers will decrese, or will be superseded to deeper situted lyers. 4) The strch dsorbs wter nd swells. Therefore this component becomes less concentrted nd is therefore less observed. In generl, these observtions strongly indicte tht the hydrtion of gluten protein plys n importnt role in the development of sticky dough. Another trend ws observed when evluting the sticky nd nonsticky dough compositions t three different smpling depths. As shown in Fig. 3A nd B, the bsorption of ft t 1,745 cm 1 (C=O stretch) nd the OH deformtion bnd of wter t 2,150 cm 1 show remrkbly opposite surfce ffinity. The obtined dt suggest tht ft is preferbly locted t the outer dough surfce, in contrst to wter. When peeling the dough from the ATR crystl the opposite effect is observed. As shown in Fig. 3G, the reltive mount of ft remining t the ATR crystl is significntly higher thn wter (Fig. 3F). Along with surfce ffinity, this observtion provides direct evidence for the reltively wek interction of ft with the dough. Other thn these two mjor trends, the remining contour plots in Fig. 3 showed no distinctive reltionships. Interprettion of Fig. 3C nd H is hmpered due to the spectrl contribution of gluten protein nd wter t 1,640 cm 1 (mide I of gluten protein nd OH bend of wter). Differences in wter nd protein concentrtions, the individul hydrtion sttes, nd the sensitivity towrd infrred rdition mkes this combined infrred bnd rther difficult to interpret. Furthermore, different distribution behvior of gluten protein nd strch is observed fter peeling the sticky nd nonsticky dough from the ATR pltes. In contrst to the distribution of these constituents in the bulk dough (Fig. 3D nd J) distinctive trends re bsent. Apprently the reltive proportions of gluten protein nd strch re severely disrupted s consequence of mnul (irreproducible) peeling. It is expected tht the ppliction of more intelligent peeling pproch will led to more reproducible mesurements (Chen nd Hoseney 1995; Dobrszcyk 1997; Hoseney nd Smewing 1999; Schwzlff et l 2001). Chemicl Chrcteriztion of Gluten Protein A series of gluten smples in D 2 O were mesured on 45 ZnSe ATR plte to investigte conformtionl chnges in the secondry gluten protein structure s n effect of kneding nd stretching. In Fig. 4A, the collected infrred spectr obviously disply significnt intensity chnges in mide I (1,640 cm 1 ) nd mide II (1,550 cm 1 ) bsorptions. Similr to the kneding experiments with dough, these effects cn be explined minly by hydrtion. Kneding nd stretching ctions strongly promote protein hydrtion, which in turn cuses 1) n incresing trnsition dipole moment of the infrred sensitive mide groups; nd 2) better physicl contct between the ZnSe ATR substrte nd the sticky gluten protein. To emphsize the spectrl fetures due to conformtionl chnges in the secondry protein structure, curve fitting ws pplied to resolve four infrred frequencies t 1,612, 1,631, 1,650, nd 1,668 cm 1 (Fig. 4B). The infrred bsorption t 1,631 cm 1 is highly chrcteristic of mide groups involved in the extended b-sheet structure, while the low-frequency mide I bnd t 1,612 cm 1 cn be ssocited with the presence of intermoleculr b-sheet networks becuse it occurs redily on ggregte proteins. The prominent bnd t 1,650 cm 1 is ssigned to the -helicl conformtion, lthough Fig. 3. Dough constituent distributions of (A)+(F) wter, (B)+(G) ft, (C)+(H) wter/gluten protein, (D)+(I) gluten protein, nd (E)+(J) strch t dough surfces depended on smpling depth nd kneding time. Mesurements were performed (A E) before nd (F J) fter peeling the dough from the ZnSe ATR pltes. Approximtion of penetrtion depth t vrious ngles of incidence re given in Tble IV. Fig. 4. A, Infrred spectr of gluten protein in D 2 O s n effect of kneding nd stretching, mesured t 45 ZnSe ATR plte. B, Curvefitted mide I bnd of gluten protein in D 2 O (Pezolet et l 1992). Vol. 80, No. 4, 2003 381
When nlyzing gluten protein in D2O insted of dough mtrix, more or less similr results were obtined. Significnt differences in the mide intensities were mesured for kneded or stretched protein gluten in comprison with the wet, untreted gluten. In ddition to chnges in bnd intensities, conformtionl chnges were lso quntified. It ppers tht kneding nd stretching results in decrese of the mount of -helices ccompnied by n increse of the content of the extended b-sheet conformtions. ACKNOWLEDGMENTS We thnk R. de Mn nd K. vn der Linden for prepring the dough smples nd Y. Nicols for ssistnce with the infrred mesurements nd the stimulting discussions. LITERATURE CITED Fig. 5. Conformtionl chnges in the secondry protein structure of gluten protein s result of stretching or kneding. Reltive pek res determined (Pezolet et l 1992) t 1,612 cm 1 of intermoleculr b-sheet (A), 1,631 cm 1 of extended b-sheet (B), 1,650 cm 1 of -helix (C), nd 1,668 cm 1 of b-turn (D). Curve-fitting ws pplied in the mide I region of infrred spectrum to resolve overlpping structure elements. the spectrl contribution from the unordered conformtion cnnot be completely ruled out t this frequency. Finlly, the 1,668 cm 1 bnd is ssigned to the b-turn conformtion (Pezolet et l 1992; Popineu et l 1994; Belton 1999). The method of Pezolet et l (1992) hs been used to quntify the secondry structure elements in the gluten protein. The results obtined with this method (Fig. 5) show tht gluten protein contins minly reltively high frctions of b-sheets. Resonble differences were observed when the gluten proteins were kneded nd stretched before infrred nlysis in comprison with freshly prepred (moistened) gluten. It ppers tht kneding nd stretching minly result in decresing mount of -helices (Fig. 5C) ccompnied by incresed content of the extended b-sheet conformtions (Fig. 5B). DISCUSSION AND CONCLUSIONS The importnt role of gluten proteins with respect to dough stickiness ws estblished by ATR FTIR spectroscopy. Using ATR FTIR, spectrl informtion ws obtined from the outer dough surfce (0.5 4 µm). Depending on the hydrtion stte of the gluten protein, reltively higher mide intensities were mesured t dough surfces compred with strch nd wter. It ws estblished experimentlly tht chnges in reltive mide bnd intensities cn be explined by chnges in the trnsition dipole moment. Bsed on these observtions, hydrted gluten might be one of the mjor contributors to dough stickiness. Along with gluten protein, ft lso shows reltively strong surfce ffinity. However, when pplying peeling test t n ATR plte, ft shows reltively wek ffinity with dough. This might indicte tht ft is probbly not the dominting fctor in dough stickiness. Belton, P. S. 1999. On the elsticity of whet gluten. J. Cerel Sci. 29:103-107. Byler, D. M., Bouillette, J. N., nd Susi, H. 1986. Quntittive studies of protein structure by FTIR spectrl deconvolution nd curve fitting. Spectroscopy 3:29-32. Chen, W. Z., nd Hoseney, R. C. 1995. Development of n objective method for dough stickiness. Lebensm. Wiss. Technol. 28:467-473. Dobrszcyk, B. L. 1997. The rheologicl bsis of dough stickiness. J. Texture Stud. 28:139-162. Dupuy, N., Duponchel, L., Amrm, B., Huvenne, J. P., nd Legrnd, P. 1994. Quntittive nlysis of ltex in pper cotings by ATR-FTIR spectroscopy. J. Chemometrics 8:333-347. Hrrick, N. 1979. Internl Reflection Spectroscopy. Hrrick Scientific Corp.: Ossining, NY. Hoseney, H. C., nd Smewing, J. 1999. Instrumentl mesurement of stickiness of doughs nd other foods. J. Texture Stud. 30:123-136. Mngvel, C., Crre, P., Snchez, A. C., Brbot, J., Popineu, Y., nd Gueguen, J. 1998. Film formtion from whet gluten proteins: A Fourier trnsform infrred spectroscopic study. Biopolym. Sci. Food Non Food Appl. 1:257-261. Mntsch, H. H., nd Csl, H. L. 1986. Biologicl pplictions of infrred spectrometry. Fresenius Z. Anl. Chem. 324:655-661. Mrtin, K. 1999. Infrred nd Rmn studies of skin nd hir: A review of cosmetic spectroscopy. Int. J. Vibr. Spectrosc. 3:1-21. Pezolet, M., Bonenfnt, S., Dousseu, F., nd Popineu, Y. 1992. Conformtion of whet gluten proteins Comprison between functionl nd solution sttes s determined by infrred spectroscopy. FEBS Lett. 3:247-250. Popineu, Y., Bonenfnt, S., Cornec, M., nd Pezolet, M. 1994. A study by infrred spectroscopy of the conformtions of gluten proteins differing in their glidin nd glutenin compositions. J. Cerel Sci. 20:15-22. Schwzlff, S. S., Uriyo, M. G., Johnson, J. M., Brbeu, W. E., nd Griffey, C. A. 2001. Apprent dough stickiness of selected 1BL/1RS trnslocted soft whet flours. Cerel Chem. 78:93-96. Tthm, S. A., Drke, A. F., nd Shewry, P. R. 1990. Conformtionl studies of synthetic peptides corresponding to the repetitive regions of the high moleculr weight (HMW) glutenin subunits of whet. J. Cerel Sci. 11:189-200. vn de Voort, F. R., nd Ismil, A. A. 1991. Proximte nlysis of foods by mid-ftir spectroscopy. Trends Food Sci. Technol. 13-17. Whrton, C. W. 1986. Infr-red nd Rmn spectroscopic studies of enzyme structure nd function. J. Biochem. 233:25-36. Wilson, R. H. 1990. Fourier trnsform mid-infrred spectroscopy for food nlysis. Trends Anl. Chem. 9:127-131. Workmn, J. 2001. Methods nd interprettions. Pges 209-242 in: Hndbook of Orgnic Compounds. Acdemic Press: London. [Received Jnury 11, 2002. Accepted November 19, 2002.] 382 CEREAL CHEMISTRY