A lipase based approach for studying the role of wheat lipids in bread making

Food Chemistry 156 (2014) 190–196

Contents lists available at ScienceDirect

Food Chemistry journal homepage: www.elsevier.com/locate/foodchem

A lipase based approach for studying the role of wheat lipids in bread making Lien R. Gerits ⇑, Bram Pareyt, Jan A. Delcour Laboratory of Food Chemistry and Biochemistry, Leuven Food Science and Nutrition Research Centre (LFoRCe), KU Leuven, Kasteelpark Arenberg 20 Box 2486, B-3001 Leuven, Belgium

a r t i c l e

i n f o

Article history: Received 12 November 2013 Received in revised form 24 December 2013 Accepted 27 January 2014 Available online 7 February 2014 Keywords: Wheat flour lipids Lipases Bread making Lipid mesomorphic phase HPLC–ELSD

a b s t r a c t While endogenous wheat lipids exert a major effect on bread quality, little is known on the way they impact on bread loaf volume (LV). Here we altered wheat flour lipid composition during bread making using lipases in situ. Lipopan F, Lecitase Ultra, and surfactants increased LV to similar extents. The increases in bread LV as a result of these enzymes were related to decreased levels of galactolipids and phospholipids and concomitant increased ‘lyso’-lipid as well as free fatty acid (FFA) levels. The FFA formed were transferred to the free lipid fraction, while the ‘lyso’-lipids remained in the bound lipid fraction. For optimal gas cell stabilisation, an equilibrium between the lipid classes present and hence, the type of mesophase formed, is essential. Sufficient levels of lipids forming lamellar mesophases and lipids forming hexagonal I mesophases, which respectively form condensed monolayers or emulsify (deleterious) non-polar lipids in dough liquor, are needed. Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction Endogenous wheat lipids, although a minor fraction of the kernel and the derived flour, are important in bread making (Chung, Ohm, Ram, Park, & Howitt, 2009; Pareyt, Finnie, Putseys, & Delcour, 2011). During dough mixing, lipids redistribute as observed from decreased levels of free and concomitant increased levels of bound lipids. The phenomena occurring have been referred to as lipid binding (Carr, Daniels, & Frazier, 1992; Chung, 1986; Olcott & Mecham, 1947; Wootton, 1966). As described recently by Finnie, Jeanotte, Morris, Giroux and Faubion (2010) and Gerits, Pareyt, and Delcour (2013), during dough development lipids are rubbed from the surface of the starch granules and thereby become trapped in (Marion, Le Roux, Akoka, Tellier, & Gallant, 1987) or interact with (McCann, Small, Batey, Wrigley, & Day, 2009) the gluten network.

Abbreviations: ASE, accelerated solvent extractor; DAG, diacylglycerols; DATEM, diacetyl tartaric esters of mono- and diglycerides; DGDG, digalactosyldiacylglycerols; DGMG, digalactosylmonoacylglycerols; dm, dry matter; EP, enzyme protein; FFA, free fatty acids; HPLC–ELSD, high performance liquid chromatography coupled with evaporative light scattering detection; LPC, lysophosphatidylcholine; LV, loaf volume; MAG, monoacylglycerols; MGDG, monogalactosyldiacylglycerols; MGMG, monogalactosylmonoacylglycerols; NAPE, N-acyl phosphatidylethanolamine; NALPE, N-acyl lysophosphatidylethanolamine; PC, phosphatidylcholine; PE, phosphatidylethanolamine; SSL, sodium stearoyl lactylate; TAG, triacylglycerols; WSB, water saturated butan-1-ol. ⇑ Corresponding author. Tel.: +32 16 32 1634; fax: +32 16 32 1997. E-mail address: [email protected] (L.R. Gerits). http://dx.doi.org/10.1016/j.foodchem.2014.01.107 0308-8146/Ó 2014 Elsevier Ltd. All rights reserved.

This can positively impact the gluten network (strength) (Köhler, 2001; Pomeranz & Chung, 1978). That way, lipids can indirectly stabilise the gas cells in dough. This is very important as dough’s gas-holding capacity is one of, if not the most important feature in bread making as it is associated with an airy crumb structure and a good bread loaf volume (LV) (Eliasson & Larsson, 1993). However, next to their impact on the gluten network, (polar) lipids also exert a direct effect on gas cell stabilisation. Several authors (Gan, Angold, Williams, Ellis, Vaughan, & Galliard, 1990; Gan, Ellis, & Schofield, 1995; Sroan, Bean, & MacRitchie, 2009; Sroan & MacRitchie, 2009) have suggested a dual mechanism whereby gas cells are stabilised during fermentation and the early baking phase. It is based on the cooperative (and successive) effect of (i) the gluten network and (ii) a liquid lamella surrounding the gas cells. Indeed, in optimally mixed dough, the gluten network holds the gas cells and, at that stage of the bread making process, is the primary (or even sole) responsible for their stabilisation. However, already after 15 min of fermentation and during early baking discontinuities appear in the gluten network (Gan et al., 1990). From that moment onwards, a thin liquid lamella around the gas cell, stabilised by adsorbed surface active proteins and/or (polar) lipids, aids in further gas cell stabilisation and, hence, provides a secondary stabilisation mechanism (Sroan & MacRitchie, 2009). Surface active proteins and (polar) lipids stabilise the gas cells in a different way: whereas the former form viscoelastic films, the latter act through the Gibbs–Marangoni mechanism within a liquid lipid membrane. When both components are present, they compete

L.R. Gerits et al. / Food Chemistry 156 (2014) 190–196

for the gas cell interface and impair each other’s ability to stabilise gas cells (Mills, Wilde, Salt, & Skeggs, 2003). Hence, not only the gluten network rheology, but also the type of components at the interface as well as their surface active properties are important for proper gas cell stabilisation. However, although both the impact of lipids on gluten network strength including that of polar lipids and surface active proteins on gas cell stabilisation have been studied profoundly (Krog, 1981; Köhler, 2001; Selmair & Koehler, 2009), little is known about the stabilisation effects exerted by the different wheat endogenous (polar) lipid classes. Lipases have gained interest in the bread making industry as alternatives for surfactants, generally (incorrectly) referred to as emulsifiers in literature dealing with bread making. Lipases hydrolyse the endogenous wheat lipids in dough to form surface active lipids (Aravindan, Anbumathi, & Viruthagiri, 2007; Colakoglu & Özkaya, 2012; Moayedallaie, Mirzaei, & Paterson, 2010). In fact, lipase addition allows modifying the lipid population in situ, without altering other flour components. We here studied how the different wheat endogenous lipid classes affect bread LV by including lipases with different specificities in a straight dough recipe. Advantages of such an approach are that (i) it is free from impacts of extraction solvent(s) on (other) flour constituents (in particular gluten proteins) and (ii) when in contact with the enzymes lipids still occur at their native (endogenous) position. The latter is not the case when using fractionation–reconstitution (i.e. defatting followed by re-addition) (MacRitchie & Gras, 1973). According to De Maria, Vind, Oxenboll, and Svendsen (2007), the ‘perfect’ lipase would be one with optimal activity on the (tri)acylglycerol, phospholipid and galactolipid substrates in flour, and result in a gas cell stability similar to that brought about by surfactants. However, it is not clear which lipase currently can be considered as ‘perfect’ and to what extent each of the lipid classes should be hydrolysed. Against this background, it seemed logical to include bread making trials with two surfactants, i.e. diacetyl tartaric esters of mono- and diglycerides (DATEM) or sodium stearoyl lactylate (SSL), both of which are regularly used in bread making to positively impact bread LV. Free and bound lipid fractions of fermented control dough pieces and dough pieces containing added lipases in their formula were analysed with high pressure liquid chromatography coupled with evaporated light scattering detection (HPLC–ELSD). Bread properties, in particular LV, were analysed as well. Taken together, the present study not only demonstrates the specific action of the different lipases during bread making, but also relates (changes in) the properties of the lipid population to bread LV. We here report on the outcome of our work.

2. Materials and methods 2.1. General Grains from soft wheat cultivar Claire were obtained from Limagrain (Rilland, The Netherlands) and conditioned to 16.0% moisture before milling with a Bühler (Uzwil, Switzerland) MLU-202 laboratory mill, of which the milling flow scheme is depicted in Delcour, Vanhamel, and De Geest (1989). Milling yield of straight grade flour was 72.7%, its moisture and protein contents were respectively 14.1% and 10.6% [on dry matter (dm) basis]. The latter were determined with an American Association of Cereal Chemists International (AACC-I) approved method 44-19.01 (AACC-I, 1999) and an adaptation of the AOAC official method (AOAC, 1995) to an automated Dumas protein analysis system (EAS Vario Max CN, Elt, Gouda, The Netherlands) with 5.7 as nitrogen to protein conversion factor. Four different lipases were kindly donated by Novozymes (Bagsvaerd, Denmark) in a purified form, and without any amylase,

191

peptidase or xylanase side activities. Lipopan F, a Fusarium oxysporum enzyme preparation, is used in the bread making industry as a source of both lipase and phospholipase activities. Lecitase Ultra is a phospolipase used in degumming of edible oils. It is a product of combining homologous genes encoding Thermomyces lanuginosus lipase and F. oxysporum phospholipase (De Maria et al., 2007). Lipolase, a recombinant T. lanuginosus lipase is used in detergents (Aravindan et al., 2007). Finally, YieldMAX is a phospholipase A1. It is, hence, active on the sn1 acyl chain of phospholipids (Aloulou, Ben Ali, Bezzine, Gargouri, & Gelb, 2012). It originates from Fusarium sp. and is widely used in the dairy industry (De Maria et al., 2007). Lipopan F (56.51 U) had the highest lipase activity (determined as described below) towards p-nitrophenyl palmitate, followed by Lecitase Ultra (0.14 U) and Lipolase (0.12 U), which had very similar activities, and YieldMAX (5.78  10 2 U), which had the lowest activity towards p-nitrophenyl palmitate. DATEM and SSL were from Puratos (Groot-Bijgaarden, Belgium). The lipid standards needed to identify the lipids in the HPLC–ELSD method were as in Gerits et al. (2013), with exception of monogalactosylmonoacylglycerol (MGMG), which was kindly donated by Novozymes. All solvents used were HPLC-grade and from VWR (Haasrode, Belgium) or Sigma Aldrich (Steinheim, Germany), unless specified otherwise. 2.2. Dough and bread making Bread was prepared in triplicate on a 10 g scale based on the straight dough method of Shogren and Finney (1984) but without using shortening. Flour (10.0 g on a 14.0% moisture base), water, sugar (6.0% on flour basis), compressed yeast (5.3% on flour basis) and salt (1.5% on flour basis) were mixed in a 10 g pin mixer (National Manufacturing, Lincoln, NE). The amount of water added and the optimal mixing time were determined by Mixograph analysis (National Manufacturing) according to AACC-I approved method 54-40.02 (AACC-I, 1999), and were respectively 4.8 ml and 165 s. Lipases were included in the dough recipes in levels ranging from 0 to 5.0 mg EP/kg flour. In a second set-up, DATEM or SSL were added in levels of 0.5%, 1.0% and 1.5% on flour basis. For lipid analyses, dough samples were prepared in duplicate, fermented (126 min) and immediately frozen with liquid nitrogen, freeze dried, milled, sieved (mesh size: 250 lm) and stored at 18 °C. They are hereafter referred to as processed dough samples. 2.3. Lipid extraction and purification Sequential free and bound lipid extraction was as in Gerits et al. (2013) with an Accelerated Solvent Extractor (ASE) 200 (Dionex, Amsterdam, The Netherlands). Processed dough sample [0.86 g dry matter (dm)] was homogenised with 26 g of sand (50–70 mesh particle size) (Sigma Aldrich, Steinheim, Germany) and subsequently poured in a 22 ml ASE extraction cell. Free lipids were extracted with hexane and bound lipids with water saturated butan1-ol (WSB). The ASE settings were as in Gerits et al. (2013). The extracts were collected in glass test tubes and the respective extraction solvents evaporated with a Rotational Vacuum Concentrator (Q-lab, Vilvoorde, Belgium). Bound lipid extracts were purified from non-lipid material (mainly protein) as in Bligh and Dyer (1959). Finally, dry lipid extracts were stored at 80 °C in amber coloured vials under nitrogen prior to further analysis. 2.4. Lipid analyses Lipid analyses were conducted with HPLC–ELSD as in Gerits et al. (2013), which itself was based on the method described by Graeve and Janssen (2009). In the present case, the Alltech Model 3300 ELSD (Grace, Lokeren, Belgium) detector allowed altering

192

L.R. Gerits et al. / Food Chemistry 156 (2014) 190–196

the signal gain along the run, hence, avoiding multiple injections. In Gerits et al. (2013), such multiple injections with different injection volumes were still needed for proper detection and analysis of the phospholipid classes, which typically had a lower response than the non-polar lipids and galactolipids. After further optimisation, in the present case, the gain was increased from 1 to 8 after 13 min (i.e. between the elution of galactolipids and phospholipids), further increased to 16 after 15 min [i.e. between the elution of N-acyl phosphatidylethanolamine (NAPE) and phosphatidylethanolamine (PE)], and finally decreased to 1 after 22 min [i.e. before the elution of lysophosphatidylcholine (LPC)]. One injection (2.0 and 5.0 ll for free and bound lipid extracts, respectively) sufficed for proper detection of the different lipid classes. Lipid levels were expressed as the area under the curve relative to that of the internal standard cholesterol, which does not occur in wheat.

Table 1 Loaf volume (cm3) of breads, the recipe of which did or did not contain added lipases and/or surfactants. Enzyme levels are expressed as the level enzyme protein (EP) added per kg flour, whereas surfactant levels are indicated as a percentage on flour basis. Loaf volumes with the same letter or number are not significantly different; capital letters indicates the significant differences within rows, small letters within columns of the loaf volumes with added lipases, small numbers within columns of the loaf volumes with added surfactants. DATEM, diacetyl tartaric esters of mono- and diglycerides; SSL, sodium stearoyl lactylate. Analyses were conducted in triplicate. Averages ± standard deviations are given. Control

YieldMAX Lipolase Lecitase Ultra Lipopan F

45.34 ± 1.77 (A,a,1) 0.5 mg EP/kg flour

1.0 mg EP/kg flour

5.0 mg EP/kg flour

43.49 ± 2.86 (A,a) 49.29 ± 0.63 (B,b) 53.24 ± 0.63 (C,c)

45.07 ± 1.32 (A,a) 49.12 ± 0.68 (B,b) 52.33 ± 0.63 (C,b)

45.83 ± 0.28 (A,bc) 47.29 ± 0.35 (AB,bc) 49.34 ± 0.63 (B,b)

55.66 ± 0.49 (BC,c)

57.40 ± 0.49 (C,c)

53.53 ± 0.87 (B,c)

0.5%

2.5. Lipase activity Lipase activity was determined based on Huggins and Lapides (1947) with p-nitrophenyl palmitate as substrate. The release of p-nitrophenol at 30 °C and pH 8.2 was measured colorimetrically at 410 nm. A long fatty acid chain ester was chosen to determine lipase activity, whereas short-chain water-soluble esters would be used to measure esterase activity (Gilham & Lehner, 2005). Also, the saturated free fatty acid (FFA) palmitic acid occurs in wheat flour (lipids) (Eliasson & Larsson, 1993). Lipases were dissolved in three different concentrations and 1.0 ml was added to 2.5 ml tris-(hydroxymethyl)-aminomethane (Tris) HCl buffer (made by adding 0.1 M HCl to 0.1 M Tris to bring the pH to 8.2), and 2.5 ml of 420 lM p-nitrophenyl palmitate. The extinction was measured every minute at 410 nm for 15 min. The levels of p-nitrophenol released were calculated with a p-nitrophenol calibration curve. The lipase activity (U) was defined as the amount of p-nitrophenol (lmol) released per minute and per mg enzyme under the conditions of the assay.

2.6. Statistical analyses Statistical analyses were performed with the Statistical Analysis System software 9.3 (SAS Institute, Cary, NC, USA). For several variables, it was verified whether mean values, based on at least three individual measurements, significantly differed (significance level a = 0.05, ANOVA analysis). Pearson’s correlation coefficients (p < 0.05) for mean values were calculated. For lipid levels and bread LV, ranges of duplicate measurements and standard deviations based on triplicate measurements were calculated, respectively.

3. Results 3.1. Impact of lipase addition on bread loaf volume Table 1 lists LV readings of bread, the recipe of which contained different levels of lipases or surfactants. Both DATEM and SSL increased bread LV. When added in similar levels, the increase in bread LV was larger with SSL than with DATEM. At the lowest enzyme dosage (0.5 mg EP/kg flour), all lipases but YieldMAX significantly increased bread LV, the effect being most pronounced for Lecitase Ultra and Lipopan F. Similar trends and LV readings were observed at enzyme dosage levels of 1.0 mg EP/kg flour. Further increasing the lipase addition level to 5.0 mg EP/kg flour further increased bread LV, but now to a lesser extent when considered on a per enzyme level dosage.

DATEM SSL

55.74 ± 1.99 (B,2) 59.39 ± 0.55 (B,3)

1.0% 54.19 ± 0.60 (B,2) 59.45 ± 2.01 (B,3)

1.5% 52.93 ± 0.80 (B,2) 59.08 ± 0.83 (B,3)

3.2. Impact of different lipases on wheat endogenous lipids during bread making Table 1 lists LV readings of control bread and breads from lipase containing recipes. Despite their similar activities towards p-nitrophenyl palmitate, Lipolase and Lecitase Ultra clearly had a different impact on bread LV. This suggested differences in their lipid specificity in bread dough systems. We next analysed both the free and bound lipid populations in fermented and ready to bake dough, prepared either with or without the different lipases (Figs. 1 and 2). A concentration of 0.5 mg EP/kg flour was chosen since most of the enzymes exerted an optimal effect on bread LV at that concentration (Table 1). Also, since bread LV decreased when further increasing the lipase concentrations, and since the relative extent of this decrease was largest for Lecitase Ultra, we also analysed the free and bound lipid fraction of dough to which 5.0 mg Lecitase Ultra EP/kg flour had been added (Figs. 1 and 2). Lipolase most efficiently hydrolysed TAG (Fig. 1). A concomitant increase in the level of FFA in the free lipid extract was a logical result. Surprisingly, although Lecitase Ultra and Lipopan F had little, if any, impact on TAG in the free lipid fraction, an even larger increase in FFA levels was noticed (Fig. 1). Despite the decrease in TAG and the increase in FFA levels, a concomitant increase in the level of mono- (MAG) and/or diacylglycerols (DAG) was not observed. In line with Gerits et al. (2013), the dough free lipid fraction contained no significant levels of galacto- and phospholipids. Fig. 2 shows the lipid profiles of the bound lipid fractions for control fermented dough and for fermented dough containing added lipase in a concentration of 0.5 mg EP/kg flour. The levels of TAG decreased to a similar extent when using Lipolase, Lecitase Ultra or Lipopan F. The decrease in TAG levels was smaller with YieldMAX. However, lipase addition caused only slight increases in the level of FFA. The level of DAG was below the detection limit whereas that of MAG decreased only upon addition of Lecitase Ultra and Lipopan F. In general, Lecitase Ultra and Lipopan F most efficiently hydrolysed the polar lipids. The chromatogram of the bound lipid extract of dough showed two peaks with an elution time almost identical to that of the monogalactosyldiacylglycerol (MGDG) standard. However, since the impact of any given lipase on the second peak (elution time: 12 min 18 s) was rather small (data not shown), and since the decrease in the first peak (elution time: 12 min 10 s) was linearly related (r = 0.97) to the increase in the level of monogalactosylmonoacylglycerols (MGMG) (Fig. 2), the first peak was judged to represent MGDG. Also, the conversion of MGDG into MGMG was most pronounced with Lecitase Ultra or Lipopan F and smaller with Lipolase. Fig. 2 shows that

193

L.R. Gerits et al. / Food Chemistry 156 (2014) 190–196 0.70

0.70

TAG

0.60

0.50

0.50

0.40

0.40

FFA

AU

AU

0.60

0.30

0.30

0.20

0.20

0.10

0.10 0.00

0.00 Control

YieldMAX

Control

Lipolase Lecitase Ultra Lipopan F

YieldMAX

Lipolase Lecitase Ultra Lipopan F

Fig. 1. Free lipid levels, expressed as relative peak areas, in dough containing 0.5 mg EP/kg flour YieldMAX, Lipolase, Lecitase Ultra or Lipopan F (j) and of dough containing 5.0 mg EP/kg flour Lecitase Ultra (h). TAG, triacylglycerols; FFA, free fatty acids. Averages of duplicate measurements are given.

0.35

1.60

TAG

0.20

FFA

0.16

1.20

0.25

0.14

1.00

0.12

AU

AU

AU

0.20 0.80

0.15

0.06

0.40

0.05

0.04

0.20

0.00

0.02

0.00 Control

YieldMAX

Lipolase

Lecitase Ultra

0.30

Lipopan F

0.00 Control

YieldMAX

Lipolase

Lecitase Ultra

0.25

MGDG

0.10 0.08

0.60 0.10

MAG

0.18

1.40

0.30

Control

Lipopan F

YieldMAX

Lipolase

Lecitase Ultra Lipopan F

1.60

MGMG

DGDG

1.40 0.25

0.20

1.20 1.00

AU

0.15

AU

AU

0.20 0.15

0.10

0.80 0.60

0.10 0.40 0.05 0.05

0.20 0.00

0.00

0.00 Control

YieldMAX

Lipolase

Control

Lecitase Ultra Lipopan F

1.40

YieldMAX

Lipolase

Lecitase Ultra

1.00

DGMG

YieldMAX

Lipolase

Lecitase Ultra

1.40

NAPE

0.90

1.20

Control

Lipopan F

Lipopan F

NALPE

1.20

0.80 1.00

0.60

0.80

AU

AU

0.70

0.80

AU

1.00

0.50

0.60

0.40

0.60

0.40

0.30

0.40

0.20 0.20

0.20

0.10

0.00

0.00 Control

YieldMAX

Lipolase

0.35

Lecitase Ultra

Lipopan F

0.00 Control

YieldMAX

Lipolase

Lecitase Ultra Lipopan F

2.50

PE

Control

YieldMAX

Lipolase

0.35

PC

0.30

Lecitase Ultra

Lipopan F

LPC

0.30 2.00

0.25

0.25 1.50

AU

0.20

AU

AU

0.20 0.15

0.15

1.00

0.10

0.10 0.50

0.05

0.05

0.00

0.00 Control

YieldMAX

Lipolase

Lecitase Ultra Lipopan F

0.00 Control

YieldMAX

Lipolase

Lecitase Ultra

Lipopan F

Control

YieldMAX

Lipolase

Lecitase Ultra Lipopan F

Fig. 2. Bound lipid levels, expressed as relative peak areas, of dough containing 0.5 mg EP/kg flour of YieldMAX, Lipolase, Lecitase Ultra and Lipopan F (j), and of dough containing 5.0 mg EP/kg flour Lecitase Ultra (h). TAG, triacylglycerols; FFA, free fatty acids; MAG, monoacylglycerols; MGDG, monogalactosyldiacylglycerols; MGMG, monogalactosylmonoacylglycerols; DGDG, digalactosyl-diacylglycerols; DGMG, digalactosylmonoacylglycerols; NAPE, N-acyl phosphatidylethanolamine; NALPE, N-acyl lysophosphatidylethanolamine; PE, phosphatidylethanolamine; PC, phosphatidylcholine, LPC, lysophosphatidylcholine. Averages of duplicate measurements are given.

the relative area of the digalactosyldiacylglycerols (DGDG) significantly decreased as a result of Lecitase Ultra or Lipopan F action. Although no (commercial) lipid standard for digalactosylmonoacylglycerols (DGMG) was available, the results logically suggested that the peak eluting at 14 min 55 s is to be assigned to DGMG, since it increased at the expense of that of DGDG (r = 0.91). The levels of the flour phospholipid fraction are generally lower than those of both the non-polar lipids and the galactolipids (results not shown). NAPE was largely affected by Lecitase Ultra, Lipopan F and YieldMAX and to a lesser extent by Lipolase. N-acyl lysophosphatidylethanolamine (NALPE) levels on the other hand

increased when adding YieldMAX and to a smaller extent upon addition of Lecitase Ultra and Lipopan F. However, the increases in NALPE contents were not to any extent comparable to the decreases in NAPE contents. PE and phosphatidylcholine (PC) levels decreased efficiently with Lecitase Ultra or Lipopan F, whereas LPC levels increased for both enzymes, with a large decrease in PC levels than the corresponding increase in LPC level. The dotted bars in Fig. 2 indicate the lipid levels of dough containing 5.0 mg EP Lecitase Ultra/kg flour in its recipe. As expected, compared to dough containing less Lecitase Ultra (0.5 mg EP/kg flour), a further decrease of TAG, MGDG, DGDG, NAPE, and PE

194

L.R. Gerits et al. / Food Chemistry 156 (2014) 190–196

levels, and a concomitant increase in FFA levels was observed. More surprisingly, a further decrease (rather than an increase) in the levels of MAG, MGMG, and NALPE was observed. Only the additional decrease of the DGDG level caused a concomitant increase in DGMG concentration. PC levels remained constant, while the levels of its corresponding lyso-phospholipid, i.e. LPC, also decreased. Finally, PE completely disappeared when adding 5.0 mg EP Lecitase Ultra/kg flour. 4. Discussion The impact of the different lipases on bread LV was dose dependent, with an optimum around 0.5 or 1.0 mg EP/kg flour. Higher lipase dosages (5.0 mg EP/kg flour) were less efficient at increasing bread LV than the lower dosages. Addition of Lipopan F (at all dosage levels) and of Lecitase Ultra (at the lowest dosage level) yielded bread with a LV similar to that obtained when the recipe contained surfactants. This indicates conversion of wheat endogenous lipids to more polar components with a mode of action similar to that of the surfactants. The present data indicate that the action patterns of the different lipases towards the lipids in dough are complex and that the lipases can hydrolyse multiple lipid classes at the same time. Hence, their in vitro activity towards p-nitrophenyl palimitate is not a good predictor for their activity in dough systems. In any case, lipases are promiscuous enzymes (Aloulou et al., 2012). In general, the total amount of FFA released, i.e. the difference between the total FFA in the free and bound lipid fraction of the lipase treated samples and that of the control, can be considered to be a measure for the (in situ) lipase activity in dough itself. Remarkably, the level of FFA increased mainly to the benefit of the free lipid fraction. Thus, hydrolysis of wheat lipids in the bound lipid fraction leads to FFA which are released into the free lipid fraction (Fig. 3). An increasing level of FFA can cause off flavours in terms of acid rancidity (by the FFA themselves) or in terms of (poly)unsaturated FFA, providing a substrate for wheat lipoxygenase (Castello, Jollet, Potus, Baret, & Nicolas, 1998). However, lipoxygenase requires oxygen, which is (completely) consumed by yeast by the end of mixing (Leenhardt et al., 2006) whereas at that moment the lipid types present upon lipase addition still strongly resemble those in flour, because the effect of added lipases in terms of increasing FFA is only visible much later, i.e. after at least 10 min of fermentation (unpublished data). Furthermore, a large part of the FFA formed can complex with amylose, thereby forming amy-

2

Free lipid fracon

DAG

MGDG DGDG

NAPE PE

Phospholipids

MAG

2

TAG

FFA 234

Galactolipids

2

PC

DAG

34 34

1 34

34

234

MGMG

MAG

234

34

DGMG 34

NALPE LPE LPC

Bound lipid fracon Fig. 3. Mode of actions of the four different lipases in dough; 1, YieldMAX; 2, Lipolase; 3, Lecitase Ultra; 4, Lipopan F. Abbreviations as in Fig. 2. DAG, diacylglycerols and LPE, lysophosphatidylethanolamine.

lose–lipid inclusion complexes (Delcour & Hoseney, 2010). Taken together, production of off flavours is expected to be limited. On a protein level, Lipopan F had the highest lipase activity towards p-nitrophenyl palmitate and also released most FFA in the dough system. Although Lecitase Ultra and Lipolase showed comparable activities towards p-nitrophenyl palmitate, the former released more FFA, and its lipid hydrolysis pattern resembled more that of Lipopan F. This probably partly explained the larger bread LV obtained with Lecitase Ultra than with Lipolase. YieldMAX had a smaller effect on the lipid population than the other lipases. The level of FFA, both in the free and the bound lipid fraction, were highly correlated (r = 0.98 and r = 0.91, respectively) with the bread LV. However, in spite of this high correlation, the addition of FFA as such in bread making is, to the best of our knowledge, not a (simple) way to increase bread LV. According to De Stefanis and Ponte (1976) and Sroan and MacRitchie (2009), adding linoleic acid [i.e. the most common fatty acid in wheat flour (Eliasson & Larsson, 1993)] to native as well as to defatted flour in a bread making trial even decreases bread LV. Hence, the levels of FFA released are only a good indication for the lipase activity in situ, but provide little if any information on the underlying mechanism responsible for the observed differences in bread LV. The different impact of the lipases used has to be found in the (degree of) hydrolysis of the different lipid classes into more polar components and the type of polar components formed. The lipases, with exception of Lipolase, did not hydrolyse TAG in the free lipid fraction. This possibly indicates that these lipids are embedded inside micelles or liposomes and therefore are not present at interfaces at which lipases typically act (Aravindan et al., 2007). In spite of the hydrolysis of TAG in the free lipid fraction by Lipolase (Fig. 1), no concomitant formation of DAG and MAG was observed. This indicates, for all practical purposes, complete hydrolysis of TAG. In the bound lipid fraction (Fig. 2), Lipopan F and Lecitase Ultra also hydrolysed MAG, while Lipolase and YieldMAX showed no such action. As observed for the free lipid fraction, DAG were not detected. Taken together, TAG in the free and bound lipid fractions were thoroughly converted into FFA and glycerol. This indicates that the lipases probably release FFA from all three positions on glycerol (Macrae & Hammond, 1985) and, hence, are non-regiospecific. Based on the changes in lipid distribution summarised in Table 2 and Fig. 3 and the increases in bread LV (Table 1), the hydrolysis of the galactolipids and, to a lesser extent, of phospholipids and the concomitant release of FFA impact bread LV. All galactolipid types strongly affect bread LV, with Pearson’s correlation coefficients between bread LV and the levels of MGDG, DGDG, MGMG and DGMG of respectively 0.95, 0.91, 0.95 and 0.92. Remarkably, especially the lyso-forms of the galactolipids positively impact bread LV. Correlations between phospholipid levels and bread LV were less clear, with only the level of PE and LPC negatively correlated with bread LV (respectively r = 0.89 and r = 0.93). The correlation coefficient between PC levels and bread LV was r = 0.83 (p < 0.10). That the concomitant increase in LPC levels due to hydrolysis of PC was lower than expected was explained by the lyso-phospholipase activity of the lipases. Remarkably, although YieldMAX had little if any effect on the other lipid classes, it hydrolysed NAPE (much as did Lecitase Ultra and Lipopan F). However, based on this, NAPE was converted into NALPE and/or NALPE was further hydrolysed due to lyso-phospholipase activity, had little if any impact on bread LV. Such lyso-phospholipase activity was observed especially for Lipopan F and Lecitase Ultra. The relative decrease in bread LV upon addition of 5.0 mg EP/kg rather than 0.5 mg EP/kg could be due to thorough lipid hydrolysis. This was confirmed by analysis of the lipid population of fermented dough containing 5.0 mg EP Lecitase Ultra/kg flour. With the exception of the levels of FFA and DGMG, all lipid levels

195

L.R. Gerits et al. / Food Chemistry 156 (2014) 190–196 Table 2 Impact of the four different lipases on the different lipid classes, TAG, triacylglycerols; FFA, free fatty acids; MGDG, monogalactosyldiacylglycerols; MGMG, monogalactosylmonoacylglycerols; DGDG, digalactosyldiacylglycerols; DGMG, digalactosylmonoacylglycerols; NAPE, N-acyl phosphatidyl-ethanolamine; NALPE, N-acyl lysophosphatidylethanolamine; PE, phosphatidylethanolamine; PC, phosphatidylcholine, LPC, lysophosphatidylcholine.

TAG FFA MAG MGDG MGMG DGDG DGMG NAPE NALPE PE PC LPC

YieldMAX

Lipolase

Lecitase Ultra

Lipopan F

+ 

++ 

+++

+++



++

+++

+++





+++

++++

++



+

+

 

 

+

+

+: lipid class increases in concentration compared to the control, : lipid class decreases in concentration compared to the control, o: no significant difference compared to the control. The magnitude of the changes in concentrations gives by the number of + or signs.

decreased, indicating a further hydrolysis of the lipids as well as of the lyso-lipids. This confirms the above observation that Lecitase Ultra (and based on a similar lipid hydrolysis pattern probably also Lipopan F), not only classifies as a lipase, phospholipase and galactolipase but also as a lyso-phospholipase (and lyso-galactolipase). The mode of action appeared to depend on the substrates available. Lecitase Ultra preferably hydrolysed (bound) TAG, (MAG), phospholipids and galactolipids, but when these are present in low levels, it also hydrolyses the remaining (earlier formed) lyso-lipids, with exception of DGMG, which appears to be unaffected by the lipase. The level of PC on the other hand remained constant, independent of the level of Lecitase Ultra. As mentioned in the Introduction, lipids redistribute to the gluten fraction during dough development. Gluten proteins are positively charged in a dough system (pH  6) (Danno & Hoseney, 1982). Anionic endogenous polar lipids (Gerits et al., 2013) and anionic surfactants, like DATEM (Köhler, 2001) and SSL (Van Steertegem, Pareyt, Brijs, & Delcour, 2013), interact with these proteins through electrostatic and hydrophobic interactions. This promotes formation of gluten aggregates due to decreased electrostatic repulsion, which in itself increases gluten network strength (Köhler, 2001). The present data show that FFA released as a result of lipase action partition to the free lipid fraction, whereas the other hydrolysis products, i.e. the remaining lyso-lipid (e.g. MGMG, DGMG and NALPE), remain in the bound lipid fraction. That the latter (polar) lipids cannot be extracted with non-polar solvents such as hexane points to mediation of their association with the gluten proteins by their hydrophilic (head) part. Indeed, hexane treatment of dough ball isolated prime starch, i.e. in the absence of gluten, also extracts polar lipids (Pauly, Pareyt, De Brier, Delcour, in press). In contrast, the released FFA either do not, or interact mainly with their hydrophobic fatty acid tail. The increase in bread LV by lipases or surfactants has been attributed to improved stabilisation of gas cells during fermentation and early baking (Köhler, 2001; Primo-Martin, Hamer, & de Jongh, 2006). At the air/water interface, surfactants form monoor single-lamellar phase systems. Three types of films can be formed: gaseous, expanded and condensed liquid films, with the latter being the most stable barrier against coalescence of gas cells (Krog, 1981; Sroan & MacRitchie, 2009). Endogenous phospholipids, DGDG and saturated FFA form condensed monolayers, whereas unsaturated FFA form expanded monolayers (Sroan & MacRitchie, 2009). When in contact with water, surface active components

Molecular shape

Mesomorphic phase

Lipid classes

Cone

Cylindrical

Inverted cone

Hexagonal II

Lamellar

Hexagonal I

MGDG PE (unsaturated) NAPE (unsaturated) MAG (unsaturated)

DGDG MGMG PC NAPE (saturated) NALPE MAG (saturated)

LPC FFA

Fig. 4. The molecular shape and mesomorphy of the different (polar) lipids present in wheat, based on Selmair (2010). Abbreviations as in Fig. 2.

form highly ordered liquid-crystal phases or mesophases (Fig. 4) (Gunstone, Harwood, & Padley, 1994). Polar lipids promoting the lamellar crystalline lipid–water phase (Fig. 4) also favour the formation of condensed monolayers (Carr et al., 1992; Gan et al., 1995). The lamellar phase spontaneously forms small, bilayer aggregates (liposomes). These bilayers interact by weak Van der Waals forces. Such forces break easily, which then leads to separation of the bilayers. This, in turn, exposes the internal hydrophobic surface for interaction with the air/water interface (Carr et al., 1992; Gan et al., 1995). DATEM and SSL can form lamellar structures in water at dough mixing temperatures (Krog, 1981), which explains their positive impact on bread LV. Other mesophases are the hexagonal II and the hexagonal I lipid–water phases (Fig. 4) (Selmair, 2010). When considering Figs. 2 and 3, it appears that the lipid classes, the levels of which increase due to the action of Lecitase Ultra and Lipopan F, mainly have an ‘inverted cone’ or a ‘cylindrical’ shape, whereas the lipid classes of which the levels decreased have predominantly a ‘cone’ or ‘cylindrical’ shape. Overall, an increase in those lipids favouring a hexagonal I mesophase and a decrease in those favouring the hexagonal II phase occurred. We speculate that lipids forming a hexagonal phase I type promote the formation of micelles that positively impact the bread LV by emulsifying the deleterious endogenous wheat (non-polar) lipids in the dough aqueous phase and thereby hindering the adsorption of such lipids at the air–water interface. There appears to be an optimal balance between the lipid classes present and, hence, the mesophases formed. In particular, a sufficient level of lipids forming a lamellar mesophase stabilising the gas cells by forming a condensed monolayer should be maintained and an adequate level of lipids forming a hexagonal I mesophase, emulsifying the deleterious non-polar lipids in the dough aqueous phase have to be present. Addition of too high lipase levels decreases the level of lipid classes with a lamellar mesophase (Figs. 2 and 3) and, hence, the ability to stabilise gas cells. 5. Conclusions Wheat flour endogenous lipids tremendously affect bread LV. Altering the composition of the endogenous lipid fraction with lipases increases bread LV, the extent depending on the specificity of the respective lipases. Analysis of the lipid population of fermented bread dough revealed that hydrolysis of the galactolipids and to a lesser extent of phospholipids is important in terms of affecting bread LV. Due to lipolysis, the FFA took place in the free lipid fraction and the remaining ‘lyso’-lipid(s) remained in the bound lipid fraction. This points to an interaction of their hydrophilic part with other flour constituents, and in particular, gluten.

196

L.R. Gerits et al. / Food Chemistry 156 (2014) 190–196

The stabilisation of the gas cells is probably based on (i) a sufficient amount of lipids favouring the lamellar phase which stabilises the interface by forming a condensed monolayer, (ii) an overall decrease in the levels of those lipids promoting hexagonal phase II, and (iii) an increase in the levels of lipids favouring the hexagonal phase I. The latter may emulsify the deleterious lipids, which prevents their adsorption at the interface. However further research is needed to confirm this hypothesis. The indirect impact of wheat endogenous lipids on gas cell stability can be analysed by studying dough extensibility, either in the presence or absence of added lipases. The analysis of the lipid composition and properties (e.g. surface tension, type of mesophases formed) of dough liquor should allow unravelling the direct impact of wheat lipids on gas cell stability. Finally, wheat lipids, either endogenously present or altered by lipases, may affect bread LV due to their impact on starch gelatinization and extension of the oven spring. However, this is not clear at present. Acknowledgements The authors are grateful to H. Van den Broeck and C. Pouillie (both this lab) for technical assistance and to Novozymes (Bagsvaerd, Denmark) for providing the enzymes. This work is part of the Methusalem programme ‘Food for the future’ (2007–2014). B. Pareyt acknowledges the Research Foundation – Flanders (FWO – Vlaanderen, Brussels, Belgium) for a position as postdoctoral researcher. J.A. Delcour is W.K. Kellogg Chair in Cereal Science and Nutrition at the KU Leuven. References AACC-I. (1999). Approved Methods of Analysis (11th ed.). Methods 44-19.01 and 54-40.02. Mixograph method. Approved November 3, 1999. St. Paul: AACC International, MN, USA. doi: http://dx.doi.org/10.1094/AACCIntMethod-5440.02; doi: http://dx.doi.org/10.1094/AACCIntMethod-44-19.01. AOAC. (1995). Official methods of analysis of the association of official analytical chemists 402 (16th ed.). Washington DC, USA. Aloulou, A., Ben Ali, Y., Bezzine, S., Gargouri, Y., & Gelb, M. H. (2012). Phospholipases: An overview. In G. Sandoval (Ed.), Lipases and phospholipases (pp. 63–85). New York: Springer. Aravindan, R., Anbumathi, P., & Viruthagiri, T. (2007). Lipase applications in food industry. Indian Journal of Biotechnology, 6, 141–158. Bligh, E. G., & Dyer, W. J. (1959). A rapid method of total lipid extraction and purification. Canadian Journal of Biochemistry and Physiology, 37, 911–917. Carr, N., Daniels, N. W. R., & Frazier, P. J. (1992). Lipid interactions in bread making. Critical Reviews in Food Science and Nutrition, 31, 237–258. Castello, P., Jollet, S., Potus, J., Baret, J. L., & Nicolas, J. (1998). Effect of exogenous lipase on dough lipids during mixing of wheat flour. Cereal Chemistry, 75, 595–601. Chung, O. K. (1986). Lipid-protein interactions in wheat-flour, dough, gluten, and protein-fractions. Cereal Foods World, 31, 242–256. Chung, O. K., Ohm, J. B., Ram, M. S., Park, S. H., & Howitt, C. A. (2009). Wheat lipids. In K. Khan & P. R. Shewry (Eds.), Wheat: Chemistry and technology (4th ed., pp. 363–399). St. Paul: AACC International. Colakoglu, A. S., & Özkaya, H. (2012). Potential use of exogenous lipases for DATEM replacement to modify the rheological and thermal properties of wheat flour dough. Journal of Cereal Science, 55, 397–404. Danno, G., & Hoseney, R. C. (1982). Effect of sodium chloride and sodium dodecyl sulfate on mixograph properties. Cereal Chemistry, 59, 202–204. De Maria, L., Vind, J., Oxenboll, K. M., & Svendsen, A. (2007). Phospholipase and their industrial applications. Applied Microbiology and Biotechnology, 74, 290–300. De Stefanis, V. A., & Ponte, J. G. (1976). Studies on the breadmaking properties of wheat-flour nonpolar lipids. Cereal Chemistry, 53, 636–642. Delcour, J. A., Vanhamel, S., & De Geest, C. (1989). Physico-chemical and functional properties of rye nonstarch polysaccharides. I. Colorimetric analysis of pentosans and their relative monosaccharide compositions in fractionated (milled) rye products. Cereal Chemistry, 66, 107–111. Delcour, J. A., & Hoseney, R. C. (2010). Principles of cereal science and technology (3rd ed.). St Paul, MN: AACC International. Eliasson, A. C., & Larsson, K. (1993). Cereals in bread making: A molecular colloidal approach. New York: Marcel Dekker.

Finnie, S. M., Jeannotte, R., Morris, C. F., Giroux, M. J., & Faubion, J. M. (2010). Variation in polar lipids located on the surface of wheat starch. Journal of Cereal Science, 51, 73–80. Gan, Z., Angold, R. E., Williams, M. R., Ellis, P. R., Vaughan, J. G., & Galliard, T. (1990). The microstructure and gas retention of bread dough. Journal of Cereal Science, 12, 15–24. Gan, Z., Ellis, P. R., & Schofield, J. D. (1995). Gas cell stabilisation and gas retention in wheat bread dough. Journal of Cereal Science, 21, 215–230. Gerits, L. R., Pareyt, B., & Delcour, J. A. (2013). Single run HPLC separation coupled to evaporative light scattering detection unravels wheat flour endogenous lipid redistribution during bread dough making. LWT – Food Science and Technology, 53, 426–433. Gilham, D., & Lehner, R. (2005). Techniques to measure lipase and esterase activity in vitro. Methods, 36, 139–147. Graeve, M., & Janssen, D. (2009). Improved separation and quantification of neutral and polar lipids classes by HPLC–ELSD using a monolithic silica phase: Application to exceptional marine lipids. Joural of Chromatography B, 877, 1815–1819. Gunstone, F. D., Harwood, J. L., & Padley, F. B. (1994). The lipid handbook (2nd ed.). London: Chapman & Hall. Huggins, C., & Lapides, J. (1947). Chromogenic substrates IV. Acyl esters of pnitrophenol as substrates for the colorimetric determination of esterase. The Journal of Biological Chemistry, 170, 467–482. Köhler, P. (2001). Study of the effect of DATEM. 3: Synthesis and characterization of DATEM components. Lebensmittel-Wissenschaft Und-Technologie – Food Science and Technology, 34, 359–366. Krog, N. (1981). Theoretical aspects of surfactants in relation to their use in bread making. Cereal Chemistry, 58, 158–164. Leenhardt, F., Lyan, B., Rock, E., Boussard, A., Potus, J., Chanliaud, E., et al. (2006). Wheat lipoxygenase activity induces greater loss of carotenoids than vitamin E during bread making. Journal of Agricultural and Food Chemistry, 54, 1710–1715. Macrae, A. R., & Hammond, R. C. (1985). Present and future applications of lipase. Biotechnology and Genetic Engineering Reviews, 193–217. MacRitchie, F., & Gras, P. W. (1973). The role of flour lipids in baking. Cereal Chemistry, 50, 292–302. Marion, D., Le Roux, C., Akoka, S., Tellier, C., & Gallant, D. (1987). Lipid–protein interactions in wheat gluten: A phosphorus nuclear magnetic resonance spectroscopy and freeze-fracture electron microscopy study. Journal of Cereal Science, 5, 101–115. McCann, T. H., Small, D. M., Batey, I. L., Wrigley, C. W., & Day, L. (2009). Protein-lipid interactions in gluten elucidated using acetic-acid fractionation. Food Chemistry, 115, 105–112. Mills, E. N. C., Wilde, P. J., Salt, L. J., & Skeggs, P. (2003). Bubble formation and stabilisation in bread dough. Transactions of the Institution of Chemical Engineers, 81, 189–193. Moayedallaie, S., Mirzaei, M., & Paterson, J. (2010). Bread improvers: Comparison of a range of lipase with a traditional emulsifier. Food Chemistry, 122, 495–499. Olcott, H. S., & Mecham, D. E. (1947). Characterization of wheat gluten. I. Proteinlipid complex formation during doughing of flours. Lipoprotein nature of the glutenin fraction. Cereal Chemistry, 24, 407–414. Pareyt, B., Finnie, S. M., Putseys, J. A., & Delcour, J. A. (2011). Lipids in bread making: Sources, interactions, and impact on bread quality. Journal of Cereal Science, 54, 266–279. Pauly, A., Pareyt, B., De Brier, N., & Delcour, J. A. (in press). Incubation of isolated wheat strach with proteolytic or lipolytic enzymes and different extraction media reveals a tight interaction between puroindolines and lipids at its granule surface. Cereal Chemistry. Pomeranz, Y., & Chung, O. K. (1978). Interaction of lipids with proteins and carbohydrates in bread making. Journal of the American Oil Chemists’ Society, 55, 285–289. Primo-Martin, C., Hamer, R. J., & de Jongh, H. H. J. (2006). Surface layer properties of dough liquor components: Are they key parameters in gas retention in bread dough? Food Biophysics, 1, 83–93. Selmair, P. L. (2010). Structure–function relationship of glycolipids in bread making (Ph.D. dissertation). München (Germany): Technische Universität München. Selmair, P. L., & Koehler, P. (2009). Molecular structure and baking performance of individual glycolipid classes from lecithins. Journal of Agricultural and Food Chemistry, 57, 5597–5609. Shogren, M. D., & Finney, K. F. (1984). Bread-making test for 10-gram of flour. Cereal Chemistry, 61, 418–423. Sroan, B. S., Bean, S. R., & MacRitchie, F. (2009). Mechanism of gas cell stabilisation in bread making. I. The primary gluten–starch matrix. Journal of Cereal Science, 49, 32–40. Sroan, B. S., & MacRitchie, F. (2009). Mechanism of gas cell stabilisation in bread making. II. The secondary liquid lamellae. Journal of Cereal Science, 49, 41–46. Van Steertegem, B., Pareyt, B., Brijs, K., & Delcour, J. A. (2013). Impact of mixing time and sodium stearoyl lactylate on gluten polymerization during baking of wheat flour dough. Food Chemistry, 141, 4179–4185. Wootton, M. (1966). Binding and extractability of wheat flour lipid after dough formation. Journal of the Science of Food and Agriculture, 17, 297–301.