Investigation of the Germination of Barley and Wheat Grains with a Design of Experiments for the Production of Hydrolases

Many enzymes from diff erent classes used for food processing (1–5) originate from genetically modifi ed microorganisms (6). People in Finland, Italy and Germany show the most positive att itude towards the use of endogenic enzymes from plants (without genetic modifi cation) and most negative towards the use of genetically modifi ed enzymes, regardless of whether they are of microbial or plant origin, as shown by a study about consumers’ attitudes to enzymes in food production (7). This att itude is driven by sociopolitical viewpoints, and even a price reduction of 25 % does not seem to change their buying intention. In a study about the degradation of gliadin peptides with peptidases from germinated wheat, barley and rye grains, the authors stated that peptidases derived from germinated grains have a higher acceptance by co eliac patients (8). Hence, it seems that non-genetically modifi ed enzymes and enzymes from natural origins (e.g. plants) and not from microorganisms are of high interest for certain consumers. Additionally, the grain-associated peptidases are more suitable for the hydrolysis of some food ingredients, e.g. gluten (9).


Introduction
Many enzymes from diff erent classes used for food processing (1)(2)(3)(4)(5) originate from genetically modifi ed microorganisms (6).People in Finland, Italy and Germany show the most positive att itude towards the use of endogenic enzymes from plants (without genetic modifi cation) and most negative towards the use of genetically modifi ed enzymes, regardless of whether they are of microbial or plant origin, as shown by a study about consumers' attitudes to enzymes in food production (7).This att itude is driven by sociopolitical viewpoints, and even a price reduction of 25 % does not seem to change their buying intention.In a study about the degradation of gliadin pep-tides with peptidases from germinated wheat, barley and rye grains, the authors stated that peptidases derived from germinated grains have a higher acceptance by co eliac patients (8).Hence, it seems that non-genetically modifi ed enzymes and enzymes from natural origins (e.g.plants) and not from microorganisms are of high interest for certain consumers.Additionally, the grain-associated peptidases are more suitable for the hydrolysis of some food ingredients, e.g.gluten (9).
Enzymes from germinating seeds have been used for food processing for several years (16).However, β-glu cosidase and β-galactosidase, and systematic germination with a design of experiments in order to extend the yield were not the focus of these research activities.
In this study, the activities of selected hydrolases (glycosidases, peptidases, esterases and lipases) of the extracts from germinating barley and wheat grains are determined.Consequently, the grains were imbibed and ger minated according to the beer-brewing process, as has been done in several studies previously (17)(18)(19).These conditions were varied and three hydrolase activities (β-glucosidase, β-galactosidase and lipase) were investigated with a design of experiments.The aim of this study is the examination of the total yields of hydrolases from grains of barley and wheat for food processes.
If not otherwise described, each experiment was carried out independently (on diff erent days) in duplicate and each sampling was also carried out in duplicate.For each sampling, 2.5 g were taken, which correlates to at least 50 grains of wheat or 60 grains of barley.A negative control (hydrolases deactivated at 90 °C for 15 min) was performed for all experiments.
The limit of detection (LOD) and the limit of quantification (LOQ) are defi ned as the sixfold and ninefold signal-to-noise ratio, respectively.

Growth stage scale
The phenological changes of barley and wheat grains were characterised as published previously (20).The following modifi cations of the growth stages were used: 08= leaf just at coleoptile tip, 09=fi rst leaf penetrates coleoptile, and 10=fi rst leaf through coleoptile.

Cereal processing for preliminary experiments
Barley and wheat grains were allowed to imbibe tap water up to 28 h at (20±2) °C and samples were taken regularly to determine the water content.Germination was carried out at 20 °C in the absence of light without moisturising aft er imbibition (for 16 h) and, in an alternative experiment, with watering twice a day with tap water for 30 s. Germination was carried out at 25 and 30 °C in the absence of light, watering twice a day with tap water for 30 s.The eff ect of watering on β-glucosidase activity was measured as described below.

Cereal processing for the main experiments and the experimental design
The infl uence of temperature on the three treatment steps: pretreatment, imbibition and germination of barley and wheat grains was investigated.The cereals were pretreated for 48 h and imbibed for 16 h.Aft er removal of the liquid, germination was carried out on a tray (the tray was fi lled with 500 mL of distilled water without contact with the grains) in the absence of light with moisturising once a day by immersing the grains for 5 min.All watering steps were performed using a mineral solution (7 ).The tray contained 350 g of grains (average specifi c mass of wheat grain was 0.05 g and of barley grain 0.04 g), which formed a monolayer.
Germination experiments were carried out until the grains reached growth stage 10 (see above).Samples were taken aft er each treatment step and every or every other day (when germinated at 6 °C) during germination, and stored at -18 °C until further analysis.Three factors were considered for the screening test: the temperature during pretreatment, imbibition and germination, respectively.The temperature was varied in each of the three treatment steps.A 2 3 factorial design was chosen with a total of 19 runs consisting of eight corner points in duplicate and three centre-point replicates, thus, nine diff erent experiments (Table 1).
The pretreatment temperature was in the range from frozen soil (-18 °C) to an overheated depot (50 °C), simulating the seeding or storage of the grain.The imbibition and germination temperatures ranged from a cold room (4 to 6 °C) to a depot in summer (37 °C).Preliminary experiments (see below) showed that a maximum of 20 °C for germination is imperative for avoiding microbial growth.
Statistical analysis (ANOVA) of the observations was performed with the soft ware Design-Expert v. 8 (Stat-Ease, Inc., Minneapolis, MN, USA).Diff erences at p≤0.05 were considered signifi cant.A calculation was made for the prediction of the infl uence of temperature on the β-glucosidase activity in the examined range.The priority of the target dimension was the maximum β-glucosidase activity.

Enzyme extraction
The frozen samples (2.5 g) were homogenised aft er the addition of 10 mL of H 2 O using Ultra-Turrax (IKA, Staufen, Germany, idle speed: 13 000 rpm) for cell disruption at 0-4 °C for 4 or 6 min (wheat and barley grains, respectively).The homogenate was centrifuged (8 000×g at 4 °C for 15 min) to obtain the enzyme extract (supernatant), which was immediately used for the determination of dry mass and enzyme activities.

Determination of dry mass
A volume of 2 mL was removed from the supernatant of the homogenates and dried at 100 °C for 48 h for the determination of dry mass (in triplicate) of the extracts.The dry mass of the grains was determined by weighing before and aft er drying at 100 °C for 48 h (at least 50 grains, in triplicate).The dry mass of the grains was constant over the germination time.The water to total mass ratio during germination was always ζ=0.12 (standard deviation <5 %).The total mass (including moisture) of the grains was determined for each process step (pretreat-ment, imbibition and germination), as well as for each sample, and was expressed in kg of grain.

Determination of hydrolytic activities
Measurements of the enzyme extracts were performed in triplicate using 0.1 M sodium phosphate buff er (pH=6.0)for barley and 33 mM sodium phosphate buff er (pH=6.3)for wheat extract.
The activity of endoglycosidases was analysed using the azo-dyed substrates azo-arabinoxylan (1.0 g/L, powder), azo-galactomannan (0.2 g/L, powder) and azo-glucan (1.0 g/L, liquid).Each assay was carried out as a multipoint reaction (six points) at 37 °C and the absorbance change of the supernatant was measured at 595 nm.All substrates and measurements were carried out according to the manufacturer's instructions (Megazyme Ltd., Bray, Ireland).The absorption change was defi ned as the increase of the absorption signal at 595 nm per second per dry mass.
Endopeptidase activity was measured with the substrate azo-casein (23).The enzyme extract (100 μL) and buff er (115 μL) were preincubated at 37 °C for 5 min.The reaction was started by the addition of 60 μL of azo-casein (25 g/L) in buff er, stopped aft er 10 min with 25 μL of 2 M trichloroacetic acid and centrifuged at 13 000×g for 5 min.The supernatant (188 μL) was added to 1 M NaOH (62 μL) and the absorption was measured at 450 nm.The absorption change was defi ned as the increase of the absorption signal at 450 nm per second and per dry mass (kg of grain).

Preliminary experiments
First it was investigated which parameters (such as water content and temperature) infl uence the production of hydrolytic enzymes during germination (preliminary experiments).The water content in both barley and wheat grains reached a plateau of 50 and 40 %, respectively, aft er 16 h of imbibition (data not shown).The experiments showed that the daily rinsing of the cereal grains resulted in a twofold higher specifi c activity of β-glucosidases (a marker enzyme in this study) per grain dry mass during the germination process.Germination temperature higher than 20 °C (25 and 30 °C) did not lead to higher enzyme activities (data not shown).In summary, 16 h of imbibition and germination at no more than 20 °C were used in all experiments.

Screening for hydrolytic enzymes in barley and wheat grains
The enzyme extracts from barley and wheat were tested for diff erent enzyme activities (Table 2).Diff erent temperatures were chosen for these screening experiments in order to cover a broad temperature range.A straight comparison of the results was done with the afterwards obtained main results (see Table 3).
By far the most abundant enzyme activity present in both barley and wheat extracts was that of α-amylase, with an almost twelvefold higher α-amylase activity in the germinated barley extract (roughly 118 000 μkat per kg of grain) than in the wheat extract (about 10 000 μkat per kg of grain).Other endoglycosidase activities were determined with azo-dyed substrates.The barley extracts The β-cellobiosidase (EC 3.2.1.91)activity was eightfold higher in the wheat (7.0 μkat per kg of grain) than in the barley extracts (0.8 μkat per kg of grain).The other exoglycosidase activities, namely β-galactosidase, β-gluco sidase and β-xylosidase, were also up to sixfold higher in the wheat than in the barley extracts.The α-ara bi nofuranosidase activity was slightly higher in barley (2.2 μkat per kg of grain) than in wheat extracts (1.6 μkat per kg of grain).
The barley extract exhibits a higher endopeptidase (1.84) absorption change than the wheat extract (0.56).Exopeptidase activities of the chromogenic peptides used in this study were lower, ranging between 0.18 and 0.36 μkat per kg of grain in the barley extracts, and they were lower than the LOQ in the wheat extracts.
Lipase and esterase activities were low in both barley and wheat extracts.Wheat showed a lipase activity of 0.23 μkat per kg of grain aft er germination at 20 °C.Barley, which germinated at 6 °C, showed a 30-fold higher (6.0 μkat per kg of grain) enzyme activity than the wheat extract.

Hydrolase activity depending on the germination time
The maximum β-glucosidase activities are shown in Fig. 1.These maximum activities were achieved at a germination temperature of 7 °C for barley extracts (duration of 18 days) and 20 °C for wheat (duration of 7 days).The β-glucosidase activity of barley extracts increased about 150 % to 100 nkat per g of dm during germination for up to eight days.The β-glucosidase activity of wheat extracts increased 30-fold to about 600 nkat per g of dm within four days.The β-galactosidase activity in barley extracts was constant over the germination time with about 20 nkat per g of dm, while in wheat grains it increased about 15-fold until day four of the germination.There was also a fi vefold increase of lipase activity with a maximum of (6.6±0.2) nkat per g of dm in barley extracts aft er 14 days of germination.Lipase activity could not be detected in the extracts of germinated wheat.These results were used as a basis for the set-up of the design of experiments.

Interaction of hydrolases with the pretreatment, imbibition and germination
The experiments were carried out under the conditions tested previously, namely 16 h of imbibition and moisturising of the grains once a day during germination at a maximum temperature of 20 °C.The design of experiments of the hydrolytic enzymes was performed using pNPGlc (β-glucosidase), pNPGal (β-galactosidase) and 4-MUL (lipase) as substrates.Instead of the α-amylase activity, which has already been extensively examined, the three hydrolases (β-galactosidase, β-glucosidase and lipase) showing the highest enzyme activities in the wheat and barley extracts were chosen.
The β-glucosidases showed the highest activity in both barley and wheat extracts, followed by β-galac to sidase and lipase activity (Table 3; see Table 1 for processing conditions).
The maximum β-glucosidase activity was up to sixfold higher in wheat ((585±151) nkat per g of dm) than in barley extracts ((109±15) nkat per g of dm).It was obtained in wheat extracts germinated at 20 °C (pretreatment: -18 °C; imbibition: 37 °C).A maximum activity in barley extracts pretreated at 50 °C and subsequently imbibed at 4 °C was (109±15) nkat per g of dm, followed by germination at 6 °C.The β-galactosidase activity in barley extracts ranged between 22 and 34 nkat per g of dm.Wheat β-galactosidase activity had a maximum of (63±26) nkat per g of dm and a minimum of (7.1±0.3)nkat per g of dm.
Most of the glycosidase activities (pNPGal and pNP-Glc) in wheat extracts increased with higher germination temperatures within all experiments.
The maximal lipase activity (4-MUL) in barley extracts ranged between 2.3 and 6.7 nkat per g of dm.A germination temperature of 16 °C led to the highest lipase activity (barley, experiment no.5: 6.7 nkat per g of dm).The lipase activity in the wheat extract (0.5-4.6 nkat per g of dm) was lower than in the extract from barley.In contrast to the glycosidases, wheat grains exhibit lipase activity when germinated at low temperatures, such as 6 °C (experiment nos. 1 to 4), but very low activity at higher temperatures (16 and 20 °C; experiment nos. 5 to 9).The screening of the hydrolytic activities (Table 2) was made fi rst.Based on the results, the hydrolytic activities were examined with a design of experiments (Table 3).In order to illustratively show the infl uence of temper-ature, the β-glucosidase, β-galactosidase and lipase activities were compared.The screening experiments in barley extracts were carried out with pretreatment, imbibition and germination temperatures of 37, 4 and 7 °C, respectively.The design of experiment no. 4 was used for comparison, so that the pretreatment temperature was diff erent (50 and 37 °C).Only the β-galactosidase and lipase activities showed slightly signifi cant diff erences.The pretreatment, imbibition and germination in the screening experiment of wheat extracts were carried out at temperatures of -18, 25 and 20 °C, respectively, while experiment no.6 was used for comparison, also using diff erent imbibition temperatures (37 and 25 °C).Only the β-glucosidase activity was slightly, but signifi cantly, diff erent.Other experiments carried out at only one of the three temperatures showed similar results (data not shown).Therefore, no changes or only slight changes were observed with only one variation at one temperature for both comparisons used illustratively.This suggests that not only one temperature, but also the temperatures of at least two parameters have a strong infl uence on the hydrolase activities.

Calculation of the β-glucosidase activity
The β-glucosidase activity was considered as a marker activity for the calculation of the infl uence of temperature on the enzyme activity of barley extracts (p=0.0002,R 2 =0.899, adjusted R 2 =0.834) and wheat extracts (p=0.003,R 2 =0.942, adjusted R 2 =0.897).The following equations (Eq. 1 for barley and Eq. 2 for wheat extracts) exhibit a three--factor interaction (each regression coeffi cient was significant), which means that the enzyme yield depends on the cumulative eff ect of the temperatures from each treatment step.
The estimated parameters for maximum β-glu co si dase activity of barley extracts were pretreatment at 40 °C followed by imbibition at 4 °C and germination at 7 °C, and of wheat extracts at -18, 25 and 20 °C, respectively.These experiments with the predicted parameters were carried out in order to prove the predictions and were not part of the experiments for the calculation.The β-glucosidase activity in barley extracts of (107.9±4.5)nkat per g of dm verifi ed the predicted value of 106.7 nkat per g of dm.The observation of (634±68) nkat per g of dm also met the prediction (587 nkat per g of dm) of the calculation for the wheat β-glucosidase.Therefore, the calculation is able to simulate the expected β-glucosidase activity in the range of the investigated parameters.3D surfaces were calculated to visualise the infl uence of temperature on diff erent treatment steps for β-glu cosidase activity in barley and wheat extracts.The pretreatment temperature in the 3D surfaces was set at 40 °C (Fig. 2).
The surfaces showed major diff erences between the temperature infl uence on β-glucosidase activity in barley and wheat extracts.The simulation showed that the β-glucosidase activity in barley extract decreased with the increase of imbibition temperature.At a germination temperature of 6 °C, the β-glucosidase activity was about 60 % lower at an imbibition temperature of 40 °C than at 4 °C.With increasing germination temperature from 6 to 20 °C, the β-glucosidase activity increased at an imbibition temperature of 40 °C, but decreased at an imbibition temperature of 4 °C.However, β-glucosidase activity at a germination temperature of 20 °C was higher at an imbibition temperature of 4 °C than at 40 °C.In barley extracts, the combination of a low imbibition temperature followed by a low germination temperature at the border of the experimental set-up led to the highest calculated β-glu cosidase activity.
The infl uence of imbibition temperature on the β-glucosidase activity was higher in barley than in wheat extract.By contrast, the germination temperature was the determining factor for wheat β-glucosidase activity.The β-glucosidase activity in germinating wheat grains increased more than threefold with the germination temperature.A trend of low imbibition temperature followed by high germination temperature can be derived from the 3D surfaces for high β-glucosidase activity in wheat extracts.

Comparison of hydrolase activities in the extracts of barley and wheat
Some hydrolase activities of barley and wheat extracts had been examined previously.The α-amylase activity in the barley extract (118 273 μkat per kg of grain) in this study was higher than the maximum activity (about 5500 μkat per kg of malt, calculated from the original data) measured in eleven barley varieties (24).The α-amylase activity in barley extracts in this study was twelvefold higher than in wheat extracts (10 000 μkat per kg of grain).Other authors obtained a maximum α-amylase activity of about 38 700 μkat per kg (calculated from the original data) in germinated wheat grains for six days, without specifying the unit of mass used for the calculation (25).
The second most abundant enzyme examined in the extracts from germinated barley and wheat was β-glucosidase.The crude extract of wheat had a more than fourfold higher β-glucosidase activity (400 μkat per kg of grain) than barley (102 μkat per kg of grain).The crude extract from the shoots of germinated wheat had a β-glucosidase activity of about 1200 μkat per kg of grain (estimation from the original data) in the whole germinated grain (26).The β-glucosidases from wheat shown in this study had roughly half the specifi c activities compared with microbial β-glucosidases from, for example, Lactobacillus brevis SK 3 or Penicillium purpurogenum KJS506 with 700-2000 μkat per kg of cell dm, respectively (27,28).
In barley extracts, endoxylanase activity expressed as absorption change (1.617 per h and per g of dm at 590 nm), activities of α-arabinofuranosidases (2.64 μkat per kg, calculated from original data) and xylosidase (6.27 μkat per kg, calculated from original data) were found (29).Compared to that, barley extracts in this study showed a 33-fold higher endoxylanase absorption change recalculated per h and per g of dm (53.7) and about the same α-arabinofuranosidase activity (2.2 μkat per kg of grain), whereas the xylosidase activity (1.6 μkat per kg of grain) was 70 % lower.
The endopeptidase activity expressed as absorption change towards azo-casein per s and 1 kg of grain was higher in barley extracts (1.84) than in wheat extracts (0.56).These results were comparable to the values of a previous study with 0.01 absorption change per h, although it is not clear which amount of germinated grains was necessary to obtain these results (30).
Most studies concerning cereal lipases are in the context of rancidifi cation of fl our during storage (31), therefore, no comparison can be performed with this study.

Infl uence of the process conditions on the hydrolase activity
It has been shown that a pretreatment of sunfl ower seeds, which are orthodox seeds (water to mass ratio ζ<0.15), with a temperature of 45 °C prior to germination  (32), thus to reduction of seed germination.This eff ect is amplifi ed if the grains are exposed to high temperatures in combination with high relative humidity.Wheat grains also showed viability loss in this study when pretreated at 50 °C (experiments nos. 2 and 4), which resulted in lower enzyme activities in comparison with the pretreatment at -18 °C (experiments nos. 1 and 3).However, the viability loss could only be observed when the grains were germinated at 6 °C.The enzyme extracts germinated at 20 °C have similar enzyme activities, independent of the pretreatment.Hence, the infl uence of the germination temperature is most important for the enzyme yield in wheat.
Barley seems to be resistant to high pretreatment temperatures, which could be due to the grain structure.Contrary to wheat grains, barley seeds are encapsulated by husks, which could have this protective eff ect on the grains.
The imbibition temperature seems to have the highest impact on the enzyme activity in barley extract.The β-glucosidase activity is almost twofold higher if imbibed at 4 °C compared to the imbibition at 37 °C under the same pretreatment and germination conditions.The β-glucosidase activities aft er germination at high (20 °C) and low (6 °C) temperatures are comparable considering the standard deviation.Although wheat and barley are closely related (33), there are major diff erences in the infl uence of the cereal processing conditions, as shown in this study.As a further step, a response surface methodology should be carried out, as shown for rice previously (34), in order to overcome the limitations of the factorial design (35).

Conclusion
To the best of our knowledge, the infl uence of temperature of diff erent cereal treatment steps on the hydrolase activities of barley and wheat grain extracts has so far not been investigated systematically.The calculation of the design of experiments showed, for example, that the maximum activity of β-glucosidase was reached with low imbibition and low germination temperatures in barley and with low imbibition and high germination temperatures in wheat extracts.However, with only one cultivar of barley and wheat used, more cultivars must be examined in order to check if these results are of general validity.The specifi c yield of the β-glucosidase activity in the extracts of barley and wheat grains was up to half of the specifi c yield of the β-glucosidase activity in the cultivation of the microorganisms Lactobacillus brevis SK 3 and Penicillium purpurogenum KJS506.
As a proof of principle, β-glucosidase could be used for food processes, which are of possible interest for consumers who desire 'green labelled' produced foods.

Fig. 2 .
Fig. 2. 3D surfaces of the β-glucosidase activity from: a) barley and b) wheat depending on the imbibition and germination tem perature with the pretreatment temperature set to 40 °C.dm=dry mass

Table 1 .
Temperatures of the nine experiments of the two-level factorial design for barley and wheat corner point, b centre point sured at 405 nm (5 min, intervals: 20 s).One katal of activity is defi ned as the amount of enzyme that releases one mole of p-nitroaniline per second. a

Table 2 .
Hydrolase activity found in the extracts from barley (pretreatment: 37 °C, imbibition: 4 °C, germination for 12 days: 7 °C) and wheat (pretreatment: -18 °C, imbibition: 25 °C, germination for fi ve days: 20 °C) Results represent the mean value±standard deviation of two independent measurements LOQ=limit of quantifi cation, n.d.=not determined.Absorption change is defi ned as the increase of A at 595 nm (for azo-casein at 450 nm) per s per 1 kg of grain dry mass showed higher endoglucanase (1.16) and galactomannase (0.52) absorption change than wheat extracts, in which both absorption changes were smaller than the LOQ.

Table 3 .
Hydrolase activities (β-glucosidase, β-galactosidase and lipase) in barley and wheat from the nine experiments of the two--level factorial design *see Table1for processing conditions of the respective experiment number, **single measurement; LOQ=limit of quantifi cation, dm=dry mass