4.1 Materials and methods
Barley, wheat and faba bean samples were studied. All sample material was provided by Viking Malt Oy, Lahti Finland. Samples were received in June 2019 and studied during a period from October 2019 to January 2020. Cereal samples were provided in the form of intact grains, whereas faba bean samples were provided in the form of flour.
Cereal grains were studied in native and malted form. Faba bean samples contained native and germinated versions. Barley malt was a commercial Pilsner malt from Viking Malt. The germinated samples of faba bean and wheat were micromalted as 2 kg batches in a laboratory scale micromalting unit. Micromalting consisted of the steps of steeping, germinating, and kilning. To initiate the germination and increase the seed moisture the seeds were steeped using a three-step steeping program consisting of two wet steeps and one dry steeping. The overall duration of the steeping was 26 hours for faba bean and 31 hours for wheat. The germination was carried out for 72 h (faba bean) and 120 h (wheat). Germinated seeds were dried using a stepwise temperature increase to final 55°C temperature for faba bean and 83°C for wheat.
Sample preparations
Cereal samples were ground with OBH Nordica 6658 (50 Hz, 300 W, 800 ml) blender. Samples were ground with short 10 second pulses; the container was shaken manually 10 seconds and ground again.
The procedure was continued until homogenous consistency was achieved and visible improvement could not be further detected. The container was rinsed with tap water and dried thoroughly between every sample. Ground samples were transferred into 50 ml Falcon tubes and sealed with Parafilm to prevent any formation of moisture. Samples were stored at room temperature away from light.
In vitro digestion
The in vitro digestion was conducted according to a protocol developed for the Food Research group at University of Eastern Finland, which was adapted from the INFOGEST model (Brodkorb et al., 2019) and model from Gómez-Gallego et al. (2016). The schema of the protocol is presented in Figure 2.
The oral phase was simulated by crushing the sample as presented previously and mixing it with water. This study was limited to protein digestion and therefore salivary amylase was not used. Pepsin
from porcine gastric mucosa (Sigma-Aldrich) was used for gastric digestion and Pancreatin, USP (MP Biomedicals) for intestinal digestion. Ox bile (Sigma-Aldrich) was used for pancreatin-bile-solution. The enzymatic activity was set to the volume of the chyme and was 2000 U/ml for pepsin and approximately 100 U/ml for pancreatic protease. Before the actual digestions, the pH adjustment was tested to evaluate the amount of required adjustment solutions. The working conditions and procedures were kept as sterile as possible. The sterility of the enzymes was tested with Plate Count Agar culturing and bacterial growth was not detected even after several days of incubation at 37℃.
The digestion was performed as follows. 20.0 g of sample was measured into a 150 ml Erlenmeyer flask. Milli-Q water and enzyme solutions were warmed to 37℃ before adding. 20 ml of Milli-Q water was added to the sample and stirred slightly with a spatula to ensure even absorption. After that, 40 ml of pepsin solution was added. The sample was stirred with magnetic stirrer while pH was adjusted to 3. The pH was adjusted with 1 N HCl and 1 1 M NaHCO3 solutions. Samples were then incubated 2 hours at 37℃ with 150 rpm shaking. After the gastric incubation, small samples were collected to 1.5 ml Eppendorf tubes and heated at 95℃ fir 15 minutes. Then 80 ml of pancreatin-bile-solution was added. PH was adjusted to 7 and samples were incubated at the same conditions as previously. After the incubation, samples were placed into a water bath and heated at 95℃ for 15 min. Samples were then transferred into 50 ml Falcon tubes and frozen to -20℃. Each sample was digested in 3 replicates.
Digested samples were dried on a petri dishes at 70℃ until the reduction of weight could not be detected. Dried samples were used for IVPD analyses and TNBSA assay. Samples used in the SDS-PAGE analyses were not dried.
SDS-PAGE analysis
The degree of protein fragmentation was investigated utilizing sodium dodecyl sulfate–
polyacrylamide gel electrophoresis (SDS-PAGE) with Bio-Rad Mini-PROTEAN® 3 equipment. The analyses were conducted according to a protocol developed for the Food Research group at University of Eastern Finland, which was based on manufacturer’s instructions (Bio-Rad Laboratories Inc., 2020). Reagents used in the gel casting are presented in Table 13. After the gel casting, 50 µl of 2-mercaptoethanol (ME) was mixed with 950 ml of 2 X sample buffer. 2 X sample buffer (Laemmli) contained 3.55 ml of Milli-Q water, 1.25 ml of upper gel buffer, 2.5 ml glycerol, 10 % sodium dodecyl sulfate (SDS) and 0.2 ml 0.5 % bromophenol blue dye solution. 7.5 ml of sample buffer/ME-solution and 7.5 ml of protein sample were mixed in a 1.5 ml Eppendorf tube, spun briefly and then placed on a heat block at 95℃ for 4 min. Cooled tubes were spun briefly again. Samples were diluted into Milli-Figure 2. Schema of in vitro digestion.
Q water to reach a protein concentration of 3000 µl/ml based on the protein content determined by Kjeldahl method, resulting approximately 22.5 µg of protein in the wells. Each product was studied with two replicates of three different batches after both gastric and intestinal phases.
Table 13. SDS-PAGE reagents.
Reagent Lower gel* Upper gel*
30% bis-acrylamide (ml) 4.7 0.65
Buffer (ml)
• Lower gel: 1.5 M Tris. pH 8.8
• Upper gel: 0.5 M Tris. pH 6.8
2.5 1.25
10% sodium dodecyl sulfate (µl) 100 50
Milli-Q water (ml) 2.7 3.05
10% ammonium persulfate (µl) 50 25
TEMED (µl) 5 10
*The amounts are sufficient for two 0,75 mm gels.
TEMED: Tetramethylethylenediamine; Tris: tris(hydroxymethyl)aminomethane.
The gels were then placed in the mini tank and the equipment was assembled according to protocol.
The chambers were filled with 1 X running buffer. 5 X running buffer consisted of 15 g of tris(hydroxymethyl)aminomethane (Tris), 72 g of glycine and 5 g of SDS diluted into 1000 ml of Milli-Q water. 1 X buffer was prepared by diluting the 5 X buffer with Milli-Q water in relation of 1:5. 3 µl of Bio-Rad Precision Plus Protein™ Dual Color Standards was used as a molecular weight marker. After the pipetting of samples, the electrophoresis was run for approximately 1 h with 150 V voltage. The gels were then stained with Coomassie blue dye solution for 15 min and washed for 2 h with a solution containing 50 % methanol and 10 % acetic acid. The washing solution was changed at least twice during the washing.
Crude protein content and in vitro protein digestibility
The protein content of the products was determined by the Kjeldahl method. The conversion factors were selected according to Jones (1931): 5.83 for wheat and barley and 6.25 for faba bean.
The TCA-precipitation of protein was conducted according to Pérez-Conesa et al. (2005) and Awolumate (1983) with some modifications. 500 mg of each sample was measured into 50 ml open top polypropylene centrifuge tubes, 13.3 ml of 0.2% NaOH was added and tubes were placed on an orbital shaker for 15 min at 150 rpm. Tubes were then centrifuged with Sorwall RC 6 plus at 6000g for 5 min at 20℃ and the supernatant was collected. The pellet was washed with 10.7 ml of 0.2 % NaOH, shaken and centrifuged at the same conditions. The supernatant was added to the previous one and pellet was discarded. 16 ml of 25 % TCA was added to the collected supernatants and tubes were placed in a refrigerator at 4℃ for 2 hours and shaken on an orbital shaker every 15 minutes.
After that samples were centrifuged at 12000 g for 20 min at 20 ℃. The supernatant was discarded, and the pellet was transferred with Milli-Q water into 50 ml Falcon tubes for weighing. The Falcon tubes were centrifuged at 3000g for 5 min at 20℃ with Jouan BR4 i centrifuge before drying in an oven at 65℃ until completely dry. The empty Falcon tubes were weighed beforehand, and the amount of dried precipitate was calculated as the difference of a tube containing the sample and an empty tube. The precipitates were transferred to Kjeldahl burning tubes with Milli-Q water and the amount of nitrogen was measured by the Kjeldahl method. Precipitates from replicates were combined to ensure a detectable amount of nitrogen.
True protein nitrogen (TPN) was acquired from the precipitates and non-protein nitrogen (NPN) was calculated by subtracting TPN from total nitrogen acquired before. NPN was determined before and after the in vitro digestion and in vitro protein digestibility (%) was calculated according to the following equation:
IVPD (%) = NPN1−NPN0
TPN
x 100
, where NPN0 means the NPN before and NPN1 after digestion.
Determination of free amino acids
The amount of free amino acids in the samples were determined with 2,4,6-Trinitrobenzene Sulfonic Acid assay according to the instructions (Thermo Fisher Scientific Inc, 2012). The samples were diluted into 0.1 M NaHCO3 to achieve concentration of 20–200 µg/ml. Different concentrations of glycine solution were used to form a standard curve. The standard curve was formed with Microsoft
Excel version 16 by Microsoft Corporation. The amount of free amino acids was calculated from the standard curve using linear trend from the mean value of two or three replicates per sample.
Statistical analyses
Statistical analysis was not conducted due to low number of samples.
4.2 Results
SDS-PAGE
The results from SDS-PAGE analyses are presented in Figures 3–7. Malting does not appear to have major effects on barley protein molecular weight. Clear large molecular weight bands cannot be detected in undigested samples, but an indistinct area of protein is present above 75 kDa as Figure 4 indicates. In digested samples, bands larger than 20 kDa cannot be detected except for one faint band between 37 and 50 kDa size after intestinal digestion in malted barley. First two analyses of wheat showed only faint bands but after another rerun bands were detected. Nevertheless, there is no clear distinction between raw and malted samples. Germination does not appear to affect faba bean either, but some protein fragments resistant to digestion can be detected. Strong bands are detected before digestion, but due to heavy aggregation, these bands cannot be detected after gastric digestion. Yet, some of the bands are clearly visible after intestinal digestion. Two bands are visible at the proximity of size 37 kDa and three bands around 20 and 25 kDa size. Bands longer than 37 kDa have disappeared after intestinal digestion.
Figure 3. SDS-PAGE analysis of barley. A: Raw, undigested; B; malted, undigested; 1: Raw after gastric digestion; 2: Malted after gastric digestion; 3: Raw after intestinal digestion; 4: Malted after intestinal digestion. Parallel wells under the same number are from different digestion batches.
Figure 4. Upper part of the SDS-PAGE analysis of barley before digestion.
Figure 5. SDS-PAGE analysis of wheat. A: Raw, undigested; B; malted, undigested; 1: Raw after gastric digestion; 2: Malted after gastric digestion; 3: Raw after intestinal digestion; 4: Malted after intestinal digestion. Parallel wells under the same number are from different digestion batches
Figure 6. SDS-PAGE analysis of faba bean. A: Raw, undigested; B; Germinated, undigested; 1: Raw after gastric digestion; 2: Germinated after gastric digestion. Parallel wells under the same number are from different digestion batches
Figure 7. SDS-PAGE analysis of faba bean after intestinal digestion: 1: Raw; 2: Germinated. Parallel wells under the same number are from different digestion batches
Protein content and in vitro protein digestibility
The determination of protein and IVPD is presented in the Table 14. The crude protein content and IVPD of wheat and barley were slightly higher in malts. Native faba bean had slightly higher IVPD than germinated one, but lower protein content. Germinated/malted products had higher amount of NPN before digestion.
Table 14. Crude protein and in vitro protein digestibility of barley, wheat and faba bean.
Crude
NPN: non-protein nitrogen; IVPD: In vitro protein digestibility; GFB: Germinated faba bean; DM: Dry mass.
*Mean value (Standard deviation) of two measurements. The protein content of faba bean is from a single measurement.
Amount of free amino acids
The results from the TNBSA assay are presented in Table 15. The table indicates an increased amount of free amino acids after digestion. Malted/germinated products contain higher amount of free amino acids before and after digestion.
Table 15. Free amino acid content of the sample products before and after in vitro digestion.
Sample Amount of protein in undigested samples
(µg/ml)
Free amino acids in undigested samples
(µg/ml)*
Free amino acids in digested samples
(µg/ml)*
Raw barley 85 1.53 (1.45–1.60) 7.71 (7.40–8.71)
Malted barley 92 4.77 (4.50–5.03) 8.28 (8.16–8.55)
Raw wheat 61 0.08 (0.04–0.12) 5.06 (4.61–5.23)
Malted wheat 64 2.60 (2.05–3.15) 6.27 (5.83–6.74)
Faba bean 213 7.99 (7.37–8.62) 11.20 (11.16–11.30)
GFB 214 8.44 (8.28–8.60) 12.21 (12.09–12.33)
GFB: Germinated faba bean. *Median (Minimum–Maximum).
5 DISCUSSION
5.1 Setting the methodology
This study piloted in vitro digestion of cereals and legumes, and setting the methodology was a large part of the experimental work of this thesis. The in vitro digestion protocol used in this work was tested with milk-based products before, which is a much simpler matrix compared to cereal products and legumes. First, cereals and legumes are in solid form and different parts of the grain, such as embryo, aleurone, endosperm and husk, have different structure and chemical composition (Fox, 2010). Second, cereals and legumes contain components such as antinutritional factors and fiber, which may weaken the access of proteolytic enzymes to proteins (Joye, 2019). The access can be weakened by intact cellular structure and tight folding of certain proteins as well. Due to these facts, different behavior of the digestion was expected.
Factors affecting results of in vitro digestion
The digestion was conducted in as sterile conditions as possible to minimize the risk of contamination in the in vitro colon fermentation to be performed after digestion. All the glassware and other equipment were sterilized. Filtration of reagents was tested, but with unsatisfactory results, and thus sterility of reagents and solutions were monitored with Plate count agar culturing. No contaminations were observed from reagents. Heating after the digestion and quick freezing to -20℃ lowered the risk of contaminations and unwanted bacterial growth as well. Grain and legume samples were non-sterile and most of the steps of the digestion could not be performed in a sterile laminar flow cabinet and thus some possible sources of contamination were still present.
A blender was used to simulate the oral phase of digestion of wheat and barley. This method was not able to produce homogenous flour, and visible bits of different sizes were left in the samples. Malted grains tented to need less treatment and the outcome was finer compared to raw grains, even though raw grains were blended longer. Durum wheat particle size has been reported to affect IVPD (Mandalari et al., 2018). Therefore, the different consistency of malted and raw flours can cause an error in the results and an additional step of sieving could be considered. Sieving, however, can alter the composition of flour so that parts of the outer layer are discarded and thus causing error. Also, a grain mill would be better equipment for grinding the grains, especially if the interest is to use these grains in the form of flour. On the other hand, increased friability due to structural modifications is a well-known property of malted grains (MacLeod & Evans, 2016). In the case of eating the grains whole, the finer consistency of malts mimics that they are easier to chew than raw grains and thus, differences/error in the IVPD caused by particle size can be accepted.
From the practical point of view, large particles cause difficulties during the digestion and analyses.
First, larger particles block the tip of a pipette and pipetting small amounts is thus difficult. Second, heavier particles tend to descend and form a sediment in the bottom of the flask during the incubation in the shaker, which can reduce the access of enzymes to some parts of the sediment. On the other hand, large particles are more difficult to digest in the GI-tract, and therefore some level of sedimentation can be accepted to mimic this phenomenon. Third, some of the particles adhere to the walls of the flask, which limits the exposure to digestive enzymes during digestion and causes loss of the matter when the digesta is transferred to other containers. Occasional manual shaking during incubation can be used to minimize the error caused by the second and the third problem. Also, the residue can be rinsed with known amount of sterilized Milli-Q water.
The pH after adding of pepsin and pancreatin solutions varied a lot, even with samples from the same batch. Therefore, the adjustment of pH is a critical step when working with cereals and legumes.
Acidic conditions caused especially faba bean samples to foam and coagulate heavily and thus, neutralization should be conducted for the samples collected after gastric phase.
The addition of protein via enzymes to the samples were not considered in the results. The amount is the same in all the digested samples, so the possible error is the same as well. However, this issue could be at least partly corrected by measuring the amount of nitrogen in the enzymes and subtracting that from the amount of N received from protein precipitation with the assumption that enzymes are still intact. With the assumption that used enzymes are pure protein, the calculated addition of protein per digested sample is 0.56 g. This is in fact a quite high amount, considering that for example 20 g or barley contains 1.69 g of protein. Thus, the determination of N from enzymes and bile should be included in the protocol.
Considerations about methods of analysis
The TCA-precipitation method was selected as it has been used with both cereals (Pérez-Conesa et al., 2005) and legumes (Awolumate, 1983) and it was a relatively simple method. Other methods available can precisely measure every protein fraction with different solubilities (Bayram & Alameen, 2018; Branlard & Bancel, 2007), but these methods are more laborious and require more chemicals.
The protein solubility is the main aspect to consider in future studies using these methods. Differences in protein solubility could cause an error, as the insoluble protein is discarded with the first pellet in the TCA-precipitation method and thus interpreted as NPN. The protein solubility is problematic especially with barley, because relatively high proportion of the protein is in insoluble form and
different types of proteins have different solubilities (Arendt & Zannini, 2013). For example, Yalçın
& Çelik (2007) reported at best 60% protein solubility at pH 11 for hull-less barley flour.
IVPD might not be the optimal method to evaluate the benefits of germination because higher amount of total N is already in the form of NPN in germinated grains compared to raw. Consequently, raw grains may demonstrate a higher increase of NPN in relation to TPN and thus higher IVPD, even though the outcome of digestion would be similar. Additionally, changes in protein solubility might be one reason for increased IVPD in germinated grains. The TCA-precipitation method used in this study might not properly express this difference with species with low protein solubility, such as barley, because insolubilized fractions are interpreted as NPN before digestion and the differences between native and malted grains are diminished. The inclusion of other analysis methods such as TNBSA assay in this study are valuable to highlight this aspect of germinated grains.
Nevertheless, piloting current methodology can be considered a satisfactory, as published methods were used, and results were in line with previous literature. For further optimization of the methods used in this study, following aspects should be considered: 1) Particle size and how it affects the digestion and analyses 2) Access to required equipment, e.g. ultra-centrifuge, and suitable tubes for used reagents. 3) Optimization of pH during TCA-precipitation for each sample material separately
Nevertheless, piloting current methodology can be considered a satisfactory, as published methods were used, and results were in line with previous literature. For further optimization of the methods used in this study, following aspects should be considered: 1) Particle size and how it affects the digestion and analyses 2) Access to required equipment, e.g. ultra-centrifuge, and suitable tubes for used reagents. 3) Optimization of pH during TCA-precipitation for each sample material separately