| | Introduction | Materials and methods | Results and discussion | Acknowledgement | References
Introduction
The walls surrounding the cells of the starchy endosperm of barley must be effectively degraded during malting if problems with extract yield, wort and beer filtration, and beer clarity are to be avoided. While it is possible to select barley varieties for malting on the basis of low levels of beta-glucans, there is no clear relationship between beta-glucan content and malt quality. Although beta-glucans are the major constituents of the endosperm cell walls, other polysaccharides may also contribute to the overall quality of malting barley. The coexistence of several biopolymers in the cell walls, their spatial organization, and the nature of interactions (cross-linking) among them might contribute to the mechanical strength, permeability, and solubility, and therefore to enzymic susceptibility of cell walls during malting. The influence of composition and properties of the endosperm cell walls on kernel hardness has not been studied in detail, although some relationships between hardness and cell walls have been suggested.
The objectives of this study were (1) to examine compositional and structural differences in endosperm cell wall components derived from barley grains varying in hardness, protein and total beta-glucan contents, (2) to investigate the enzymic degradation of isolated barley endosperm cell walls, and (3) to determine how the differences in composition and morphology of the cell walls influence their solubility, susceptibility to enzymatic hydrolysis, and degradation patterns.
Materials and Methods
Three malting barley samples (cv. Metcalfe) were grown in 2003 in 3 different locations in Canada (A: Davidson, SK; B: Hythe, AB; C: Hamiota, MB). Grain hardness was measured with the SKCS 4100 (Perten Instruments Inc., IL). Endosperm cell walls were isolated from a fiber rich fraction obtained by roller milling of pearled barley, followed by pin milling and dry sieving. Wet sieving (with 1% sodium dodecyl sulfate in 70% ethanol), homogenization and sonication were used to purify the endosperm cell wall material (CWM). Monosaccharide and phenolic acid composition was determined by high-performance reverse phase and anion exchange chromatography (HPAEC), respectively (Izydorczyk et al. 1998 and Cyran et al. 2002).
The endosperm cell walls were sequentially extracted with water at 65oC (WE), saturated barium hydroxide (BaE), water (Ba/WE), and 1N sodium hydroxide (NaE) at 25oC. The residue remaining after the sequential extraction was designated RES. The fine structure of beta-glucan was investigated by lichenase digestion and HPAEC (Izydorczyk et al. 1998). Monosaccharide and glycosidic linkage composition was determined by HPLC and GC- MS (Izydorczyk et al., 1998), respectively. Samples were prepared for SEM by mounting them onto aluminum stubs covered with double-sided carbon adhesive discs and allowed to set for 24 h. The mounted samples were placed in a Hummer VII (Anatech, Ltd.) sputter coater and coated with 50 nm of gold and examined with a JEOL JSM-6400 SEM at 10 KV.
Results and Discussion
The endosperm cell walls were obtained from barley samples differing in grain hardness, protein and beta-glucan contents (Table 1). The general morphological features of the inner surface of the walls can be seen in Figure 1. The inner wall surface of sample A appeared deeply pitted with indentations made by small and large starch granules. These indentations were less pronounced in the walls of samples B, and C. Sample C also contained areas with an uneven and folded surface, possibly representing imprints of dense protein matrix rather than starch granules. The thickness of the cell walls ranged from 0.5 to 1.6, 0.5-1.7 and 0.8-2.3 µm for samples A, B, and C, respectively.
Table 1. Composition and hardness of barley grains
Sample | Protein % | Starch % | Beta-glucan
% | Soluble Beta-glucan % | Arabinoxylans % | Hardness index |
| A | 10.8 | 61.3 | 4.2 | 2.3 | 5.8 | 59.25 |
| B | 11.8 | 58.9 | 4.6 | 1.7 | 5.6 | 68.88 |
| C | 17.1 | 54.8 | 4.8 | 2.6 | 5.4 | 61.70 |
Figure 1. SEM photographs of cell walls isolated from Metcalfe grown in Davidson, SK (A), Hythe, AB (B), and Hamiota, MB (C).
The isolated cell walls contained very little starch (< 1 %) and proteins (~6%) and were built up mainly from glucose, xylose, arabinose, mannose and small amounts of galactose (Fig. 2). The intact endosperm cell walls of sample A contained the least amount of beta-glucans and the highest amount of arabinoxylans and mannose-containing polysaccharides. The walls of sample C contained the highest amount of beta-glucans, in agreement with the highest content of these polymers in the barley grain. The endosperm cell walls of sample A contained the highest amount of phenolic acids (ferulic, coumaric and diferulic), but the arabinoxylans in the walls of sample B were more cross-linked than those in samples A and C (Table 2).
The treatment of the CWM with water at 65oC solubilized mostly beta-glucans, whereas the treatment with saturated barium hydroxide extracted mostly arabinoxylans (Fig. 3). The extract obtained with water after the barium hydroxide treatment contained about 60% beta-glucans and ~35% arabinoxylans. The least soluble extract obtained with NaOH (NaE) and the residue remaining after all extractions (RES) contained approximately equal parts of beta-glucans, arabinoxylans and mannose-containing polysaccharides.
Table 2. Distribution of phenolic acids in CWM
| | A | B | C |
| Total phenolic acids, g/100g CWM | 0.34 | 0.24 | 0.17 |
| Total FAa, g / 100g CWM | 0.28 | 0.20 | 0.15 |
| Total DFAb, g / 100g CWM | 0.0096 | 0.0092 | 0.0055 |
| FA/Xyl x 10000c | 119 | 128 | 110 |
| (DFA/Xyl)x10000d | 2.0 | 3.0 | 2.0 |
aferulic acid; bdehydrodiferulic acid; cmoles of FA per 10,000 moles of Xyl; dmoles of DFA per 10,000 moles of xylose residues
Following the extraction of the endosperm cell walls with water, the surface indentations due to starch granules could no longer be seen. It appears that the water-soluble beta-glucans may be layered onto the surface of the walls rather than being distributed throughout the wall structure as suggested by Fincher (1975). Following the extraction of arabinoxylans from the water-extracted CWM the definition of endosperm cells disappeared. The NaOH extraction caused further pitting and corrosion of the wall material.
A detailed analysis of oligosaccharides released by lichenase digestion of beta-glucans revealed some differences in the structural features of these polymers among the samples. Beta-glucans in the cell walls of sample B had the highest ratio of 3-O- -D-cellobiosyl-D-glucose (DP3) to 3-O- -D-cellotriosyl-D-glucose (DP4) and a slightly lower level of oligosaccharides of DP 5-9, representing a more cellulose-like region of the beta-glucans. On the other hand, the DP of longer cellulosic fragments in this sample was higher than in sample A and C. Sample B had the lowest ratio of β-(1->4) to β-(1->3) linkages. Beta-glucans originating from sample C clearly had the greatest amount of glucose residues linked via β-(1->4) linkages, which was confirmed by the highest ratio of β-(1->4) to β-(1->3) linkages and the lowest DP3/DP4 ratio.
Table 3. Structural features of beta-glucans
| Sample | Ratio
DP3/DP4a | DP(5-9)a % | Longer cellulosic fragmentsb |
| A | 2.11 | 7.8 | DP 10-24 |
| B | 2.28 | 7.7 | DP 10-28 |
| C | 2.10 | 8.2 | DP 10-25 |
a found in water-soluble digests from lichenase treatment
b found in water-insoluble precipitates released after lichenase treatment
Monosaccharide and glycosidic linkage analyses revealed that arabinoxylans in the cell walls of sample A were less substituted than those in sample B and C (Table 4). Arabinoxylans in the cell walls of sample A had the highest amount of unsubstituted but the smallest amount of doubly substituted xylose residues. The least soluble polysaccharide of the endosperm cell walls, present in the NaE and remaining in the residue, differed substantially from those found in the WE, BaE, and BaWE (Table 5). The linkage analysis revealed the presence of lowly substituted arabinoxylans, beta-glucans with a high ratio of β-(1->4) to β-(1->3) linkages, and the presence of glucomannans and/or(galacto)glucomannans. The highest amounts of lowly substituted arabinoxylans and mannose-containing polysaccharides were found in sample A whereas the highest amount of beta-glucans with cellulose-like features was found in sample C.
Table 4. Structural features of arabinoxylans
Sample | Xyl/Ara | Unsub/Sub Xylb | Doubly/Sing Xylc |
A | 1.74 | 1.76 | 0.97 |
B | 1.54 | 1.46 | 1.28 |
C | 1.44 | 1.14 | 2.43 |
a Ratio of xylose to arabinose residues
b Ratio of unsubstituted xylose to substituted Xylp
c Ratio of doubly to singly substituted xylose residues
Table 5. Structural features of NaE fraction in CWM of various samples
| | A | B | C |
| Ratio Unsub/Sub Xylp | 2.54 | 1.98 | 1.40 |
| Ratio Xyl/Ara | 2.8 | 2.4 | 2.0 |
| ->4 Manp 1-> (%mol) | 32 | 20 | 12.5 |
| Ratio (1->4)/(1->3) Glc | 3.20 | 3.4 | 3.9 |
The enzymic degradation of endosperm cell walls was investigated by treating buffered suspensions of isolated cell wall fragments (previously extracted with water at 45oC) with malt extracts and determining the amount, monosaccharide composition and molecular size distribution of the soluble carbohydrate products. The water solubility of the CWM in 45oC ranged from 24% for sample B, through 27% for sample A, to 41% for sample C. Approximately 20% of the CWM, remaining after the initial water extraction, was solubilized with the malt enzymes. Substantially more beta-glucans than arabinoxylans or mannose-containing polysaccharides were solubilized with water at 45oC (Figure 4). Interestingly, the majority of mannose containing polymers was solubilized during digestion of CWM with the malt enzymes. Almost equal amounts of arabinoxylans (~10%) were solubilized with water and with the malt enzymes. Overall, the solubility and digestibility was lower for sample B than for A and C. Both high- and low-molecular weight (HMW and LMW) materials were released from the CWM during digestion with malt extracts. The monosaccharide analysis revealed the HMW malt digest contained mostly glucose, xylose and arabinose, whereas the LMW malt digest contained glucose and mannose. Differences in the molecular structure of beta glucans and arabinoxylans extracted with water compared to those solubilized with the malt enzymes were also observed.
Figure 4. Amount of solubilized carbohydrates during solubilization with water and digestion with malt extract of the CWM.
Acknowledegment
The financial support from NSERC is gratefully acknowledged.
References
Izydorczyk, M. S., Macri, L.J., MacGregor, A.W. (1998). Carbohydr. Polym., 35, 249-258.
Cyran, M., Izydorczyk, M. S., and MacGregor, A. W. (2002). Cereal Chem., 79, 359-366.
Fincher, G.B. (1975). J. Ins. Brew., 81, 116-122.
M.S. Izydorczyk* (1), A. Lazaridou (2), T. Chornick (1), L. Dushnicky (1)
*Corresponding author: mizydorczyk@grainscanada.gc.ca
(1) Grain Research Laboratory, CGC, Winnipeg, MB, Canada;
(2) Department of Food Science, University of Manitoba, Winnipeg, MB, Canada
Presented at the 18th North American Barley Researchers Workshop, July 17-20, 2005 |
|