The endoproteinases of barley and malt and their endogenous inhibitors

Introduction

During the malting and brewing processes, a portion of the barley proteins must be degraded to amino acids and small peptides. Among other problems, if too little protein hydrolysis occurs there will be insufficient low molecular weight nitrogenous compounds in the wort for optimal yeast nourishment. Alternatively, too much hydrolysis will deplete the wort of proteins that are necessary for good beer foam formation, mouth feel, etc. The endoproteinases of barley and malt are the enzymes that initially catalyze the hydrolysis of the insoluble barley storage proteins, and thus play major roles in determining whether or not a given barley variety will be useful for malting and brewing. It is imperative that we understand how, when, and why these enzymes function if new barleys with improved malting quality are to be developed or if malting or brewing methods are to be efficiently altered to produce worts that have improved soluble/ insoluble protein levels. The mixture of proteins, peptides and amino acids that ends up in a wort due to these endoproteolytic activities is termed its ‘soluble protein’ or ‘SP’ level.

To understand the protein hydrolysis that occurs during malting and mashing it is necessary to study not only the endoproteinases but also groups of proteins, called proteinase inhibitors, that interact with the proteinases and control their activities. Ungerminated barley contains only relatively small amounts of both endoproteinase enzymes and their protease inhibitors. During the germination process the proteolytic activities increase many fold and the endogenous proteinase inhibitor content increases to a lesser extent.

Barley malt contains multiple representatives of each of the four commonly occurring proteinase types, i.e. those belonging to the aspartic, cysteine, serine and metalloproteinase classes. Early studies on barley and malt indicated that the cysteine and metalloproteinases were probably the main ones responsible for hydrolyzing protein during seed germination and that the grain probably contained endogenous inhibitors that could inhibit the activities of members of both of these enzyme classes. The members of each of the four proteinase types can be specifically inhibited by one of four chemicals. These chemical inhibitors and their target proteases are: E-64 (cysteine proteinases), o-phenanthroline or o-phen (metalloproteinases), pepstatin A (aspartic proteases) and PMSF (serine enzymes).

Results and Discussion

Enzyme purifications and analyses
Several laboratories have purified and characterized different barley and malt endoproteinases, as listed in Table 1. Of these purified enzymes, some of the cysteine and metalloproteinases were able to hydrolyze hordein protein preparations. The hordeins are the main storage proteins of barley and are presumably the major proteins that need to be hydrolyzed during seed germination and malt mashing. None of the purified aspartic or serine class enzymes hydrolyzed these storage proteins. Using the substrates gelatin and edestin, over 40 proteinase activities were detected using 2-D IEF x PAGE gels.

During these and other purification and characterization studies, nearly all of the enzyme activity analyses were carried out at abnormally low pH values and in the presence of added reducing agents, because these conditions usually yielded maximal activity values. It is now known that these low pH and reducing conditions lead to artificially high cysteine and aspartic proteinase activities and underestimate the serine and metalloproteinase activities. In addition, many of the activity measurements did not yield true initial enzyme reaction rates and often the proteins that were used as substrates were not readily hydrolyzed by members of all of the proteinase classes. For these reasons, many of the endoproteinase activity measurements that have been made with both crude and purified barley/malt endoproteinases have not really been very relevant to what actually occurs during the malting of barley grains and in brewery mashes.


Table 1. Barley/Malt enzymes that have been purified

Enzyme class	# purified, (# distinct enzymes)1	hydrolyze hordeins?
Cysteine	5, (3)	Yes
Aspartic	1, plus 4 processing variants	No
Serine	2, (2)	No
Metallo	A group, 3 major and 6 minor forms	Yes

¹Some enzymes were purified by multiple researchers

The effects of pH and redox agents on malt proteinases
After it was discovered that the pH values and redox states of mashes strongly affected their proteolytic activities, quantitative studies were carried out to measure how these variables affected the protein solubilization that occurred during mashing (1). Mashing is carried out at about pH 5.9 in North America and the pH inside germinating barley grains is around 4.8, so experimental mashes were carried out at pH values that varied from 5.0 to 6.6. Over this pH range the mash proteolytic activity varied by over 7 fold and the SP levels of the final worts ranged from 4.8% to 7.0%, with the pH 5.9 value being 5.7%, which is a normal value for the varieties tested. This demonstrated several things; 1) that the SP level of a wort can be strongly and easily varied by adjusting the pH of its mash, 2) that any extract or mash proteinase activity measurements that are made at pH values below 5.9 are not relevant to what really happens in a mash and 3) that the rate of protein hydrolysis during malting, at pH 4.8, is probably much slower than what it is during mashing.

The addition of cysteine, a weak reducing agent, to mashes increased their proteolysis rates by over 3 fold and their wort SP levels were raised from 5.5% to 7.3% (1). These effects were negated when oxidizing agents were added to the mashes together with the cysteine, and similar effects were found when stronger reducing agents were added to mashes. This shows that, as with pH adjustments, the presence of redox agents in mashes can strongly shift the SP levels of their worts. Many of the previously measured proteolysis rates, determined in the presence of reducing agents, were probably incorrect. It has been proposed that redox reactions may naturally occur in seeds during germination. If so, that would probably influence the rate at which protein solubilization occurs during both malting and mashing.

How the various malt proteinase classes affect wort soluble protein levels
Based on the early proteinase activity measurements, the dogma had become accepted that the great majority of the protein solubilization that occurred during malting and mashing was due to the cysteine class proteinase activities, with possibly some contribution from the aspartic and metalloproteinases. When the effect of pH and reducing agents on soluble protein levels was noticed, however, it seemed possible that this perceived preponderance of cysteine proteinase activity was an artifact, because these enzymes are the ones that would have been most strongly activated by both the low pH and strongly reducing conditions. The serine and metalloproteinase activities would actually have been reduced under the low pH conditions.

When is the soluble protein of worts released?
In order to ascertain the relative contributions of the malting and mashing steps to the release of SP to worts, barleys and malts were extracted and mashed in the presence and absence of chemical proteinase inhibitors (2). With both Morex (6-rowed) and Harrington (2-rowed) barleys about 32% of the wort SP was already soluble in ungerminated barley grains, 46% was solubilized during malting and the final 22% was released during mashing at pH 6.0. This indicates that while the majority of the protein hydrolysis occurred during malting, almost a quarter of the wort SP was solubilized during mashing, when conditions for altering the proteolytic activities that control its release can be readily regulated.

The contributions of the various endoproteinase classes to wort soluble protein levels
To determine what portion of the final wort SP content was contributed by each of the four protease classes, mashes were carried out in the presence of each of the four class-specific chemical inhibitors (2). The results are shown in Table 2. These show that, under these standard mashing conditions, the cysteine class proteinases are not the only ones that release SP into the worts. Because of the complexity of these experiments there was substantial variation in the results obtained, but it is obvious that the metalloproteinases released as much SP as the cysteine proteinases, that the aspartic enzymes released SP, but at a slower rate, and that the serine proteases hydrolyzed little or no protein.

Table 2. The inhibition of soluble protein formation by class-specific inhibitors during mashing

Protease class	% inhibition1				% inhibition, ASBC mash2
	Morex	Harrington	Average		Morex	Harrington
Cysteine3	12	12	12		12	11
Aspartic	7	9	8		5	6
Serine	1	4	3		0	3
Metallo	9	14	12		13	16

¹ Average of mashing with 3 malt concentrations.
² Average of 3 experiments.
³ Inhibitors were, respectively, E-64, pepstatin A, PMSF and o-phen.

These findings correlate fairly well with the results obtained by the researchers who purified and characterized the various barley/malt proteinases. The two enzyme classes whose purified members hydrolyzed barley storage proteins, the cysteine and metalloproteinases, apparently catalyze the majority of the hydrolysis of the storage proteins into SP and the serine proteinases, neither of whose purified forms hydrolyzed hordein preparations, also did not release SP during mashing. On the other hand, even though the one aspartic proteinase that has been studied in depth has only been shown to hydrolyze one non-plant protein, enzymes of this class apparently do hydrolyze proteins during mashing, but at a slower rate than either the cysteine or metalloproteinases. This indicates that there are probably other, still unpurified, aspartic class proteases in malt that carry out this hydrolysis. This hypothesis is strengthened by the fact that electrophoretic studies have demonstrated that several aspartic class proteinases occur in malt. These SP-releasing aspartic protease forms still need to be purified and characterized.

From these recent findings, as well as from the 1970 report that showed that malt contained substantial levels of metalloproteinases, it is obvious that the question of how proteins are solubilized during mashing needs to be reconsidered. The large apparent contribution of the metalloproteinases to the formation of SP shows that these enzymes need to be studied in detail. To date, only a single in-depth study of the barley/malt metalloproteinases has been carried out and very few metalloproteases from any plants have been studied.

Barley and malt proteins that inhibit their endogenous endoproteinases
In the early 1960s it was noted that the addition of unmalted cereal flours to mashes led to worts that would not ferment. This problem was traced to the fact that the worts did not contain enough low molecular weight nitrogenous compounds to support good yeast growth. This low wort nitrogen problem was in turn traced back to the fact that ungerminated wheat and barley both contained compounds that inhibited the abilities of certain of the malt endoproteinases to hydrolyze storage proteins into SP. It was eventually ascertained that these endogenous proteinase inhibitors interfered with the activities of the cysteine class and metalloproteinases.

Metalloproteinase inhibitors. Because the significance of the contribution of the metalloproteinases to SP formation during mashing has only recently been demonstrated, the metalloproteinase inhibitors have scarcely been studied. However, in view of the facts discussed above, which show that these enzymes probably play a major part in SP production, both these proteinases and their endogenous inhibitors obviously deserve to be studied in detail. The malt metalloproteinases have proven to be particularly recalcitrant to purification and characterization and one reason for this could be that they bind to their endogenous inhibitors as soon as extracts are prepared, and are thus rendered inactive.

Cysteine protease inhibitors. Conversely, the inhibitors of the cysteine proteases, the other main enzymes responsible for SP formation during mashing, have been studied in depth and two of them have been purified, identified and characterized in detail (2). Both are proteins that belong to the lipid transfer protein (LTP) family. The major form is apparently a form of LTP1 that has been slightly modified and the other is LTP2. Two other inhibitory fractions have been isolated from barley malt and they also appear to be contain altered forms of LTP1. The amount of the inhibitory LTP1 increases about 3 fold during malting and the protein binds strongly to the cysteine endoproteinases, whose concentrations increase even more strikingly during malting. The LTP1-enzyme complex is readily broken when heated to 100^oC, upon which the enzyme is inactivated and precipitated. The heating does not, however, affect the inhibitory LTP1 molecule, and this characteristic has been used to develop an LTP1-enzyme ‘affinity’ method for concentrating and partially purifying the endogenous inhibitors of the cysteine and serine proteinases of malt. It has not been possible to dissociate the LTP1- cysteine endoproteinase complex without inactivating the enzymes that are involved.

When added to mashes the purified LTP1 inhibitor strongly inhibited their endoproteolytic activities and lowered the SP content of the resulting worts, so increasing or reducing its concentration in mashes could provide a ‘natural’ method for altering the SP contents of worts. It is not known whether or not the LTP molecules interact with the endoproteinases inside germinating barley seeds. If so, then altering the natural concentrations of these molecules would presumably also affect the amount of SP that is released during malting. Because the cysteine proteases are very active in producing SP, adjusting the levels of their natural inhibitors should have a significant effect on wort compositions.

Serine proteinase inhibitors. It has been known for many years that barleys and malts both contained a family of proteins that have been called the ‘chloroform-methanol’, or ‘CM’, proteins. Because one of these proteins clearly acted as an inhibitor of the bovine serine proteinase called trypsin, these proteins have also been called ‘trypsin/a-amylase inhibitors’ and it had been proposed that some of them might inhibit the activities of barley serine proteinases. However, none had been shown to affect the barley enzymes. In my laboratory we purified a barley serine proteinase called SEP-1 and showed that barley contained proteins that inhibited its activity. Using the proteinase-inhibitor affinity method discussed above, we purified, isolated and characterized four of the most effective SEP-1 inhibitors and showed that they were all members of the CM protein family. The barley contained additional SEP-1 inhibitors that were apparently less potent and these have not yet been studied.

Because the serine endoproteinases do not appear to directly contribute to the hydrolysis of proteins during mashing, these inhibitors would presumably not affect the SP content of mashes. However, they could affect wort compositions by indirectly controlling the proteolysis that occurs during the germination/malting process, since their true functions in the grain are still unknown.

Summary

Due to the use of inaccurate analytical methods, most of the past dogma that related to the solubilization of proteins during malting and mashing is probably incorrect. The cysteine proteinases do not uniquely control the formation of SP, but share this duty with the aspartic and, to an even greater extent, the metalloproteinases. A good, inclusive study of the malt metalloproteinases needs to be carried out, including a determination of what the metalloproteinase inhibiting compounds of malt are that were reported many years ago. In addition, it always needs to be remembered that it is not simply the presence of the barley/malt endoproteinases that controls the amount, and probably the type, of SP that occurs in worts. By controlling the activities of these enzymes, the various proteinaceous endogenous inhibitors could, and probably do, determine what proteins are really hydrolyzed during mashing and, possibly, during malting.

This report is a greatly condensed version of two papers that are in press/submitted to J. Cereal Sci. (3,4) and any interested parties should consult them when they are published.

References

(1) Jones, B.L. and Budde, A.D. The effect of reducing and oxidizing agents and pH on malt endoproteolytic activities and on malt mashes. J. Agric. Food Chem. 2003, 51, 7504-7512.

(2) Jones, B.L. and Budde, A.D. How various malt endoproteinase classes affect wort soluble protein levels. J. Cereal Sci. 2005, 41, 95-106.

(3) Jones, B.L. Endoproteases of barley and malt. J. Cereal Sci. 2005, In Press.

(4) Jones, B.L. The endogenous endoproteinase inhibitors of barley and malt. J. Cereal Sci. 2005, Submitted.

Berne L. Jones
RR1, Box 6, Kooskia, Idaho
Presented at the 18th North American Barley Researchers Workshop, July 17-20, 2005