University of California–Davis Davis, California, U.S.A ... · activities that have been detected in milk but which have not been isolated and have no known signiﬁcance in milk

Marcel Dekker, Inc. New York • Basel

University of California–DavisDavis, California, U.S.A.

Wageningen UniversityWageningen, The Netherlands

U.S. Department of AgricultureAlbany, California, U.S.A.

Copyright 2003 by Marcel Dekker, Inc. All Rights Reserved.

19

Significance of Indigenous Enzymes in Milk and Dairy Products

Patrick F. Fox

University College, Cork, Ireland

I. INTRODUCTION

About 60 indigenous enzymes have been reported innormal bovine milk (1); with the exception of �-lactal-bumin, most of these have no obvious physiologicalrole. The indigenous enzymes are constituents of themilk as excreted and arise from three principal sources:(a) the blood via defective mammary cell membranes;(b) secretory cell cytoplasm, some of which is occasion-ally entrapped within fat globules by the encircling fatglobule membrane (MFGM); and (c) the MFGMitself, the outer layers of which are derived from theapical membrane of the secretory cell, which in turnoriginates from the Golgi membranes; this is probablythe principal source of the enzymes. Thus, mostenzymes enter milk owing to peculiarities of themechanism by which milk constituents, especially thefat globules, are excreted from the secretory cells. Milkdoes not contain substrates for many of the enzymespresent, while others are inactive in milk owing tounsuitable environmental conditions such as pH.

Many indigenous milk enzymes are technologicallysignificant from five viewpoints:

1. Deterioration (lipase [potentially the most signif-icant enzyme in milk], proteinase, acid phosphataseand xanthine oxidase) or preservation (lactoperoxi-dase, sulfhydryl oxidase, superoxide dismutase) ofmilk quality.

2. As indices of the thermal history of milk; theseinclude alkaline phosphatase, �-glutamyl transpepti-dase, lactoperoxidase, and perhaps others.

3. As indices of mastitic infection; the concentrationof several enzymes increases on mastitic infection,especially catalase, N-acetyl-�-glucosaminidase andacid phosphatase.

4. Antimicrobial activity, such as lysozyme and lac-toperoxidase (which is exploited as a component of thelactoperoxidase–H2O2–thiocyanate system for the coldpasteurization of milk).

5. As commercial source of enzymes; these includeribonuclease and lactoperoxidase.

With a few exceptions (e.g., lysozyme and lactoperox-idase), the indigenous milk enzymes do not have abeneficial effect on the nutritional or organolepticattributes of milk, and hence their destruction byheat is one of the objectives of many dairy processes.

The technologically significant indigenous enzymesin milk and their catalytic activities are listed inTable 1. All these enzymes have been isolated andwell characterized. Other enzymes that have been iso-lated and characterized but which have little or nosignificance in milk are listed in Table 2. Enzymatic

The following abbreviations are used: MFGM, milk fat glo-

bule membrane; �-CN, �-casein; �s2-CN, �s2-casein; LPL,

lipoprotein lipase; UHT, ultra high temperature; HTST,

high temperature short time; LTLT, low temperature long

time; PE, pasteurization equivalent; HML, human milk lyso-

zyme; BML, bovine milk lysozyme; EWL, egg white lyso-

zyme; NAGase, N-acetyl-�-D-glucosaminidase; GGTP, �-glutamyl transpeptidase; LPO, lactoperoxidase; XO,

xanthine oxidase; SO, sulfhydryl oxidase; SOD, superoxide

dismutase.


activities that have been detected in milk but whichhave not been isolated and have no known significancein milk are listed in Table 3; it is possible that some ofthese enzymes are secreted by contaminating bacteriain milk.

The indigenous enzymes in milk have attracted theattention of researchers for > 100 years, mainlybecause of their potential to cause defects in milkand dairy products, especially lipase, and their useful-ness as indicators of the thermal treatment of milk.More recently, they have assumed importance as

indices of animal health and of the mechanismsinvolved in the synthesis and secretion of milk. Avery extensive literature has accumulated. The generaltopic has been the subject of several general reviews(1–10); in addition, the literature on the principal tech-nologically significant enzymes has been reviewedseparately (see below).

In this chapter the occurrence, distribution, isola-tion, and characterization of the principal indigenousenzymes in bovine milk will be discussed, with empha-sis on their commercial significance in milk and dairy

Table 2 Other Enzymes That Have Been Isolated From Milk and Partially Characterized but Which Are of No Known

Significance in Milk

Enzyme Reaction catalyzed Comment

Glutathione peroxidase EC 1.11.1.92 GSHþH2OÐ GSSH Contains Se

Ribonuclease EC 3.1.27.5 Hydrolysis of RNA Milk is a very rich source;

similar to pancreatic

RNase

�-Amylase EC 3.2.1.1 Hydrolysis of starch

�-Amylase EC 3.2.1.2 Hydrolysis of starch

�-Mannosidase EC 3.2.1.24 Hydrolysis of mannan Contains Zn2þ

�-Glucuronidase EC 3.2.1.31 Hydrolysis of glucuronides

5 0-Nucleotidase EC 3.1.3.5 5 0- Nucleotides þH2OÐ ribonucleosidesþPi Diagnostic test for mastitis

Adenosine triphosphatase EC 3.6.1.3 ATPþH2OÐ ADPþ Pi

Aldolase EC 4.1.2.13 Fructose, 1,6 diPÐ glyceraldehyde-3-

Pþ dihydroxyacetone-P

Source: Ref. 70.

Table 1 Technologically Significant Indigenous Enzymes in Milk

Enzyme Reaction Significance

Lipase TriglyceridesþH2O! fatty acidsþ partial

glyceridesþ glycerol

Off flavor in milk; flavor development in blue

cheese

Proteinase (plasmin) Hydrolysis of peptide bonds, particularly in

�-caseinReduced storage stability of UHT products;

cheese ripening

Alkaline

phosphomonoesterase

Hydrolysis of phosphoric acid esters Index of pasteurization

Acid phosphomonoesterase Hydrolysis of phosphoric acid esters Cheese ripening; reduced heat stability of

milk;

Lysozyme Hydrolysis of mucopolysaccharides Bacteriocidal agent

�-Glutamyl transpeptidase Transfer of �-glutamyl residues Index of heat treatment

N-Acetylglucosaminidase Hydrolysis of glycoproteins Index of mastitis

Xanthine oxidase AldehydeþH2OþO2 ! AcidþH2O2 Pro-oxidant; cheese ripening

Sulfhydryl oxidase 2RSHþO2 ! RSSRþH2O2 Amelioration of cooked flavor

Superoxide dimutase 2O2 þ 2Hþ ! H2O2 þO2 Antioxidant

Catalase 2H2O2 ! O2 þ 2H2O Index of mastitis; pro-oxidant

Lactoperoxidase H2O2 þAH2 ! 2H2OþA Index of pasteurization; bactericidal agent;

index of mastitis; pro-oxidant


products. The available information indicates that themilks of other species have an enzyme profile similar tobovine milk, although very considerable interspeciesdifferences exist in the level of certain enzymes, e.g.,the very high level of lysozyme in human and equinemilks. Human milk and that of other primates containa bile salts-activated lipase, in addition to the ubiqui-tous lipoprotein lipase, which is not present in the milkof other species. The indigenous enzymes in humanmilk have been described by Hamos et al. (11) andHernell and Lonnerdal (12).

II. PROTEINASES (EC 3.4.–.–)

The presence of an indigenous proteinase in milk wassuggested by Babcock and Russel in 1897 but becauseit occurred at a low concentration or had low activityin milk, it was believed until the 1960s that the protei-nase in milk might have been of microbial origin.Recent changes in the dairy industry, e.g., improvedhygiene in milk production, extended storage of milkat a low temperature at the farm and/or factory, andaltered product profile—e.g., UHT processing ofmilk—have increased the significance of indigenousmilk proteinase which has, consequently, been thefocus of considerable research.

Milk contains at least two proteinases, plasmin(alkaline milk proteinase) and cathepsin D (acid milkproteinase) and possibly several other proteolyticenzymes, e.g., two thiol proteinases, thrombin, andan aminopeptidase. In terms of activity and technolo-gical significance, plasmin is the most important of theindigenous proteolytic enzymes and has been the sub-ject of most attention. The relevant literature has beenreviewed by Grufferty and Fox 13) and Bastian andBrown (14).

A. Plasmin (EC 3.4.21.7)

The physiological function of plasmin (fibrinolysin) isto dissolve blood clots. It is part of a complex systemconsisting of plasmin, its zymogen (plasminogen),plasminogen activators, plasmin inhibitors, and inhi-bitors of plasminogen activators (Fig. 1). In milk,there is about four times as much plasminogen asplasmin, and both, as well as plasminogen activators,are associated with the casein micelles, from whichthey dissociate when the pH is decreased to 4.6; theinhibitors of plasmin and of plasminogen activatorsare in the milk serum. It has been reported that thereis a low level of plasmin activity in the milk fat glo-bule membrane but this appears to be due to theadsorption of plasmin to casein micelles which areadsorbed on the membrane (15). The concentrationof plasmin and plasminogen in milk increase withadvancing lactation, mastitic infection, and numberof lactations. The conversion of plasminogen to plas-min in milk increases with advancing lactation, andthere is a positive correlation between plasmin activityand the level of plasminogen activator, which itself ispositively correlated with somatic cell count (16). Thelevel of plasmin in milk is also affected by diet andmanagement practices (16). No activation of plasmi-nogen to plasmin is reported (17) to occur duringstorage of milk at 4C for 6 days; in fact, plasminand potential plasmin (plasminogen) activitydecreased under these conditions.

Bovine plasminogen contains 786 amino acids witha mass of 88.092 kDa. Its primary structure is arrangedin five loops (called kringles), each stabilized by threeintramolecular disulfide bonds. Plasminogen is acti-vated in a two-step process: it is first cleaved atArg557-Ile558 (bovine) by plasmin (a trace of whichoccurs in blood) to yield Lys-plasminogen which isinactive but undergoes a conformational change

Figure 1 Schematic representation of the plasmin system in milk.


Table 3 Partial List of Minor Enzymes in Milk

Enzyme Reaction catalyzed Source

Distribution

in milk

EC 1.1.1.1 Alcohol dehydrogenase EthanolþNADþ Ð acetaldehyde þNADHþHþ

EC 1.1.1.14 L-Iditol dehydrogenase L-IditolþNADþ Ð L-sorboseþNADHþHþ — SM

EC 1.1.1.27 Lactate dehydrogenase Lactic acidþNADþ Ð pyruvate acidþNADHþHþ SM

EC 1.1.1.37 Malate dehydrogenase MalateþNADþ Ð oxaloacetateþNADHþHþ Mammary gland SM

EC 1.1.1.40 Malic enzyme MalateþNADPþ Ð pyruvateþ CO2þNADHþHþ Mammary gland SM

EC 1.1.1.42 Isocitrate dehydrogenase IsocitrateþNADPþ Ð 2-oxoglutarateþ CO2

þNADHþHþMammary gland SM

EC 1.1.1.44 Phosphoglucuronate dehydrogenase

(decarboxylating)

6-Phospho-D-gluconate þNADPþ Ð D-ribose-5-

Pþ CO2 þNADPHþHþMammary gland SM

EC 1.1.1.49 Glucose-6-phosphate dehydrogenase D-Glucose-6-P þNADPþ Ð D-glucono-1,5-lactone-6-

PþNADPHþHþMammary gland SM

EC 1.4.3.6 Amine oxidase (Cu-containing) RCH2NH2 þH2OþO2 Ð RCHOþNH3 þH2O2 — SM

— Polyamine oxidase Spermine O2

! spermidine O2

! putrescine — SM

— Fucosyltransferase Catalyzes the transfer of fucose from GDP L-fucose to

specific oligosaccharides and glycoproteins

— SM

EC 1.6.99.3 NADH dehydrogenase NADH þ acceptorÐ NADþ þ reduced acceptor — FGM

EC 1.8.1.4 Dihydrolipoamide dehydrogenase

(diaphorase)

DihydrolipoamideþNADþ Ð lipoamideþNADH — SM/FGM

EC 2.4.1.22 Lactose synthetase (A protein: UDP-

galactose: D-glucose, 1-

galactosyltransferase; B protein: �-lactalbumin)

UDP galactoseþD-glucoseÐ UDPþ lactose Golgi apparatus SM

EC 2.4.1.38 Glycoprotein 4-�-galactosyltransferase UDP galactoseþN-acetyl D-glucosaminyl-

glycopeptideÐ UDPþ 4; �-D-galactosyl-N-acetyl-D-

glucosaminyl glycopeptide

— FGM

EC 2.4.1.90 N-Acetyllactosamine synthase UDP galactoseþN-acetyl-D-glucosamineÐ UDP

N-acetyllactosamine

Golgi apparatus —

EC 2.4.99.6 CMP-N-acetyl-N-acetyl-lactosaminide

�-2,3-sialyltransferaseCMP-N-acetylneuraminate+�-D-galactosyl-1,4-N-acetyl

D-glucosaminyl glycoproteinÐ CMPþ�-N-acetylneuraminyl 1-2, 3-�-D-galactosyl-1,4-N-

acetyl-D-glucosaminyl-glycoprotein

— SM


EC 2.5.1.3 Thiamine-phosphate pyrophosphorylase 2-Methyl-4-amino-5-hydroxymethyl/pyrimidine

diphosphateþ 4-methyl-5-(2-phosphonooxyethyl)-

thiazoleÐ pyrophosphateþ thiamine monophosphate

— FGM

EC 2.6.1.1 Aspartate aminotransferase L-aspartate þ 2-oxoglutarateÐ oxaloacetateþ L-

glutamate

Blood SM

EC 2.6.1.2 Alanine aminotransferase L-alanineþ 2-oxoglutarateÐ pyruvateþ L-glutamate Blood SM

EC 2.7.7.49 RNA-directed DNA polymerase n Deoxynucleoside triphosphateÐn

pyrophosphateþDNAn

—

EC 2.8.1.1 Thiosulfate sulfur transferase Thiosulfateþ cyanideÐ sulfiteþ thiocyanate — SM

EC 3.1.1.8 Cholinesterase AcylcholineþH2OÐ cholineþ carboxylic acid anion Blood FGM

EC3.1.3.9 Glucose-6-phosphatase D-Glucose 6-PþH2OÐ D-glucoseþ Pi — FGM

EC 3.1.4.1 Phosphodiesterase Phosphodiester þH2O! phosphomonoesterþ alcohol —

EC 3.1.6.1 Arylsulfatase Phenol sulfateþH2OÐ phenolþ sulfate — —

EC 3.2.1.21 �-Glucosidase Hydrolysis of terminal non-reducing �-D-glucose residues Lysosomes FGM

EC 3.2.1.23 �-Galactosidase Hydrolysis of terminal nonreducing �-D-galactose

residues in �-D-galactosides

Lysosomes FGM

EC 3.2.1.51 �-Fucosidase An �-L-fucosideþH2OÐ an alcoholþ L-fucose Lysosomes —

EC 3.4.11.1 Cytosol aminopeptidase (leucine

aminopeptidase)

Aminoacyl-peptide þH2OÐ amino acidþ peptide — SM

EC 3.4.11.3 Cystyl-aminopeptidase (Oxytocinase) Cystyl-peptides þH2OÐ amino acidþ peptide — SM

EC 3.4.21.4 Trypsin Hydrolyzes peptide bonds, preferentially Lys-X, Arg-X — SM

EC 3.6.1.1 Inorganic pyrophosphatase Pyrophosphate þH2OÐ 2 orthophosphate — SM/FGM

EC 3.6.1.9 Nucleotide pyrophosphate A dinucleotideþH2OÐ 2 mononucleotides — SM/FGM

EC 4.2.1.1 Carbonate dehydratase H2CO3 Ð CO2 þH2O — SM

EC 5.3.1.9 Glucose-6-phosphate isomerase D-glucose-6-PÐ fructose-6-P — SM

EC 6.4.1.2 Acetyl-CoA carboxylase ATPþ acetyl-

CoAþHCO�3 Ð ADPþ orthophosphateþmalonyl-

CoA

— FGM

Source: Ref. 70.

SM ¼ skin milk; FGM ¼ fat globule membrane; P ¼ phosphate:


which exposes the bond Lys77-Arg78 to hydrolysis byurokinase or tissue-type plasminogen activator.Hydrolysis of this bond yields a mature enzyme of� 81 kDa, consisting of two polypeptide chains heldtogether by a single disulfide bond. The five kringlesof plasminogen are retained in plasmin; they arerequired for activity and are conserved in plasminsfrom different species. Bovine plasmin is also cleavedat Arg342-Met343 to yield midi plasmin (15). Bovineplasminogen cDNA has been cloned (18).

Plasmin is a serine proteinase (inhibited by di-iso-propylfluorophosphate, phenylmethyl sulfonyl fluor-ide, and trypsin inhibitors) with a high specificity forpeptide bonds to which lysine or arginine supplies thecarboxyl group. The active site is in the smaller of thetwo chains and consists of His598, Asp641, and Ser736. Itis optimally active at � pH7:5 and � 35C; it exhibits� 20% of maximum activity at 5C and is stable overthe pH range 4–9.

Plasmin is usually extracted from casein by washingwith water at pH 3.5 and purified by precipitation withðNH4Þ2SO4 and various forms of chromatography,including affinity chromatography. Plasmin is quiteheat stable; it is partially inactivated by heating at72C for 15 sec, but its activity in milk increasesfollowing HTST pasteurization, probably owing toinactivation of the indigenous inhibitors of plasminor, more likely, inhibitors of plasminogen activators.It partly survives UHT sterilization and is inactivatedby heating at 80C for 10 min at pH 6.8; its thermalstability decreases with increasing pH in the range3.5–9.2.

The inactivation of plasmin in milk follows first-order kinetics. Arrhenius plots show that inactivationis not linear with increasing temperature. In the tem-perature range 63–110C, Ea for inactivation is 52.75and 74:44 kJmol�1 for low and high somatic cellcount (SCC) milk, respectively, while in the range100–130C, Ea is 22.03 and 24:70 kJmol�1 for low andhigh SCC milk, respectively (19). Plasmin is more heatstable to low-temperature treatments, e.g., thermizationandHTST pasteurization, in high than in low SCCmilkbut the reverse is true in the UHT range (19). Milk witha high SCChas high plasminogen activator activity (19).

1. Assay of Plasmin Activity

Plasmin activity may be assayed on a wide range ofsubstrates, including proteins, but the most widelyused substrate is the synthetic fluorogenic peptide,N-succinyl-L-Ala-L-Phe-L-Lys-7-amido-4-methyl cou-marin; the liberated 7-amido-4-methyl coumarin is

quantified by determining the intensity of fluorescence,with excitation at 380 nm and emission at 460 nm (20).The chromogenic substrate, Val-Leu-Lys-p-nitroani-lide, is also used (16). Plasminogen activity is measuredby assaying plasmin activity before and after activationof indigenous plasminogen by an excess of added uro-kinase (16). Plasminogen activator activity may beassayed by the ability of an ultracentrifugal caseinmicelle pellet to activate exogenous (added) plasmino-gen (16).

2. Activity of Plasmin on Milk Proteins

�-Casein is the most susceptible milk protein to plas-min action; it is hydrolyzed rapidly at Lys28-Lys29,Lys105-His106, and Lys107-Glu108 to yield �1 ð�-CNf29-209), �2 (�-CN f106-209), and �3 (�-CN f108-209) caseins and proteose peptone (PP)5 (�-CN f1-105/7), and PP8 slowly (�-CN f29-105/7) and PP8fast (�-CN f1-29). In vitro (in solution), �-casein isalso hydrolyzed fairly rapidly at Lys113-Tyr114 andLys183-Asp184, but it is not known if these bonds arehydrolyzed in milk. �-Caseins normally represent � 3% of total N in milk but can be as high as 10% in late-lactation milk; as a percent of total N, the value forproteose peptones is about half that of the �-caseins.�s2-Casein in solution is also hydrolyzed very

rapidly by plasmin at bonds Lys21-Gln22, Lys24-Asn25,Arg114-Asn115, Lys149-Lys150, Lys150-Thr151, Lys181-Thr182, and Lys188-Ala189 (see 14), but its hydrolysisin milk has not been characterized. Although less sus-ceptible than �s2- or �-caseins, �s1-casein in solution isalso readily hydrolyzed by plasmin (14) but it does notappear to be hydrolyzed to a significant extent in milk,although it has been suggested that �-casein is pro-duced from �s1-casein by plasmin. Although �-caseincontains several Lys and Arg residues, it appears tobe quite resistant to plasmin, presumably owing to arelatively high level of secondary and tertiary struc-tures. The whey proteins are quite resistant to plasmin,probably owing to their compact, globular structures;in fact, �-lactoglobulin, especially when denatured,inhibits plasmin, presumably via sulfhydryl-disulfideinteractions which rupture the structurally importantkringles.

3. Significance of Plasmin Activity in Milk

There is sufficient plasmin in milk to cause very exten-sive proteolysis, but this is not realized owing to thepresence of inhibitors. If the casein micelles are sedi-mented by ultracentrifugation and redispersed in buf-fer, very extensive proteolysis occurs on storage since


the inhibitors are removed in the ultracentrifugalserum. According to Guinot-Thomas et al. (17), micro-bial proteinases cause more proteolysis in milk thanplasmin during storage at 4C for 6 days.

Plasmin and plasminogen accompany the caseinmicelles on the rennet coagulation of milk and areconcentrated in cheese in which plasmin contributesto primary proteolysis of the caseins, especially incheeses that are cooked to a high temperature, e.g.,Swiss and some Italian varieties, in which the coagu-lant is totally or extensively inactivated (21). �-Caseinis the principal substrate and even in low-cookedcheeses (e.g., Cheddar, Gouda), in which the coagulantis the principal primary proteinase, proteolysis of �-casein is due mainly to plasmin action; some hydrolysisof �s1-casein by plasmin also occurs.

The level of plasmin activity in cheese varies sub-stantially with the variety; Emmental, Parmesan, andDutch-type cheeses contain about three times as muchplasmin activity as Cheddar-type cheeses (22). This isprobably due to the greater activation of plasminogenin the former as a result of inactivation of inhibitors ofplasminogen activators in high-cooked cheese andtheir removal from Dutch-type cheese curd on washing(when whey is removed and replaced by water).Plasmin activity is decreased in cheeses made fromUF-concentrated milk because of the retention of inhi-bitors of plasmin and plasminogen activators and �-lactoglobulin in the curd. Proteolysis is less in UFcheeses than in their conventional counterparts; pre-sumably, the lower level of plasmin activity is a con-tributory factor.

It has been suggested that an elevated level of plas-min activity in late-lactation milk contributes to itspoor cheese-making properties; however, an elevatedlevel of somatic cells in milk does not appear to lead todefects in cheese (13). The casein micelles in bovinemilk are capable of binding � 10 times as much plas-min as occurs naturally in milk. Exogenous plasminadded to milk is incorporated and uniformly distribu-ted in the cheese curd, in which it accelerates proteo-lysis and maturation (23). When other exogenousproteinases are added to cheese milk, much of theadded enzyme is lost in the whey, increasing cost andcreating potential problems for whey processors. Theyield of cheese may also be decreased owing to earlyhydrolysis of casein in the vat. For Cheddar-typecheese, exogenous proteinases may be added to themilled curds at salting, but the enzyme is concentratedat the surface of the curd chips. Activation of indigen-ous plasminogen by added urokinase also acceleratesproteolysis (14, 23a).

Cheese analogues are usually produced from rennetcasein which may contain active plasmin. Hydrolysisof �-casein and undesirable changes in the rheologicalproperties of cheese analogs have been attributed toplasmin action (14).

Plasmin activity may contribute to the age gelationof UHT milk produced from high-quality raw milk(which contains a low level of Pseudomonas protei-nase). The acid precipitability of casein from late-lacta-tion milk is poor, but evidence for the involvement ofplasmin is lacking. Reduced yields of cheese and caseincan be expected to result from plasmin action in milksince the proteose peptones are, by definition, solubleat pH 4.6 and are not incorporated into acid- orrennet-produced casein curd.

4. Proteinase Inhibitors in Milk

Milk contains several broad-specificity plasma-derivedproteinase inhibitors: �1-proteinase inhibitor, �2-anti-plasmin, C1 inhibitor, antithrombin-III, �2-macroglo-bulin, inter-�-trypsin inhibitor, and two inhibitorsanalogous to human �1-antichymotrypsin which pos-sess inhibitory activity against trypsin and elastase,respectively. Inhibitory activity is highly elevated inmastitic milk owing to increased leakage of blood pro-teins into milk. It has been proposed that trypsin inhi-bitory activity in milk may be a useful index of mastitisin cows. Bovine colostrum contains colostrum-specificproteinase inhibitors, including trypsin-inhibitory andthiol proteinase-inhibitory activities, which protectimmunoglobulins and other biologically active pro-teins and peptides against proteolysis by gastrointest-inal enzymes in the newborn (24, 25).

Apart from the colostrum-specific inhibitors, theplasma-derived inhibitors are present in milk as aresult of membrane leakage (mastitis or late lactation)and therefore might be expected to be without signifi-cance. However, they are probably quite significance:(a) The level of plasmin in milk is sufficient to causevery extensive hydrolysis of the caseins, as can be read-ily demonstrated by dispersing ultracentrifugally sedi-mented casein micelles in buffer (plasmin andplasminogen accompany the casein micelles while theinhibitors are in the milk serum). Plasmin activity isincreased by pasteurization apparently becauseproteinase inhibitors are inactivated by heating.(b) The inhibitors are retained in cheese made frommilk concentrated by ultrafiltration (UF) whereasthey are lost in the whey during conventional cheesemaking. Consequently, proteolysis is retarded in UFcheese (14, 25).


Denatured �-lactoglobulin also inhibits plasmin,apparently owing to sulfhydryl-disulfide interchangereactions. Since �-lactoglobulin is concentrated byUF, it probably contributes to the inhibition of pro-teolysis in cheese made from UF-concentrated milkduring ripening.

5. Plasminogen Activators

There are two types of plasminogen activators, uroki-nase-type plasminogen activator (uPA) and tissue-typeplasminogen activator (tPA). Both types occur in milk;uPA is confined to cells in milk while tPA is associatedwith the casein micelles (26, 27). Both activators havebeen characterized and cloned (28). Lu and Nielsen(29) reported that there are five plasminogen activatorsin milk, with molecular weights of 93, 57, 42, 35, and27 kDa; most were uPA-type activators. PA levelincreases with mastitic infection and advancing lacta-tion (30), which explains the greater conversion ofplasminogen to plasmin in such milks.

B. Cathepsin D (EC 3.4.23.5)

It has been known for � 30 years that milk also con-tains an acid proteinase (optimum pH � 4:0) which isnow known to be cathepsin D, a lysosomal enzyme. Itis relatively heat labile (inactivated by 70C for10min). Its activity in milk has not been studied exten-sively, and its significance is unknown. At least some ofthe indigenous acid proteinase is incorporated intocheese curd; its specificity on �s1- and �-caseins isquite similar to that of chymosin, but it has verypoor milk clotting activity (31). It may contribute toproteolysis in cheese, but its activity is probably nor-mally overshadowed by chymosin, which is present at amuch higher level in cheese.

C. Other Proteinases

The presence of other minor proteolytic enzymes inmilk, including thrombin and a lysine aminopeptidase,has been reported (32). In addition to cathepsin D,other proteolytic enzymes from somatic cells are prob-ably present in milk. Verdi and Barbano (33), whostudied the degradation of caseins in milk by somaticcells or plasmin, found that somatic cell proteinasesand plasmin produced distinctly different peptides,and that the plasmin inhibitor 6-aminohexanoic acidwas suitable for studying the action of somatic cellproteinases, without interference from plasmin.Somatic cell proteinases are capable of activating plas-

minogen (34), and this may influence proteolysis incheese by elevating plasmin levels. Although leucocyteproteinases were more active on �-casein at pH 6.6than at pH 5.2, their activity at the lower pH wassuch as to suggest that they may be active in cheeseduring ripening (35). Suzuki and Katoh (36) found twocysteine proteinases in milk (45 kDa and > 150 kDa).The authors suggested that these proteinases origi-nated in somatic cells and their level increased duringmastitic infection.

Grieve and Kitchen (37) compared the action ofleucocyte proteinases, plasmin, and some psychotrophproteinases on the caseins. Leucocyte extracts hydro-lyzed the caseins in order �s1 > �� . Although theseauthors considered that neutral proteinases from leu-cocytes (isolated from blood) were unlikely to beimportant for proteolysis in milk, other authors havefound considerably lower proteolytic activity in leuco-cytes isolated from blood than from milk (34, 35).

Kelly et al. (39), who compared proteolysis inGouda cheeses made from milks with the same totalsomatic cell count but different levels of polymorpho-nuclear (PMN) leucocytes, found more rapid produc-tion of �s1-CN f24-199 and total free amino acids incheese made from milk with high PMN levels (�s1-CNis �s1-casein).

The significance of minor indigenous proteins dur-ing mastitic infections was discussed by Fang andSandholm (40) who investigated the possibility ofusing specific proteinase inhibitors as therapeuticagents. Although the minor proteinases are probablyless significant technologically than plasmin, morework on the subject is warranted.

III. LIPASES AND ESTERASES (EC 3.1.1.–)

Lipases catalyze the development of hydrolytic rancid-ity which is a serious defect in milk and some milkproducts, and, consequently, lipases and lipolysis inmilk have been studied extensively. Milk containsthree types of esterase: (a) A-type carboxylic esterhydrolase (arylesterase; EC 3.1.1.2), which hydrolyzesaromatic esters, e.g., phenylacetate. It shows littleactivity on tributyrin, and is not inhibited by organo-phosphates. (b) B-type esterase (glycerol tricarboxylesterase, aliphatic esterase, lipase; EC 3.1.1.3). Suchenzymes are most active on aliphatic esters althoughthey show some activity on aromatic esters; they areinhibited by organophosphates. (c) C-type esterase(cholinesterase; EC 3.1.1.7; EC 3.1.1.8). These enzymesare most active on choline esters but hydrolyze some


aromatic and aliphatic esters slowly; they are inhibitedby organophosphates.

In normal milk, the ratio of A : B :C types of ester-ase activity is about 3 : 10 : 1 but the level of A-esteraseactivity increases considerably on mastic infection. Aand C esterases are of little technological significancein milk. Lipases hydrolyze ester bonds in emulsifiedesters, i.e., at a water/oil interface, although somemay have limited activity on soluble esters. They areusually activated by blood serum albumin and Ca2þ

which bind free fatty acids, which are inhibitory.Milk lipase was first isolated and characterized by

Fox and Tarassuk (41) and Patel et al. (42). Theenzyme was optimally active at pH 9.2 and 37C andwas found to be a serine enzyme (inactivated by orga-nophosphates). A lipoprotein lipase (LPL; activated bylipoprotein cofactors) was demonstrated in milk byKorn in 1962 (42a) and isolated by Egelrud andOlivecrona (43). The lipase isolated by Fox andTarassuk (41) is, in fact, an LPL which is the principal,probably the only, indigenous lipase in bovine milk. Ithas been the focus of considerable research and hasbeen characterized at the molecular, genetic, enzy-matic, and physiological levels (44).

Under optimum conditions, the kcat for milk LPL is� 3000 s�1 and milk contains sufficient enzymes (1–2mg=L; i.e., 10–20 nM) to theoretically cause rancidityin 10 sec. However, this does not occur in practicebecause the triglycerides are protected by the MFGMwhile the lipase is naturally associated with the caseinmicelles. Also, environmental conditions, e.g., pH, arenot optimal. However, if the MFGM is damaged byagitation (e.g., by milking machines, bulk tanks,pumps, etc.), homogenization or temperature fluctua-tions, lipolysis occurs rapidly and rancidity ensues.Milk LPL appears to be derived from blood plasmaand hence any condition that increases the permeabil-ity of mammary cell membranes, e.g., physiologicalstress, mastitic infection, or late lactation, increasesthe level of LPL in milk and hence the risk of lipolysis.Some individual cows produce milk which becomesrancid spontaneously, i.e., without apparent activa-tion. Apparently, spontaneous rancidity occurs whenmilk contains a high level of lipoprotein (co-lipase)from blood serum which activates the LPL. Normalmilk will become spontaneously rancid if bloodserum is added, suggesting that ‘‘spontaneous milks’’contain a higher than normal level of blood serum.Dilution of spontaneous milk with normal milk pre-vents spontaneous rancidity, which, consequently, isnot normally a problem with bulk herd milks.Presumably, dilution with normal milk decreases the

lipoprotein content of the bulk milk to below thethreshold necessary for lipase activation. Natural var-iations in the level of free fatty acids in normal milkand the susceptibility of normal milks to induced lipo-lysis may be due to variations in the level of bloodserum components in milk.

In addition to LPL, human milk contains a bilesalts–activated lipase, which probably contributes tothe metabolism of lipids by breastfed babies whohave limited pancreatic lipase activity. Bovine milkand milks from other dairy animals do not containthis enzyme.

A. Assay of Lipolytic Activity

Lipase can be assayed by incubating the sample withan emulsified lipid substrate—e.g., milk fat, olive oil,tributyrin, etc.—extraction of liberated fatty acids withdiethyl ether-petroleum ether, and titration with etha-nolic KOH. This method is rather tedious and tribu-tyrin agar diffusion assays may be used when rapidscreening is required. Chromogenic substrates, e.g.,�-naphthyl- or p-nitrophenyl derivatives of fattyacids or fluorogenic substrates, e.g., coumarin deriva-tives of fatty acids, are very sensitive and satisfactory,especially when the enzyme has been at least partiallypurified.

B. Significance of Lipase

Technologically, LPL is, arguably, potentially the mostsignificant indigenous enzyme in milk. Although LPLmay play a positive role in cheese ripening, undoubt-edly the most industrially important aspect of milklipase is its role in hydrolytic rancidity which rendersliquid milk and dairy products unpalatable and even-tually unsaleable. Lipolysis in milk has been reviewedextensively (45). It appears to occur mainly at the farmlevel and the problem may be minimized by good man-agement practices on the farm:

1. Proper installation, maintenance, and operationof milking machines.

2. Avoidance of excessive agitation by pumps oragitators in bulk tanks or risers in milk pipe-lines.

3. Avoidance of freezing on the walls of bulktanks.

4. Avoidance of cooling and warming cycles in thebulk tank.

5. Culling of cows with high somatic cell counts.


The first three factors damage the MFGM, makingthe core triglycerides accessible to lipase. Some caseinmicelles probably adsorb on exposed fat surfaces.Pipeline milking machines cause more damage to theMFGM than hand milking or bucket milkingmachines. Suction of air at teat cups (which causesfoaming), risers, and change of dimensions in the pipe-line and the hose connecting the clawpiece to the receiv-ing jar are major potential sites for damage.Homogenization causes total replacement of the nat-ural MFGM by a membrane composed of caseinmicelles or submicelles and whey proteins. Unless indi-genous LPL is inactivated by pasteurization before orimmediately after homogenization, rancidity willdevelop very rapidly. Minimal high temperature shorttime (HTST; 72C for 15 sec) pasteurization of milkcauses extensive inactivation of LPL but pasteurizationat 80C for 10 sec is required for complete inactivation.

Freezing of milk on the walls of bulk tanks willdamage the MFGM, inducing lipolysis. Temperaturefluctuations, e.g., cooling to � 5C, rewarming to� 30C, and recooling, also activate lipolysis. Suchtemperature fluctuations may occur if bulk tanks arenot completely emptied at each milk collection andwarm milk is added at the subsequent milking. Themechanism of thermal activation is not clear, butdamage to the MFGM by fat crystals seems likely.

The propensity of milk to lipolysis increases whenthe producing animals are under any type of stress.The mammary cell membranes become more perme-able to blood constituents as a result of mastitic infec-tion in late lactation and as the animal ages. Thenumber of somatic cells in milk is a good index ofsuch stress or damage, and upper limits for somaticcell count (SCC) are now frequently prescribed bymilk processors. Although the potential for hydrolyticrancidity always exists in raw milk, the problem can bereduced to insignificant levels by good managementpractices at the farm.

IV. ALKALINE PHOSPHATASES (EC 3.1.3.1)

Milk contains several phosphatases, the principal onesbeing alkaline and acid phosphomonoesterases, whichare of technological significance, and ribonuclease,which has no known function or significance in milk.The alkaline and acid phosphomonoesterases havebeen studied extensively (1, 8, 9).

The occurrence of a phosphatase in milk was firstrecognized in 1925. Subsequently characterized as analkaline phosphatase, it became significant when it

was shown that the time–temperature combinationsrequired for the thermal inactivation of alkaline phos-phatase were slightly more severe than those required tokill Mycobacterium tuberculosis, then the target micro-organism for pasteurization. The enzyme is readilyassayed, and a test procedure based on alkaline phos-phatase inactivation was developed as a routine qualitycontrol test for the HTST pasteurization of milk.

A. Assay Methods

Several major modifications of the original assay havebeen developed. The usual substrates are phenyl phos-phate, p-nitrophenyl phosphate, or phenolphthaleinphosphate which are hydrolyzed to inorganic phos-phate and phenol, p-nitrophenol, or phenolphthalein,respectively:

where XOH ¼ phenol, p-nitrophenol, or phenolphtha-lein.

The release of inorganic phosphate may be assayedbut the other product is usually determined. Phenol iscolorless but forms a colored complex on reaction withone of several reagents, e.g., 2,6-dichloroquinonechlor-oimide, with which it forms a blue complex. p-Nitrophenol is yellow while phenolphthalein is red atthe alkaline pH of the assay (� 10) and hence these areeasily quantified.

A fluorogenic aromatic orthophosphoric monoe-ster, Fluorophos (Advanced Instruments, NeedhamHeights, MA), has been developed and approved forthe determination of alkaline phosphatase in milk andmilk products. Hydrolysis of the ester yields a fluores-cent compound, Fluoroyellow, the concentration ofwhich is determined fluorometrically (excitation,439 nm; emission, 560 nm).

Fluorometric methods are about 100–1000 times moresensitive than colorimetric assays. A simple fluorom-eter has been developed for the analysis (AdvancedInstruments). Comparative studies on the fluorometricand the standard colorimetric methods include those ofRocco (46) and Eckner (47).


B. Isolation and Characterization of Alkaline

Phosphatase

Alkaline phosphatase is concentrated in the fat globulemembrane and hence in cream. The membrane isreleased into the buttermilk on phase inversion; conse-quently, buttermilk is the starting material for mostpublished methods for the purification of alkalinephosphatase. Later methods have used chromatogra-phy on various media to give a homogeneous prepara-tion; up to � 7500-fold purification with a yield of� 30% have been reported (1). The characteristics ofmilk alkaline phosphatase are summarized in Table 4.The enzyme appears to be similar to the alkaline phos-phatase of mammary tissue, which is the presumedsource of the enzyme in milk.

The enzyme is a dimer of two identical 85-kDa sub-units. It contains 4 Zn2þ per mole which are requiredfor activity; Mg2þ are also strong activators, and1mMMg2þ is usually added to the assay mixture.The enzyme is strongly but reversibly inhibited bymetal chelators; the apoenzyme is reactivated byZn2þ, Mg2þ, and other metal ions. Reaction of apoen-zyme may in fact be used as a sensitive assay for avail-able Zn2þ in foods. Inorganic orthophosphates arestrong competitive inhibitors of the hydrolysis of p-nitrophenylphosphate. Alkaline milk phosphatase candephosphorylate phosphoproteins, including casein,with a pH optimum of 6.5–7.0; however, it does notappear to do so in milk, probably owing to inhibitionby the high level of orthophosphate.

C. Reactivation of Phosphatase

Much work has been focused on a phenomenonknown as ‘‘phosphatase reactivation,’’ first recognizedby Wright and Tramer in 1953, who observed thatUHT-treated milk was phosphatase-negative immedi-ately after processing but became positive on standing;microbial phosphatase was shown not to be responsi-ble. Bulk HTST milk never showed reactivation,although occasional individual cow samples did.HTST pasteurization after UHT treatment usually pre-vented reactivation which was never observed in in-container sterilized milk. Reactivation can occur fol-lowing heating at a temperature as low as 84C formilk or 74C for cream. The optimum storage tem-perature for reactivation is 30C, at which reactivationis detectable after 6 h and may continue for up to 7days. The greater reactivation in cream than in milkmay be due to protection by fat, but this has not beensubstantiated.

A number of attempts have been made to explainthe mechanism of reactivation (8). There is evidencethat the form of the enzyme which becomes reactivatedis membrane bound, and several factors which influ-ence reactivation have been established. Mg2þ andZn2þ strongly promote reactivation; Sn2þ, Cu2þ,Co2þ and EDTA are inhibitory, while Fe2þ has noeffect. Sulfhydryl (SH) groups appear to be essentialfor reactivation; perhaps this is why phosphatasebecomes reactivated in UHT milk but not in HTSTmilk. The role of -SH groups, supplied by denaturedwhey proteins, is considered to be chelation of heavymetals, which would otherwise bind to -SH groups ofthe enzyme (also activated on denaturation), thus pre-venting renaturation. It has been proposed that Mg2þ

and Zn2þ cause a conformational change in the dena-tured enzyme, necessary for renaturation (1).

According to Murthy et al. (48), maximum reactiva-tion occurs in products heated at � 104C; incubatedat 34C, adjusted to pH 6.5, and containing0:064MMg2þ; homogenization of products beforeheat treatment reduces the extent of reactivation.There are reports that raw milk contains three isoen-zymes of alkaline phosphatase and that the zymogrampatterns of raw and reactivated milk or cream are dif-ferent (1, 48); however, these different forms probablyrepresent free and bound forms of the enzyme. Amechanism for the reactivation of alkaline phospha-tase was proposed by Copius-Peereboom (49); how-ever, this mechanism of reactivation is based on aputative structure of the milk fat globule membranewhich is now known to be incorrect. Linden (50)

Table 4 Characteristics of Milk Alkaline Phosphatase

Characteristic Conditions

pH optimum Casein: 6.8

p-nitrophenylphosphate: 9.9

Temperature optimum 37CKm 0.69 mM on p-

nitrophenylphosphate

Activators Ca2þ, Mn2þ; Zn2þ, Co2þ,Mg2þ

Inhibitors Metal chelators (EDTA,

EGTA, etc.);

orthophosphates

Native molecular weights 170–190 kDa

Quaternary structure 2 subunits each of molecular

weights 85 kDa

Zn content 4 mol mol�1 enzyme

Thermal stability:

D-value at 60C, pH 9 27.2 min

63C, pH 9 8.3 min


proposed that reactivation occurs in stages and fullreactivation requires Zn2þ, Mg2þ, inorganic phosphate(Pi), and substrate (S).

Reactivation of alkaline phosphatase is of consider-able practical significance since regulatory tests forpasteurization assume the absence of phosphataseactivity. Methods for distinguishing between renaturedand residual native alkaline phosphatase are based onthe increase in phosphatase activity resulting fromaddition of Mg2þ to the reaction mixture; various ver-sions of the test have been proposed (48, 51, 52). Theofficial AOAC method (53) is based on that of Murthyand Peeler (52). However, difficulties are experiencedin the interpretation of this test applied to cream orbutter (54, 55).

D. Significance

Alkaline phosphatase in milk is significant mainlybecause of its use as an index of HTST pasteurizationand is used universally for this purpose. However, theenzyme may not be the most appropriate for thispurpose (56) becomes (a) reactivation of alkaline phos-phatase under certain conditions complicates inter-pretation of the test; (b) the enzyme appears to befully inactivated by subpasteurization conditions(70C for 16 sec); and (c) the relationship betweenlog10% initially activity and pasteurization equivalent(PE) is less linear than the relationship of lactoperox-idase or �-glutamyl transpeptidase (57).

Although alkaline phosphatase can dephosphory-late casein under suitable conditions, as far as isknown, it has no direct technological significance inmilk. Perhaps its pH optimum is too far removedfrom that of milk, especially acid milk products,although the pH optimum on casein is reported to be� 7. It is also inhibited by inorganic phosphate.

Proteolysis is a major contributor to the develop-ment of flavor and texture of cheese during ripening(58). Most of the small water-soluble peptides in cheeseare from the N-terminal half of �s1- or �-casein; manyare phosphorylated but show evidence of phosphataseactivity (i.e., they are partially dephosphorylated (59–62). In cheese made from pasteurized milk, both indi-genous acid phosphatase and bacterial phosphataseare probably responsible for dephosphorylation(which is the more important is not clear), but in rawmilk cheese, e.g., Parmigiano Reggiano or GranaPadano, milk alkaline phosphatase appears to be themost important (63, 64). Further work on the signifi-cance of indigenous alkaline and acid phosphatases in

the dephosphorylation of phosphopeptides in cheese iswarranted.

V. ACID PHOSPHOMONOESTERASE

(EC 3.1.3.2)

Milk contains an acid phosphatase which has a pHoptimum at 4.0 and is very heat stable. (LTLT pasteur-ization causes only 10–20% inactivation and 30 min at88C is required for full inactivation; when heated inmilk at pH 6.7, the enzyme retains significant activityfollowing HTST pasteurization but it does not survivein-bottle sterilization or UHT treatment.) The enzymeis not activated by Mg2þ (as is alkaline phosphatase),but it is slightly activated by Mn2þ and is very stronglyinhibited by fluoride. The level of acid phosphataseactivity in milk is only � 2% that of alkaline phospha-tase; activity reaches a maximum 5–6 days postpartum,then decreases and remains at a low level to the end oflactation (65).

A. Isolation and Characterization

Acid phosphatase is found free in skim milk, in mem-brane material in skim milk and in the fat globulemembrane. Kitchen (9) appears to believe that a singleenzyme is involved and reported that the membrane-bound enzyme is strongly attached and is not releasedby nonionic detergents. The enzyme has been purifiedto homogeneity by various forms of chromatography,including affinity chromatography. Purification factorsof 10,000 to 1 million have been reported (1).Adsorption onto Amberlite IRC50 resin is a very effec-tive first step in purification. Andrews (65) stated thatall the acid phosphatase activity in skim milk isadsorbed by Amberlite IRC50, but this is not indicatedin the original papers. In an unpublished study, N.A.Flynn and P.F. Fox found that only � 50% of thetotal acid phosphatase in skim milk was adsorbed byAmberlite IRC50 even after reextracting the skim milkwith fresh batches of Amberlite, suggesting that skimmilk may contain two acid phosphatases. About 40%of the acid phosphatase in skim milk partitioned intothe whey on rennet coagulation and this enzyme didnot adsorb on Amberlite IRC50. The enzyme waspartly purified from whey.

Flynn and Fox attempted to purify the alkalinephosphatase from MFGM by gel permeation chroma-tography; however, sonication and nonionic detergentsfailed to disassociate the enzyme from the membrane.The MFGM enzyme, which does not adsorb on


Amberlite IRC50, was much less heat stable than theacid phosphatase isolated from whey or from skimmilk by adsorption on Amberlite IRC50. Attemptswere made to confirm that the MFGM, whey, andAmberlite-adsorbed enzymes are different by studyingthe effects of inhibitors, but the results have been equi-vocal. Overall, it appears that milk contains more thanone acid phosphatase. Using a zymogram technique,Andrews and Alichanidis (66) reported that milk fromhealthy cows contained one acid phophatase while thatfrom mastitic cows contained two additional acidphosphatases which were of leucocyte origin. It is unli-kely that the heterogeneity observed by Flynn and Foxwas due to a high proportion of mastitic milk.

The acid phosphatase isolated from skim milk byadsorption on Amberlite IRC50 has been well charac-terized. It is a glycoprotein with a molecular weight of� 42 kDa and a pI of 7.9. It is inhibited by many heavymetals, F�, oxidizing agents, orthophosphates, andpolyphosphates and is activated by thiol-reducingagents and ascorbic acid. It is not affected by metalchelators. Its amino acid composition indicates ahigh level of basic amino acids and no methionine (65).

The enzyme is quite active on phosphoproteins,including caseins. It has been suggested that it is aphosphoprotein phosphatase. Although casein is asubstrate for milk acid phosphatase, the major caseins,in the order �s ð�s1 þ �s2Þ > � > �, also act as compe-titive inhibitors of the enzyme when assayed on p-nitrophenylphosphate, probably owing to binding ofthe enzyme to the casein phosphate groups (the effec-tiveness of the caseins as inhibitors is related to theirphosphate content).

B. Assay

Acid phosphatase may be assayed, at pH � 5, on thesame substrates as used for alkaline phosphatase. If p-nitrophenyl phosphate or phenolphthalein phosphateis used, the pH must be adjusted to > 8 at the end ofthe enzymatic reaction in order to induce the color ofthe product, i.e., p-nitrophenol or phenolphthalein.

C. Significance

Although acid phosphatase is present in milk at amuch lower level than alkaline phosphatase, its greaterheat stability and lower pH optimum may make ittechnologically significant. Dephosporylation of caseinreduces its ability to bind Ca2þ, to react with �-casein,to form micelles, and its heat stability. As discussed inSection IV.D, several small partially dephosphorylated

peptides have been isolated from Cheddar andParmigiano Reggiano and Grana Padano cheeses.However, it is not known whether indigenous or bac-terial acid phosphatase is mainly responsible fordephosphorylation in cheese made from pasteurizedmilk. It is claimed (62–64) that alkaline phosphataseis mainly responsible for dephosphorylation in rawmilk cheese. Dephosphorylation may be rate limitingfor proteolysis in cheese ripening since most protei-nases and peptidases are inactive on phosphoproteinsor phosphopeptides. It has been suggested that phos-phatase activity should be included in the criteria forstarter selection.

The acid phosphatase activity in milk increasesfourfold to 10-fold during mastitic infection. Threeisoenzymes are then present, only one of which is indi-genous milk acid phosphatase, the other two being ofleucocyte origin (66). The latter isoenzymes are morethermolabile than the indigenous enzyme and are inac-tivated by HTST pasteurization.

The suitability of acid phosphatase as an indicatorenzyme for superpasteurization of milk has beenassessed (67, 68). It is not as useful for this purposeas some alternatives, e.g., �-glutamyl transpeptidase orlactoperoxidase.

VI. LYSOZYME (EC 3.2.1.17)

Lysozyme (muramidase, mucopeptide N-acetyl-mura-myl hydrolase) is a widely distributed enzyme whichlyses certain bacteria by hydrolyzing the �(1-4) linkagebetween muramic acid and N-acetylglucosamine ofmucopolysaccharides in the bacterial cell wall.Lysozyme activity is normally assayed by the lysis ofcultures of Micrococcus lysodeikticus measured by adecrease in turbidity.

Lysozyme was isolated from human milk by Jollesand Jolles (69), who believed that bovine milk wasdevoid of lysozyme. Milks of many species, includingbovine, have since been shown to contain lysozyme,and several have been isolated and characterized (70).Human and equine milks are exceptionally richsources, containing 130 mg/L (3000 times the level ofbovine milk) and � 800mg=L, respectively.

The pH optima of human milk lysozyme (HML),bovine milk lysozyme (BML), and egg-white lysozyme(EWL) are 7.9, 6.35, and 6.2, respectively (70). BMLhas a molecular weight of 18 kDa compared with15 kDa for HML and EWL. The amino acid composi-tion of BML is considerably different from that ofHML or EWL. The amino acid sequence of BML is


highly homologous with that of �-lactalbumin, a wheyprotein that is an enzyme modifier in the biosynthesisof lactose. All lysozymes are relatively stable to heat atacid pH values (3–4) but are relatively labile at pH > 7.More than 75% of the lysozyme activity in bovine milksurvives heating at 75C for 15 min or 80C for 15 secand is therefore little affected by HTST pasteurization.Low concentrations of reducing agents increase theactivity of BML and HML by � 330% (70).

Presumably, the physiological role of lysozyme is toact as a bacteriocidal agent. In the case of milk, it maysimply be a ‘‘spillover’’ enzyme, or it may have a defi-nite protective role. If the latter is true, then the excep-tionally high level of lysozyme in human and equinemilk may be significant. Breastfed babies generally suf-fer fewer enteric problems than bottle-fed babies.While there are many major compositional and physi-cochemical differences between bovine and humanmilks which may be responsible for the observed nutri-tional characteristics, the disparity in lysozyme contentmay be significant. Fortification of bovine milk-basedinfant formulae with EWL, especially for prematurebabies, has been recommended but feeding studiesare equivocal on the benefits of this practice, andrecent trials failed to demonstrate any beneficial effectdue to inactivation of EWL in the human stomach(71, 72).

One might expect that, owing to this bacteriocidaleffect, indigenous milk lysozyme would have a benefi-cial effect on the shelf life of milk. Such effects do notappear to have been reported. Exogenous lysozymemay be added to milk for many cheese varieties, e.g.,Gouda, Edam, Emmental, and Parmigiano Reggiano,as an alternative to KNO3 to prevent the growth of Cl.tyrobutyricum which causes late gas blowing and off-flavors. At present, lysozyme is not widely used incommercial cheesemaking (71, 73). Since indigenousmilk lysozyme is in the serum phase, very little is incor-porated into cheese.

Addition of lysozyme to milk decreases the heatstability of milk, but the level of indigenous lysozymeis probably too low to contribute to the natural varia-tions in the heat stability of milk.

VII. N-ACETYL-�-D-GLUCOSAMINIDASE

(EC 3.2.1.30)

N-Acetyl-�-D-glucosaminidase (NAGase) hydrolyzesterminal, nonreducing N-acetyl-�-D-glucosamine resi-dues from glycoproteins. It is a lysosomal enzymewhich originates principally from mammary gland

epithelial cells and, to a lesser extent, from somaticcells. Consequently, NAGase activity correlates highlywith the intensity of mastitis. A field test for mastitisbased on NAGase activity has been developed, usingchromogenic p-nitrophenyl N-actyl-�-D-glucosamineas substrate. Hydrolysis yields p-nitrophenol, whichis yellow at alkaline pH. NAGase activity is alsohigh in colostrum. NAGase is optimally active at50C and pH 4.2 and is inactivated by HTST pasteur-ization (70–71C for 15–18 sec). Andrews et al. (68)proposed that NAGase would be a suitable indicatorenzyme for assessing heat treatments in the range 65–75C for 15 sec. NAGase occurs mainly in the wheyfraction, from which it has been isolated by variousforms of chromatography. Two forms, A and B, differ-ing in molecular weight, 120 and 240 kDa, respectively,and charge were obtained. Each form is dissociated totwo dissimilar subunits on treatment with 2-mercap-toethanol and SDS (9, 70).

VIII. �-GLUTAMYL TRANSPEPTIDASE

(TRANSFERASE) (EC 2.3.2.2)

�-Glutamyl transpeptidase (GGTP) catalyzes thetransfer of �-glutamyl residues from �-glutamyl-con-taining peptides:

�-glutamyl-peptideþX! peptideþ �-glutamyl-X

ð3Þwhere X is an amino acid.

The enzyme is membrane bound, being found in themembrane material in skim milk (� 70%) or in theMFGM, from which it can be dissociated by deter-gents or solvents. The enzyme, which has been purifiedfrom the MFGM, has a molecular weight of � 80 kDaand consists of two subunits of 57 and 25 kDa, both ofwhich are glycoproteins (74, 75). The enzyme isoptimally active at pH 8.5–9 and � 45C and has anisoelectric point of 3.85. It is strongly inhibited by di-isopropylfluorophosphate, iodoacetamide, and metals,e.g., Cu2þ and Fe3þ (9, 70).

A. Assay

GGTP is usually assayed using �-glutamyl-p-nitroani-lide as substrate; the liberated p-NA can be determinedby measuring the absorbance at 410 nm or by reactionwith naphthylethylenediamine and measuring theabsorbance at 540 nm (76).


B. Significance

GGTP functions in the regulation of cellular glu-tathione and amino acid transport via the �-glutamylcycle; it may be involved in the biosynthesis of milkproteins.

From a dairy technologist’s viewpoint, GGTP is ofinterest mainly because of its heat stability character-istics. As discussed in Section IV.D, alkaline phospha-tase is the test enzyme usually used to evaluate theefficiency of HTST pasteurization; however, as dis-cussed, reactivation of this enzyme in UHT-treatedproducts poses problems with the interpretation ofthe test. Based on a comparative study on the heatstability characteristics of a number of indigenousenzymes in milk, Andrews et al. 68) concluded thatGGTP was appropriate for monitoring heat treatmentsin the range of 70–80C for 16 sec. This conclusion hasbeen confirmed in pilot-scale studies (77, 78). In wholeor skim milk, GGTP is completely inactivated by heat-ing at 78C for 15 sec (77) or 77C for 16 sec (76). Noreactivation was found under a variety of conditions,and little seasonal variation occurs. As little as 0.1%raw milk could be detected in skim milk or 0.25% inwhole milk (76).

Linear models for the thermal inactivation ofGGTP and lactoperoxidase (LPO) in a HTST pasteur-izer were developed by McKellar et al. (57). The equa-tion for GGTP was: log10% initial activity ¼ 2:004–0:281� PE0:75. For LPO the equation was: log10%initial activity ¼ 2:122–0:096� PE0:75.

PE ¼ t

t0¼

ðe

Ea

R

� �1

T� 1

T0

� �dt ð4Þ

where Ea ¼ activation energy; R ¼ 8:314 J=mol K;T0 ¼ 345K (72C reference temperature); T ¼experimental temperature; t0 ¼ 15 sec (reference hold-ing time); t ¼ experimental holding time, sec; and PE¼ pasteurization equivalent (71:6C for 15 sec).

These equations indicate that 1 log decrease inenzyme activity can be achieved with PE values of5.46 and 2.65 for GGTP and LPO, respectively, indi-cating that LPO is considerably more heat stable thanGGTP. The assay for LPO (using ABTS; see Sec.XIII.A) is subject to greater variation and interferencefrom milk proteins than the GGTP assay. The relation-ship between log10% initial activity and PE was morelinear than the relationship for alkaline phosphatase,possibly due to more than one form of alkaline phos-phatase (56). GGTP was about nine times more stablein ice cream mix than in whole milk, which had a much

smaller stabilizing effect on Listeria innocua (79). Thus,it appears that GGTP is a suitable enzyme for estimat-ing the intensity of heat treatment in the range72–77C for 15 sec to which milk was subjected.

GGTP is absorbed from the gastrointestinal tract,resulting in high levels of GGTP activity in the bloodserum of newborn animals fed colostrum or earlybreast milk. Since GGTP is inactivated by the heattreatment to which infant formulae are subjected, thelevel of GGTP activity in infants can be used to dis-tinguish breastfed from formula-fed infants (70).�-Glutamyl peptides have been isolated from Comte

cheese (80). Since casein contains no �-glutamyl bonds,the presence of these peptides in cheese suggests GGTPactivity in cheese, but there appear to be no data onthis.

IX. XANTHINE OXIDASE (EC 1.2.3.2)

It has been recognized for � 90 years that milk con-tains an enzyme capable of oxidizing aldehydes andpurines with the concomitant reduction of O2 toH2O2. The enzyme is now generally referred to asxanthine oxidase (XO). Milk is a very good source ofXO, at least part of which is transported to the mam-mary gland via the bloodstream. A similar enzyme isfound in various animal tissues and several bacterialspecies (8, 9, 70).

A. Assay

Xanthine oxidase activity can be assayed manometri-cally (uptake of O2), potentiometrically using a plati-num electrode, or spectrophotometrically. This lastinvolves the conversion of the xanthine to uric acidwhich is quantified by measuring absorbance at290 nm (75).

B. Isolation

The enzyme is concentrated in the MFGM, in which itis one of the principal proteins. Therefore, all isolationmethods use cream as starting material, using a dis-sociating agent to liberate XO from membrane lipo-proteins and some form of chromatography forpurification.

Milk XO has a molecular weight of � 300 kDa andconsists of two identical subunits. The pH optimum is� 8:5 and the enzyme requires flavin adenine dinucleo-tide (FAD), Fe and Mo cations, and an acid-labilesulfur compound as cofactors. Cows deficient in Mo


cations have low XO activity. The amino acid compo-sition of XO has been determined by a number ofworkers; at least five polymorphic forms have beenreported (70).

XO can be converted to an NAD-dependent dehy-drogenase by treatment with thiol-reducing agents.The enzyme reverts to an oxidase on aerobic storage,treatment with sulfhydryl oxidase, or with sulfhydryloxidizing agents.

C. Activity in Milk

XO activity in milk varies substantially—2.5-fold (67).Various processing treatments which damage or alterthe MFGM affect the XO activity of milk. Activity isincreased by � 100% on storage at 4C for 24 h, by50–100% on heating at 70C for 5 min, and by 60–90% on homogenization. These treatments cause therelease of XO from the MFGM into the aqueousphase, rendering the enzyme more active. The heatstability of XO is very dependent on whether it is acomponent of the MFGM or is dissolved in the aqu-eous phase. Aging and homogenization increase heatsusceptibility and explain the inconsistency of earlywork in which the history of the sample was unknownor unrecorded. XO is most heat stable in cream andleast in skim milk. Homogenization of concentratedmilk prepared from heated milk (90:5C for 15 sec)partially reactivates XO, which persists on drying theconcentrate, but no reactivation occurs following moresevere heating (93C for 15 sec). Apparently, homoge-nization releases potentially active, undenatured XOfrom the MFGM. All the major milk proteins canact as either activators or inhibitors of XO, dependingon their concentration, and may have some signifi-cance in the activation, inactivation, and reactivationof the enzyme. Studies on the heat stability of XO havebeen reviewed by Griffiths (67), who investigated itsstability in a pilot scale HTST pasteurizer. The enzymewas not completely inactivated after 120 sec at 80C,and a Z-value of 6.8 was calculated.

D. Significance of Xanthine Oxidase

1. As an Index of Heat Treatment

The inactivation of XO parallels the conditions neces-sary for the production of low-heat skim milk powder(82). Andrews et al. (68) considered XO as a suitableindicator of milk heated in the temperature range 80–90C but Griffiths (67) considered the natural variabil-ity in the level of XO activity in milk to be too high foruse as a reliable index of heat treatment.

2. Lipid Oxidation

XO, which can excite stable triple oxygen (3O2) to sin-gle oxygen (1O2Þ, is a pro-oxidant. Some individual-cow milks, which undergo spontaneous oxidative ran-cidity (i.e., without contamination with metals or expo-sure to light) contain � 10 times the normal level ofXO, and spontaneous oxidation can be induced in nor-mal milk by the addition of XO to about four timesnormal levels. Heat-denatured or flavin-free enzyme isineffective in oxidation; the susceptibility of unsatu-rated fatty acids to oxidation increases with the degreeof unsaturation (8).

3. Atherosclerosis

It has been suggested that XO from homogenized milkenters the vascular system and may be involved inatherosclerosis via oxidation of plasmalogens in cellmembranes; this aspect of XO attracted considerableattention in the early 1970s (83). However, the experi-mental evidence in support of this view is very weakand the hypothesis has been disclaimed (8, 70, 84, 85).

4. Reduction of Nitrate in Cheese

Sodium nitrate is added to milk for Dutch, Swiss, andother cheese varieties to prevent the growth ofClostridium tyrobutyricum which causes flavor defectsand late gas blowing in these cheeses. Xanthine oxidasereduces nitrate to nitrite which is necessary for thebacteriocidal effect of nitrate.

Various oxidation-reduction reactions occur in milkwhich affect its flavor. Perhaps XO is significant inthese reactions, either directly or indirectly.

5. Production of H2O2

The H2O2 produced by the action of xanthine oxidaseon oxidation of xanthine, hypoxanthine, or other sub-strates can serve as a substrate for lactoperoxidase inits action as a bacteriocidal agent (see Sec. XIII.B).

X. SULFHYDRYL OXIDASE (EC 1.8.3.–)

Milk contains an enzyme, sulfhydryl oxidase (SO),capable of oxidizing sulfhydryl groups of cysteine, glu-tathione, and proteins to the corresponding disulfide(for reviews, see 8, 70). The enzyme is an aerobicoxidase which catalyzes the following reaction:

2RSHþO2! RSSRþH2O2 ð5Þ


It undergoes marked self-association and can bepurified readily by chromatography on porous glass.The enzyme has a molecular weight of � 89 kDa, a pHoptimum of 6.8–7.0, and a temperature optimum of35C. Its amino acid composition, its requirement foriron but not for molybdenum and FAD, and the cat-alytic properties of the enzyme indicate that sulfhydryloxidase is a distinct enzyme from xanthine oxidase andthiol oxidase (EC 1.8.3.2).

SO is capable of oxidizing reduced ribonuclease andrestoring enzymatic activity, suggesting that its physio-logical role may be the nonrandom formation of pro-tein disulfide bonds, e.g., during protein biosynthesis.

SO immobilized on glass beads has the potential toameliorate the cooked flavor arising from sulfhydrylgroups exposed by protein denaturation on UHT pro-cessing of milk, but the commercial viability of thissystem is not known. In any case, sulfhydryl groupsare effective antioxidants and may serve a beneficialrole in UHT milk.

The production of sulfur compounds is believed tobe very important in flavor development in Cheddarand other varieties of cheese. Residual sulfhydryl oxi-dase activity may play a role in reoxidizing sulfhydrylgroups exposed upon heating cheesemilk; the sulfhy-dryl groups thus protected may be reformed during theripening process.

XI. SUPEROXIDE DISMUTASE

(EC 1.15.1.1)

Superoxide dismutase (SOD) scavenges superoxideradicals, O�2, according to the reaction:

2O�2 þ 2Hþ ! H2O2 þO2 ð6ÞThe H2O2 formed may be reduced to H2OþO2 bycatalase, peroxidase or suitable reducing agents. SODhas been identified in many animals and bacterial cells.Its biological function is to protect tissue against oxy-gen free radicals in anaerobic systems (8, 9, 70).

SOD, isolated from bovine erythrocytes, is a blue-green protein due to the presence of copper, removal ofwhich by treatment with EDTA results in loss of activ-ity which is restored by adding Cu2þ; it also containsZn2þ, which does not appear to be at the active site.The enzyme, which is very stable in 9 M urea at neutralpH, consists of two identical subunits of molecularweight 16 kDa held together by one or more disulfidebonds.

Milk contains trace amounts of SOD which is pre-sent exclusively in the skim milk fraction. This enzyme

has been isolated and characterized. It appears to beidentical to the bovine erythrocyte enzyme. Assaymethods for SOD are described by Stauffer (86).

SOD inhibits lipid oxidation in model systems. Thelevel of SOD in milk parallels that of XO (but at alower level), suggesting that SOD may be excreted inmilk in an attempt to offset the pro-oxidant effect ofXO. Attempts have been made to correlate the stabilityof milk to oxidative rancidity with the SOD activity inthe milk, but these results have been equivocal. Milkcontains several pro- and antioxidants, including exo-genous factors such as light, the precise balance ofwhich, rather than any single factor, determines overalloxidative stability. The possibility of using exogenousSOD to retard or inhibit lipid oxidation in dairy pro-ducts has been considered. A marked improvement inthe oxidative stability of milk with a high level of lino-leic acid was achieved by adding low levels of SOD.

SOD is more heat stable in milk than in purifiedpreparations. In milk it is stable at 71C for 30 min(i.e., it is not affected by HTST pasteurization) butloses activity rapidly at even slightly higher tempera-tures. Slight variations in pasteurization temperatureare therefore critical to the survival of SOD in heatedmilk products and may contribute to variations in thestability of milk to oxidative rancidity.

XII. CATALASE (EC 1.11.1.6)

An indigenous catalase in milk was first recognized in1907. The catalase activity of whole milk is associatedeither with membranes in the skim milk phase or theMFGM. The pellet obtained from buttermilk on cen-trifugation at 10,000 g is a particularly rich source,from which catalase has been highly purified and crys-tallized (8, 9, 70).

Milk catalase is a heme protein with a molecularweight of 200 kDa and an isoelectric pH of 5.5. It isstable between pH 5 and 10 but rapidly loses activityoutside the range. Heating at 70C for 1 h at pH 7.0causes complete inactivation. Like other catalases, it isstrongly inhibited by Hg2þ, Fe2þ, Cu2þ, Sn2þ, CN�,and NO�3 . Analytical methods for catalase aredescribed by Stauffer (86).

Catalase activity in milk varies with feed, stage oflactation, and especially with mastitic infection, forwhich it may be used as an index. However, it is notusually used for this purpose, since somatic cell countand N-acetylglucosaminidase activity are superiorindices of mastitis. Catalase may act as a lipid pro-


oxidant via its heme iron (i.e., nonenzymatically), butit is probably not very significant.

There is general agreement that cheese made fromraw milk ripens more quickly and develops a moreintense (although not always a more desirable) flavorthan cheese made from pasteurized milk (87).However, for public health reasons and in the interestof producing a consistent product, pasteurized milk isnow generally used for cheese making. Many cheesesare still made from raw milk, especially in southernEurope. Many countries require that if raw milk isused for cheese making, the cheese must be ripenedfor at least 60 days during which it was presumedthat pathogens die off, a presumption that is now con-sidered to be invalid. Subpasteurized or thermized milk(e.g., heated at 63–65C for 16 sec) has been consideredas a compromise between raw and pasteurized milk forcheese making. The acceptance of such a practicewould require an appropriate validation test, but nosuitable test is available at present for identifying ther-mized milk. The possibility of using the inactivation ofcatalase as an index of thermized milk was investigatedby Hirvi and Griffiths (88). Although catalase was auseful index of thermization of milk (it was almostcompletely inactivated by heating at 65C for 16 sec),it was unsuitable as an index of cheese made fromthermized milk owing to the production of catalasein the cheese during ripening, especially by yeasts.The thermal inactivation of catalase was also studiedby Hirvi et al. (89).

XIII. LACTOPEROXIDASE (EC 1.11.1.7)

The occurrence of a peroxidase, lactoperoxidase(LPO), in milk was recognized as early as 1881. It isone of the most heat-stable enzymes in milk. Itsinactivation was used as an index of flash pasteuriza-tion (now very rarely used) and is now used as anindex of super-HTST pasteurization, e.g., tempera-tures > 75C for 15 sec. LPO was first isolated in1943; several isolation procedures have since beenpublished (8, 90).

LPO is a heme protein containing � 0:07% Fe, withan absorbance peak (Soret band) at 412 nm(A412=A280 � 0:9). The pH optimum is � 8:0, itsmolecular weight is 77.5 kDa, and it consists of twoidentical subunits. Two principal forms (A and B)occur, each of which exhibits microheterogeneity withregard to amide groups (glutamine and/or asparagine)and carbohydrate content, giving a total of 10 variants.

A. Assay of Lactoperoxidase

The principle generally used in assays for peroxidasesis the use of a chromogenic or fluorogenic reducingagent, AH2, a number of which have been used (86).A highly recommended substrate for lactoperoxidase is2; 2 0-azinobis(3-ethylbenzylthiazoline-6-sulfonic acid)[ABTS] (91).

B. Significance

1. As for many other indigenous enzymes, thelevel of LPO in milk increases on mastitic infectionand is therefore a possible index of mastitis; however,it is not well correlated with somatic cell count, andsuperior methods, including enzyme-based methods(Sec. VII) are available to monitor mastitis.

2. LPO causes nonenzymic oxidation of un-saturated lipids, probably due to its heme group.The heat-denatured enzyme is more active than thenative enzyme. Compared with other pro-oxidants,LPO is probably not significant in milk and dairyproducts.

3. LPO has been used in the Storch test for flashpasteurized milk (67), but this process is not used inmodern milk processing. However, the pasteurizationof milk at a temperature higher than the HTST mini-mum, i.e., > 72C for 15 sec, has recently becomequite common. The assay based on the inactivationof alkaline phosphatase is not applicable under suchcircumstances. Griffiths (67), who evaluated the suit-ability of several indigenous enzymes as indices of thesuper-pasteurization of milk, concluded that assay ofLPO activity was the most promising method fordetecting heat treatments in the order of 76C for15 sec. Andrews et al. (68) reported the results of agenerally similar study, in which LPO was notincluded. They conclude that N-acetylglucosamini-dase, �-glutamyl transpeptidase (�-GGTP), and �-mannosidase or xanthine oxidase may be the mostsuitable indicators of heat treatment of 65–75, 70–80, and 80–90C, respectively. The relative suitabilityof GGTP and LPO were compared by McKellar et al.(57). The kinetics of the thermal denaturation of LPOwas reported by Martin-Hernandez et al. (92).

4. Milk contains bacteriostatic or bacteriocidalsubstances, referred to as lactenins. One of these isLPO, which requires H2O2 and thiocyanate (SCN�)to cause inhibition. The nature, mode of action, andspecificity of the LPO-SCN�-H2O2 system have beenwidely studied. LPO and thiocyanate, which is pro-duced in the rumen by enzymatic hydrolysis of thio-


glycosides from Brassicae plants, occur naturally inmilk although at variable and probably suboptimallevels. Milk does not contain indigenous H2O2,which can be generated metabolically by catalase-negative bacteria, or produced in situ through theaction of exogenous glucose oxidase on glucosewhich may be added to milk or produced in situfrom lactose by exogenous �-galactosidase or theaction of indigenous XO on added xanthine orhypoxanthine, or from added sodium percarbonateor it may be added directly.

Immobilized glucose oxidase has been used to gen-erate H2O2 in situ in thiocyanate- and glucose-enrichedmilk or whey. A self-contained LPO-H2O2-SCN

� sys-tem using coupled �-galactosidase and glucose oxidase,immobilized on porous glass beads, to generate H2O2

in situ from lactose in milk containing 0.25 mM thio-cyanate has been developed.

The bacteriocidal effect of the LPO-H2O2-SCN�

system has several applications which have beenreviewed extensively (90, 91, 93, 94):

1. Sanitization of immobilized enzyme columns inwhich LPO is coimmobilized with the enzyme of inter-est. The requisite H2O2 could be generated by immo-bilized enzymes—e.g., xanthine oxidase, glucoseoxidase, or �-galactosidase/glucose oxidase.

2. As a bacteriocidal agent in toothpaste.3. In the therapy of mastitis during the non-lactat-

ing period.4. For the preservation of milk in regions lacking

refrigeration or pasteurization facilities.5. To reduce the incidence of enteritis in calves or

piglets fed milk replacers. The indigenous LPO is inac-tivated during the manufacture of these products andLPO isolated from milk or whey is added.

LPO (and lactoferrin) is cationic at the natural pHof milk at which all the principal proteins are anionic.LPO and lactoferrin can therefore be readily isolatedfrom milk or whey by using cation exchange resins.The ion exchangers may be used as columns (95),batchwise (96), or as membranes (97). At least someof these methods are applicable on an industrial scale,making it possible to isolate LPO for use as a foodingredient.

C. Biochemistry of Lactoperoxidase System

LPO catalyzes the peroxidation of �SCN to productswhich are nontoxic to mammalian cells but which killor inhibit the growth of many species of microorgan-isms. The net reaction is:

H2O2 þ� SCNðthiocyanate anionÞ

�!LPO �OSCNþH2O

(hypothiocyanite anion)

ð7ÞMicrobial membranes have low permeability for�OSCN but are quite permeable to HOSCNðpKa ¼ 5:3). HOSCN or �OSCN oxidizes sulfhydrylgroups:

RSHþ� OSCN�!R-S-SCNþ �OH

(sulfenyl thiocyanate) ð8Þ

R-S-SCNþH2O�!R-S-OHþHþ þ �SCN(sulfonic acid) ð9Þ

Any reaction involving a sulfhydryl group, e.g., thiolenzymes, will be inhibited by this oxidation. The effectmay be reversed by thiol compounds such as glu-tathione or cysteine.

XIV. OTHER ENZYMES

In addition to the enzymes described above, severalother indigenous enzymes have been isolated and par-tially characterized (Table 2) (70). Although a fairlyhigh level of some of these enzymes occurs in milk,they have no apparent function or significance inmilk which contains no substrate for many of them.It is possible that some of these enzymes will assumeimportance in the future as indices of animal health orof product quality. These enzymes will not be discussedfurther.

Nearly 40 other enzymatic activities have beendetected in milk (Table 3) but have not been isolated,and only limited information on their molecular andbiochemical properties in milk is available (70). Someof these enzymes have been evaluated as indices of theheat treatment of milk (67, 68). Perhaps further studywill identify some important or useful attributes ofsome of these enzymes.

XV. SUMMARY

Although milk contains � 60 indigenous enzymes,only two, lipoprotein lipase (LPL) and plasmin, arereally technologically significant. LPL has the potentialto cause serious problems but these can be avoidedthrough good milking practices on the farm. Theenzyme is relatively heat labile and hence does notcause problems in heat-treated products. Althoughmilk contains considerable potential plasmin, relatively


little is expressed owing to the presence of inhibitors.The significance of plasmin is mainly negative but itsaction is probably positive in cheese during ripening.Acid phosphatase may be significant in the depho-sphorylation of phosphopeptides in cheese. Severaloxidoreductases—xanthine oxidase, lactoperoxidase,catalase, superoxide dimutase, and sulfhydryl oxi-dase—are, at least potentially, significant in milk sta-bility. Lysozyme and lactoperoxidase may haveantimicrobial effects in the intestine of the consumer.

Perhaps the most significant and most useful aspectof the indigenous enzymes in milk is as indicators ofmastitis, especially N-acetylglucosaminidase, and ofthermal treatments. The thermal stabilities of the indi-genous enzymes cover quite a wide range of tempera-tures which make it possible to determine at whattemperature milk has been heat-treated in the range65–90C for 15 sec. The principal advantage ofenzymes compared with other heat-induced changesin analytical applications is the relative ease withwhich their activity can be quantified.

Most dairy products should undergo no change dur-ing storage, and hence the indigenous enzymes capableof causing undesirable changes are inactivated by heat-ing. Cheese is an exception; a very complex series ofmicrobiological, biochemical, and perhaps chemicalreactions occur which lead to desirable characteristictaste, aroma, and texture of the finished cheese (98,99). At least four indigenous enzymes contribute tocheese ripening—plasmin, lipoprotein lipase, acidphosphatase, and xanthine oxidase. Perhaps otherscontribute, but information is lacking.

Most of the indigenous enzymes in milk remain tobe isolated and characterized. As for the principalenzymes, these minor enzymes probably originatefrom the mammary tissue. Most of these enzymes areunlikely to be technologically significant as milk con-tains no substrates. However, they may be significantas indicators of animal health or thermal history; if so,they will be isolated and characterized.

REFERENCES

1. AT Andrews, T Olivecrona, S Vilaro, G Bengtsson-

Olivecrona, PF Fox, L Bjorck, NY Farkye.

Indigenous enzymes in milk. In: PF Fox, ed.

Advanced Dairy Chemistry, Vol. 1, Proteins.

London: Elsevier Applied Science, 1992, pp 285–267.

2. RM Whitney. The minor proteins of milk. J Dairy Sci

41:1303–1323, 1958.

3. KM Shahani. Milk enzymes: their role and signifi-

cance. J Diary Sci 49:907–920, 1966.

4. ML Groves. Minor milk proteins and enzymes. In:

HA McKenzie, ed. Milk Proteins: Chemistry and

Molecular Biology, Vol. 1. New York: Academic

Press, 1971, pp 367–418.

5. KM Shahani, WJ Harper, RJG Jensen, RM Parry Jr,

CA Zittle. Enzymes of bovine milk. J Dairy Sci

56:531–543, 1973.

6. BK Dwivedi. The role of enzymes in food flavors. Part

I. Dairy products. CRC Crit Rev Food Technol

3:457–478, 1973.

7. HA Johnson. The composition of milk. In: BH Webb,

AH Johnson, JA Alford, eds. Fundamentals of Dairy

Chemistry, 2nd ed. Westport, CT: AVI Publishing,

1974, pp 1–57.

8. PF Fox, PA Morrissey. Indigenous enzymes of bovine

milk. In: GG Birch, N Blakeborough, KJ Parker, eds.

Enzymes and Food Processing. London: Applied

Science, 1981, pp 213–238.

9. BJ Kitchen. Indigenous milk enzymes. In: PF Fox, ed.

Developments in Dairy Chemistry. 3. Lactose and

Minor Constituents. London: Elsevier Applied

Science, 1985, pp 239–279.

10. AT Andrews, T Olivecrona, G Bengtsson-Olivecrona,

PF Fox, L Bjorck, NY Farkye. Indigenous enzymes in

milk. In: PF Fox ed. Food Enzymology. London:

Elsevier Applied Science, 1991, pp 53–129.

11. M Hamos, LM Freed, JB Jones, SE Berkow, J

Bitman, NR Mehta, B Happ, P Hamos. Enzymes in

human milk. In: RG Jensen, MC Neville, eds. Human

Lactation: Milk Components and Methodologies.

New York: Plenum Press, 1985, pp 251–266.

12. O Hernell, B Lonnerdal. Enzymes in human milk. In:

SA Atkinson, B Lonnerdal, eds. Proteins and Non-

protein Nitrogen in Human Milk. Boca Raton, FL:

CRC Press, 1989, pp 67–75.

13. MB Grufferty, PF Fox. Milk alkaline proteinase. J

Dairy Res 55:609–630, 1988.

14. ED Bastian, RJ Brown. Plasmin in milk and dairy

products: an update. Int Dairy J 6:435–457.

15. C Benfeldt, LB Larsen, JT Rasmussen, PA Andersen,

TE Petersen. Isolation and characterization of plasmi-

nogen and plasmin from bovine milk. Int Dairy J

5:577–592, 1995.

16. A Baldi, G Savoini, F Chali, F Fantuz, E Senatare, L

Bertocchi, I Politis. Changes in plasmin-plasminogen-

plasminogen activator system in milk from Italian

Friesian herds. Int Dairy J 6:1045–1053, 1996.

17. P Guinot-Thomas, M Al Ammoury, Y La Roux, F

Laurent. Study of proteolysis during storage of milk

at 4C: effect of plasmin and microbial proteinases.

Int Dairy J 5:685–697, 1995.

18. L Berglund, MD Andersen, TE Petersen. Cloning and

characterization of the bovine plasminogen cDNA.

Int Dairy J 5:593–603, 1995.

19. A Kennedy, AL Kelly. The influence of somatic cell

count on the heat stability of bovine plasmin activity.

Int Dairy J 7:717–721, 1997.


20. BC Richardson, KN Pearce. The determination of

plasmin in dairy products. NZ J Dairy Sci Technol

16:209–220, 1981.

21. P Fox, PLH McSweeney. Proteolysis in cheese during

ripening. Food Rev Int 12:457–509, 1996.

22. NY Farkye, PF Fox. Observations on plasmin activity

in cheese. J Dairy Res 57:413–418, 1990.

23. N Farkye, PF Fox. Contribution of plasmin to

Cheddar cheese ripening: effect of added plasmin. J

Dairy Res 59:209–216, 1992.

23a. FM Barrett, AL Kelly, PLH McSweeney, PF Fox.

Use of exogeneous urokinase to accelerate proteolysis

in Cheddar cheese during ripening. Int Dairy J 9:473–

479, 1999.

24. S Christensen, T Wiegers, J Hermansen, L Sottrup-

Jensen. Plasma-derived protease inhibitors in bovine

milk. Int Dairy J 5:439–449, 1995.

25. C Benfeldt, J Sorensen, TE Petersen. Inhibitory activ-

ity against plasmin, trypsin, and elastase in rennet

whey and in cheese fortified with whey protein. J

Dairy Sci 81:615–620, 1998.

26. CW Heegaard, LK Rasmussen, PA Andreasen. The

plasminogen activation system in bovine milk.

Differential localization of tissue-type plasminogen

activator and urokinase in milk fractions is caused

by binding to casein and urokinase receptor.

Biochim Biophys Acta 1222:45–55, 1994.

27. JH White, B Zavizion, K O’Hare, J Gilmore, MR

Guo, P Kindstedt, I Politis. Distribution of plasmino-

gen activator in different fractions of bovine milk. J

Dairy Res 62:113–122, 1995.

28. P Ravn, L Berglund, TE Petersen. Cloning and char-

acterization of the bovine plasminogen activators uPA

and tPA. Int Dairy J 5:605–617, 1995.

29. DD Lu, SS Nielsen. Isolation and characterization of

native bovine plasminogen activators. J Dairy Sci

76:3369–3383, 1993.

30. JA Gilmore, JH White, B Zavizion, I Politis. Effects

of stage of lactation and somatic cell count in plasmi-

nogen activator in bovine milk. J Dairy Res 62:141–

145, 1995.

31. PLH McSweeney, PF Fox. Proteolysis of bovine case-

ins by cathepsin D: preliminary observations and com-

parison with chymosin. Int Dairy J 5:321–336, 1995.

32. EH Reimerdes. New aspects of naturally occurring

proteases in bovine milk. J Dairy Sci 66:1591–1600,

1983.

33. RJ Verdi, DM Barbano. Preliminary investigation of

the properties of somatic cell proteases. J Dairy Sci

72:534–538, 1988.

34. RJ Verdi, DM Barbano. Effect of coagulants, somatic

cell enzymes, and extracellular bacterial enzymes on

plasminogen activation. J Dairy Sci 74:772–782, 1991.

35. RJ Verdi, DM Barbano. Properties of proteases from

milk somatic cells and blood leukocytes. J Dairy Sci

74:2077–2091, 1991.

36. J Suzuki, N Katoh. Cysteine protease in bovine milk

capable of hydrolyzing casein as the substrate and

elevation of the activity during the course of mastitis.

Jpn J Vet Sci 52:947–954, 1990.

37. PA Grieve, BJ Kitchen. Proteolysis in milk: the sig-

nificance of proteinases originating from milk leuco-

cytes and a comparison of the action of leucocyte,

bacteria and natural milk proteinases on the caseins.

J Dairy Res 52:101–112, 1985.

38. AL Kelly. Variations in total and differential milk

somatic cell counts and plasmin levels and their role

in proteolysis and quality of milk and cheese. PhD

thesis, National University of Ireland, Cork, 1995.

39. A Kelly, J Sheehan, D Tiernan, A Shattock, P Joyce, J

Foley. Bovine milk with PMN content and its effects

on cheese quality. Proc 3rd Mastitis Seminar, Tel

Aviv, Section S-3, 1995, pp 31–35.

40. W Fang, M Sandholm. Inhibition of the proteinase

activity in mastitic milk. J Dairy Res 62:61–68, 1995.

41. PF Fox, NP Tarassuk. Bovine milk lipase. I. Isolation

from skimmilk. J Dairy Sci 51:826–833, 1968.

42. CV Patel, PF Fox, NP Tarassuk. Bovine milk lipase.

II. Characterization. J Dairy Sci 51:1879–1886, 1968.

42a. ED Korn. The lipoprotein lipase of cow’s milk. J

Lipid Res 3:246–250, 1962.

43. T Egelrud, T Olivecrona. The purification of a lipo-

protein lipase from bovine milk. J Biol Chem

247:6212–6217, 1972.

44. T Olivecrona, S Vilaro, G Bengtsson-Olivecrona.

Indigenous enzymes in milk: lipases. In: PF Fox, ed.

Advanced Dairy Chemistry. 1. Proteins. London:

Elsevier Applied Science, 1992, pp 292–310.

45. HC Deeth, CH Fitz-Gerald. Lipolytic and hydrolytic

rancidity in milk and milk products. In: PF Fox, ed.

Advanced Dairy Chemistry. 2. Lipids, 2nd ed.

London: Chapman & Hall, 1995, pp 247–308.

46. RM Rocco. Fluorometric analysis of alkaline phos-

phatase in fluid dairy products. J Food Protect

53:588–592, 1990.

47. KF Eckner. Fluorometric analysis of alkaline phos-

phatase inactivation correlated to Salmonella and

Listeria inactivation. J Food Protect 55:960–963,

1992.

48. GK Murthy, S Cox, L Kaylor. Reactivation of alka-

line phosphatase in ultra high-temperature, short-time

processed liquid milk products. J Dairy Sci 59:1699–

1710, 1976.

49. JW Copius Peereboom. Studies on alkaline milk phos-

phatase III: a theory for the mechanism of reactiva-

tion of alkaline milk phosphatase in a model system.

Fette Seifen Anstrichmittel 72:299–308, 1970.

50. G Linden. Biochemical study of some aspects of milk

alkaline phosphatase reactivation. Milchwissenschaft

34:329–332, 1979.

51. EF McFarren, RC Thomas, LA Black, JE Campbell.

Differentiation of reactivated from residual phospha-


tase in high-temperature short-time pasteurized milk

and cream. J Assoc Off Anal Chem 43:414–426, 1960.

52. GK Murthy, JT Peeler. Rapid methods for differen-

tiating reactivated from residual phosphatase in milk

and cream: collaborative study. J Assoc Off Anal

Chem 62:822–827, 1979.

53. AOAC. Official method 961.08: phosphatase

(reactivated and residual) in milk. Official Methods

of Analysis, 16th ed., Vol. 2. Arlington, VA:

Association of Official Analytical Chemists, 1995, p 36.

54. WS Kwee. Phosphatase reactivation in cream samples.

Aust J Dairy Technol 38:160–162, 1983.

55. R Karmas, DH Kleyn. Determination and interpreta-

tion of alkaline phosphatase activity in experimental

and commercial butters. J Dairy Sci 73:584–589, 1990.

56. RC McKellar, HW Modler, H Couture, A Hughes, P

Mayers, T Gleeson, WH Ross. Predictive modeling of

alkaline phosphatase inactivation in a high tempera-

ture short time pasteurizer. J Food Protect 57:424–

430, 1994.

57. RC McKellar, S Liou, HW Modler. Predictive model-

ing of lactoperoxidase and �-glutamyl transpeptidase

inactivation in a high-temperature short-time pasteur-

izer. Int Dairy J 6:295–301, 1996.

58. PF Fox, PLH McSweeney. Proteolysis in cheese dur-

ing ripening. Food Rev Int 12:457–509, 1996.

59. TK Singh, PF Fox, A Healy. Water soluble peptides

in Cheddar cheese: isolation and identification of pep-

tides in the diafiltration retenate of the water-soluble

fraction. J Dairy Res 62:629–640, 1995.

60. TK Singh, PF Fox, A Healy. Isolation and identifica-

tion of further peptides in the diafiltration retentate of

the water-soluble fraction of Cheddar cheese. J Dairy

Res 64:433–443, 1997.

61. F Addeo, L Chianese, R Sacchi, S Spagna Musso, P

Ferranti, A Malorni. Characterization of the oligo-

peptides of Parmigiano Reggiano cheese soluble in

120 g trichloroacetic acid/L. J Diary Res 61:365–374,

1994.

62. P Ferranto, F Barone, L Chianese, F Addeo, A

Scaloni, L Pellegrino, P Resmini. Phosphopeptides

from Grana Padano cheese: nature, origin and

changes during ripening. J Dairy Res 64:601–615,

1997.

63. L Pellegrino, G Batteli, P Resmini, P Ferranti, F

Barone, F Addeo. Effects of heat load gradient occur-

ring in moulding on characterization and ripening of

Grana Padano. Lait 77:217–228, 1997.

64. I de Noni, L Pellegrino, P Resmini, P Ferranti. About

presence of free phosphoserine in ripened cheese and

in enzymatic hydrolysate of casein. Nahrung 41:268–

273, 1997.

65. AT Andrews. Indigenous milk enzymes: phospha-

tases. In: PF Fox, ed. Advanced Dairy Chemistry. 1.

Proteins. London: Elsevier Applied Science, 1992, pp

322–331.

66. AT Andrews, E Alichanidis. The acid phosphatases of

bovine leucocytes, plasma and the milk of healthy and

mastitic cows. J Dairy Res 42:391–400, 1975.

67. MW Griffiths. Use of milk enzymes as indices of heat

treatment. J Food Protect 49:696–705, 1986.

68. AT Andrews, M Anderson, PW Goodenough. A

study of the heat stabilities of a number of indigenous

milk enzymes. J Dairy Res 54:237–246, 1987.

69. P Jolles, J Jolles. Lysozyme from human milk. Nature

192:1187–1188, 1961.

70. NY Farkye. Indigenous enzymes in milk: other

enzymes. In: PF Fox, ed. Advanced Dairy

Chemistry. 1. Proteins. London: Elsevier Applied

Science, 1992, pp 339–367.

71. PF Fox, MB Grufferty. Exogenous enzymes in dairy

technology. In: PF Fox, ed. Food Enzymology, Vol. 1.


72. PF Fox. Exogenous enzymes in dairy technology: a

review. J Food Biochem 17:173–199, 1993.

73. PF Fox, L Stepaniak. Enzymes in cheese technology.

Int Dairy J 3:509–530, 1993.

74. CR Baumrucker. �-Glutamyl transpeptidase of

bovine milk membranes: distribution and characteri-

zation. J Dairy Sci 62:253–258, 1979.

75. CR Baumrucker. Purification and identification of �-glutamyl transpeptidase of milk membranes. J Dairy

Sci 63:49–54, 1980.

76. RC McKellar, DB Emmons, J Farber. Gamma-gluta-

myl transpeptidase in milk and butter as an indicator

of heat treatment. Int Dairy J 1:241–251, 1991.

77. SS Patel, RA Wilbey. Heat exchanger performance:

gamma-glutamyl transpeptidase assay as a heat treat-

ment indicator for dairy products. J Soc Dairy

Technol 42:79–80, 1989.

78. HD Carter, CF Cavanagh, JL Higgins, RA Wilbey.

Assessment of the heat treatment of ice cream mixes

by enzyme assay. J Soc Dairy Technol 43:67–68, 1990.

79. RC McKellar. Influence of ice-cream mix components

on the thermal stability of �-glutamyl transpeptidase

and Listeria innocua. Int Dairy J 6:1181–1189, 1996.

80. Roudot-Algaron, L Kerhoas, D Le Bars, J Einhorn, J-

C Gripon. Isolation of �-glutamyl peptides from

Comte cheese. J Dairy Sci 77:1161–1166, 1994.

81. J Cerbulis, HM Farrell Jr. Xanthine oxidase activity

in dairy products. J Dairy Sci 60:170–176, 1977.

82. GF Edwards. A new method for the detection of

heated or reconstituted milk. J Assoc Off Agric

Chem 33:855–960, 1950.

83. KA Oster. Plasmalogen diseases: a new concept of the

atherosclerotic process. Am J Clin Res 2:30–35, 1971.

84. AJ Clifford, CY Ho, H Swenerton. Homogenized

milk xanthine oxidase: a critique of the hypothesis

relating to plasminogen depletion and cardiovascular

disease. Am J Clin Nutr 38:327–332, 1983.

85. HC Deeth. Homogenized milk and atherosclerotic dis-

ease: a review. J Dairy Sci 66:1419–1435, 1983.


86. CE Stauffer, ed. Enzyme Assays for Food Scientists.

New York: Van Nostrand Reinhold, 1989.

87. PF Fox, PLH McSweeney, CM Lynch. Significance of

non-starter lactic acid bacteria in Cheddar cheese.


88. Y Hirvi, MW Griffiths. Milk catalase activity as an

indicator of thermization treatments used in the man-

ufacture of Cheddar cheese. J Dairy Sci 81:338–345,

1998.

89. Y Hirvi, MW Griffiths, RC McKellar, HW Modler.

Linear transform and non-linear modelling of bovine

catalase inactivation in a high-temperature short-time

pasteurizer. Food Res Int 29:89–93, 1996.

90. L Bjorck. Lactoperoxidase. In: PF Fox, ed. Advanced

Dairy Chemistry. 1. Proteins. London: Elsevier

Applied Science, 1992, pp 332–338.

91. KM Pruitt, DN Kamau. Quantitative analysis of

bovine lactoperoxidase system components and of

the effects of the activated system on bacterial growth

and survival. In: Indigenous Antimicrobial Agents of

Milk: Recent Developments, Special Issue 9404.

Brussels: International Dairy Federation, 1994,

pp 73–87.

92. MC Martin Hernandez, BW van Markwijk, HJ

Vreeman. Isolation and properties of lactoperoxidase

from bovine milk. Neth Milk Dairy J 44:213–231,

1990.

93. B Reiter. The biological significance of the non-immu-

noglobulin protective proteins in milk: lysozyme,

lactoferrin, lactoperoxidase. In: PF Fox, ed.



Science, 1985, pp 281–336.

94. J Stadhouders, RR Beumer. Actual and potential

applications of the natural antimicrobial agents of

milk in the dairy industry. In: Indigenous

Antimicrobial Agents of Milk: Recent developments,

Special Issue 9404. Brussels: International Dairy

Federation, 1994, pp 175–197.

95. PF Fox, DM Mulvihill. Developments in milk protein

processing. Food Sci Technol Today 7:152–161, 1993.

96. R Mitchell, GW Smithers, DA Dionysius, PA Grieve,

GORegester, EA James. Extraction of lactoperoxidase

and lactoferrin from cheese whey using membrane

cation exchangers. In: Indigenous Antimicrobial

Agents of Milk: Recent Developments, Special Issue

9404. Brussels: International Dairy Federation, 1994,

pp 89–95.

97. DA Dionysius, JB Herse, PA Grieve. Extraction of

lactoperoxidase and lactoferrin from cheese whey

using batch ion exchange techniques. Aust J Dairy

Technol 46:72–75, 1991.

98. PF Fox, TP O’Connor, PLH McSweeney, TP Guinee,

NM O’Brien. Cheese: physical, biochemical and nutri-

tional aspects. Adv Food Nutr Res 39:163–328, 1996.

99. PF Fox, JMWallace. Formation of flavor compounds

in cheese. Adv Appl Microbiol 45:17–85, 1997.


20

Exogenous Enzymes in Dairy Technology

Patrick F. Fox

University College, Cork, Ireland

I. INTRODUCTION

Milk from domesticated animals and products formedtherefrom have been components of the human dietfor � 10,000 years. Some dairy products are con-sumed in most or all regions of the world and theyare major dietary items in many regions, e.g., Europe,North America, and South America. Total milkproduction is � 530� 106 metric tons per year, ofwhich � 85% is bovine milk; sheep, goats, water buf-falo, camels, and mares are important dairy animalsin some regions. Milk is a very perishable commodityand hence there has been a strong incentive toconvert it to more stable products, which is facilitatedby certain properties of milk. Classically, milk hasbeen preserved by fermentation, usually with saltaddition (cheese, fermented milks, butter). Newerpreservation methods include drying, pasteurization/sterilization, and freezing. Some characteristics ofmilk also render it very amenable to modificationby enzymes; that milk is a liquid facilitates enzymeaddition. Cheese production is probably the oldest,and is still the largest, application of exogenousenzymes in food processing.

Since the principal components of milk are proteins(� 3:5%), lipids (� 3:6%), and lactose (� 4:8%; a dis-accharide containing galactose and glucose), the prin-cipal enzymes used in dairy technology are proteinasesand peptidases, lipases, and �-galactosidase (lactase).However, several oxidoreductases have significantapplications.

The principal applications of enzymes in dairy tech-nology will be considered in this review. Earlier reviewsinclude Fox and Grufferty (1), Fox (2), Fox andStepaniak (3), Brown (4), and Desmazeaud andSpinnler (5).

II. PROTEINASES

Bovine milk contains � 3:5% proteins which can beresolved into two groups based on solubility at pH4.6 and 20C: the caseins, which are insoluble underthese conditions and which represent � 80% of thetotal protein; and the whey (serum) proteins, whichare soluble. The casein fraction of bovine milk andthat of the other main dairying species comprisesfour proteins, �s1-, �s2-, �-, and �-caseins, which,although they have certain features in common, e.g.,insolubility at pH 4.6, are distinctly different proteins.The whey protein fraction also comprises four mainproteins—�-lactoglobulin, �-lactalbumin, bloodserum albumin, and immunoglobulins—and severalminor proteins, including � 60 indigenous enzymes.Readers are referred to Fox (6) for a detailed descrip-tion of the milk protein system.

As discussed in Section II.A.1, the colloidal stabilityof the caseins is extensively changed by limited proteo-lysis, leading to gelation of the milk system, which isthe first step in the production of many cheese vari-eties. The milk proteins can be easily separated fromthe other milk constituents and the caseins and whey


proteins readily separated from each other on anindustrial scale. Both the caseins and whey proteinshave good and diverse functional properties; conse-quently, milk proteins are the preferred functionalfood proteins for a wide range of applications (7).Some functional properties can be improved and/ormodified by limited proteolysis.

Proteolytic enzymes are the most widely usedenzymes in dairy technology and will be discussedbelow under three headings: cheese manufacture,modification of functional properties, and productionof protein hydrolyzates for nutritional and otherapplications.

A. Cheese Manufacture

The manufacture of all cheese varieties essentiallyinvolves concentrating the protein and fat of milk six-fold to 12-fold, depending on the variety.Concentration is achieved by (a) coagulating the prin-cipal milk proteins, i.e., the caseins (if present, the milkfat is occluded in the coagulum); (b) cutting or break-ing the coagulum and inducing it to synerese under theinfluence of heat and acid; (c) separation of curds andwhey; and (d) acidification, pressing, and salting of de-wheyed curd.

Coagulation of the casein is induced by one of threemethods:

1. Limited proteolysis by a crude proteinase(rennet), which is exploited in the manufacture of allripened and some fresh cheeses (� 75% of total pro-duction).

2. Isoelectric precipitation at � pH4:6, used forfresh cheeses, usually by in situ production of lacticacid by a culture (starter) of lactic acid bacteria andless frequently by direct acidification with preformedacid, usually HCl, or acidogen, usually gluconic acid–�-lactone.

3. Acid plus heat, i.e., acidification to � pH5:2 withacid whey, acid milk, citrus juice, vinegar, or aceticacid at 80–90C; this method is used to produce asmall number of relatively minor varieties, e.g.,Ricotta.

Concentration of the total colloidal phase of milk(i.e., fat and total protein) to the level present in cheese,i.e., to a ‘‘pre-cheese,’’ by ultrafiltration is now usedcommercially for the manufacture of several cheesevarieties. Additions of rennet and starter to theconcentrate are still necessary for texture and flavordevelopment.

1. Enzymatic Coagulation of Milk

The principal gastric enzyme of neonatal ruminants ischymosin rather than pepsin. Chymosin has low gen-eral proteolytic activity but high milk-clotting activity.Presumably, it evolved to coagulate milk in the sto-mach and thus delay its discharge into the intestineand increase the efficiency of digestion. Shortly afterthe domestication of dairy animals (� 8000 B.C.),humans learned to exploit the ability of chymosinand some other proteinases, collectively referred to asrennets, to coagulate milk for the production of cheese,which was probably the first application of enzymes infood processing.

The rennet coagulation of milk is a two-stage pro-cess. The first (primary) phase involves the enzymaticproduction of ‘‘para-casein’’ and TCA-soluble pep-tides (glycomacropeptides), while the secondaryphase involves the Ca-induced gelation of para-caseinat a temperature in the range of 30–35C. Proteolysis isessentially complete before the onset of coagulation.

The enzymatic coagulation of milk exploits certainproperties of the caseins. As discussed above, bovinecasein consists of four proteins—�s1-, �s2-, �-, and �-caseins—in the approximate ratio of 40 : 10 : 35 : 12.These contain 8–9, 10–13, 4–5, and 1–2 mol of P permol, respectively. Owing to their high phosphate con-tent, �s1-, �s2-, and �-caseins bind Ca2þ strongly andprecipitate at Ca2þ > 6mM. However, �-casein bindsCa2þ weakly and is soluble at high Ca2þ. It also reactshydrophobically with �s1-, �s2-, and �-caseins and canstabilize up to 10 times its weight of these Ca2þ-sensi-tive caseins against precipitation by forming colloidalaggregates, called micelles.

In milk, > 95% of the casein exists as micelles,which consist, on a dry weight basis, of � 94% proteinand 6% of other species, mainly Ca2þ and PO3�

4 withsome Mg2þ and citrate, collectively called colloidal cal-cium phosphate (CCP). The micelles are spherical, 50–600 nm (mean, � 120 nm) in diameter, with particleweights of � 108 daltons; i.e., a typical micelle contains� 5000 monomers (Mr ¼ 20–24 kDa). The micellestypically bind � 2 g H2O=g protein.

Within the micelles, the caseins are held together byCCP bridges, hydrophobic interactions, and hydrogenbonds. There is a widely held view that the caseinmonomers are organized as submicelles (spherical par-ticles, Mr � 5� 106 daltons). The micelles dissociatewhen CCP is removed (e.g. by Ca2þ chelators or acid-ification/dialysis), or when the pH is increased above� 9, or on addition of detergents (e.g., sodium dodecyl


sulfate) or urea. In most models of the casein micelle, itis envisaged that the Ca2þ-sensitive �s1- �s2-, and �-caseins interact hydrophobically to form the core ofthe micelles, with �-casein located predominantly onthe surface. The N-terminal two-thirds of �-casein ishydrophobic and reacts hydrophobically with the coreproteins, leaving the hydrophilic C-terminal regionprojecting into the surrounding environment. It hasbeen proposed that submicelles contain variableamounts of surface �-casein and aggregate such that�-casein-rich submicelles predominate at the surface ofthe micelles with the �-casein-deficient submicelles bur-ied within. The micelles are stabilized by a zeta poten-tial of � �20mV and by steric factors caused by theprotruding C-terminal segments of �-casein whichform a ‘‘hairy’’ layer, 7–10 nm thick, on the surfaceof the micelles, preventing close approach.

a. Primary Phase of Rennet Action. During theprimary phase of rennet action, �-casein is the onlyprotein hydrolyzed to a significant extent. It is cleavedby chymosin, and for most of the other proteinasesused as rennets, at the bond Phe105-Met106, which ismany times more susceptible to hydrolysis by acid pro-teinases (which include all commercial rennets) thanany other bond in the milk protein system. �-Caseinf1–105, which is referred to as para-�-casein, remainsattached to the rennet-altered micelle but the hydro-philic peptide, �-CN f106–169, referred to as the case-ino(glyco)macropeptide, diffuses into the whey andconsequently the micelle-stabilizing properties of �-casein are lost.

A number of attempts have been made to explainthe unique sensitivity of the Phe-Met bond. Di-, tri-, ortetrapeptides containing a Phe-Met bond are nothydrolyzed. However, the Phe-Met bond is hydrolyzedin the pentapeptide, H.Leu-Ser-Phe-Met-Ala-OMe,i.e., a derivative of �-CN f103–107. The length of thepeptide and the sequence around the cleavage site areimportant determinants of enzyme-substrate interac-tion. Ser104 is particularly important and its replace-ment by Gly or Ala in the above pentapeptiderenders the Phe-Met bond very resistant to hydrolysisby chymosin but not by pepsins. Even substituting D-Ser for L-Ser markedly reduces the sensitivity of theadjacent Phe-Met bond. Extension of the above penta-peptide from the N- and/or C-terminal to reproducethe sequence of �-casein increases the efficiency ofhydrolysis of the Phe-Met bond by chymosin; the tet-radecapeptide, �-CN f98–111, is hydrolyzed as effi-ciently as intact �-casein and � 66,000 faster than the

parent pentapeptide, �-CN f103–107, with a kcat=KM

of � 2M�1 sec�1.The two residues, Phe-Met, are not intrinsically

essential for chymosin action. Replacement of Phe byPheNO2 or cyclohexylalanine decreases kcat=KM aboutthreefold and � 50-fold, respectively. Oxidation ofMet106 decreases kcat=KM � 10-fold but substitutionof Ile for Met increases this ratio about threefold. Infact, the chymosin-susceptible bond in porcine orhuman �-caseins is Phe-Ile, which is readily hydrolyzedby calf chymosin. Thus, the sequence around the Phe-Met bond, rather than the bond itself, contains theimportant determinants of hydrolysis. The sequenceLeu103-Ser-Phe-Met-Ala-Ile108 of �-casein, which mayexist as a �-structure, fits into the active site cleft ofacid proteinases. The hydrophobic residues, Leu103,Phe105, Met106, and Ile108, probably interact withhydrophobic residues along the active site cleft whilethe hydroxyl group of Ser104 forms a hydrogen bridgewith a counterpart on the enzyme. Residues 98–102and 109–111 probably form �-turns around the edgesof the active site cleft in the enzyme � substrate com-plex; this conformation is stabilized by Pro residues atpositions 99, 101, 109, and 110. One or more of thethree His residues, 98, 100, 102, and Lys111, are prob-ably involved in electrostatic bonding.

Pepsins and most other acid proteinases used asrennets hydrolyze �-casein at Phe105-Met106 but theacid proteinase of Cryphonectria parasitica hydrolyzesSer104-Phe105. Although the specificity of cathepsin Don �s1- and �-caseins is generally similar to that ofchymosin, it has very poor milk-clotting activity. Theaggregation characteristics of micelles vary with therennet used, suggesting differences in the extent and/or specificity of the hydrolysis of �-casein or perhaps ofthe other caseins. Furthermore, the commonly usedrennets have markedly different specific activities onsynthetic �-casein-related peptides.

Reviews on the enzymatic coagulation of milkinclude Dalgleish (8, 9), Fox (10–12), Fox andMulvihill (13), and Fox and McSweeney (14).

b. Rennets. The rennets used to coagulate milkare crude preparations of selected proteinases. Manyproteinases can coagulate milk but most are too pro-teolytic relative to their milk clotting activity andhydrolyze the coagulum too quickly, causing reducedcheese yield and/or defective, e.g., bitter cheese.Traditionally, rennets were prepared from calves’,kids’, or lambs’ stomachs; the principal proteinase insuch rennets is chymosin. The molecular and enzy-


matic properties of chymosins have been studied exten-sively (15–17).

Owing to increasing world production of cheese(� 3% per year over the past 30 years), concomitantwith a reduced supply of calf vells, the supply of vealrennet has been inadequate for many years, which hasled to a search for rennet substitutes. Although manyproteinases can coagulate milk, only six have beenfound to be more or less acceptable as rennets: bovine,porcine, and chicken pepsins and the acid proteinasesfrom Rhizomucor miehei, R. pusillus, and C. parasitica.Chicken pepsin is the least suitable of these and is usedonly in special circumstances. Bovine pepsin gives gen-erally satisfactory results with respect to cheese yieldand quality; many commercial ‘‘calf rennets’’ contain asubstantial proportion of bovine pepsin. Although theproteolytic specificity of the three commonly used fun-gal rennets on �s1- and �-caseins is considerably differ-ent from that of calf chymosin, they generally yieldacceptable cheese and were widely used in the UnitedStates before the introduction of microbial recombi-nant chymosin. Acid proteinases from flowers of thegenus Cynara are used to coagulate sheep’s milk forsome artisinal cheeses in Portugal and Spain, especiallySerra d’Estrala cheese; these proteinases are not gen-erally suitable as rennets. The extensive literature onrennet substitutes has been reviewed (18–21).

The gene for calf chymosin has been cloned inselected bacteria, yeasts, and molds. Chymosin fromgenetically engineered Kluyveromyces marxianus var.lactis (Gist-brocades), Escherichia coli (Pfizer), andAspergillus nidulans (Hansens) is commercially avail-able and used extensively, with excellent results; how-ever, these products are not yet permitted in allcountries. Reviews on microbial recombinant chymo-sin include Teuber (22) and IDF (23).

Microbial recombinant chymosin preparations con-tain no pepsin whereas 5–50% of the milk-clottingactivity of calf rennets may be due to pepsin; hence,some minor differences in the pattern of proteolysis incheese made with microbial recombinant chymosin orcalf rennet are observed, most notably the formationof the peptide �s1-CN f110–199 in cheese made usingcalf rennet (owing to the action of pepsin). For thosewishing to simulate the action of calf rennet more clo-sely, blends of microbial recombinant chymosin andbovine pepsin are commercially available.

Calf chymosin contains three isoenzymes—A, B,and C. A and B are gene products that differ fromeach other by one amino acid; A has Asp at position243 while B has Gly at this position. Chymosin Cappears to be a degradation product of chymosin A

which lacks three residues, Asp244-Phe246.Commercially available microbial recombinant chy-mosins contain only chymosin A or B; it is notknown if the different forms of chymosin differ in spe-cificity, but it is claimed by rennet manufacturers thatthe cheesemaking properties of chymosin A and B arenot equivalent.

The primary, and probably higher, structures ofcommercially available microbial recombinant chymo-sins are identical to that of calf chymosin. However,several modified chymosins have been produced fromgenetically engineered microorganisms. At present, theobjective of these investigations is to study the mechan-ism of chymosin action at the molecular level, but it isprobable that chymosin with improved cheese-makingproperties will emerge from such studies, e.g., enzymeswith increased activity on certain bonds shown to pro-duce cheese with improved quality or reduced activityon other bonds, cleavage of which results in flavor ortextural defects. It should be remembered that chymo-sin evolved to coagulate milk in the neonatal stomach(to improve the efficiency of digestion) and not to pro-duce cheese. It is fortuitous that chymosin is the bestproteinase for cheese production, not just for milk coa-gulation, but it is highly probable that it can beimproved. For reference on genetically engineered chy-mosins, see (14, 17).

Most (70–90%) of the rennet added to cheese milkis lost in the whey. Therefore, the possibility of immo-bilizing rennet has been investigated as a means ofextending its working life. Several rennets have beenimmobilized but their efficacy as milk coagulants hasbeen questioned. There is widespread support for theview that properly immobilized enzymes can not coa-gulate milk owing to inaccessibility of the Phe-Metbond of �-casein and that the apparent coagulatingactivity of immobilized rennets is due to leaching ofenzyme from the support. Even if immobilized rennetscould hydrolyze micellar �-casein, operational difficul-ties would exist at the cheese factory level.Furthermore, as discussed in Section II.A.2.d, the resi-dual rennet in cheese curd plays an essential role incheese ripening and it would be necessary to addsome chymosin or similar enzyme to the curd aftercoagulation, which would be difficult or impossible(see 8, 9, 24 for reviews).

c. Factors Affecting the Hydrolysis of �-Casein. The pH optimum for chymosin and bovinepepsin is � 4:7 on small Phe-Met-containing peptidesand 5.3–5.5 on �-CN f98–111 or on whole �-casein.The pH optimum for the first stage of rennet action


in milk is � 6:0. Milk for most cheese varieties isrenneted at about pH 6.5.

Increasing ionic strength (0.01–0.11) decreases therate of hydrolysis of �-CN f98–112, especially if thepH is also increased. At 1 mM, NaCl, CaCl2, andMgCl2 stimulate the hydrolysis of �-casein in isolatedform or in sodium caseinate.

The optimum temperature for the coagulation ofmilk by calf rennet at pH 6.6 is � 45C. The tempera-ture coefficient (� rate/10C) for the hydrolysis of �-casein in solutions of Na-caseinate is � 1:8, the Ea is� 10,000 cal mol�1 and the activation entropy is� �39 cal deg�1 mol�1; generally similar values havebeen reported for the hydrolysis of isolated �-casein.

The efficacy of �-casein as a substrate for chymosindecreases as its level of glycosylation increases. At pH6.6, kcat decreases from � 43 sec�1 for carbohydrate-free �-casein to � 25 sec�1 for �-casein containing 6moles N-acetyl neuraminic acid (NANA) per mol.However, Km is lowest for the �-casein componentcontaining 3 moles NANA per mol. Polymerization(aggregation) markedly increases Km with little effecton kcat.

d. Secondary (Nonenzymatic) Phase of RennetCoagulation. Hydrolysis of �-casein removes itshighly charged, hydrophilic C-terminal segment fromthe surface of the casein micelles, thereby reducingtheir zeta potential from � �20mV to � �7mV andremoving the steric stabilizing layer. When � 85% ofthe total �-casein has been hydrolyzed, the caseinmicelles begin to aggregate and eventually form a gel.Reducing the pH or increasing the temperature fromthe normal (� 6:6 and � 31C, respectively) inducescoagulation at a lower degree of �-casein hydrolysis.

The mechanism involved in the coagulation ofrennet-altered micelles is not known precisely.Coagulation is dependent on Ca2þ and on colloidalcalcium phosphate, which are exchangeable to a cer-tain extent. Ca2þ may function by neutralizing nega-tive charges on the caseins. Coagulation is highlytemperature dependent; rennet-altered micelles do notcoagulate below � 20C, above which the Q10C forcoagulation is � 16. The high temperature dependenceof coagulation suggests that hydrophobic bonds maybe involved or that multiple bonds are formed.

The rennet coagulability of milk is adverselyaffected by heat treatments at temperatures > 65Cand is prevented by very severe heat treatments(> 90C for 10 min). Although changes in the equili-bria of milk salts, especially calcium phosphate, arecontributory factors, complexation of �-casein with

�-lactoglobulin and/or �-lactalbumin is primarilyresponsible for the increased rennet coagulation timeof heated milk; both the primary phase and especiallythe secondary phase are adversely affected. Theadverse effects of heating can be reversed by acidifica-tion before or after heating or by addition of CaCl2.

Rennet-coagulated milk gels are relatively stable ifleft undisturbed but synerese strongly if cut or broken.The rate and extent of syneresis are promoted by redu-cing the pH, increasing the temperature, and applyingpressure, e.g., agitation. By controlling the extent ofsyneresis, the cheese maker can control the moisturecontent of cheese which is a major factor affecting therate and pattern of ripening and the stability of cheese.Differences in moisture content are, in fact, a majorfactor responsible for the diversity of cheese flavor andtexture.

2. Proteolysis During Cheese Ripening

Acid-coagulated cheeses are usually consumed fresh,but the vast majority of rennet-coagulated cheesesare ripened (matured) for a period ranging from � 3weeks to > 2 years; the rate of ripening is directlyrelated to the moisture content of the cheese. Duringripening, numerous microbiological, biochemical, andchemical events occur, as a result of which the princi-pal constituents of the cheese—the proteins, lipids, andlactose—are transformed to primary, and later to sec-ondary, products. Among the principal flavor com-pounds present in most cheese varieties are: peptides,amino acids, amines, acids, thiols, and thioesters(derived from proteins); fatty acids, methyl ketones,lactones, esters, and thioesters (derived from lipids);organic acids (lactic, acetic, and propionic); carbondioxide; esters; and alcohols (derived from lactose).At appropriate concentrations and combinations,these compounds are responsible for the characteristicflavor of the various cheese varieties.

The biochemistry of cheese ripening has beenreviewed by Fox et al. (25, 26) and Fox and Wallace(27); only proteolysis is discussed here.

a. Significance of Proteolysis. Proteolysis isessential in all rennet-coagulated cheese varieties, espe-cially internal- and surface-bacterially ripened cheesesin which it is probably the principal biochemical eventduring ripening. Proteolysis contributes to cheeseripening in at least four ways: (a) makes a direct con-tribution to flavor, or off-flavor, e.g., bitterness, orindirectly since free amino acids are catabolized toamines, acids, thiols, thioesters, etc.; (b) facilitates therelease of sapid compounds during mastication; (c) the


production of NH3 from amino acids released by pro-teolysis affects flavor and texture; (d) changes in tex-ture due to breakdown of the protein network, increasein pH, and greater water binding by the newly formedamino and carboxyl groups. There is a good correla-tion between the intensity of Cheddar cheese flavorand the extent and depth of proteolysis.

Considerable information is available on the leveland type of proteolysis in the principal cheese groups(14, 25, 26, 28–33).

b. Proteolytic Agents in Cheese. Four, and insome varieties five, agents contribute to proteolysis incheese during ripening: rennet or rennet substitute;indigenous milk enzymes, especially plasmin; starterbacteria and their enzymes, released on cell lysis; non-starter bacteria, which either survive pasteurization ofthe cheese milk or gain access to the pasteurized milkor curd during manufacture; and secondary inocula,e.g., propionic acid bacteria, Brevibacterium linens,yeasts, and molds (Penicillium roqueforti andGeotricum candidum), and their enzymes are of majorimportance in some varieties. Techniques have beendeveloped which permit quantitation of the contribu-tion of each of these five agents to the primary aspectsof cheese ripening (25, 31).

Studies using these techniques have shown that pri-mary proteolysis, as determined by gel electrophoresisor the formation of water- or pH 4.6-soluble N, is duemainly to the coagulant except in those varieties, e.g.,Emmental, Parmesan, and mozzarella, cooked to ahigh temperature (52–55C) in which the coagulant isextensively or completely denatured. In these lattervarieties, primary proteolysis is relatively limited andis due mainly to plasmin, which also contributes toproteolysis, especially of �-casein, in low-cookedcheese. The peptides produced by the coagulant andplasmin are further hydrolyzed by the proteinases andpeptidases of starter and nonstarter bacteria, leadingto the formation of many small peptides and freeamino acids. Proteolysis varies widely among varieties,from very limited in short-ripened cheeses such as moz-zarella, to very extensive in extramature Cheddar,Parmesan, or blue cheeses. Proteolysis in cheddarcheese has been well characterized and considerableinformation is also available on Gouda, Emmental,and Parmesan cheeses (33).

c. Contribution of Coagulant to Proteolysis inCheese. Most of the rennet added to cheese milk iseither denatured or lost in the whey. The amountretained is influenced by the type of rennet; e.g., por-cine pepsin is extensively denatured during cheese

making, pH (low pH favors retention of chymosinbut not microbial rennets and reduces denaturation,e.g., of pepsins), and cooking temperature, e.g., little,if any, coagulant survives the cooking conditions usedfor Swiss-type cheeses. About 6% of the added chymo-sin is retained in Cheddar cheese and up to 20–30% inhigh-moisture, low-cook, low-pH cheeses, such asCamembert.

The proteolytic specificity of calf chymosin on �s1-,�s2-, and �-caseins in solution has been establishedand these findings can, largely, be extended to cheese(14, 33).

The principal chymosin cleavage sites on �s1-caseinin cheese are: Phe23-Phe24, which is hydrolyzed rapidlyand completely in cheese (e.g., within � 3 months inCheddar), and Leu101-Lys102, which is cleaved fairlyextensively; Phe33-Gly34 and Leu98-Leu99 are alsocleaved to some extent in cheese. Surprisingly, theTrp164-Tyr165 bond, which is the second most suscep-tible bond in �s1 casein in solution, does not appear tobe hydrolyzed in cheese. The small peptide, �s1-CN f1–23, does not accumulate in cheese but is hydrolyzedrapidly by the cell envelope-associated lactococcal pro-teinase, with a specificity dependent on the strain (34).

Some peptide bonds in �-casein in solution arehydrolyzed quite rapidly by chymosin in the order:Leu192-Tyr193, Ala189-Phe190, Leu163-Ser164, andLeu139-Leu140. In cheese, these bonds are hydrolyzedto a very limited extent or not at all, probably becausethe C-terminal region of �-casein is very hydrophobicand undergoes hydrophobically driven interactions incheese. These interactions appear to be accentuated byNaCl; even in solution, the hydrolysis of �-casein isstrongly inhibited by 5% NaCl, which is the typicalNaCl concentration in the aqueous phase of manycheese varieties. Inhibition of the hydrolysis of �-caseinin cheese is desirable since the peptide �-CN f193–209and fragments thereof are very bitter.

Although �s2-casein in solution is fairly readilyhydrolyzed by chymosin, its fate in cheese is notclear, and para-�-casein (�-CN f1–105) is very resistantto chymosin (and to other proteinases in cheese).

The specificity of pepsins is generally similar to thatof chymosin but has not been established precisely.Bovine pepsin cleaves the Leu109-Glu110 bond of �s1-casein quite rapidly, a bond which is cleaved veryslowly by chymosin. The specificity of the fungalrennet substitutes is quite different from that of chy-mosin (35). The principal cleavage sites of R. mieheiproteinase in �s1-casein in solution are Phe23-Phe24,Met123-Lys124, and Tyr165-Tyr166 while those in �-casein are Glu31-Lys32, Val58-Val59, Met93-Gly94, and


Phe190-Leu191 (35). C. parasitica proteinase is muchmore active on �-casein in cheese than chymosin,pepsin, or Rhizomucor proteinases, but its specificityon the caseins has not been determined; since it doesnot cause significant bitterness in cheese, its primarycleavage sites in �-casein may be in the N-terminalrather than in the C-terminal region (the N-terminalregion of �-casein is quite hydrophilic, and smallpeptides produced from it would be expected to benonbitter).

d. Significance of Secondary CoagulantProteolysis. The proteolytic activity of the coagulantin cheese influences quality in four ways:

1. Some rennet-produced peptides may have a posi-tive influence on flavor but excessive or unbalancedproteolysis, e.g., too much or excessively proteolyticrennet or unsuitable environmental conditions, e.g.,too much moisture or too little NaCl, leads to bitter-ness. The peptide �-CN f193–209 derived from the C-terminal of �-casein and fragments thereof are particu-larly bitter.

2. Rennet-produced peptides serve as substrates formicrobial proteinases and peptidases which producesmall peptides and amino acids. These contribute atleast to background flavor and perhaps to bitternessif the activity of such enzymes is excessive. Catabolismof amino acids by microbial enzymes and perhapsalterations via chemical mechanisms leads to a rangeof sapid compounds—amines, acids, NH3, thiols—which are major contributors to characteristic cheeseflavors.

3. Alterations in cheese texture appear to influencethe release of flavorful and aromatic compounds, aris-ing from proteolysis, lipolysis, glycolysis, and second-ary metabolic changes, from cheese duringmastication, and this may be the most significant con-tribution of proteolysis to cheese flavor.

4. Texture is an important attribute of all cheesesand is critical in some varieties. Protein forms a con-tinuous solid matrix in cheese, and its hydrolysis leadsto a softening of the texture. Chymosin is primarilyresponsible for textural changes during the early stagesof ripening. The functionality (stretchability and melt-ability) of mozzarella is strongly influenced by proteo-lysis; a low level of proteolysis improves functionality,but quality deteriorates on further proteolysis.

3. Acceleration of Cheese Ripening

The original objective of cheese manufacture wasconservation of the principal nutrients in milk (i.e.,lipids and proteins) by a combination of acidification,

dehydration, low Eh, and salting. Chemical andbiochemical changes do occur during storage butstability was the prime objective. While still important,stability is no longer the primary objective of cheesemanufacture, and since ripening is expensive, itsacceleration, especially in low-moisture, slow-ripeningvarieties, is desirable, at least under certain circum-stances, provided the whole process can be maintainedin balance.

Some high-moisture cheeses develop an intense fla-vor through a very active secondary flora, e.g., internalblue mold, external white mold, or a bacterial surfacesmear; these cheeses ripen quickly, e.g., 4–16 weeks.High-moisture internal bacterially ripened cheesesalso mature rapidly but develop a low flavor intensity;if the ripening of such cheeses is extended, they willprobably develop off-flavors. It is possible to developan intense flavor in internal bacterially ripened cheesesonly if the moisture content is low and they are ripenedfor a long period, e.g., Parmesan, extramatureCheddar, or extramature Gouda, which are ripenedfor 2–3 years. Owing to the high cost of ripening facil-ities and stocks, the ripening of extramature cheeses isexpensive. Consequently, there is commercial interestin accelerating the ripening of these cheeses, providedquality can be maintained. It might also be possible toapply similar techniques to medium-moisture, med-ium-flavor cheeses, e.g., regular Cheddar and Gouda,with the objective of accentuating their flavor. Most ofthe work on accelerating cheese ripening has in factbeen on Cheddar. Techniques for accelerating ripeningmay also be applicable to reduced-fat cheeses, whichtend to ripen slowly. A substantial literature onattempts to accelerate cheese ripening has accumulatedand has been reviewed regularly (36).

Glycolysis is rapid in all cheeses and does notrequire acceleration. Lipolysis is limited in mostcheeses and excessive lipolysis is undesirable.Consequently, most studies on accelerated ripeningof cheese have focused on proteolysis, which contri-butes to flavor and is mainly responsible for changesin texture.

Methods for accelerating cheese ripening fall intosix categories: elevated ripening temperature, exogen-ous enzymes, chemically or physically modified cells,genetically modified starters, adjunct starters, andenzyme-modified cheeses. These methods either seekto make the conditions under which indigenousenzymes function more favorable (i.e., elevated tem-perature) or to increase the level of certain key enzymeswhich are considered to be particularly important incheese ripening.


A complex cascade of enzymes is involved in cheeseripening, and most key enzymes have not been identi-fied. Not surprisingly, the use of single enzymes, e.g.,additional coagulant (which is mainly responsible forprimary proteolysis), plasmin (for which some benefitsare claimed), or Neutrase (from B. subtitis), while accel-erating proteolysis, does not accelerate flavor develop-ment and may cause off-flavors. It has been claimedthat a combination of exogenous proteinases and lacto-coccal cell-free extracts (rich in peptidases) accelerateripening, but the results are equivocal. Uniform incor-poration of the enzyme preparation into cheese curdposes problems. For Cheddar, the enzyme preparation,diluted with salt, may be added to milled curd, but thismethod is not applicable to most varieties. Addition ofmicroencapsulated enzyme(s) to cheese milk is tech-nically feasible, but microencapsulation techniquescurrently available are not very efficient. At present,exogenous enzymes are not being used commerciallyto accelerate the ripening of natural cheese.

Exogenous lipase, traditionally pregastric esterase,is added to certain hard Italian cheeses, e.g., Romanoand provolone. It has been claimed that inclusion ofselected lipases in the blend of exogenous enzymesaccelerates the ripening of other cheeses, e.g.,Cheddar and Ras.

Most of the enzymes involved in cheese ripening areproduced by microorganisms that grow in or on thecheese. By increasing the number of these organisms itshould be possible to accelerate ripening. The use ofwhole cells should have two advantages over isolatedenzymes: they contain the ‘‘natural’’ cocktail ofenzymes found in cheese and they are cheaper to pro-duce. Three approaches have been considered in theuse of bacterial cells to accelerate cheese ripening: atte-nuated starter cells, genetically modified cultures,adjunct cultures.

The starter, in addition to its essential role in curdacidification, is also essential for secondary proteolysisand flavor development. Therefore, it might beexpected that increasing the number of starter cells incheese would accelerate ripening. However, increasingthe level of active starter added to the cheese milkcauses an excessively rapid rate of acid productionwhich has undesirable effects, e.g., a crumbly textureand overacid flavor. The acid-producing ability of star-ter cells may be attenuated or destroyed by heat-shock-ing, freeze-shocking, or solvent treatment with verylittle effect on their enzyme activities. These attenuatedcells are in effect packages of enzymes and their use hasbeen reported to accelerate the ripening of a number ofcheese varieties.

A simpler and probably more effective approach isthe use of lactose-negative strains of Lactococcus,which cannot grow in milk or cheese curd but serveas a balanced ‘‘natural’’ source of enzymes importantin cheese ripening. Such cultures are available commer-cially and are claimed to give satisfactory results.

Cells of selected non-lactic-acid bacteria, which donot grow in a particular cheese, either because they areaerobic, e.g., Pseudomonas spp. or Brevibacterim spp.,or because they have a high growth temperature, e.g.,Propionibacterium, might also serve as suitable‘‘packages of enzymes’’ for addition to cheese curd;however, such cultures do not appear to be used com-mercially at present.

It is probable that certain enzymes are rate limitingin cheese ripening. At present, the key limiting enzymesare unknown, but studies are in hand with the objec-tive of identifying these enzymes using deficient oroverproducing mutants. It is possible to geneticallymodify Lactococcus to overproduce certain desirableenzymes, to delete undesirable genes, or to introduceforeign genes for putitively important enzymes. It ishighly probable that genetically modified bacteriawith the ability to accelerate ripening and improvecheese quality will become available in due course foruse as primary or secondary cultures.

All cheese acquires an adventitious nonstartermicroflora, predominantly mesophilic lactobacilli,which grow from low numbers initially (e.g.,< 103 cfu=gÞ to 107–108 cfu=g and which dominatethe microflora of cheese after � 3 months owing tothe death of the primary starter. These nonstarter lac-tic acid bacteria (NSLAB) probably affect cheeseripening, and there is considerable interest in inoculat-ing cheese milk with selected strains of NSLAB toaccelerate or modify cheese flavor development (37).

An extreme form of accelerated ripening is practicedin the production of enzyme-modified cheese (EMC);the subject has been reviewed by Kilcawley et al. (38).EMCs are produced by adding a cocktail of enzymes(proteinases, peptidases, lipases) and perhaps bacterialcultures to homogenized, pasteurized fresh curd oryoung cheese. The mixture is incubated for a requisiteperiod and repasteurized to terminate the microbiolo-gical and enzymatic reactions. The preparation may bespray-dried or commercialized as a paste.

Although the flavor of EMCs does not approximatethat of natural cheese, they have the ability to potenti-ate cheeselike flavor in various food products, e.g.,processed cheese, cheese analogs, cheese sauces, anddips, and products containing cheese, e.g., crackers,crisps, etc. For such applications, EMCs may replace


20–50 times their weight of natural cheese and arecheaper. Cheddar EMCs are the most important com-mercially, but EMCs that stimulate several varietieshave been developed, e.g., blue, Swiss, and Romano.

B. Other Applications of Proteinases

In comparison with their use in cheese making, theother applications of proteinases in dairy technologyare quite small but some have considerable growthpotential. The more important of these are discussedbelow.

1. Dietary Products

Protein hydrolyzates for use in soups, gravies, flavor-ings, and dietetic foods are generally prepared fromsoy proteins, gluten, milk proteins, meat, or fish pro-tein by acid hydrolysis. Neutralization results in a highsalt content which is acceptable for certain applicationsbut may be unsuitable for dietetic foods and food sup-plements. Furthermore, acid hydrolysis causes total orpartial destruction of some amino acids. Enzymatichydrolysis is a viable alternative (39), but bitternessdue to hydrophobic peptides is frequently encountered.Caseins are strongly hydrophobic and yield very bitterhydrolyzates, but bitterness may be eliminated, or atleast reduced, by one of several treatments (39, 40).

There is increasing interest in the production ofcasein-derived peptides with special nutritional or phy-siological properties; some of the possibilities havebeen reviewed (41, 42). Apart from the interest incasein hydrolyzates for the nutrition of patients withdigestive problems, interest has been focused recentlyon phosphopeptides derived from casein which it isclaimed stimulate the absorption of calcium and iron,but views on this are not unanimous (43). Methods forthe production of caseinophosphopeptides for nutri-tional and/or medical applications have been devel-oped (44–46).

The casein(glyco)macropeptide (CMP), �-CN f106–169, produced from �-casein during the enzymatic coa-gulation of milk for cheese or rennet casein, is lost intothe whey. CMP is devoid of aromatic amino acids andhence is a suitable nutrient for patients suffering fromphenylketonuria. Several biological activities have beenattributed to the CMP, including inhibition of adhesionof oral actinomyces and streptococci to erythrocytes,effects on gastrointestinal motility, growth factors forBifidobacterium spp., inhibition of the binding of cho-lera toxin, inhibition of influenza virus hemagglutinin,stimulation of cholecystokinin release from intestinal

cells, and inhibition of acid secretion in the stomach(47, 48). Methods have been developed for the indus-trial production of CMP from whey (48, 49).

There is considerable interest in the fortification ofperformance-enhancing drinks with protein hydroly-sates; bitterness is a problem here, and whey proteinhydrolysates are preferred to casein hydrolysates.

2. Physiologically Active Peptides from MilkProteins

Peptides with various physiological activities have beenisolated from milk protein hydrolysates; at least someof these peptides are produced in vivo and may play aphysiological role (50). The best-studied of these arethe �-caseinomorphins, a family of peptides containing4–7 amino acids (representing �-CN f60–63/7) withopioid activity. These peptides are produced in theintestine in vivo but it is still unclear whether or notthey reach the brain. Peptides with opiate propertieshave also been isolated from hydrolysates of �s1- and�-caseins, lactotransferrin, �-lactalbumin, and �-lacto-globulin.

Other biologically active peptides that have beenisolated from hydrolysates of milk proteins include:immunomodulating peptides, an inhibitor of angioten-sin-converting enzyme, blood platelet modifiers, andstimulators of DNA synthesis (50). Whether any ofthese peptides are active in vivo remain to be estab-lished, but their formation has led to casein beingreferred to as a pro-hormone (51).

The release of biologically active peptides requiresprecise hydrolysis of the parent molecules at specificbonds. If the application of these peptides developsas predicted, very interesting applications for protei-nases in the dairy industry will emerge.

3. Modification of Protein Functionality

Milk proteins are among the principal functional pro-teins used in food products (52). In general, milk pro-teins possess very good functional properties but suffersome limitations, notably the insolubility of casein inthe pH range 3.0–5.5. The functional properties of milkproteins may be improved by limited proteolysis (1, 40,53, 54). An acid-soluble casein, free from off-flavorand suitable for incorporation into beverages andother acid foods, has been prepared by limited proteo-lysis. The antigenicity of casein is destroyed by proteo-lysis, and the hydrolysate is suitable for use in milkprotein-based foods for infants allergic to cow’s milk.Controlled proteolysis improves the meltability ofdirectly acidified cheese but excessive proteolysis


causes bitterness. Casein solutions are very viscous atconcentrations > 20%, w/v, which increases the cost ofdrying caseinates. Viscosity may be reduced by limitedproteolysis, but only a small increase in solids is pos-sible owing to bitterness after even moderate levels ofproteolysis.

The surface activity of sodium caseinate can beincreased considerably by the treatment with plasmin,apparently owing to the formation of �2- and �3-case-ins (�-CN f106–209, �-CN f108–209) which are verysurface active. Generally, the emulsifying and foamingproperties of small peptides are poor, since they formvery thin interfacial layers.

Limited proteolysis of lactalbumin (heat-denaturedwhey protein), which is insoluble and has very poorfunctional properties, yields a product with greatlyimproved solubility and functionality. Limited proteo-lysis of whey protein concentrate (WPC) reduces itsemulsifying capacity and increases its specific foamvolume but reduces foam stability. The heat stabilityof WPC may be improved considerably by limitedhydrolysis without concomitant impairment of otherfunctional properties or off-flavor development.

The plastein reaction has been proposed as amechanism by which the functional and nutritionalproperties of proteins may be improved. The reactionis of little relevance to milk proteins because theyalready have good functional and nutritional proper-ties and yields of plastein are low.

One of the principal food applications of whey pro-tein (WP) and whey protein isolates (WPI) is in theproduction of thermo-set gels. The gelation of WPCsat low temperatures has been achieved by limited pro-teolysis by certain enzymes (55); perhaps gelationoccurs through the plastein reaction.

III. LIPASES

Lipases have a number of relatively low-volume appli-cations in the dairy industry. Some of these applica-tions are traditional and essential for the manufactureof particular products: others are emerging but holdconsiderable potential.

A. Lipases in Cheese Production

The principal application of lipases in dairy technologyis in cheese manufacture, particularly, some Italianvarieties, e.g., Romano and provolone. The character-istic ‘‘piquant’’ flavor of these cheeses is due primarilyto short-chain fatty acids resulting from the action of

lipase(s) in the rennet paste traditionally used in theirmanufacture.

Rennet pastes are prepared from the stomachs ofcalves, lambs, or kids slaughtered after suckling; thestomachs and contents are held for a considerable per-iod prior to maceration. Because of possible risks topublic health, the use of rennet pastes, which haveproteolytic and lipolytic activities, is prohibited insome countries. The lipase in rennet paste is of oralorigin and its secretion is stimulated by suckling: thesecreted lipase is washed into the stomach with theingested milk. Oral (lingual) lipase, commonly referredto as a pregastric esterase (PGE), is secreted by severalspecies and probably makes a significant contributionto the digestion of lipids by the neonate in which theactivity of pancreatic lipase is limited. The consider-able literature on PGE has been comprehensivelyreviewed (3, 56). PGE shows a high specificity forshort chain fatty acids, especially butyric, esterifiedon the Sn-3 position of glycerol, although some inter-species differences in specificity have been reported.They are maximally active at 32–42C, pH 4.8–5.5,and in the presence of 0.5 M NaCl. Calf, kid, andlamb PGEs have been partially purified from commer-cial preparations (57). Calf PGE has been isolatedfrom oral tissue and characterized with respect to pI(7.0), molecular weight (� 49 kDa), and amino acidcomposition (58, 59). The secondary structures of ratlingual lipase and pancreatic lipase were studied andcompared (60). The gene for rat lingual lipase has beencloned and sequenced, and the amino acid sequence ofthe enzyme has been deduced (61).

PGE-containing extracts from calf, kid, or lamb tis-sue are commercially available as alternatives forrennet pastes. Slight interspecies differences in specifi-city render one or another more suitable for particularapplications. Particularly large differences in the abil-ity of lipases to release 4-methyloctanoic acid, whichexhibits a goat-muttony aroma, have been found (62).Such differences in specificity permit the generation ofa range of flavors in cheese products. The propensity ofPGE to synthesize triglycerides is increased in the awrange 0.75–0.90 or by the addition of ethanol to thereaction mixture (63); this property may be exploitedto modify cheese flavors.

Although PGE extract is now widely used in themanufacture of hard Italian cheese varieties, connois-seurs of Italian cheese claim that rennet paste givessuperior results. Perhaps rennet paste containsenzymes in addition to chymosin and PGE. A secondlipase, termed gastric lipase, was identified in anextract of cleaned gastric tissue and partially charac-


terized. A combination of calf gastric lipase and goatPGE gave Cheddar and provolone of superior qualityto cheese made with PGE alone (64). However, Nelsonet al. (56) expressed reservations on the occurrence of agastric lipase distinct from PGE. Traditional rennetpaste would be expected to contain both PGE andgastric lipase.

R. miehei secretes a lipase that is reported to givesatisfactory results in Italian cheese manufacture (65).The enzyme has been characterized (66) and is com-mercially available as Piccantase. The lipases secretedby selected strains of Penicillium roqueforti, P. candi-dum, or A. niger are considered to be potentially usefulfor the manufacture of Romano, provolone, and othercheese varieties (67, 68). Unlike PGE, the fungallipases do not preferentially release short-chain fattyacids. The use of different lipase preparations or blendsof lipases opens possibilities for producing Italiancheeses with different degrees of sharpness.

A. oryzae secretes a lipase which has as exception-ally high specificity for C6–C8 acids (69). Anotherinteresting characteristic of this enzyme is that itforms micelles, � 0:2�m in diameter, in aqueousmedia, as a result of which � 94% of the enzymeadded to milk is recovered in cheese curds. The forma-tion of short-chain fatty acids was reported to parallelflavor intensity in Cheddar cheese containing thisenzyme. In contrast to the FFA profile caused bycalf PGE, which liberated high concentrations ofC4:0, the FFA profile in the cheese containing A. oryzaelipase was similar to that in the control cheese, but thelevel of FFA was much higher.

Extensive lipolysis also occurs in blue cheese vari-eties, and in addition to making a direct contributionto flavor, the free fatty acids serve as substrates forfungal enzyme systems in the biosynthesis of methylketones, which are the principal contributors to thetypical flavor of blue cheese (70). P. roqueforti lipasepredominates the ripening of blue cheeses. However,blue cheese ripening may be accelerated and qualityimproved by the addition of exogenous lipases (71, 72).

Blue cheese is a popular ingredient for salad dres-sings and cheese dip. High-quality natural cheese is notnormally required for these applications and there isconsiderable interest in the production of cheaper sub-stitutes. Various methods have been developed for theproduction of blue cheese flavor concentrates; most ofthese methods involved the use of fungal lipases andusually P. roqueforti spores (1, 2, 73).

The low level of lipolysis that occurs in most othervarieties is catalyzed by lipases/esterases derived fromthe starter or nonstarter lactic acid bacteria. Consider-

ably more lipolysis occurs in cheeses made from rawmilk than in pasteurized milk cheeses, possibly owingto the indigenous milk lipoprotein lipase and/or a morediverse nonstarter microflora in the former. It has beenclaimed that the flavor of many cheeses can be intensi-fied or their ripening accelerated by incorporating PGEor fungal lipases, although their use appears to be verylimited (1–3). Relatively extensive lipolysis occurs inParmigiano-Reggiano, which can probably be attribu-ted to the use of raw milk (which contains an indigen-ous lipase) and the long ripening time (2 years).

Lipases, probablymainly of fungal origin, are used inthe production of some enzyme-modified cheeses (38).

B. Other Applications of Lipases

Lipases are used to hydrolyze milk fat for a variety ofuses in the confectionery, candy, chocolate, sauce, andsnack food industries. The partially hydrolyzed fatimparts a greater intensity of butterlike flavor to theproducts and delays stalling, presumably as a result ofthe emulsifying effect of di- and monoglycerides (56,71, 74).

An important new application of lipases is in thetrans/interesterification of fatty acids on triglycerides(75). This approach can be used to modify the meltingpoint of triglycerides, and hence their rheological prop-erties. An important application of this technology isin the production of cocoa butter substitutes for cho-colate manufacture. Mono- or polyunsaturated fattyacids may also be introduced to relatively saturatedmilk lipids to improve their nutritional qualities.Immobilized lipase systems have been developed forthese applications (74).

IV. �-GALACTOSIDASE

Lactose is a reducing disaccharide containing galactoseand glucose linked by a �-1-4 O-glycosidic bond (O-�-D-galactopyranosyl-(1-4)-�- or �-D-glucopyranose, �-and �-lactose, respectively). Lactose is by far the domi-nant carbohydrate in milks which are, in turn, the onlysignificant natural sources of lactose. The concentra-tion of lactose in milk ranges from 0% in the milk ofmarine mammals to � 10% in milk from some speciesof monkey; bovine and human milk contain � 4:8%and 7% lactose, respectively. Lactose is an importantsource of energy for the newborn mammal, but when avery energy-dense milk is required (mammals in aqua-tic or polar environments), the lipid rather than thelactose content is increased. In fact, there is a fairly


good inverse relationship between the level of fat andthat of lactose.

Among sugars, lactose possesses many fairly uniqueproperties: low solubility (� 18 g=100mL H2O at20C); a marked tendency to form supersaturated solu-tions which are difficult to crystallize; when crystalliza-tion does occur, the crystals are hard and sharp and,unless kept to dimensions < 20�m, cause a sandy tex-ture in foods; crystallization is complicated by itsmutarotation characteristics since �- and �-lactose dif-fer considerably in solubility and degree of hydration;it has low sweetness (16% as sweet as sucrose at 1%,w/v); owing to its crystallization and mutarotationcharacteristics, it is hygroscopic and may cause ‘‘cak-ing’’ of dairy powders; it has a strong tendency toabsorb flavors and odors.

At 20C, �-lactose is considerably less soluble(� 7 g=100 g H2O) than �-lactose (� 50 g=100 g H2O).However, the solubility of the �-anomer increasesmore sharply with increasing temperature than thatof the �-anomer and the solubility curves intersect at� 93:5C; therefore, when lactose is crystallized at atemperature < 93:5C, �-lactose is obtained. �-Lactose crystallizes as a monohydrate while �-lactosecrystals are anhydrous.

When milk is spray-dried, there is insufficient timefor lactose to crystallize and an amorphous lactoseglass is formed. If the moisture content of the powderis low, the glass is stable, but if the moisture contentincreases, the glass become hygroscopic, and lactosecrystallizes as �-lactose monohydrate, leading to cak-ing. In practice, the problem is solved by precrystalliz-ing the lactose, usually induced by seeding withpowdered lactose crystals.

When milk is frozen, the lactose usually does nothave time to crystallize and the concentration of inor-ganic solutes in the liquid aqueous phase increases. Onholding in the frozen state, calcium phosphate crystal-lizes as Ca3ðPO4Þ2, releasing Hþ and reducing the pHto � 5:8, and lactose crystallizes as �-lactose monohy-drate, reducing the amount of solvent water, furtherincreasing the concentration of inorganic solutes. Thecombination of low pH and high Ca2þ destabilizes thecasein micelles which aggregate when the milk isthawed. Unless properly controlled, these characteris-tics of lactose may cause defects in concentrated, dehy-drated, and frozen dairy products. However, some ofthe same characteristics may be exploited to make lac-tose an interesting and useful food additive—e.g., as afree-flowing agent, an agglomerating (instantizing)agent, an additive to stabilize color, flavor, and texture,especially when concomitant sweetness is undesirable,

and as a reducing sugar in products in which Maillardbrowning is desirable as a source of color and flavor.The chemistry, properties, problems, modifications,and applications of lactose and lactose derivativeshave been reviewed (76–85).

It is claimed that lactose promotes the intestinalabsorption of calcium and phosphorus and henceshould be nutritionally beneficial, especially in infantnutrition. However, lactose is involved in two enzymedeficiency diseases: lactose intolerance and galactose-mia. There are in fact two forms of galactosemia, botharising from the congenital deficiency of an enzyme intheir Leloir pathway for galactose metabolism (86).Classical galactosemia is due to a deficiency of galac-tose-1-phosphate:uridyl transferase. Ingested galactose(from lactose or other source) is phosphorylated togalactose-1-phosphate which is not metabolizedfurther, leading to the accumulation of galactose andgalactose-1-phosphate. Galactosemic infants appearnormal at birth but develop various symptoms, includ-ing mental retardation, unless put on a galactose-freediet within 2–3 months. The second form, galactoki-nase-deficient galactosemia, results in the failure tophosphorylate galactose, some of which is metabolizedto galactitol which accumulates in the eye, causing cat-aracts.

Disaccharides must be hydrolyzed to monosacchar-ides in the intestine prior to absorption, in the case oflactose by �-galactosidase. The vast majority of infantssecrete adequate levels of �-galactosidase in the brushborder of the small intestine to hydrolyze ingested lac-tose; however, a small minority of infants secrete aninadequate level of �-galactosidase. The level of intest-inal �-galactosidase reaches a maximum shortly afterbirth and declines thereafter to a low level. With theexception of northwestern Europeans and a fewAfrican tribes, the level of intestinal �-galactosidasebecomes so low within 6–8 years as to render the sub-ject incapable of hydrolyzing ingested lactose at anadequate rate, leading to lactose intolerance. Theunhydrolyzed lactose passes to the large intestinewhere it leads to flatulence, cramps, diarrhea, and pos-sibly death. Thus, only a minority of the world’s popu-lation can consume large quantities of lactose-containing foods with impunity. The subject of lactoseintolerance has been reviewed extensively (87, 88).

The third feature of the lactose problem is the devel-opment of economic outlets for the large quantities(� 5� 109 kg=year) available from cheese and caseinwheys. The technology for lactose production is welldeveloped but < 10% of the potentially available lac-tose is recovered as such; although lactose has a num-


ber of special food applications, the market appears tobe limited. Increasing concern about environmentalpollution has focused attention on more completeutilization of whey for which there are many options(76–85).

The dairy industry has developed methods for con-trolling the physicochemical problems posed by lac-tose, and lactose intolerance may not be too seriousif lactose-containing products are introduced graduallyto the diet. However, all the various problems posed bylactose (technological, nutritional, utilization) may besolved via hydrolysis by �-galactosidase (lactase) to theless problematic sugars, glucose and galactose.�-Galactosidase (�-D-galactoside galactohydrolase;

EC 3.2.1.23) catalyzes the hydrolysis of lactose to itscomponent monosaccharides, glucose and galactose.The use of �-galactosidase in dairy technology hasbeen considered as one of the most promising appli-cations of exogenous enzymes in food processing.However, although many interesting applicationshave been demonstrated, the use of �-galactosidasehas not yet become commercially successful foreconomic reasons. The voluminous literature on thepreparation, properties, and uses of �-galactosidaseshas been the subject of several reviews (76, 81, 83,89, 90).

Although �-galactosidase is widely distributed inplant, animal, and microbial sources, only the enzymesfrom Aspergillus niger, A. oryzae, Kluyveromyces marx-ianus var. lactis, K. fragilis, Bacillus stearothermophilus,and Escherichia coli are commercially available. �-Galactosidases from several sources have been isolatedand characterized; the important properties have beensummarized by Mahoney (89, 90). In general, �-galac-tosidases produced by molds have an acid pH opti-mum (2.5–4.5) and are therefore best suited for usein acid wheys, while yeast and bacterial �-galactosi-dases, with a pH optimum in the range 6–7.5, aremore suitable for use in milk or rennet wheys. �-Galactosidases with high thermal stability have beenisolated and are attractive because they can be used athigh temperatures at which microbial growth is slow orabsent. The heat stability of �-galactosidase from K.lactis and Streptococcus thermophilus is considerablygreater in milk than in buffer systems owing to thecombined effects of casein and lactose.�-Galactosidases from several sources have been

immobilized by encapsulation; entrapment in fibers,gels, or semipermeable membranes; adsorption orcovalent attachment by a variety of techniques to var-ious supports, e.g., porous glass, collagen, cellulosederivatives, and various resins (76, 83, 89, 90).

The principal applications of �-galactosidase indairy technology are: (a) production of low-lactosemilk and dairy products for �-galactosidase-deficientpatients; (b) modification of dairy products for use inice cream, baked goods, yogurt, etc.; (c) production ofsyrups and sweeteners for food applications; and (d)pretreatment of milk for freezing.

In spite of the widespread incidence of �-galactosi-dase deficiency among non-Caucasians, such subjectsadjust to lactose-containing diets if lactose is intro-duced gradually and the response to lactose is moder-ated by other components in the diet. Researchers aredivided as to the desirability of including milk in thediets of lactose-intolerant subjects (87, 88). However, itappears to be generally agreed that treatment with �-galactosidase would enhance the nutritional value ofdairy products in such cases and render protein-richdairy products suitable for supplementation of nutri-tionally deficient diets. Lactose-hydrolyzed milk iscommercially available but is more expensive than nor-mal milk. �-Galactosidase is also available in powderor liquid form for home use (89, 90). Direct addition of�-galactosidase to milk at mealtime (‘‘enzyme replace-ment therapy’’) has also given satisfactory results. Apromising method for reducing treatment cost is theaddition of a very low level of soluble �-galactosidaseto UHT milk which, during prolonged storage, inducesextensive hydrolysis. This approach has been commer-cialized by the Tetra-Pak Company (Lund, Sweden).The increased sweetness of lactose-hydrolyzed milkdoes not appear to be a problem and may actually bepreferred by some individuals.�-Galactosidase can act as a transferase resulting in

the synthesis of several oligosaccharides (galacto-oligo-saccharides), the range and concentration of whichappear to vary with the source of the �-galactosidaseand the duration of treatment. Many of the oligosac-charides are �-(1! 6Þ galactosides which are nothydrolyzed in the small intestine and pass into thelarge intestine where they are acted on by bacterialeading to intestinal disturbances, mainly flatulence.However, galacto-oligosaccharides stimulate thegrowth of Bifidobacterium spp. in the lower intestine,which is believed to be beneficial. A product (oligo-nate, 6 0-galactosyl lactose) is produced commerciallyby the Yokult Company in Japan for addition to infantformulae. Some galacto-oligosaccharides have interest-ing functional properties and may find commercialapplications (91, 92).

Hydrolysis of lactose in milk for yogurt has beensuggested as a means of increasing the sweetness ofyogurt without a concomitant increase in calories; it


is reported to reduce fermentation time (89, 90). Someauthors have reported that lactose-hydrolyzed yogurthas superior texture and consistency, with less wheyingoff, than controls. Off-flavors have been reported insome lactose-hydrolyzed yogurts, perhaps owing tothe action of contaminating proteinases. Yogurt andother cultured milks are well tolerated by lactose-intol-erant subjects, apparently owing to the secretion of �-galactosidase by the thermophilic culture used or toslower gastric emptying (88).

It is also claimed that the manufacturing time forCheddar and other cheeses is reduced by pretreatmentof cheese milk with �-galactosidase, and, possibly moresignificantly, the quality of the cheese is improved andripening accelerated. However, a proteinase present inthe commercial �-galactosidase preparations used,rather than to the action of �-galactosidase, was prob-ably responsible for the accelerated ripening.

Hydrolysis of lactose improves the functionality ofmilk powder in bakery products. The glucose moiety isfermentable by baker’s yeast (Saccharomyces cerevi-siae), leading to increased loaf volume, while the non-fermentable galactose contributes to flavor and crustcolor through Maillard browning. Prehydrolysis of lac-tose using �-galactosidase renders milk stable to freez-ing (89, 90, 93).

Dulce de Leche, a sweetened concentrated dairyproduct, is popular in Latin America as a dessert orspread. Lactose crystallization is a major problem inthis highly concentrated product. Growth of K. marx-ianus var. lactis in milk for the preparation of Dulce deLeche has been recommended for the control of lactosecrystallization (94). About 50% of the lactose washydrolyzed in 12 h (at 5:2� 107 cells/mL) or 100% in24 h. When the delactosed milk was mixed in propor-tions of 1 : 2 with normal milk and concentrated, nolactose crystallization and no significant changes inflavor were noted. The use of permeabilized K. marx-ianus var. lactis cells for the same application wasfound to be very satisfactory (95). Presumably, isolated�-galactosidase could be used successfully in the pre-paration of Dulce de Leche.

Lactose crystallization may also be a problem inconventional sweetened condensed milk, although theproblem can be controlled by appropriate manufactur-ing steps. Prehydrolysis with �-galactosidase offers analternative solution.

Treatment with �-galactosidase prevented lactosecrystallization in a whey retentate–buttermilk powderspread (96). Acid (mold) �-galactosidase was prefer-able to neutral (yeast) enzyme, and hydrolysis of30% of the lactose was sufficient. Lactose-hydrolyzed

whey appears to be suitable for incorporation into icecream in which up to 25% of the skim milk solids maybe replaced by hydrolyzed whey syrup without adverseeffects on quality; 50% of the sucrose may also bereplaced by whey syrup (89, 90, 97). Lactose-hydro-lyzed whey may be fed in larger amounts than normalwhey to animals, especially pigs.

Probably the principal commercial interest in �-galactosidase is for the production of syrups as a prof-itable outlet for lactose. Glucose-galactose syrups are� 70% as sweet as sucrose and about four timessweeter than lactose. It has been known for a longtime that lactose can be hydrolyzed to glucose andgalactose by strong mineral acids, and there has beenrenewed commercial interest in the process, using freeacid or ion exchange resins, as a means of producingglucose-galactose syrups. Production costs arereported to be lower than for enzymatic hydrolysis,but acid hydrolysis is applicable only to purified lac-tose solutions or perhaps ultrafiltration permeates.

Numerous applications have been reported for glu-cose-galactose syrups (in addition to the use of lactose-hydrolyzed whey) (83, 89, 90). Although effective sys-tems for the production of glucose-galactose syrupsand other lactose-hydrolyzed products are availableand continued research will undoubtedly lead toimproved systems, the cost of such syrups vis-a-visalternatives, mainly glucose syrups from starch,remains a problem.

The sweetness of glucose-galactose syrups may beincreased by isomerizing the glucose to fructose (whichis about twice as sweet as glucose) using glucose iso-merase, which is now widely used in the commercialproduction of high-fructose syrups from starch. Such asystem has been patented and a glucose-fructose-galac-tose-lactose syrup with a sweetness equal to sucrose(both at 10% solution) has been prepared by treatinga glucose-fortified lactose hydrolyzate with glucoseisomerase. Conversion of lactose to lactulose offersanother avenue for the production of novel sugar mix-tures. Lactulose is a disaccharide consisting of galac-tose and fructose, which can be produced from lactoseby mild alkaline treatment. At least some �-galactosi-dases can hydrolyze lactulose, although more slowlythan lactose.

It will be apparent that �-galactosidase has numer-ous applications in the dairy industry. However, inspite of very considerable research and demonstratedtechnological feasibility in pilot-scale experiments, �-galactosidase is not yet exploited commercially on asignificant scale, except for lactose-hydrolyzed pasteur-ized and UHT milks. Zadow (98) concluded that the


markets for lactose-hydrolyzed syrups are likely toremain limited for economic reasons.

V. LYSOZYME

Lysozyme (muramidase, EC 3.2.1.17), an enzymewhich causes lysis of certain bacteria by hydrolyzingcell wall polysaccharides, is widely distributed in ani-mal tissues and secretions. Some bacteria and bacter-iophage secrete similar enzymes, lysins. Egg white is aparticularly rich source of lysozyme, which is the bestcharacterized of these enzymes. The molecular andbiological properties of lysozyme and its use in foodpreservation and as a pharmaceutical have been com-prehensively reviewed by Proctor and Cunningham(99) and Cunningham et al. (100).

The milks of most species contain an indigenouslysozyme; human and equine milks are particularlyrich in this enzyme. Various aspects of indigenousmilk lysozyme were reviewed by Farkye (101, 102).

In view of its antibacterial activity, the large differ-ence in lysozyme content between human and bovinemilks may have significance in infant nutrition. It isclaimed that supplementation of baby food formulaebased on cows’ milk with egg-white lysozyme givesbeneficial results, especially with premature babies;however, results are equivocal (1).

Lysozyme has some other minor applications indairy technology, but most current interest is focusedon its use in Dutch, Swiss, Italian, and other cheesevarieties to prevent late gas blowing and/or off-flavorscaused by the growth of Clostridium tyrobutyricum.Contamination of cheese milk with Clostridium spp.can be reduced by good hygienic practices, and popu-lations may be further reduced by bactofugation ormicrofiltration. However, in most countries it is nor-mal practice to add sodium nitrate as a further precau-tion. The use of nitrate in foods is suspect because itleads to nitrosamine formation, and many countrieshave reduced permitted levels or prohibited its use.

Lysozyme is effective in killing Clostridium cells andpreventing the outgrowth of their spores. It has beenshown to be an effective alternative to nitrate in pre-venting the butyric acid fermentation and late gasblowing in several cheese varieties (1, 2, 99, 100, 103,104). It was concluded (103, 104) that

the published information indicates that for somecheese types, lysozyme is a suitable substance forthe control of late blowing, provided the numberof clostridial spores is low. For these cheesetypes, which may be considered less sensitive to

late blowing, the level of lysozyme already per-mitted by regulatory authorities in a number ofcountries appears to be satisfactory until furtherevidence is obtained. Lysozyme appears to be ofvalue in the control of late blowing in countrieswhich prohibit nitrate. Published informationindicates that in some cheese types which arevery sensitive to late blowing, such as Gouda,lysozyme used at the current normal additionunder normal manufacturing and storage condi-tions is less effective than the usual amount ofnitrate. In this case, lysozyme can not be consid-ered a suitable alternative to nitrate at present.More information will become available for thevarious cheese types about the critical number ofspores in the raw milk to cause defects whenlysozyme is used. Combinations of lysozymeaddition and other control measures can thenbe evaluated further.

Lysozyme appears to be quite effective againstListeria monocytogenes and other bacteria involved infoodborne diseases and food spoilage (105).Considering the widespread attention now focused onListeria in dairy products, especially cheeses, it is likelythat this application of lysozyme will be the subject offurther research. The preservative effects of lysozymein other foods were reviewed by Proctor andCunningham (99) and Cunningham et al. (100). It ispossible that some of these may be applicable to dairyproducts.

VI. GLUCOSE OXIDASE

Glucose oxidase (GO) catalyzes the oxidation of glu-cose to gluconic acid (via gluconic acid-�-lactone)according to the following reactions (106):

ð1Þ��

��

��

The hydrogen peroxide formed is normally reduced bycatalase present as a contaminant in commercial pre-parations of GO (from P. notatum, P. glaucum, or A.niger) or added separately. GO, which has a pH opti-mum � 5:5, is highly specific for D-glucose and is usedto assay specifically for D-glucose in the presence ofother sugars, blood, urine, etc.


In the food industry, GO has four principal applica-tions, which are not commercially very significant,especially in the dairy industry (1, 106, 107):

1. Removal of residual, trace levels of glucose. Thisapplication, which is particularly useful for the treat-ment of egg white prior to dehydration (although analterative procedure using yeast fermentation is morecommonly used), is of little if any significance in dairytechnology.

2. Removal of trace levels of oxygen. Traces of oxy-gen in wines and fruit juices cause discoloration and/oroxidation of ascorbic acid. Chemical reducing agentsmay be used to scavenge oxygen, but enzymatic treat-ment with GO may be preferred. The GO system hasbeen proposed as an antioxidant for high-fat products,such as mayonnaise, butter, and whole milk powder,but it does not appear to be used commercially for thispurpose, probably because of cost vis-a-vis chemicalantioxidants (if permitted) and the relative effective-ness of inert gas flushing of canned milk powder.

3. Generation of hydrogen peroxide in situ. Thehydrogen peroxide generated by glucose oxidase hasa direct bacteriocidal effect (which is a useful side effectof GO applied to egg products), but its bacteriocidalproperties can be much more effectively exploited as acomponent of the lactoperoxidase/hydrogen peroxide/thiocyanate system (see Sec. X). Two components ofthis system occur naturally in milk: lactoperoxidase ispresent at � 30mg/L, and thiocyanate, produced inthe rumen by hydrolysis of thioglucosides from mem-bers of the Brassicae family, varies from 0.017 to0.26mM. Hydrogen peroxide does not occur naturallyin bacteria-free milk, but it can be generated metabo-lically by catalase-negative bacteria, added directly(which is usually preferred) or produced in situ fromsodium percarbonate or by the action of xanthine oxi-dase on added hypoxanthine or by the action of glu-cose oxidase on glucose, either added or produced insitu from lactose by �-galactosidase. Activation of thelactoperoxidase–hydrogen peroxide–thiocyanate sys-tem suppresses the growth of psychrotrophs in milkstored at 5C and has given promising results as amilk preservative in tropical regions where refrigera-tion is lacking. It would appear that in such applica-tions the use of exogenous hydrogen peroxide is thesimplest and most appropriate.

4. Production of acid in situ. Direct acidification ofdairy products, particularly cottage, feta-type, andmozzarella cheeses, is now fairly common.Acidification is normally performed by addition ofacid or acidogen (usually gluconic acid-�-lactone) orby a combination of acid and acidogen. In situ produc-

tion of gluconic acid from added glucose or from glu-cose produced in situ from lactose by �-galactosidaseor from added sucrose by invertase has been proposed(108). It is possible to produce gluconic acid from glu-cose using immobilized glucose oxidase. However, it isdoubtful whether immobilized glucose oxidase couldbe applied to the acidification of milk because of thehigh probability of fouling by precipitated protein evenif low temperatures at which less extensive casein pre-cipitation occurs were used.

VII. SUPEROXIDE DIMUTASE

Superoxide dismutase (SOD) catalyzes the reduction ofsuperoxide anions, O�2, according to the following:

2O�2 þ 2Hþ�!H2O2 þO2 ð2ÞThe hydrogen peroxide formed may be reduced bycatalase, peroxidase, or a suitable reducing agent.SOD occurs widely in tissues where it plays a majorantioxidant role in scavenging superoxide radicalswhich are produced through the action of severalenzymes, e.g., XO and peroxidases.

Milk contains a low level of indigenous SOD whichhas been isolated and characterized (101, 102); the milkenzyme appears to be identical to SOD from bovineerythrocytes. The level of indigenous SOD in milk,which is entirely in the skim phase, varies considerablyamong individual cows, with stage of lactation andmastitic infection. It may play a role in the oxidativestability of milk, but attempts to correlate stabilitywith SOD activity have been inconclusive, presumablyowing to the interaction of various pro- and antioxi-dants. The tendency of fat in milk to oxidize is directlycorrelated with increases in XO and negatively withincreases in SOD activity. A low level of exogenousSOD, together with catalase, is a very effective antiox-idant in dairy products, and it has been suggested thattreatment of milk with SOD may be effective in pre-serving the flavor of UHT milk which is prone to lipidoxidation (1). It has been reported that a combinationof SOD and catalase is a more effective antioxidantthan butylated hydroxyanisole. However, it appearsto be ineffective as an antioxidant in the presence of0:1 ppm Cu2þ, apparently because Cu2þ effectivelycompetes with SOD for O�2 and converts it to lipid-reactive species such as OH. The commercial feasi-bility of using SOD as an antioxidant depends oncost, particularly vis-a-vis chemical antioxidants, ifpermitted. As far as is known, SOD is not usedcommercially as an antioxidant in the dairy industry.


VIII. SULFHYDRYL OXIDASE

Milk contains an enzyme capable of oxidizing sulfhy-dryl groups to disulfides:

2RSHþO2�!RSSRþH2O ð3ÞThe enzyme, which has molecular and catalytic proper-ties distinctly different from thioloxidase (EC 1.8.3.2),glutathione:protein disulfide oxidoreductase (EC1.8.4.2), and protein disulfide isomerase (EC 5.3.4.1),has been isolated and well characterized (101, 102).Sulfhydryl oxidase (EC 1.8.3.–) also has been demon-strated in human milk.

It has been proposed that sulfhydryl oxidase has aphysiological role in the formation of disulfide bondsin vivo to give proteins the correct three dimensionalstructure. Industrially, the potential of the enzyme liesin its ability to ameliorate the cooked flavor of UHT-treated milk (1, 109). The milk enzyme occurs exclu-sively in the serum (whey) from which it may be easilyisolated by exploiting its marked tendency to aggre-gate, thus facilitating its isolation by chromatographyon porous glass. The enzyme has been immobilized onporous glass and titanium oxide, and its effectivenessin ameliorating the cooked flavor of UHT-treated milkhas been demonstrated on a pilot scale using immobi-lized enzyme columns. For industrial-scale use, an ade-quate supply of the enzyme from whey may be alimitation but production of the enzyme by geneticallyengineered microorganisms might be feasible.

IX. CATALASE

2H2O2�!2H2OþO2 ð4ÞHydrogen peroxide is used for the cold-sterilization ofmilk in regions lacking refrigeration and perhaps insome developed countries also; e.g., it may be used totreat cheese milk in the United States (1, 107). Inunderdeveloped regions, treatment of milk with H2O2

is performed at ambient temperature and excess H2O2

is reduced by indigenous milk catalase or by chemicalinteraction with milk proteins in which it causes somephysicochemical changes, principally oxidation ofmethionine, with adverse effects on cheese quality.Side effects can be decreased by short exposure toH2O2 at � 65C, after which the residual H2O2 isreduced by added catalase, usually from beef liver orAspergillus niger. Most of the recent interest in the useof H2O2 as a preservative for milk has focused on thelactoperoxidase-H2O2-thiocyanate system in which

low concentrations of H2O2 are required (see Sec. X).However, at 400 mg/kg, H2O2 alone is a more effectivelong-term bactericidal agent than 8.5 mg/kg H2O2 inthe lactoperoxidase system. There was no significantdifference in the quality of cultured milk made fromeither lactoperoxidase-activated or H2O2-treated milk.The activity of lactic starter was significantly lower inthe former, but the rennet coagulability of the latterwas extended.

There is interest in using immobilized catalase reac-tors for milk pasteurization or for glucose oxidase-cat-alase reactors. Although catalase may be readilyimmobilized, it is inactivated rapidly on exposure tohydrogen peroxide. Hydrogen peroxide appears to bean effective agent for inactivation of aflatoxin M1. Asdiscussed above, catalase, usually as a contaminant, isnecessary for many of the applications of glucose oxi-dase and also to optimize the antioxidative propertiesof SOD.

X. LACTOPEROXIDASE

H2O2 þ 2AH�!2H2Oþ 2A ð5ÞBovine milk is rich in lactoperoxidase (� 30�g=mL),which is distinct from the myloperoxidase of leucocytesand salivary peroxidase. Lactoperoxidase has beenpurified by several investigators and well characterized(109–112).

The most important physiological and technologicalfeature of lactoperoxidase is its ability, in the presenceof H2O2 and thiocyanate, to inhibit the growth of sev-eral bacteria. Since lactoperoxidase is relatively heatresistant, there is adequate lactoperoxidase activityeven in milk that has been moderately to severelyheat-treated. Thiocyanate occurs in many animal tis-sues and fluids. It is produced endogenously during thedetoxification of thiosulfates and metabolic productsof sulfur amino acids and cyanide and from foods con-taining thioglucosides. Cows on pastures containingclover (rich in RCN) and nongrasses (e.g., Cruciferaecontaining thioglucosides) yield milk containing higherconcentrations of thiocyanide than cows on winter feedor lay pastures. Saliva contains high levels of thiocya-nate which is also secreted by gastric mucosal cells.Milk does not contain indigenous H2O2 but it maybe added or produced in situ (see Sec. VI). Thus,bovine milk possesses an effective antibacterialsystem which probably affects the intestinal micro-flora of calves (and presumably other species).Lactoperoxidase also appears to protect the mammary


gland against mastitic infection, especially during thedry period. The antibacterial significance of lactoper-oxidase in vivo has been reviewed (113, 114). Humanmilk appears to contain a low level of lactoperoxidase(< 0:1�g=mL) as well as myloperoxidase and eosino-phil peroxidase. The lactoperoxidase system may beexploited in vitro to extend the shelf life of milkunder conditions where refrigeration and pasteuriza-tion facilities are lacking (111–114).

While most interest in lactoperoxidase has focusedon the indigenous enzyme, it may also acquire signifi-cance as an exogenous enzyme. Techniques have beendeveloped for the commercial-scale purification of lac-toperoxidase from milk (42). Large-scale trials haveshown that addition of purified lactoperoxidase tocalf milk replacers markedly reduces the incidence ofdiarrhea and increases weight gain. Although the anti-bacterial properties of human milk do not depend onlactoperoxidase, its addition to bovine milk-basedinfant formulae may have beneficial effects (111–114).

XI. TRANSGLUTAMINASE

Transglutaminase (TGase; protein-glutamine �-gluta-myltransferase; EC 2.3.2.13) catalyzes an acyl transfer-ase reaction between the �-carboxyamide group ofpeptide-bound glutamine residues (acyl donors) and avariety of primary amines (acyl acceptors), includingamino acids, and the "-amino group of lysine residuesin certain proteins. In the absence of an amine, TGasecatalyzes the deamination of glutamine residues, withwater molecules acting as acyl acceptors. Thus, TGasecan modify proteins by incorporation of (a) an amine,(b) crosslinking, or (c) deamination:

The amino group might be from a simple aliphatic oraromatic amine, an amino acid, an amino sugar, or aphospholipid (�-aminoethanol moiety).

TGase is present in several animal tissues and bodyfluids and is involved in several biological phenomena,

including blood clotting, wound healing, keratinizationof epidermal tissue, and stiffening of cell membranes.An enzyme with a similar activity has been found inthe bacterium Streptoverticillium mobaraense; this isreferred to as MTGase (M ¼ microbialÞ. However,TGase and MTGase differ in a number of respects,most notably in molecular weight (� 77 and� 38 kDa, respectively) and dependence on Ca2þ;TGase requires Ca2þ, MTGase does not.

The activity of (M)TGase suggests several possibleapplications in the food industry:

1. Gelation of proteins by crosslinking2. Formation of restructured meat and fish pro-

ducts from smaller pieces or comminuted meats3. Stabilization of protein-stabilized foams or

emulsions (by crosslinking proteins in the stabi-lizing layer)

4. Modification of the functionality, e.g., solubilityor water binding, of proteins by converting glu-tamine to glutamic acid residues or binding ofamino sugars or phospholipids

5. Modification of the nutritional value of foodsby enzymatically attaching essential aminoacids to proteins

6. Conjugation of the same or different proteins7. Conjugation of amino sugars with proteins8. Conjugation of phospholipids with proteins9. Immobilization of enzymes

The feasibility of these applications has been demon-strated in laboratory studies, but the limited availabil-ity and cost of the enzyme to date have restrictedlarger-scale trials (guinea pig liver is the usual sourceof TGase). However, the gene for TGase has beeninserted into microorganisms (115) which would beexpected to increase the availability and reduce thecost of the enzyme.

Various aspects of the chemistry of TGase and itsapplications in foods have been reviewed by Motokiand Seguro (116) and Dickinson (117).

XII. �-LACTAMASE

ð9ÞResidues of penicillin in milk, resulting from treatmentof mastitis-infected cows with antibiotics, inhibit or


prevent the growth of lactic acid bacteria used in theproduction of cheese and fermented milks. They mayalso evoke an allergic response in susceptible consu-mers and create conditions for the selection of penicil-lin-resistant pathogens. The usual approach taken toprevent contamination of the milk supply with antibio-tics is to withhold milk from treated cows for an ade-quate period after treatment; obviously, this results ineconomic losses to farmers.

Procedures developed for the removal of �-lactam-type antibiotics from milk include columns of activatedcharcoal, ultrafiltration or diafiltration (1). While thesemethods may be suitable under certain circumstances,they have obvious limitations. An alternative approachis the inactivation of penicillin by �-lactamase [Eq. (9)].

Korycka-Dahl et al. (118) demonstrated the abilityof �-lactamase from Bacillus cereus to inactivate peni-cillin G such that Cheddar or Swiss cheeses of normalquality could be produced from the decontaminatedmilk. The enzyme was active at 4C so that penicillinin milk could be inactivated during cold storage on thefarm or at the factory. The enzyme lost only 50% of itsactivity following heating at 63C for 30 min and hencewould retain considerable activity in pasteurized milk,which may be illegal. These problems may be solved byusing immobilized �-lactamase (119). This procedureappears promising and awaits further development.

An alternative approach to solving the problemscaused by penicillin in fermented products is the selec-tion of �-lactamase-producing mutants of lactic acidbacteria (120).

XIII. PERSPECTIVES

Owing to the physicochemical properties of its consti-tuents, milk is very amenable to enzymatic modifica-tion and it will be apparent from the foregoing thatmany potential applications of enzymes in dairy tech-nology exist. However, in spite of numerous studies,only a few of these applications are commercially sig-nificant. Many of the potential applications are unsuc-cessful commercially because the proposed applicationis not sufficiently significant economically or alterna-tive solutions are available.

By far the most important application is the use ofrennets in cheese making, for which no alternativeexists. It is likely that rennets will remain the principalenzyme in dairy technology for the foreseeable future.An unlimited supply of chymosin from geneticallyengineered microorganisms is now available; it is pos-sible that the cheese-making properties of these micro-

bial recombinant chymosins will be improved throughgenetic engineering. The production of cheese isexpanding worldwide, and the market for rennets willtherefore continue to increase. Development of cheese-like products as food ingredients is a growth area, andit is likely that the application of proteinases andlipases in the production of such products will increase.The use of proteinases and peptidases to produce pro-tein hydrolysates with specific functional, nutritional,or physiological properties appears to be very promis-ing and has been attracting increasing attention.

Lipases have some traditional and novel applica-tions in cheese technology, but the growth potentialof these applications is probably limited. However,new and more effective lipases may be identified.Perhaps the greatest potential application of lipases isin the production of tailor-made lipids with superiornutritional or functional properties.

Although �-galactosidase has applications in dairytechnology, most of these are not commercially viableat present, mainly for economic reasons. Niche mar-kets exist for lactose-hydrolyzed milks and these mayexpand. Exploitation of the transferase activity of �-galactosidase in the production of new oligosacchar-ides with novel functional and/or nutritional propertiesappears to hold potential.

Transglutaminase should have several interestingapplications in dairy technology. The availability of acheaper enzyme should encourage work on the appli-cation of this enzyme.

With the possible exception of lactoperoxidase andlysozyme, the other enzymes discussed in this chapterappear to have very limited application in dairy tech-nology, at least under present circumstances.

REFERENCES

1. PF Fox, MB Grufferty. Exogenous enzymes in dairy

technology. In: PF Fox, ed. Food Enzymology, Vol. 1.

London: Elsevier Science Publishers, 1991, pp 219–

269.

2. PF Fox. Exogenous enzymes in dairy technology—a

review. J Food Biochem 17:173–199, 1993.

3. PF Fox, L Stepaniak. Enzymes in cheese technology.

Int Dairy J 3:509–530, 1993.

4. RJ Brown. Dairy products. In: T Nagodawithana, G

Reed, eds. Enzymes in Food Processing. New York:

Academic Press, 1993, pp 347–361.

5. M Desmazeaud, E Spinnler. Laits et produits laitieres.

In: V Larreta-Garde, ed. Enzymes en Agroalimentaire.

Paris: Lavoisier, 1997, pp 48–76.


6. PF Fox, ed. Advanced Dairy Chemistry. 1. Proteins.

London: Elsevier Science Publishers, 1992.

7. DM Mulvihill. Production, functional properties and

utilization of milk proteins. In: PF Fox, ed. Advanced

Dairy Chemistry. 1. Proteins. London: Elsevier


8. DG Dalgleish. Enzymatic coagulation of milk. In: PF

Fox, ed. Advanced Dairy Chemistry. 1. Proteins.

London: Elsevier Science, 1992, pp 579–619.

9. DG Dalgleish. The enzymatic coagulation of milk. In:

PF Fox, ed. Cheese: Chemistry, Physics and

Microbiology, Vol. 1. London: Elsevier Applied

Science, 1987, pp. 63–96.

10. PF Fox. Proteolysis and protein-protein interactions

in cheese manufacture. In: BFJ Hudson, ed.

Developments in Food Proteins 3. London: Elsevier


11. PF Fox. Coagulants and their action. Proceedings of

the 22nd International Dairy Congress, The Hague.

Dordrecht: Reidel Publishing, 1987, pp 61–73.

12. PF Fox. Rennets and their action in cheese manufac-

ture and ripening. Biotechnol Appl Biochem 10:522–

535, 1988.

13. PF Fox, DMMulvihill. Casein. In: P Harris, ed. Food

Gels. London: Elsevier Applied Science, 1990, pp 121–

173.

14. PF Fox, PLH McSweeney. Rennets and their role in

milk coagulation and cheese ripening. In: BA Law, ed.

Microbiology and Biochemistry of Cheese and

Fermented Milk, 2nd ed. London: Blackie Academic

& Professional, 1997, pp 1–49.

15. B Foltmann. General and molecular aspects of

rennets. In: PF Fox, ed. Cheese: Chemistry, Physics

and Microbiology, Vol. 1, 2nd ed. London: Chapman

& Hall, 1992, pp 37–68.

16. B Foltmann. Chymosin: a short review on foetal and

neonatal gastric proteases. Scand J Clin Lab Invest

52(suppl 210):65–79, 1992.

17. S Chitpinityol, MJC Crabbe. Chymosin and aspartic

proteinases. Food Chem 61:395–418, 1998.

18. JH Nelson. Impact of new milk clotting enzymes on

cheese technology. J Dairy Sci 58:1739–1750, 1975.

19. M Sternberg. Microbial rennets. Adv Appl Microbiol

20:135–157, 1976.

20. ML Green. Review of the progress of dairy science:

milk coagulants. J Dairy Res 44:159–188, 1977.

21. JA Phelan. Milk coagulants: an evaluation of alterna-

tives to standard calf rennet. PhD thesis, National

University of Ireland, Cork, 1986.

22. M Teuber. Production of chymosin (EC 3.4.23.4) by

microoganisms and its use for cheesemaking. Bulletin


pp 3–15.

23. International Dairy Federation. Fermentation-pro-

duced enzymes and accelerated ripening in cheesemak-

ing. Bulletin 269. Brussels: Int Dairy Federation, 1992.

24. A Carlson, GC Hill, NF Olson. The coagulation of

milk with immobilized enzymes: a critical review. Enz

Microbiol Technol 8:642–650, 1986.

25. PF Fox, J Law, PLH McSweeney, J Wallace.

Biochemistry of cheese ripening. In; PF Fox, ed.

Cheese: Chemistry, Physics and Microbiology, 2nd

ed. London: Chapman and Hall, 1993, pp 389–438.

26. PF Fox, TP O’Connor, PLH McSweeney, TP Guinee,

NM O’Brien. Cheese: physical, biochemical and

nutritional aspects. Adv Food Nutr Res 39:163–328,

1996.

27. PF Fox, JMWallace. Formation of flavor compounds

in cheese. Adv Appl Microbiol 45:17–85, 1997.

28. R Grappin, TC Rank, NF Olson. Primary proteolysis

of cheese proteins during ripening. J Dairy Sci

68:531–540, 1985.

29. TC Rank, R Grappin, NF Olson. Secondary proteo-

lysis of cheese during ripening: a review. J Dairy Sci

68:801–805, 1985.

30. BA Law. Proteolysis in relation to normal and accel-

erated cheese ripening. In: PF Fox, ed. Cheese:

Chemistry, Physics and Microbiology, Vol. 1.


31. PF Fox. Proteolysis in cheese during manufacturing

and ripening. J Dairy Sci 72:1379–1400, 1989.

32. IDF. Chemical methods for evaluating proteolysis in

cheesematuration. Bulletin 261. Brussels: International

Dairy Federation, 1991.

33. PF Fox, PLH McSweeney. Proteolysis in cheese dur-

ing ripening. Food Rev Int 12:457–509, 1996.

34. ERS Kunji, I Mierau, A Hagting, B Poolman, WN

Konings. The proteolytic systems of lactic acid bac-

teria. Antonie van Leeuwenhoek 70:187–221, 1996.

35. DM Rea. Comparison of Cheddar cheese made

with chymosin, Rhizomucor miehei proteinase or

Cryphonectria parasitica proteinase. MSc thesis,

National University of Ireland, Cork, 1997.

36. PF Fox, JM Wallace, S Morgan, CM Lynch, EJ

Niland, J Tobin. Acceleration of cheese ripening.

Antonie van Leeuwenhoek 70:271–279, 1996.

37. PF Fox, PLH McSweeney, CM Lynch. Significance of

non-starter lactic acid bacteria in Cheddar cheese.


38. KN Kilcawley, MG Wilkinson, PF Fox. Enzyme-

modified cheese: a review. Int Dairy J 8:1–10, 1998.

39. J Adler-Nissen. Enzymatic Hydrolysis of Food

Proteins. London: Elsevier Applied Science, 1986.

40. PF Fox, PA Morrissey, DM Mulvihill. Chemical and

enzymatic modification of food proteins. In: BJF

Hudson, ed. Developments in Food Proteins 1.

London: Applied Science Publishers, 1982, pp 1–60.

41. JL Maubois. Separation, extraction and fractionation

of milk protein components. Le Lait 64:485–495,

1984.

42. PF Fox, DM Mulvihill. Developments in milk protein

processing. Food Sci Technol Today 7:152–161, 1993.


43. DD Kitts, YV Yuan. Caseinophosphopeptides and

calcium bioavailability. Trends Food Sci Technol

3:31–35, 1992.

44. JL Maubois, J Leonil. Peptides du lait a activite bio-

logique. Le Lait 69:245–269, 1989.

45. MA Juillerat, R Baechler, R Berrocal, S Chanton, J-C

Scherz, R Jost. Tryptic phosphopeptides from whole

casein: preparation and analysis by fast protein liquid

chromatography. J Dairy Res 56:603–611, 1989.

46. D McDonagh, RJ Fitzgerald. Production of caseino-

phosphopeptides (CCPs) from sodium caseinate using

a range of commercial protease preparations. Int

Dairy J 8:39–45, 1998.

47. J Dziuba, P Minkiewiez. Influence of glycosylation on

micelle-stabilizing ability and biological properties of

C-terminal fragments of cow’s �-casein. Int Dairy J

6:1017–1044, 1996.

48. MH Abd El-Salam, S El-Shibiny, W Buchheim.

Characteristics and potential uses of casein macropep-

tide. Int Dairy J 6:327–341, 1996.

49. SC Marshall. Casein macropeptide from whey—a new

product opportunity. CSIRO Food Res Q 51:86–91,

1991.

50. PF Fox, A Flynn. Biological properties of milk pro-

teins. In: PF Fox, ed. Advanced Dairy Chemistry. 1.

Proteins. London: Elsevier Science, 1992, pp 255–284.

51. D Migliore-Samour, P Jolles. Casein, a prohormone

with an immuno-modulating role for the newborn?

Experimentia 44:188–199, 1988.

52. PF Fox, ed. Developments in Dairy Chemistry. 4.

Functional Milk Proteins. London: Elsevier Applied

Science, 1989.

53. A Kilara. Enzyme-modified protein food ingredients.

Process Biochem 20(5):149–157, 1985.

54. S Nakai, E Li-Chan. Chemical and enzymatic

modification of milk proteins. In: PF Fox, ed.

Developments in Dairy Chemistry. 4. Functional

Milk Proteins. London: Elsevier Applied Science,

1989, pp 347–376.

55. ZY Ju, J Otte, M Zakora, KB Qvist. Enzyme-induced

gelation of whey proteins: effect of protein denatura-

tion. Int Dairy J 7:71–78, 1997.

56. JH Nelson, RG Jensen, RE Pitas. Pregastric esterase

and other oral lipases—a review. J Dairy Sci 60:327–

362, 1977.

57. HJ Lee, NF Olson, DS Ryan. Characterization of

pregastric esterase. J Dairy Sci 63:1834–1838, 1980.

58. BJ Sweet, LC Matthews, T Richardson. Purification

and characterization of pregastric esterase from calf.

Arch Biochem Biophys 234:144–150, 1984.

59. S Bernback, O Hernell, L Blackberg. Purification and

molecular characterization of bovine pregastric lipase.

Eur J Biochem 148:233–238, 1985.

60. IM Roberts, P Jacobson, J Cornette. Secondary struc-

tures of rat lipolytic enzymes: circular dichroism and

relation to hydrophobic enzymes: circular dichroism

and relation to hydrophobic moments. Biochem

Biophys Res Commun 162:95–101, 1989.

61. AJP Docherty, MW Bodiner, S Angal, R Verger, C

Riviere, PA Lowe, A Lyons, JS Emtage, TJR Harris.

Molecular cloning and nucleotide sequence of rat lin-

gual lipase cDNA. Nucleic Acids Res 13:1891–1903,

1985.

62. JK Ha, RC Lindsay. Release of volatile branched-

chain and other fatty acids from ruminant milk fats

by various lipases. J Dairy Sci 76:677–690, 1993.

63. JK Ha, RC Lindsay. Influence of aw on volatile free

fatty acids during storage of cheese bases lipolyzed by

kid goat pregastric esterase. Int Dairy J 2:179–195,

1992.

64. RV Chaudhari, GH Richardson. Lamb gastric lipase

and protease in cheese manufacture. J Dairy Sci

54:467–471, 1971.

65. HJ Peppler, JG Dooley, HT Huang. Flavor develop-

ment in Fontina and Romano cheese by fungal ester-

ase. J Dairy Sci 59:859–862, 1976.

66. GT Moskowitz, SS Noelck. Enzyme-modified cheese

technology. J Dairy Sci 70:1761–1769, 1987.

67. K Kornacki, L Stepaniak, I Adamiec, J Grabska, K

Wroma. Characteristics of lipolytic mould prepara-

tions as compared to hog pancreatic lipase.

Milchwissenschaft 34:340–343, 1979.

68. M De Felice, T Gomes, T de Leonardis. Addition of

animal and microbial lipases to curd. Effects on free

fatty acid composition during ripening. Le Lait

71:637–643, 1991.

69. MW Arbige, PR Freund, SC Silver, JT Zelro. Novel

lipase for Cheddar cheese development. Food Technol

40(4):91–98, 1986.

70. J-C Gripon. Mould-ripened cheeses. In: PF Fox,

ed. Cheese: Chemistry, Physics and Microbiology,

Vol. 2, 2nd ed. London: Chapman & Hall, 1993,

pp 111–136.

71. RS Arnold, KM Shahani, BK Dwivedi. Application

of lipolytic enzymes to flavor development in dairy

products. J Dairy Sci 58:1127–1143, 1975.

72. RC Jolly, FV Kosikowski. Flavor development in

pasteurized milk blue cheese by animal and micro-

bial lipase preparations. J Dairy Sci 58:846–852,

1975.

73. A Tomasini, G Bustillo, J-M Lebault. Production of

blue cheese flavour concentrates from different sub-

strates supplemented with lipolyzed cream. Int Dairy

J 5:247–257, 1995.

74. A Kilara. Enzyme-modified lipid food ingredients.

Process Biochem 20(2):35–45, 1985.

75. J Grailli. Les enzymes en lipotechnie. In: V Larreta-

Garde, ed. Enzymes en Agroalimentaire. Paris:

Lavoisier, 1997, pp 246–257.

76. TP Shukla. �-Galactosidase technology: a solution to

the lactose problem. CRC Crit Rev Food Technol

5:325–356, 1975.


77. PG Hobman. Review of processes and products for

utilization of lactose in deproteinized milk serum. J

Dairy Sci 67:2630–2653, 1984.

78. JG Zadow. Lactose: properties and uses. J Dairy Sci

67:2654–2679, 1984.

79. HW Modler. Functional properties of nonfat dairy

ingredients—a review. Modification of lactose and

products containing whey proteins. J Dairy Sci

68:2206–2214, 1985.

80. PF Fox, ed. Developments in Dairy Chemistry. 3.

Lactose and Minor Constituents. London: Elsevier

Applied Science, 1985.

81. International Dairy Federation. Trends in Whey

Utilization. Bulletin 212. Brussels: Int Dairy

Federation, 1987.

82. VH Holsinger. Lactose. In: NP Wong, ed.

Fundamentals of Dairy Chemistry, 3rd ed. New

York: Van Nostrand Reinhold, 1988, pp 279–342.

83. International Dairy Federation. Proceedings of the

IDF Workshop on Lactose Hydrolysis. Bulletin 289.

Brussels: Int Dairy Federation, 1993.

84. ST Yang, EM Silva. Novel products and new technol-

ogies for use of a familiar carbohydrate, milk lactose.

J Dairy Sci 78:2541–2562, 1995.

85. PF Fox, ed. Advanced Dairy Chemistry. 3. Lactose,

Water, Salts and Vitamins, 2nd ed. London: Chapman

& Hall, 1997.

86. A Flynn. Nutritional significance of lactose. II.

Metabolism and toxicity of galactose. In: PF Fox,

ed. Developments in Dairy Chemistry. 3. Lactose

and Minor Constituents. London: Elsevier Applied

Science, 1985, pp 133–141.

87. DM Paige, LR Davis. Nutritional significance of lac-

tose. I. Nutritional aspects of lactose digestion. In: PF

Fox, ed. Developments in Dairy Chemistry. 3. Lactose

and Minor Constituents. London: Elsevier Applied

Science, 1985, pp 111–132.

88. A Mustapha, SR Hertzler, DA Savaiano. Lactose:

nutritional significance. In: PF Fox, ed. Advanced

Dairy Chemistry. 3. Lactose, Water, Salts and

Vitamins. London: Chapman & Hall, 1997, pp 127–

154.

89. RR Mahoney. Modification of lactose and lactose

containing dairy products with �-galactosidase. In:

PF Fox, ed. Developments in Dairy Chemistry. 3.

Lactose and Minor Constituents. London: Elsevier


90. RR Mahoney. Lactose: enzymatic modification. In:

PF Fox, ed. Advanced Dairy Chemistry. 3. Lactose,

Water, Salts and Vitamins. London: Chapman &

Hall, 1997, pp 77–125.

91. JB Smart. Transferase reactions of �-galactosidases—new product opportunities. In: Proceedings of the

IDF Workshop on Lactose Hydrolysis, Bulletin No.


pp 16–22.

92. RR Mahoney. Galactosyl-oligosaccharide formation

during lactose hydrolysis: a review. Food Chem

63:147–154, 1998.

93. DD Muir. Reviews of the progress of dairy science:

frozen concentrated milk. J Dairy Res 51:649–664,

1984.

94. JG Sabioni, DO Silva, AJR Pinheiro, AC Borges, JBP

Chaves. Control of lactose crystallization in ‘‘Dulce de

Leche’’ by Kluyveromyces lactis fermentation. J Dairy

Sci 67:1694–1698, 1984.

95. JG Sabioni, AJR Pinheiro, DO Silva, JBP Chaaves,

AC Borges. Control of lactose crystalization in

‘‘Dulce de Leche’’ by beta-D-galactosidase activity

from permeabilized cells of Kluyveromyces lactis

cells. J Dairy Sci 67:2210–2215, 1984.

96. J Patocka, P Jelen. Enzymatic lactose hydrolysis for

prevention of lactose crystallization in a whey spread.

J Food Sci 53:1370–1372, 1988.

97. SB Martinez, RA Speckman. �-Galactosidase treat-

ment of frozen dairy product mixes containing whey.

J Dairy Sci 71:893–900, 1988.

98. JG Zadow. Economic considerations related to the

production of lactose and lactose by-products. In:

Proceedings of the IDF Workshop on Lactose

Hydrolysis, Bulletin No. 289. Brussels: International

Dairy Federation, 1993, pp 10–15.

99. VA Proctor, FE Cunningham. The chemistry of lyso-

zyme and its use as a food preservative and a pharma-

ceutical. CRC Crit Rev Food Sci Nutr 26:359–395,

1988.

100. FE Cunningham, VA Proctor, SJ Goetsch. Egg-white

lysozyme as a food preservative: an overview. World’s

Poultry Sci J 47:141–163, 1991.


enzymes. In: PF Fox, ed. Food Enzymology,

Vol. 1. London: Elsevier Applied Science, 1991, pp

107–129.


enzymes. In: PF Fox, ed. Advanced Dairy


Science, 1992, pp 338–367.

103. International Dairy Federation. The use of lysozyme

in the preservation of late blowing in cheese. Bulletin

216. Brussels: author, 1987, pp 1–16.

104. International Dairy Federation. The use of lysozyme

to control butyric acid fermentation. A-Doc 112.

Brussels: author, 1988, pp 40–43.

105. VL Hughey, EA Johnson. Antimicrobial activity of

lysozyme against bacteria involved in food spoilage

and food-borne disease. Appl Environ Microbiol

53:2165–2170, 1987.

106. FE Hammer. Oxidoreductases. In: T Nagodawithana,

G Reed, eds. Enzymes in Food Processing, 3rd ed.

New York: Academic Press, 1993, pp 221–277.

107. T Szalkucki. Applications of oxidoreductases. In: T

Nagodawithana, G Reed, eds. Enzymes in Food


Processing, 3rd ed. New York: Academic Press, 1993,

pp 279–291.

108. AG Rand, JA Hourigan. Direct conversion of lactose

in milk to acid. J Dairy Sci 58:1144–1150, 1975.

109. PF Fox, PA Morrissey. Indigenous enzymes of bovine

milk. In: GG Birch, N Blakeborough, KJ Parker, eds.

Enzymes and Food Processing. London: Applied

Science Publishers, 1981, pp 213–238.

110. BJ Kitchen. Indigenous milk enzymes. In: PF Fox, ed.



Science, 1985, 239–279.

111. L Bjorck. Indigenous enzymes in milk. V.

Lactoperoxidase. In: PF Fox, ed. Food Enzymology,

Vol 1. London: Elsevier Applied Science, 1991,

pp 100–106.

112. L Bjorck. Indigenous enzymes in milk. V.

Lactoperoxidase. In: PF Fox, ed. Advanced Dairy


Science, 1992, pp 332–338.

113. B Reiter. The biological significance of the non-immu-

noglobulin protective proteins in milk: Lysozyme, lac-

toferrin, lactoperoxidase. In: PF Fox, ed.



Science, 1985, pp 281–336.

114. J Stadhouders, RR Beumer. Actual and potential

applications of the natural antimicrobial agents of

milk in the dairy industry. In: Indigenous

Antimicrobial Agents of Milk, Special Issue 9404.

Brussels: International Dairy Federation, 1994,

pp 175–197.

115. K Seguro, N Nio, M Motoki. Some characteristics of

a microbial protein cross-linking enzyme: transgluta-

minase. In: N Parris, A Kato, LK Creamer, and J

Pearce, eds. Macromolecular Interactions in Food

Technology. Washington: American Chemical

Society, 1996, pp 271–280.

116. M Motoki, K Seguro., Transglutaminase and its use

for food processing. Trends Food Sci Technol 9:204–

210, 1998.

117. E Dickinson. Enzymatic crosslinking of milk proteins.

Proc 25th Intern Dairy Congr (Aarhus), II:209–215,

1998.

118. M Korycka-Dahl, T Richardson, RL Bradley. Use of

microbial �-lactamase to destroy penicillin added to

milk. J Dairy Sci 68:1910–1916, 1985.

119. N-ZP Lee, T Richardson. Preparation and char-

acterization of immobilized �-lactamase for destruc-

tion of penicillin in milk. J Dairy Sci 70:2032–2039,

1987.

120. L Khosravi, WE Sandine. �-Lactamase-producing

mutants of Streptococcus cremoris. J Dairy Sci

70:1529–1543, 1987.


Documents

University of California–Davis Davis, California, U.S.A ... · activities that have been detected in milk but which have not been isolated and have no known signiﬁcance in milk