9 Heat-induced changes in milk
9.1 Introduction
In modern dairy technology, milk is almost always subjected to a heat
treatment; typical examples are:
Thermization
Pasteurization
e.g. 65°C x 15 s
LTLT (low temperature, long time) 63°C x 30 min
HTST (high temperature, short time) 72°C x 15 s
e.g. 90°C x 2-10 min, Forewarming (for sterilization)
Sterilization
120°C x 2 min
UHT (ultra-high temperature) 130-140°C x 3-5s
In-container 110-115°C x 10-20min
The objective of the heat treatment varies with the product being
produced. Thermization is generally used to kill temperature-sensitive
micro-organisms, e.g. psychrotrophs, and thereby reduce the microflora of
milk for low-temperature storage. The primary objective of pasteurization
is to kill pathogens but it also reduces the number of non-pathogenic
micro-organisms which may cause spoilage, thereby standardizing the milk
as a raw material for various products. Many indigenous enzymes, e.g.
lipase, are also inactivated, thus contributing to milk stability. Forewarming
(preheating) increases the heat stability of milk for subsequent sterilization
(as discussed in section 9.7.1). Sterilization renders milk shelf-stable for very
long periods, although gelation and flavour changes occur during storage,
especially of UHT-sterilized milks.
Although milk is a very complex biological fluid containing complex
protein, lipid, carbohydrate, salt, vitamins and enzyme systems in soluble,
colloidal or emulsified states, it is a very heat-stable system, which allows it
to be subjected to severe heat treatments with relatively minor changes in
comparison to other foods if subjected to similar treatments. However,
numerous biological, chemical and physico-chemical changes occur in milk
during thermal processing which affect its nutritional, organoleptic and/or
technological properties. The temperature dependence of these changes
348 DAIRY CHEMISTRY AND BIOCHEMISTRY
h
min.
T("C)
Figure 9.1 The time needed (1') at various temperatures (T) to inactivate some enzymes and
cryoglobulins; to kill some bacteria and spores; to cause a certain degree of browning; to
convert 1% of lactose to lactulose; to cause heat coagulation; to reduce available lysine by 1%;
and to make 10% and 75% of the whey proteins insoluble at pH 4.6 (from Walstra and Jenness,
1984).
Table 9.1 Approximate values for the temperature dependence of some reactions in heated milk
(modified from Walstra and Jennes, 1984)
Reaction Activation energy (kJ mol- ') Qlo at 100°C
Many chemical reactions
Many enzyme-catalysed reactions
Autoxidation of lipids
Maillard reactions (browning)
Dephosphorylation of caseinate
Heat coagulation of milk
Degradation of ascorbic acid
Heat denaturation of protein
Typical enzyme inactivation
Inactivation of milk proteinase
Killing vegetative bacteria
Killing of spores
(plasmin)
80-130
40 - 60
40-100
100-180
110-120
150
60-120
200-600
450
75
200-600
250-330
2.0-3.0
1.4-1.7
1.4-2.5
2.4-5.0
2.6-2.8
3.7
1.7-2.8
6.0-175.0
50.0
1.9
6.0-175.0
9.0- 17.0
varies widely, as depicted in general terms in Figure 9.1 and Table 9.1. The
most significant of these changes, with the exception of the killing of
bacteria, will be discussed below. In general, the effect(s) of heat on the
principal constituents of milk will be considered individually, although there
are interactions between constituents in many cases.
HEAT-INDUCED CHANGES IN MILK 349
9.2 Lipids
Of the principal constituents, the lipids are probably the least affected by
heat. However, significant changes do occur in milk lipids, especially in their
physical properties, during heating.
9.2.1 Physicochemical changes
Creaming. The chemical and physicochemical aspects of the lipids in milk
were discussed in Chapter 3. The principal effect of heat treatments on milk
lipids is on creaming of the fat globules. As discussed in Chapter 3, the fat
in milk exists as globules, 0.1-20pm in diameter (mean, 3-4pm). The
globules are stabilized by a complex membrane acquired within the secre-
tory cell and during excretion from the cell. Owing to differences in density
between the fat and aqueous phases, the globules float to the surface to form
a cream layer. In cows’ milk, the rate of creaming is far in excess of that
predicted by Stokes’ law, owing to aggregation of the globules which is
promoted by cryoglobulins (a group of immunoglobulins). Buffalo, ovine or
caprine milks do not undergo cryoglobulin-dependent agglutination of fat
globules and cream very slowly with the formation of a compact cream
layer.
When milk is heated to a moderate temperature (e.g. 70°C x 15 min), the
cryoglobulins are irreversibly denatured and hence the creaming of milk is
impaired or prevented; HTST pasteurization (72°C x 15 s) has little or no
effect on creaming potential but slightly more severe conditions have an
adverse effect (Figure 9.2).
Homogenization,, which reduces mean globule diameter to below 1 pm,
retards creaming due to the reduction in globule size but, more importantly,
to the denaturation of cryoglobulins which prevents agglutination. In fact,
there are probably two classes of cryoglobulin, one of which is denatured by
heating, the other by homogenization.
Changes in the fat globule membrane. The milk fat globule membrane
(MFGM) itself is altered during thermal processing. Milk is usually agitated
during heating, perhaps with foam formation. Agitation, especially of warm
milk in which the fat is liquid, may cause changes in globule size due to
disruption or coalescence; significant disruption occurs during direct
UHT processing. Foaming probably causes desorption of some membrane
material and its replacement by adsorption of skim-milk proteins. In these
cases, it may not be possible to differentiate the effect of heating from the
total effect of the process.
Heating per se to above 70°C denatures membrane proteins, with the
exposure and activation of various amino acid residues, especially cysteine.
3 50 DAIRY CHEMISTRY AND BIOCHEMISTRY
50 60 70 80
Temperature ("C)
Figure 9.2 Time-temperature curves for the destruction of M. tuberculosis (. . .), inactivation
of alkaline phosphatase (-) and creaming ability of milk (---) (from Webb and Johnson,
1965).
This may cause the release of H,S (which can result in the development of
an off-flavour) and disulphide interchange reactions with whey proteins,
leading to the formation of a layer of denatured whey proteins on the fat
globules at high temperatures (> l0OT). The membrane and/or whey
proteins may participate in Maillard browning with lactose and the cysteine
may undergo p-elimination to dehydroalanine, which may then react with
lysine to form lysinoalanine or with cysteine residues to form lanthionine,
leading to covalent cross-linking of protein molecules (section 9.6.3). Mem-
brane constituents, both proteins and phospholipids, are lost from the
membrane to the aqueous phase at high temperatures. Much of the
indigenous copper in milk is associated with the MFGM and some of it is
transferred to the serum on heat processing. Thus, severe heat treatment of
cream improves the oxidative stability of butter made from it as a result of
the reduced concentration of pro-oxidant Cu in the fat phase and the
antioxidant effect of exposed sulphydryl groups.
The consequences of these changes in the MFGM have been the subject
of little study, possibly because severely heated milk products are usually
homogenized and an artificial membrane, consisting mainly of casein and
some whey proteins, is formed; consequently, changes in the natural mem-
brane are not important. Damage to the membrane of unhomogenized
products leads to the formation of free (non-globular) fat and consequently
to 'oiling-off and the formation of a 'cream plug' (Chapter 3).
HEAT-INDUCED CHANGES IN MILK 351
Severe heat treatment, as is encountered during roller drying and to a
lesser extent spray drying, results in at least some demulsification of milk
fat, with the formation of free fat, which causes (Chapter 3):
the appearance of fat droplets when such products are used in tea or
coffee;
increased susceptibility of the fat to oxidation, since it is not protected by
a membrane;
reduced wettability/dispersibility of the powder;
a tendency of powders to clump.
9.2.2 Chemical changes
Severe heat treatments, e.g. frying, may convert hydroxyacids to lactones,
which have strong, desirable flavours and contribute to the desirable
attributes of milk fat in cooking.
Release of fatty acids and some interesterification may also occur, but
such changes are unlikely during the normal processing of milk.
Naturally occurring polyunsaturated fatty acids are methylene-interrup-
ted but may be converted to conjugated isomers at high temperatures. Four
& R1
12 10 9
R2 12 in
- Rl
R2
9.c- 11.1
9.1-1 1, t
13 12 10 9
Linoleic acid
- -
Rl
water
9,C-12.C
Liirolric acid
Rl
R1
12 10
R2
11
R2
I(l,t - 12.c
111,t - 12, t
Figure 9.3 Isomers of conjugated linoleic acid.
352 DAIRY CHEMISTRY AND BIOCHEMISTRY
Table 9.2 Concentration of conjugated linoleic acid (CLA) isomers in selected foods (modified
from Ha, Grimm and Parka, 1989)
Sample
mg CLA/kg Fat content CLA in fat
food (YO.) (mg kg-')
Parmesan cheese
Cheddar cheese
Romano cheese
Blue cheese
Processed cheese
Cream cheese
Blue spread
Cheese whiz
Milk
pasteurized whole
non-pasteurized whole
Ground beef
grilled
uncooked
622.3 f 15.0
440.6 f 14.5
356.9 f 6.3
169.3 f 8.9
574.1 f 24.8
334.5 f 13.3
202.6 & 6.1
1815.0 90.3
28.3 f 1.9
34.0 f 1.0
994.0 f 30.9
561.7 f 22.0
32.3 f 0.9
32.5 f 1.7
32.1 f 0.8
30.8 f 1.5
31.8 f 1.1
35.5 f 1.0
20.2 0.8
20.6 & 1.1
4.0 f 0.3
4.1 i 0.1
10.7 f 0.3
27.4 f 0.2
1926.7
1355.7
1111.9
549.8
1805.3
942.3
1003.0
8810.7
707.5
829.3
9289.7
2050.0
isomers of conjugated linoleic acid (CLA) are shown in Figure 9.3. It is
claimed that CLA has anticarcinogenic properties. The mechanism of CLA
formation in foods in general is not clear but heat treatment, free radical-
type oxidation and microbial enzymatic reactions involving linoleic and
linolenic acids in the rumen are thought to be major contributors. Rather
high concentrations of CLA have been found in heated dairy products,
especially processed cheese (Table 9.2). It has been suggested that whey
proteins catalyse isomerization.
9.3 Lactose
The chemistry and physicochemical properties of lactose, a reducing disac-
charide containing galactose and glucose linked by a p( l-4)-bond, were
described in Chapter 2.
When severely heated in the solid or molten state, lactose, like other
sugars, undergoes numerous changes, including mutarotation, various
isomerizations and the formation of numerous volatile compounds, includ-
ing acids, furfural, hydroxymethylfurfural, CO, and CO. In solution under
strongly acidic conditions, lactose is degraded on heating to monosacchar-
ides and other products, including acids. These changes do not normally
occur during the thermal processing of milk. However, lactose is relatively
unstable under mild alkaline conditions at moderate temperatures where it
undergoes the Lobry de Bruyn- Alberda van Ekenstein rearrangement of
aldoses to ketoses (Figure 9.4).
I
I
+
I
m
s
0 P
0
0 c
.-
I
3 54 DAIRY CHEMISTRY AND BIOCHEMISTRY
Lactose undergoes at least three heat-induced changes during the pro-
cessing and storage of milk and milk products.
9.3. I Formation of lactulose
On heating at low temperatures under slightly alkaline conditions, the
glucose moiety of lactose is epimerized to fructose with the formation of
lactulose, which does not occur in nature. The significance of lactulose has
been discussed in Chapter 2. Lactulose is not formed during HTST process-
ing but is formed during UHT sterilization (more during indirect than direct
heating) and especially during in-container sterilization; therefore, the con-
centration of lactulose in milk is a useful index of the severity of the heat
treatment to which the milk has been subjected (see Figure 2.19). The
concentration of lactulose is probably the best index available at present for
differentiating between UHT and in-container sterilized milks and a number
of assay procedures have been developed, using HPLC or enzymatic/
spectrophotometric principles.
9.3.2 Formation of acids
Milk as secreted by the cow contains about 200 mg CO, 1-'. Owing to its
low concentration in air, CO, is rapidly and, in effect, irreyersibly lost from
milk on standing after milking; its loss is accelerated by heating, agitation
2
m
u
u
.r -
9
0 1 2 3
Heating period at 116°C (h)
Figure 9.5 Changes in titratable acidity (O), lactic acid (0) and lactose (0) on heating
homogenized milk in sealed cans at 116°C. Titratable acidity expressed as mg lactic acid/100 g
milk (from Gould, 1945.)
HEAT-INDUCED CHANGES IN MILK 355
7-
6-
c
2 .d
3
8
5-
2 4-
5 3-
u CJ
z
-
E
2-
I I 1
90 100 110 120
Temperature of heating ("C)
Figure 9.6 Effect of temperature on the rate of heat-induced production of acid in milk (from
Jenness and Patton, 1959).
and vacuum treatment. This loss of CO, causes an increase in pH of about
0.1 unit and a decrease in the titratable acidity of nearly 0.02%, expressed
as lactic acid. Under relatively mild heating conditions, this change in pH is
more or less offset by the release of H+ on precipitation of Ca,(PO,),, as
discussed in section 9.4.
On heating at temperatures above lOO"C, lactose is degraded to acids
with a concomitant increase in titratable acidity (Figures 9.5, 9.6). Formic
acid is the principal acid formed; lactic acid represents only about 5% of the
acids formed. Acid production is significant in the heat stability of milk, e.g.
when assayed at 130"C, the pH falls to about 5.8 at the point of coagulation
(after about 20min) (Figure 9.7). About half of this decrease is due to the
formation of organic acids from lactose; the remainder is due to the
precipitation of calcium phosphate and dephosphorylation of casein, as
discussed in section 9.4.
In-container sterilization of milk at 115°C causes the pH to decrease to
about 6 but much of this is due to the precipitation of calcium phosphate;
the contribution of acids derived from lactose has not been quantified
accurately. Other commercial heat treatments, including UHT sterilization,
cause insignificant degradation of lactose to acids.
356 DAIRY CHEMISTRY AND BIOCHEMISTRY
r
E
.-
L
0
5.6
0 10 20 30 40
Heating period at 130°C (min)
Figure 9.7 The pH of samples of milk after heating for various periods at 130°C with air (O),
0, (0) or N, (A) in the headspace above the milk; T, coagulation time (from Sweetsur and
White, 1975).
9.3.3 Maillard browning
The mechanism and consequences of the Maillard reaction were discussed
in Chapter 2. The reaction is most significant in severely heat-treated
products, especially in-container sterilized milks. However, it may also occur
to a significant extent in milk powders stored under conditions of high
humidity and high temperature, resulting in a decrease in the solubility of
the powder. If cheese contains a high level of residual lactose or galactose
(due to the use of a starter unable to utilize galactose; Chapter lo), it is
susceptible to Maillard browning, especially during cooking on pizza, e.g.
Mozzarella (Pizza) cheese. Browning may also occur in grated cheese during
storage if the cheese contains residual sugars; in this case, the water activity
of the cheese (a, - 0.6) is favourable for the Maillard reaction. Poorly
washed casein and especially whey protein concentrates (which contain
30-60% lactose) may undergo Maillard browning when used as ingredients
in heat-treated foods.
Maillard browning in milk products is undesirable because:
1. The final polymerization products (melanoidins) are brown and hence
dairy products which have undergone Maillard browning are discoloured
and aesthetically unacceptable.
HEAT-INDUCED CHANGES IN MILK 357
2. Some of the by-products of Maillard browning have strong flavours (e.g.
furfural, hydroxymethylfurfural) which alter the typical flavour of milk.
3. The initial Schiff base is digestible but after the Amadori rearrangement,
the products are not metabolically available. Since lysine is the amino
acid most likely to be involved and is an essential amino acid, Maillard
browning reduces the biological value of proteins. Interaction of lysine
with lactose renders the adjacent peptide bond resistant to hydrolysis by
trypsin, thereby reducing the digestibility of the protein.
4. The polymerized products of Maillard browning can bind metals, especi-
ally Fe.
5. It has been suggested that some products of the Maillard reaction are
toxic and/or mutagenic but such effects are, at most, weak and possibly
due to other consequences of browning, e.g. metal binding.
6. The attachment of sugars to the protein increases its hydrophilicity;
however, solubility may be reduced, probably due to cross-linking of
protein molecules.
7. The heat stability of milk is increased by the Maillard reaction, probably
via the production of carbonyls (section 9.7).
The formation of brown pigments via the Maillard reaction, especially
in model systems (e.g. glucose-glycine), usually follows zero-order kinetics,
but the loss of reactants has been found to follow first- or second-order
kinetics in foods and model systems. Activation energies of 109, 116 and
139 kJ mo1-l have been reported for the degradation of lysine, the forma-
tion of brown pigments and the production of hydroxymethylfurfural
(HMF), respectively.
Browning can be monitored by measuring the intensity of brown colour,
the formation of hydroxymethylfurfural (which may be measured spectro-
photometrically, after reaction with thiobarbituric acid, or by HPLC, but
which is not regarded as a very good indicator of Maillard browning), loss
of available lysine (e.g. by reaction with 2,4-dinitrofluorobenzene) or by the
formation of furosine. Furosine is formed on the acid hydrolysis of lactulosyl
lysine (the principal Maillard product formed during the heating of milk).
During acid hydrolysis, lactulosyl lysine is degraded to fructosylysine which
is then converted to pyridosine, furosine and carboxymethyl lysine (Figure
9.8). Furosine may be determined by ion-exchange chromatography, GC or
HPLC, and is considered to be a very good indicator of Maillard browning
and the severity of heat treatment of milk (Erbersdobler and Dehn-Miiller,
1989). The effects of time and temperature on the formation of furosine are
shown in Figure 9.9. The concentration of furosine is highly correlated with
the concentrations of HMF and carboxymethyl lysine. The concentration of
furosine in commercial UHT milks is shown in Figure 9.10.
Dicarbonyls, which are among the products of the Maillard reaction, can
react with amines in the Strecker reaction, producing a variety of flavourful
358
.~.~..~...~..... .~~...~........~~~~....~..
/ Oxidative
cleavage
rCOOH I
DAIRY CHEMISTRY AND BIOCHEMISTRY
t
1. Addition compound
I
2. N-substituted glycosylamine
3a. Schiff base
3b. Enol form
> Fructosylysh <
(Galactose)
(Glucose)
H 0’ OCH3 II
0
Pvrldoslne
J
HC -OH
HC -OH
HZC -OH
I
I
Ervthroluc
&id
OR
RI C-CH-NH
I
11 /
W2)4
I
NH
I
y+2
CGQH
Carboxymethvl
BROWNING
Figure 9.8 Initial steps of the Maillard reaction with the formation of furosine (after hydrolysis
with 7.8 M HCI) as well as of N-E-carboxymethyl lysine and erythronic acid (from Erbersdobler
and Dehn-Muller, 1989).
HEAT-INDUCED CHANGES IN MILK
3 59
Figure
80
60
40
20
0
I2HS
93s
56s
32s
16s
4s
2s
-
100 110 120 130 140 150
Temperature ("C)
9.9 Effect of heating temperature and time on the concentration of furosine
heated UHT milks (from Erbersdobler and Dehn-Muller, 1989).
in directly
7 14 71 28 35 42 49 56 63 70 77 84 97
Furosine (mg/l)
Figure 9.10 Relative distribution of the furosine concentrations in 190 commercial UHT milks
in increments of 7 mg furosine (from Erbersdobler and Dehn-Muller, 1989).
360 DAIRY CHEMISTRY AND BIOCHEMISTRY
compounds (Figure 2.32). The Maillard and especially the Strecker reactions
can occur in cheese and may be significant contributors to flavour; in this
case, the dicarbonyls are probably produced via biological, rather than
thermal, reactions.
9.4 Milk salts
Although the organic and inorganic salts of milk are relatively minor
constituents in quantitative terms, they have major effects on many aspects
of milk, as discussed in Chapter 5. Heating has little effect on milk salts with
two exceptions, carbonates and calcium phosphates. Most of the potential
carbonate occurs as CO, which is lost on heating, with a consequent
increase in pH. Among the salts of milk, calcium phosphate is unique in that
its solubility decreases with increasing temperature. On heating, soluble
calcium phosphate precipitates on to the casein micelles, with a concomitant
decrease in the concentration of calcium ions and pH (Chapter 5). These
changes are reversible on cooling if the heat treatment was not severe.
Following severe heat treatment, the heat-precipitated calcium phosphate is
probably insoluble but some indigenous colloidal calcium phosphate dis-
solves on cooling to partly restore the pH. The situation becomes rather
complex in severely heated milk due to the decrease in pH caused by
thermal degradation of lactose and dephosphorylation of casein.
The cooling and freezing of milk also cause shifts in the salts equilibria
in milk, including changes in pH, as discussed in Chapters 2, 5 and 11.
9.5 Vitamins
Many of the vitamins in milk are relatively heat labile, as discussed in
Chapter 6.
9.6 Proteins
The proteins of milk are probably the constituents most affected by heating.
Some of the changes involve interaction with salts or sugars and, although
not always fully independent of changes in other constituents, the principal
heat-induced changes in proteins are discussed in this section.
9.6.1 Enzymes
As discussed in Chapter 8, milk contains about 60 indigenous enzymes
derived from the secretory cells or from blood. Stored milk may also contain
enzymes produced by micro-organisms. Both indigenous and bacterial
HEAT-INDUCED CHANGES IN MILK 361
h
c
.-
E
v
E
F
Xanthine oxidase
0.1 :
0.01 ! I 1 I 1
60 70 80 90 100
Temperature ("C)
Figure 9.11 Time-temperature combinations required for which milk must be heated to a
certain temperature to inactivate some indigenous milk enzymes (from Walstra and Jenness,
1984).
enzymes can have undesirable effects in milk and dairy products. Although
not the primary objective of thermal processing, some of the indigenous
enzymes in milk are inactivated by the commercially used heat processes,
although many are relatively heat stable (Figure 9.1 1).
The thermal denaturation of indigenous milk enzymes is important from
two major viewpoints:
1. To increase the stability of milk products. Lipoprotein lipase is probably
the most important in this regard as its activity leads to hydrolytic
rancidity. It is extensively inactivated by HTST pasteurization but
heating at 78°C x 10 s is required to prevent lipolysis. Plasmin activity is
actually increased by HTST pasteurization due to inactivation of inhibi-
tors of plasmin and/or of plasminogen activators.
2. The activity of selected enzymes is used as indices of thermal treatments,
e.g. alkaline phosphatase (HTST pasteurization), y-glutamyl transpep-
tidase (index of heating in the range 72-80°C) or lactoperoxidase
(80-90°C).
Microbial enzymes. The widespread use of refrigerated storage of milk at
farm and factory for extended periods has led to psychrotrophs, especially
362 DAIRY CHEMISTRY AND BIOCHEMISTRY
Pseudomonas Juorescens, becoming the dominant micro-organisms in raw
milk supplies. Psychrotrophs are quite heat labile and are readily killed by
HTST pasteurization and even by thermization. However, they secrete
extracellular proteinases, lipases and phospholipases that are extremely heat
stable - some are not completely inactivated by heating at 140°C for 1 min
and thus partially survive UHT processing. If the raw milk supply contains
high numbers of psychrotrophs (> lo6 per ml), the amounts of proteinase
and lipase that survive UHT processing may be sufficient to cause off-
flavours, such as bitterness, unclean and rancid flavours, and perhaps
gelation.
One of the very curious characteristics of the proteinases and lipases
secreted by many psychrotrophs is that they have relatively low stability in
the temperature range 50-65"C, Figure 9.12 (the precise value depends on
the enzyme). Thus, it is possible to reduce the activity of these enzymes in
milk by a low temperature inactivation (LTI) treatment (e.g. 60°C x 5-
10 min) before or after UHT processing. Inactivation of the proteinase by
LTI appears to be due mainly to proteolysis; in the native state, the enzyme
is tightly folded and resistant to proteolysis by other proteinase molecules
in its neighbourhood but at about 60"C, some molecules undergo confor-
mational changes, rendering them susceptible to proteolysis by proteinase
molecules which are still active. On increasing the temperature further, all
proteinase molecules are denatured and inactive but they can renature on
40 60 80 100 120 140 160
Temperature ("C)
Figure 9.12 Thermal inactivation of Ps. jluorescens AFT 36 proteinase on heating for 1 min in
0.1 M phosphate buffer, pH 6.6 (0) or in a synthetic milk salts buffer, pH 6 (0) (from
Stepaniak, Fox and Daly, 1982).
HEAT-INDUCED CHANGES IN MILK 363
cooling. Since this mechanism does not apply to purified lipase, the
mechanism of LTI of lipase is not clear (for reviews on enzymes from
psychrotrophs see Driessen (1989) and McKellar (1989)).
9.6.2 Denaturation of other biologically active proteins
Milk contains a range of biologically active proteins, e.g. vitamin-binding
proteins, immunoglobulins, metal-binding proteins, antibacterial proteins
(lactotransferrin, lysozyme, lactoperoxidase), various growth factors and
hormones (Chapters 4 and 8). These proteins play important nutritional and
physiological functions in the neonate. All these proteins are relatively heat
labile - some are inactivated by HTST pasteurization and probably all are
inactivated by UHT and more severe heat treatments. Inactivation of these
biologically active proteins may not be particularly important when milk is
used in the diet of adults but may be highly significant in infant formulae;
consequently, supplementation of infant formulae with some of these pro-
teins is advocated.
9.6.3 Denaturation of whey proteins
The whey proteins, which represent about 20% of the proteins of bovine
milk, are typical globular proteins with high levels of secondary and tertiary
structures, and are, therefore, susceptible to denaturation by various agents,
including heat. The denaturation kinetics of whey proteins, as measured by
loss of solubility in saturated NaCl at pH4.6, are summarized in Figure
50 i
0 20 40 60 80 100
Heating time (rnin)
Figure 9.13 Heat denaturation of whey proteins on heating skim milk at various temperatures
("C) as measured by precipitability with saturated NaCl (from Jenness and Patton, 1959).
364 DAIRY CHEMISTRY AND BIOCHEMISTRY
9.13. Thermal denaturation is a traditional method for the recovery of
proteins from whey as 'lactalbumin'; coagulation is optimal at pH 6 and
about 90°C for 10 min (Chapter 4).
The order of heat stability of the whey proteins, measured by loss of
solubility, is: a-lactalbumin (a-la) > P-lactoglobulin (p-lg) > blood serum
albumin (BSA) > immunoglobulins (Ig) (Figure 9.14). However, when
measured by differential scanning calorimetry, quite a different order is
observed: Ig > p-lg > r-la > BSA. In the case of a-la, the discrepancy
appears to be explained by the fact that it is a metallo (Ca)-protein which
renatures quite readily following thermal denaturation. However, the Ca-
free apoprotein is quite heat labile, a fact which is exploited in the isolation
of a-la. The Ca2+ is bound in a pocket to the carboxylic acid groups of three
Asp residues and the carbonyls of an Asp and a Lys residue (Chapter 4).
The carboxylic acid groups become protonated below about pH 5 and lose
their ability to bind Ca; the apoprotein can be aggregated by heating to
about 55"C, leaving mainly p-lg in solution. Apo-lactoferrin is also consider-
ably less stable than the intact protein.
The denaturation of &-la and p-lg in milk follows first- and second-order
kinetics, respectively (Figure 9.15). Both proteins show a change in the
temperature-dependence of denaturation at about 90°C (Figure 9.15).
The mechanism of the thermal denaturation of p-lg has been studied
extensively; the sequence of events is shown schematically in Figure 9.16. At
about 20°C in the pH range 5.5-7.0, p-lg exists as an equilibrium between
-
Y
E
.-
e,
c
0.3
x
c
rn 0.2
2
0.0
I
30 40 50 60 70 80 90 100
Temperature ("C)
Figure 9.14 The denaturation of the total (0) and individual whey proteins in milk, heated at
various temperatures for 30 min; /?-lactoglobulin (m), r-lactalbumin (O), proteose peptone (O),
immunoglobulins (A), and serum albumin (A) (from Webb and Johnson, 1965).
HEAT-INDUCED CHANGES IN MILK
1-
2-
3-
4-
365
s; I I 1 I 1
175 150 12s I00 7s SO
Temperature ("C)
Figure 9.15 Arrhenius plot of the rate constant for the heat treatment of a-lactalbumin (0)
and P-lactoglobulin (0) (from Lyster, 1970).
its dimeric (NJ and monomeric (2N) forms. Between pH7 and 9, it
undergoes a reversible conformational change, referred to as the N P R
transition. Both equilibria are pushed to the right as the temperature is
increased, i.e. N, + 2N + 2R. Above about 65"C, p-lg undergoes reversible
denaturation (R P D) but at about 70"C, denaturation becomes irreversible
via a series of aggregation steps. The initial type I aggregation involves the
formation of intermolecular disulphide bonds while the later type I1 aggre-
gation involves non-specific interactions, including hydrophobic and elec-
trostatic bonding. Type I11 aggregation involves non-specific interactions
and occurs when the sulphydryl groups are blocked.
Some of the most important consequences of the heat denaturation of
whey proteins are due to the fact that these proteins contain sulphydryl
and/or disulphide residues which are exposed on heating (Figure 9.17). They
are important for at least the following reasons:
1. The proteins can participate in sulphydryl-disulphide interchange reac-
tions at temperatures above about 75°C at the pH of milk, but more
rapidly at or above pH 7.5. Such interactions lead to the formation of
disulphide-linked complexes of p-lg with K-casein, and probably cc,,-
casein and cc-la, with profound effects on the functionality of the milk
protein system, such as rennet coagulation and heat stability.
N2 - 2N - -2R - - 2D-Al -A2 - A,,
Dimer monomer Ionization-linked Reversible Type I aggregation Type I1 aggregation
dissociation transition denaturation
Irreversible denaturationlaggregation
Figure 9.16 Stages in the thermal denaturation of /I-lactoglobulin (from Mulvihill and Donovan, 1987).
HEAT-INDUCED CHANGES IN MILK 367
0 20 40 60 80 100 120 140
Time of heating (min)
Figure 9.17 Exposure of sulphydryl groups by heating milk at 75 (O), 80 (O), 85 (A) or 95
(A) "C; de-aerated milk heated at 85°C (a) (from Jenness and Patton, 1959).
Products
unknown
y2
Histidine
Tryptophan
Lysinoalanine
CH2 0
CH3 0 [HI II I/
H~N-C-LLOH I 7 H2N-C-C-OH - y3
c=o
I
H
Alanine
/ 3 1 :\ k-H
Ornithine pyruvic acid
THz
CH2 0 I I/ HO
H
I 1. H2N- C - C- OH I I/
H2N-C-C-OH I H2N- C- C-OH
I 7H2 I
p2
S
p-aminoalnnine
I
CH2 0
I //
I
H
H2N-C-C-OH
Lanthionine
NH
Figure 9.18 Interaction of dehydroalanine with amino acids.
368 DAIRY CHEMISTRY AND BIOCHEMISTRY
2. The activated sulphydryls may decompose with the formation of H,S and
H,C-S-CH,, which are responsible for the cooked flavour of severely
heated milk, including UHT milk. These compounds are volatile and
unstable and disappear within about 1 week after processing so that the
flavour of UHT milk improves during the first few weeks after processing.
3. Serine, serine phosphate, glycosylated serine, cysteine and cystine residues
can undergo 8-elimination with the formation of dehydroalanine. Dehy-
droalanine is very reactive and can react with various amino acid
residues, especially lysine, leading to the formation of lysinoalanine, and
to a lesser extent with cysteine with the formation of lanthionine (Figure
9.18). These reactions lead to intra- or intermolecular cross-linking which
reduce protein solubility, digestibility and nutritive value (because the
bonds formed are not hydrolysed in the intestinal tract and lysine is an
essential amino acid). Although there are reports to the contrary, lysino-
alanine is not normally found in UHT milk or cream.
9.6.4 Efect of heat on caseins
As discussed in Chapter 4, the caseins are rather unique proteins. They are
rather small (20-25 kDa), relatively hydrophobic molecules, with little
higher structure, few disulphide bonds (present only in the two minor
caseins, tls2 and K) and no sulphydryl groups. All the caseins are phos-
phorylated (8-9, 10-13,4-5 and 1 mole P per mole protein for uSl-, tlS2-, p-
and K-casein, respectively); due to their high levels of phosphorylation, ctsl-,
usz- and 8-caseins bind calcium strongly, causing them to aggregate and
precipitate, and affecting their general stability, including heat stability.
Within the strict sense of the term, the caseins are not susceptible to
thermal denaturation, e.g. sodium caseinate (pH 6.5-7.0) may be heated at
140°C for more than 1 h without any visible physicochemical changes.
However, severe heat treatments do cause substantial changes, e.g. dephos-
phorylation (about 100% in 1 h at 140"C), aggregation (as indicated by
changes in urea-PAGE or gel permeation chromatography), possibly due
to the formation of intermolecular disulphide and intermolecular isopeptide
bonds, cleavage of peptide bonds (formation of peptides soluble at pH 4.6
or in 12% TCA). 8-Elimination of serine, serine phosphate and cysteine
residues may also occur, especially at pH values above 7. Such heat-induced
changes are evident in commercial sodium caseinate.
The remarkably high heat stability of the caseins allows heat-sterilized
dairy products to be produced without major changes in physical properties
(reviewed by Fox, 1982; Singh and Creamer, 1992). The heat stability of
unconcentrated milk is almost always adequate to withstand the tempera-
ture treatments to which it is normally subjected; only rarely is a defect
known as the 'Utrecht phenomenon' encountered, when milk coagulates on
HTST heating. This defect is due to a very high Caz+ concentration owing
HEAT-INDUCED CHANGES IN MILK 369
to a low concentration of citrate, arising from poor feed. However, the heat
stability of milk decreases sharply on concentration and is usually inad-
equate to withstand in-container or UHT processing unless certain adjust-
ments and/or treatments are made. Although the heat stability of
concentrated milk is poorly correlated with that of the original milk, most
of the research on the heat stability of milk has been done on unconcen-
trated milk.
9.7 Heat stability of milk
Studies on the heat stability of milk date from the pioneering work of
Sommer and Hart, which commenced in 1919. Much of the early work
concentrated on attempts to relate heat stability to variations in milk
composition, especially the concentrations of milk salts. Although the heat
coagulation time (HCT) of milk is inversely related to the concentrations of
divalent cations (Ca2+ and Mg2+) and positively with the concentrations of
polyvalent anions (i.e. phosphate and citrate), the correlations are poor and
unable to explain the natural variations in HCT. This failure was largely
explained in 1961 by Rose who showed that the HCT of most milks is
extremely sensitive to small changes in pH in the neighbourhood of 6.7. In
effect, the influence of all other factors on the HCT of milk must be
considered against the background of the effect of pH.
For the majority of individual-cow and all bulk milks, the HCT increases
with increasing pH from 6.4 to about 6.7, then decreases abruptly to a
minimum at around pH 6.9 but increases continuously with further in-
creases in pH (Figure 9.19). The HCT decreases sharply below pH 6.4. Milks
which show a strong dependence of heat stability on pH are referred to as
type A milks. Occasionally, the HCT of individual-cow milks increases
continuously with increasing pH, which is as would be expected due to
increasing protein charge with increasing pH; these are referred to as type
B milks.
The maximum HCT and the shape of the HCT-pH profile are influenced
by several compositional factors, of which the following are the most
significant:
1. Ca2+ reduces HCT throughout the pH range 6.4-7.4.
2. Ca-chelators, e.g. citrate, polyphosphate, increase stability.
3. p-Lg, and probably a-la, increase the stability of casein micelles at
pH6.4-6.7 but reduce it at pH6.7-7.0; in fact, the occurrence of a
maximum-minimum in the HCT-pH profile depends on the presence of
4. Addition of K-casein to milk increases stability in the pH range of the
B-k.
HCT minimum.
3 70 DAIRY CHEMISTRY AND BIOCHEMISTRY
401 P
o!. 8. I. I. I - 1
6.2 6.4 6.6 6.8 7.0 7.2
PH
Figure 9.19 Effect of pH on the heat stability of type A milk (A), type B milk (0) and whey
protein-free casein micelle dispersions (0) (from Fox, 1982).
5. Reducing the level of colloidal calcium phosphate increases stability in
the region of the HCT maximum.
6. Natural variations in HCT are due mainly to variations in the concen-
tration of indigenous urea due to changes in the animals’ feed.
The current explanation for the maximum-minimum in the HCT-pH
profile is that on heating, x-casein dissociates from the micelles; at pH values
below about 6.7, 8-1s reduces the dissociation of Ic-casein, but at pH values
above 6.7, it accentuates dissociation. In effect, coagulation in the pH range
of minimum stability involves aggregation of Ic-casein-depleted micelles, in
a manner somewhat analogous to rennet coagulation, although the mech-
anism by which the altered micelles are produced is very different.
As would be expected, heating milk at 140°C for an extended period
causes very significant chemical and physical changes in milk, of which the
following are probably the most significant:
1. Decrease in pH. After heating at 140°C for 20 min, the pH of milk has
decreased to about 5.8 due to acid production from pyrolysis of lactose,
precipitation of soluble calcium phosphate as Ca,(PO,),, with the
release of H +, and dephosphorylation of casein with subsequent precipi-
tation of the liberated phosphate as Ca,(PO,), with the release of H+.
The heat-induced precipitation of Ca,(PO,), is partially reversible on
cooling so that the actual pH of milk at 140°C at the point of
coagulation is much lower than the measured value and is probably
below 5.0.
HEAT-INDUCED CHANGES IN MILK 371
2. Precipitation of soluble calcium phosphate as Ca,(PO,), with the
release of H'. After heating at 140°C for 5-10min, most (>90%) of the
soluble phosphate has been precipitated.
3. Dephosphorylation of casein, which follows first-order kinetics. After
heating at 140°C for 60min, >90% of the casein phosphate groups
have been hydrolysed.
4. Maillard browning, which occurs rapidly at 140°C. Since Maillard
browning involves blocking of the &-amino group of proteins with a
concomitant reduction in protein charge, it would be expected that
Maillard browning would reduce HCT, but in fact the Maillard reaction
appears to increase heat stability, possibly owing to the formation of
low molecular weight carbonyls.
5. Hydrolysis of caseins. During heating at 140°C there is a considerable
increase in non-protein N (12% TCA-soluble), apparently following
zero-order kinetics. K-Casein appears to be particularly sensitive to
heating and about 25% of the N-acetylneuraminic acid (a constituent
of K-casein) is soluble in 12% TCA at the point of coagulation.
6. Cross-linking of proteins. Covalent cross-linking of caseins is evident
(by gel electrophoresis) after even 2 min at 140°C and it is not possible
to resolve the heat-coagulated caseins by urea- or SDS-PAGE.
7. Denaturation of whey proteins. Whey proteins are denatured very
rapidly at 140°C; as discussed in section 9.6.3, the denatured proteins
associate with the casein micelles, via sulphydryl-disulphide interac-
tions with K-casein, and probably with a,,-casein, at pH values below
6.7. The whey proteins can be seen in electron photomicrographs as
appendages on the casein micelles.
8. Association and shattering of micelles. Electron microscopy shows that
the casein micelles aggregate initially, then disintegrate and finally
aggregate into a three-dimensional network.
9. Changes in hydration. As would be expected from many of the changes
discussed above, the hydration of the casein micelles decreases with the
duration of heating at 140°C. The decrease appears to be due mainly to
the fall in pH - if samples are adjusted to pH 6.7 after heating, there is
an apparent increase in hydration on heating.
10. Surface (zeta) potential. It is not possible to measure the zeta potential
of casein micelles at the assay temperature but measurements on heated
micelles after cooling suggest no change in zeta potential, which is
rather surprising since many of the changes discussed above would be
expected to reduce surface charge.
All the heat-induced changes discussed would be expected to cause major
alterations in the casein micelles, but the most significant change with
respect to heat coagulation appears to be the decrease in pH - if the pH is
readjusted occasionally to pH 6.7, milk can be heated for several hours at
140°C without coagulation. The stabilizing effect of urea is, at least partially,
3 72 DAIRY CHEMISTRY AND BIOCHEMISTRY
due to the heat-induced formation of NH, which reduces or delays the fall
in pH; however, other mechanisms for the stabilizing effect of urea have been
proposed.
9.7.1 Efect of processing operations on heat stability
Concentration. Concentration by thermal evaporation markedly reduces
the heat stability of milk, e.g. concentrated skim milk containing about 18%
total solids coagulates in roughly 10min at 130°C. The stability of the
concentrate is strongly affected by pH, with a maximum at around pH 6.6,
but stability remains low at all pH values above about 6.8 (Figure 9.20).
Concentration by ultrafiltration has a much smaller effect on HCT than
thermal evaporation, due to a lower concentration of soluble salts in the
retentate.
Homogenization. Homogenization of skim milk has no effect on HCT but
it destabilizes whole milk, the extent of destabilization increasing with fat
content and the severity of homogenization (Figure 9.21). Destabilization
probably occurs because the fat globules formed on homogenization are
stabilized by casein and consequently they behave as ‘casein micelles’, in
effect increasing the concentration of coagulable material.
Forewarming (preheating). Heating an unconcentrated milk, especially at
90°C x 10 min, before a heat stability assay, reduces its heat stability,
T’
50
6.4 6.6 6.8 7.0 7.2
PH
(2)
70 1
60 -
50 -
6.4 6.6 6.8 7.0 7.2
(h) PH
Figure 9.20 Effect of total solids (TS) content on the heat stability at 130°C of skim milk 0,
9.3% TS; 0, 12.0% TS; 0, 15.0% TS; ., 18.4% TS. (a) Concentrated by ultrafiltration, (b)
concentrated by evaporation (from Sweetsur and Muir, 1980).
HEAT-INDUCED CHANGES IN MILK 373
40 1
6.6 6.8 7.0 7.2 7.4
PH
Figure 9.21 Effect of pressure (Rannie homogenizer) on the heat coagulation time (at 140°C)
of milk, unhomogenized (0) or homogenized at 3.5 MPa: (A); 10.4 MPa (W) or 20.7/3.5 MPa
(+) (from Sweetsur and Muir, 1983).
mainly by shifting its natural pH; maximum heat stability is affected only
slightly or not at all. However, if milk is preheated before concentration, the
heat stability of the concentrate is increased. Various preheating conditions
are used, e.g. 90°C x lOmin, 120°C x 2min or 140°C x 5s; the last is
particularly effective but is not widely used commercially. The stabilizing
effect is probably due to the fact that the heat-induced changes discussed
previously are less detrimental if they occur prior to concentration rather
than in concentrated milk which is inherently less stable.
Additives. Orthophosphates, and less frequently citrates, have long been
used commercially to increase the stability of concentrated milk. The
mechanism was believed to involve Ca-chelation but pH adjustments may
be the principal mechanism.
Numerous compounds increase heat stability (e.g. various carbonyls,
including diacetyl, and ionic detergents) but few are permitted additives.
Although added urea has a major effect on the stability of unconcentrated
milk, it does not stabilize concentrated milks, although it does increase the
effectiveness of carbonyls.
9.8 Effect of heat treatment on rennet coagulation of milk and
related properties
The primary step in the manufacture of most cheese varieties and rennet
casein involves coagulation of the casein micelles to form a gel. Coagulation
374 DAIRY CHEMISTRY AND BIOCHEMISTRY
involves two steps (phases), the first of which involves enzymatically
hydrolysing the micelle-stabilizing protein, k--casein, by selected proteinases,
referred to as rennets. The second step of coagulation involves coagulation
of rennet-altered micelles by Ca2+ above 20°C (Chapter lo).
The rate of rennet coagulation is affected by many compositional factors,
including the concentrations of Ca2 +, casein and colloidal calcium phos-
phate and pH. Coagulation is adversely affected by heat treatment of the
milk at temperatures above about 70°C due to interaction of denatured p-lg
(and a-la) with K-casein. The primary and, especially, the secondary phases
of rennet coagulation are adversely affected by the interaction and, if the
heat treatment is sufficiently severe (e.g. 80°C x 5-10min), the milk does
not coagulate on renneting. The effect on the primary phase is presumably
due to blockage of the rennet-susceptible bond of k--casein following
interaction with ,/?-lg. The adverse effect of heating on the second phase
arises because the whey protein-coated micelles are unable to interact
properly because the aggregation sites, which are unknown, are blocked.
The adverse effects of heat treatment on the rennetability of milk can be
offset by acidifying or acidifying-reneutralizing the heated milk or supple-
menting it with Ca2+. The mechanism by which acidification offsets the
adverse effects of heating is not known but may involve changes in Ca2+
concentration.
The strength of the rennet-induced gel is also adversely affected by heat
treatment of the milk, again presumably because the whey protein-coated
micelles are unable to participate properly in the gel network. Gels from
severely heat-treated milk have poor syneresis properties, resulting in
high-moisture cheese which does not ripen properly. Syneresis is undesirable
in fermented milks, e.g. yoghurt, the milk for which is severely heat-treated
(e.g. 90°C x 10 min) to reduce the risk of syneresis.
9.9 Age gelation of sterilized milk
Two main problems limit the shelf-life of UHT sterilized milks: off-flavour
development and gelation. Age gelation, which also occurs occasionally with
in-container sterilized concentrated milks, is not related to the heat stability
of the milk (provided that the product withstands the sterilization process)
but the heat treatment does have a significant influence on gelation, e.g.
indirectly heated UHT milk is more stable to age gelation than the directly
heated product (the former is the more severe heat treatment). Plasmin may
be responsible for the gelation of unconcentrated UHT milk produced from
good-quality milk, while proteinases from psychrotrophs are probably
responsible if the raw milk was of poor quality. It is possible that physico-
chemical phenomena are also involved, e.g. interaction between whey
proteins and casein micelles.
HEAT-INDUCED CHANGES IN MILK 375
Table 9.3 Substances making a strong contribution to the flavour of indirectly heated UHT
milk, those contributing to differences in flavour of milk heat-treated in different ways, and
those used in a synthetic UHT flavour preparation (from Manning and Nursten, 1987)
Synthetic
UHT flavourd
UHT-ia UHT-i-LPb UHT-i-UHT-d' (mg per kg LP)
Dimethyl sulphide
3-Methylbutanal
2-Methylbutanal
2-Methyl-1-propanethiol
Pen tanal
3-Hexanone
Hexanal
2-Heptanone
Styrene
2-4-Heptenale
Heptanal
2-Acet ylfuran
Dimethyl trisulphide
Cyanobenzene
I-Heptanol
I-Octen-3-one'
Octanal
p-Cymene
Phenol
Indene
2-Ethyl-1-hexanol
Benzyl alcohol
Unknown
Acetophenone
I-Octanol
2-Nonanone
Nonanal
p-Cresol
rn-Cresol
E-Z,Z-&Nonadienal
E-2-Nonenal
3-Methylindene
Methylindene
E thyldimethyl benzene
Decanal
Tetraethylthiourea
Benzothiazole
y-Octalactone
2,3,S-Trimethylanisole
6-Octalactone
1-Decanol
2-Undecanone
2-Methylnaphthalene
Indole
&Decalactone
Hydrogen sulphide
Diacetyl
Dimethyl disulphide
2-Hexanone
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
0 1
1 1
0 1
1 1 0.008
1 1
1 1
4 2 0.40
1 0
2 0
1 1
1
4
0
2 0.2 1
0.005
0.025
0.18
0.650
0.03
0.005
0.002
376 DAIRY CHEMISTRY AND BIOCHEMISTRY
Table 9.3 (Continued)
Synthetic
UHT flavourd
UHT-i" UHT-i-LPb UHT-i-UHT-d' (mg per kg LP)
;$-Dodecalactone
6-Dodecalactone
Methanethiol
2-Pentanone
Methyl isothiocyanate
Ethyl isothiocyanate
Furfural
Benzaldehyde
2-Octanone
Naphthalene
y-decalact one
2-Tridecanone
Acetaldehyde
I-Cyano-4-pentene
2-Methyl- 1-butanol
Ethyl butyrate
3-Buten-1-yl isothiocyanate
E-Z,E-4-nonadienal
2,CDithiapentane
Maltol
2
2
1
1
1
1
1
1
1
1
1
1
-1
-1
-1
-1
-1
-1
1 0.025
1 0.1
1 0.002
1 0.29
1 0.01
1 0.01
1
0
0
0
0
0
0
0
1
0
0
0
1
10.m
'Indirectly heated UHT milk; + indicates a component that makes a strong contribution to
the flavour. In addition to the components listed, a further 12 unknowns made strong
contributions.
bComponents contributing to a difference in flavour between indirectly heated UHT milk and
low temperature pasteurized (LP) milk. Scale for difference: 1, slight; 2, moderate; 3, strong; 4,
very strong.
'Components contributing to a difference in flavour between indirectly and directly heated
UHT milks. Scale for difference as in '.
dComposition of synthetic UHT flavour.
'Tentative identification.
In the case of concentrated UHT milks, physicochemical effects appear
to predominate, although proteolysis also occurs, e.g. the propensity of
UHT concentrated milk reconstituted from high-heat milk powder to age
gelation is less than those from medium- or low-heat powders, although the
formation of sediment is greatest in the concentrate prepared from the
high-heat powder (see Harwalkar, 1992).
9.10 Heat-induced changes in flavour of milk
Flavour is a very important attribute of all foods; heating/cooking makes a
major contribution to flavour, both positively and negatively. Good-quality
fresh liquid milk products are expected to have a clean, sweetish taste and
essentially no aroma; any departure therefrom can be considered as an
HEAT-INDUCED CHANGES IN MILK 377
off-flavour. Heat treatments have a major impact on the flavour/aroma of
dairy foods, either positively or negatively.
On the positive side, thermization and minimum pasteurization should
not cause the formation of undesirable flavours and aromas and should, in
fact, result in improved flavour by reducing bacterial growth and enzymatic
activity, e.g. lipolysis. If accompanied by vacuum treatment (vacreation),
pasteurization removes indigenous off-flavours, i.e. those arising from the
cow’s metabolism or from feed, thereby improving the organoleptic qualities
of milk.
Also on the positive side, severe heat treatment of cream improves the
oxidative stability of butter produced therefrom due to the exposure of
antioxidant sulphydryl groups. As discussed in section 9.2.2, lactones formed
from hydroxyacids are major contributors to the desirable cooking quality
of milk fats but contribute to off-flavours in other heated products, e.g. milk
powders.
UHT processing causes substantial deterioration in the organoleptic
quality of milk. Freshly processed UHT milk is described as ‘cooked and
‘cabbagy’, but the intensity of these flavours decreases during storage, giving
maximum flavour acceptability after a few days. These off-flavours are due
to the formation of sulphur compounds from the denatured whey proteins,
as discussed in section 9.6.3. After this period of maximum acceptability,
quality deteriorates, the milk being described as stale. At least 400 volatiles
have been detected in UHT milk, about 50 of which (Table 9.3) are
considered to make a significant contribution to flavour (Manning and
Nursten, 1987). The shelf-life of UHT milk is usually limited by gelation
and/or bitterness, both of which are due to proteolysis, as discussed in
section 9.6.1.
Since sulphur compounds are important in the off-flavour of UHT milk,
attempts to improve its flavour have focused on reducing the concentration
of these, e.g. by adding thiosulphonates, thiosulphates or cystine (which
react with mercaptans) or sulphydryl oxidase, an indigenous milk enzyme
(which oxidizes sulphydryls to disulphides; Chapter 8).
.The products of Maillard browning have a significant negative impact on
the flavour of heated milk products, especially in-container sterilized milks
and milk powders.
References
Driessen, F.M. (1989) Inactivation of lipases and proteinases (indigenous and bacterial), in
Heat-induced Changes in Milk (ed. P.F. Fox), Bulletin 238, International Dairy Federation,
Brussels, pp. 71-93.
Erbersdobler, H.F. and Dehn-Miiller, B. (1989) Formation of early Maillard products during
UHT treatment of milk, in Heat-induced Changes in Milk (ed. P.F. Fox), Bulletin 238,
International Dairy Federation, Brussels, pp. 62-7.
378 DAIRY CHEMISTRY AND BIOCHEMISTRY
Fox, P.F. (1982) Heat-induced coagulation of milk, in Deuelopments in Dairy Chemistry, Vol. 1:
Proteins (ed. P.F. Fox), Applied Science Publishers, London, pp. 189-228.
Gould, LA. (1945) Lactic acid in dairy products. 111. The effect of heat on total acid and lactic
acid production and on lactose destruction. J. Dairy Sci., 28, 367-77.
Ha, Y.L., Grimm, N.K. and Pariza, M.W. (1989) Newly recognized anticarcinogenic fatty acids.
Identification and quantification in natural and processed cheeses. J. Agric. Food Chem., 37,
Harwalkar, V.R. (1992) Age gelation of sterilized milks, in Advanced Dairy Chemistry, Vol. 1:
Jenness. R. and Patton, S. (1959) Principles ofDairy Chemistry, John Wiley & Sons, New York.
Lyster, R.L.J. (1970) The denaturation of a-lactalbumin and fi-lactoglobulin in heated milk.
J. Dairy Res., 37, 233-43.
McKellar, R.C. (ed.) (1989) Enzymes of Psychrotrophs in Raw Food, CRC Press, Boca Raton,
FL, USA.
Manning, D.J. and Nursten, H.E. (1987) Flavour of milk and milk products, in Developments
in Dairy Chemistry, Vol. 3: Lactose and Minor Constituents (ed. P.F. Fox), Elsevier Applied
Science, London, pp. 217-38.
Mulvihill, D.M. and Donovan, M. (1987) Whey proteins and their thermal denaturation: A
review. Irish J. Food Sci. Techno/., 11, 43-75.
Singh, H. and Creamer, L.K. (1992) Heat stability of milk, in Advanced Dairy Chemistry, Vol.
1: Proteins, 2nd edn (ed. P.F. Fox), Elsevier Applied Science, London, pp. 621-56.
Stepaniak, L., Fox, P.F. and Daly, C. (1982) Isolation and general characterization of a
heat-stable proteinase from Pseudomonas Juorescens AFT 36. Biochim. Biophys. Acta, 717,
Sweetsur, A.W.M. and Muir, D.D. (1980) Effect of concentration by ultrafiltration on the heat
Sweetsur, A.W.M. and Muir, D.D. (1983) Effect of homogenization on the heat stability ofmilk.
Sweetsur, A.W.M. and White, J.C.D. (1975) Studies on the heat stability of milk proteins. 111.
Walstra, P. and Jenness, R. (1984) Dairy Chemistry and Physics, John Wiley & Sons, New York.
Webb, B.H. and Johnson, A.H. (1965) Fundamentals of Dairy Chemistry, AVI Publishing,
75-81.
Proteins, 2nd edn (ed. P.F. Fox), Elsevier Applied Science, London, pp. 691-734.
376-83.
stability of skim milk. J. Dairy Res. 47, 327-35.
J. Dairy Res., 50, 291-300.
Effect of heat-induced acidity in milk. J. Dairy Res., 42, 73-88.
Westport, CT.
Suggested reading
Fox, P.F. (ed.) (1982) Developments in Dairy Chemistry, Vol. 1: Proteins (ed. P.F. Fox), Applied
Fox, P.F. (ed.) (1989) Heat-induced Changes in Milk, Bulletin 238, International Dairy
Fox, P.F. (ed.) (1995) Heat-induced Changes in Milk, 2nd edn, Special Issue 9501, International
Walstra, P. and Jenness, R. (1984) Dairy Chemistry and Physics, John Wiley & Sons, New York.
Wong, N.P. (ed.) (1980) Fundamentals of Dairy Chemistry, 3rd edn, AVI Publishing, Westport,
Science Publishers, London.
Federation, Brussels.
Dairy Federation, Brussels.
CT