442 Part IV: Milk
unconcentrated milk, they collide frequently due to
Brownian, thermal, and mechanical motion.
The micelles scatter light; the white appearance
of milk is due mainly to light scattering by the ca-
sein micelles, with a contribution from the fat glob-
ules. The micelles are generally stable to most
processes and conditions to which milk is normally
subjected. The micelles may be reconstituted from
spray-dried or freeze-dried milk without major
changes in their properties. Freezing milk destabi-
lizes the micelles, due to a decrease in pH and an
increase in [Ca^2 ]. On very severe heating, for ex-
ample, at 140°C, the micelles shatter initially, then
aggregate, and eventually, after about 20 minutes,
the system coagulates (see Chapter 20).
The micelles can be sedimented by centrifuga-
tion: approximately 50% are sedimented by cen-
trifugation at 20,000 g for 30 minutes and approx-
imately 95% by centrifugation at 100,000 g for 60
minutes. The pelleted micelles can be redispersed by
agitation, for example, by ultrasonication, in natural
or synthetic milk ultrafiltrate; the properties of the
redispersed micelles are not significantly different
from those of native micelles.
The casein micelles are not affected by regular (
20 MPa) or high-pressure (up to 250 MPa) homoge-
nization (Hayes and Kelly 2003). The average size is
increased by high-pressure treatment at 200 MPa,
but micelles are disrupted (i.e., average size reduced
by 50%) by treatment at a somewhat higher pres-
sure, that is, 400 MPa (Huppertz et al. 2002,
2004).
The microstructure of the casein micelle has been
the subject of considerable research during the past
40 years, that is, since the discovery and isolation of
the micelle-stabilizing protein, -casein; however,
there is still a lack of general consensus. Numerous
models have been proposed, the most widely sup-
ported being the submicelle model, first proposed by
Morr (1967) and refined several times since (Fox
and Kelly 2003). Essentially, this model proposes
that the micelle is built up from submicelles (molec-
ular weight 500 kDa) held together by CCP and
surrounded and stabilized by a surface layer, ap-
proximately 7 nm thick, rich in -casein but contain-
ing some of the other caseins also (Fig. 19.3a). It is
proposed that the hydrophilic C-terminal region of
-casein protrudes from the surface, creating a hairy
layer around the micelle and stabilizing it through a
zeta potential of about 20 mV and by steric stabi-
lization. The principal direct experimental evidence
for this model is provided by electron microscopy,
which indicates a nonuniform electron density; this
has been interpreted as indicating submicelles.
However, several authors have expressed reserva-
tions about the subunit model, and three alternative
models have been proposed recently. Visser (1992)
suggested that the micelles are spherical agglomer-
ates of casein molecules, randomly aggregated and
held together partly by salt bridges in the form of
amorphous calcium phosphate and partly by other
forces (e.g., hydrophobic interactions), with a sur-
face layer of -casein. Holt (1992) proposed that the
Ca-sensitive caseins are linked by microcrystals of
CCP and surrounded by a layer of -casein, with its
C-terminal region protruding from the surface (Fig.
19.3b). In the dual-binding model of Horne (2002),
it is proposed that individual casein molecules inter-
act via hydrophobic regions in their primary struc-
tures, leaving the hydrophilic regions free and with
the hydrophilic C-terminal region of-casein pro-
truding into the aqueous phase (Fig. 19.3c). Thus,
the key structural features of the submicelle model
are retained in the three alternatives, that is, the inte-
grating role of CCP and a-casein-rich surface layer.
The micelles disintegrate- When the CCP is removed, for example, by
acidification to pH 4.6 and dialysis in the cold
against bulk milk or by addition of trisodium
citrate, also followed by dialysis; about 60% of
the CCP can be removed without disintegration
of the micelles. - By raising the pH to about 9.0, which does not
solubilize the CCP and presumably causes dis-
integration by increasing the net negative charge. - By urea at 5 M, which suggests that hydrogen
and/or hydrophobic bonds are important for
micelle integrity. - When the micelles are precipitated by ethanol or
other low molecular weight alcohols at about
35% at 20°C, but if the temperature is increased
to 70°C, surprisingly, the precipitated casein
dissolves and the solution becomes quite clear,
indicating dissociation of the micelles
(O’Connell et al. 2001a,b). Micelle-like particles
reform on cooling and form a gel at about 4°C. It
is not known if the subparticles formed by any of
these treatments correspond to casein
submicelles.