Báo cáo khoa học: Casein phosphopeptide promotion of calcium uptake in HT-29 cells ) relationship between biological activity and supramolecular structure ppt

Casein phosphopeptide promotion of calcium uptake
in HT-29 cells
)
relationship between biological activity
and supramolecular structure
Claudia Gravaghi
1
, Elena Del Favero
1
, Laura Cantu’
1
, Elena Donetti
2
, Marzia Bedoni
2
,
Amelia Fiorilli
1
, Guido Tettamanti
1
and Anita Ferraretto
1
1 Department of Medical Chemistry, Biochemistry and Biotechnology, University of Milan, Italy
2 DMU, Department of Human Morphology, University of Milan, Italy
It is known that milk is an excellent source of bioavail-
able calcium, due to the presence of caseins, which
bind calcium, keeping it in a soluble and absorbable
state [1–5]. In bovine milk, about two-thirds of the
calcium and one-half of the inorganic phosphate
are bound to various species of caseins, a
S1

-casein,
a
S2
-casein, b-casein, and k-casein, forming colloidal
micelles with a calcium ⁄ phosphate ⁄ casein molar ratio
of 30 : 21 : 1 [6]. The casein micelles, of about 100 nm
radius, are stable structures composed of hundreds of
smaller aggregates, named calcium phosphate nanocl-
usters, or nanocomplexes, having a core of calcium
phosphate surrounded by a shell of casein molecules
[7–10]. The portion of the casein molecule responsible
for the ability to maintain calcium and phosphate ions
in a soluble form are amino acid sequences containing
the common motif Ser(P)-Ser(P)-Ser(P)-Glu-Glu (the
‘cluster sequence’ or ‘acidic motif’). Peptides contain-
ing this sequence (casein phosphopeptides, CPPs) are
produced in vivo from the digestion of a
S1
-casein,
a
S2
-casein and b-casein by gastrointestinal proteases
[11–13], and in vitro by tryptic and chimotryptic
fragmentation of casein followed by precipitation
[14]. Calcium phosphate nanoclusters (or complexes)
were also prepared and physicochemically character-
ized using CPPs, namely b-CN(1–25)4P and b-CN(1–
42)5P, corresponding to the first 25 or 42 amino
acids of b-casein, respectively, and a
S1

-CN(59–79)5P,
Keywords
Ca
2+
uptake; casein phosphopeptides;
casein phosphopeptide–Ca
2+
aggregates;
HT-29 cells; laser light scattering
Correspondence
A. Ferraretto, Department of Medical
Chemistry, Biochemistry and Biotechnology,
University of Milan, L.I.T.A. via F. Cervi 93,
20090 Segrate, Italy
Fax: +39 02 50330365
Tel: +39 02 50330374
E-mail:
(Received 22 May 2007, revised 6 July
2007, accepted 27 July 2007)
doi:10.1111/j.1742-4658.2007.06015.x
Casein phosphopeptides (CPPs) form aggregated complexes with calcium
phosphate and induce Ca
2+
influx into HT-29 cells that have been shown
to be differentiated in culture. The relationship between the aggregation of
CPPs assessed by laser light scattering and their biological effect was stud-
ied using the CPPs b-CN(1–25)4P and a
s1
-CN(59–79)5P, the commercial
mixture CPP DMV, the ‘cluster sequence’ pentapeptide, typical of CPPs,

and dephosphorylated b-CN(1–25)4P, [b-CN(1–25)0P]. The biological effect
was found to be: (a) maximal with b-CN(1–25)4P and null with the ‘cluster
sequence’; (b) independent of the presence of inorganic phosphate; and (c)
maximal at 4 mmolÆL
)1
Ca
2+
. The aggregation of CPP had the following
features: (a) rapid occurrence; (b) maximal aggregation by b-CN(1–25)4P
with aggregates of 60 nm hydrodynamic radius; (c) need for the concomi-
tant presence of Ca
2+
and CPP for optimal aggregation; (d) lower aggrega-
tion in Ca
2+
-free Krebs ⁄ Ringer ⁄ Hepes; (e) formation of bigger aggregates
(150 nm radius) with b -CN(1–25)0P. With both b-CN(1–25)4P and
CPP DMV, the maximum biological activity and degree of aggregation
were reached at 4 mmolÆL
)1
Ca
2+
.
Abbreviations
ALP, alkaline phosphatase; BrdU, bromodeoxyuridine; [Ca
2+
]
i
, intracellular free calcium concentration; [Ca
2+

]
o
, extracellular free calcium
concentration; CN, casein; CPP, casein phosphopeptide; CPP DMV, CPP of commercial origin; KRH, Krebs ⁄ Ringer ⁄ Hepes.
FEBS Journal 274 (2007) 4999–5011 ª 2007 The Authors Journal compilation ª 2007 FEBS 4999
corresponding to the sequence 59–79 of a
S1
-casein
[8,9,14–16].
A few years ago, we showed that a CPP mixture of
commercial origin with five main components, as well
as pure b-CN(1–25)4P and even, to a lesser extent,
a
S1
-CN(59–79)5P elicited a marked and transient rise
of intracellular free Ca
2+
concentration ([Ca
2+
]
i
)in
human intestinal tumor HT-29 cells differentiated in
culture [17]. The intracellular Ca
2+
rise caused by CPP
was due to uptake of extracellular calcium ions, with
no involvement of the intracellular calcium stores [17].
A subsequent study, performed with b-CN(1–25)4P
and some chemically synthesized peptides correspond-

ing to precise fragments of the b-CN(1–25)4P
sequence, clarified that a well-defined primary structure
is required for the bioactive response [18]. This struc-
ture includes the N-terminal portion characterized by
the presence of a loop and a b-turn, and the ‘cluster
sequence’. However, and notably, the ‘cluster sequence’
alone does not exhibit the Ca
2+
uptake effect, suggest-
ing that a particular supramolecular structure of
CPP–Ca
2+
complexes is required for the observed bio-
logical effect in vitro, by analogy with the relationship
between calcium phosphate–CPP aggregation as
nanoclusters and the capacity to bind and maintain
calcium in a bioavailable form.
The present investigation addressed the question
whether a supramolecular structure of CPP–Ca
2+
is
needed to stimulate Ca
2+
uptake by differentiated
HT-29 cells. To this end, we first tested whether, under
the conditions used to prepare calcium phosphate–CPP
nanoclusters [16], the [Ca
2+
]
i

-increasing effects of CPP
on HT-29 cells could be detected. Unfortunately, these
conditions were not suitable for the growth of HT-29
cells in culture. Therefore, we adopted the same
experimental conditions previously used to detect the
biological effects of CPPs, that is: (a) the individual
CPPs b-CN(1–25)4P and a
s1
-CN(59–79)5P, and the
commercial mixture CPP DMV; (b) HT-29 human
colon carcinoma cells, differentiated in culture; (c) a
Krebs ⁄ Ringer ⁄ Hepes (KRH) solution buffering the
cells at pH 7.4, containing given concentrations of
Ca
2+
(as CaCl
2
), with or without phosphate (as
KH
2
PO
4
), compatible with normal cell viability; and
(d) CPP concentrations that have been shown to affect
Ca
2+
uptake by the cells [17,18]. The possible occur-
rence under these conditions of a supramolecular
structural organization (aggregation) of CPP and
Ca

2+
was studied by a laser light scattering technique
capable of establishing the dimensions (hydrodynamic
radius) and the relative amounts of aggregates in
solution. Care in exactly matching the experimental
conditions for laser light scattering experiments with
those providing the mentioned biological effect of CPP
was of central importance.
Results
In our previous work [17], we demonstrated that CPPs
are able to promote Ca
2+
uptake by human intestinal
HT-29 tumor cells differentiated in culture (RPMI)
with a consequent transient rise of [Ca
2+
]
i
. In order to
address the question whether a supramolecular struc-
tural organization of CPP–Ca
2+
is needed to promote
this biological effect, we first verified the differentiation
state of HT-29 cells in culture. It is known that HT-29
cells cultured in DMEM with a high d-glucose content
(25 mmolÆL
)1
) do not present signs of spontaneous dif-
ferentiation towards intestinal-like cells [19]. Instead,

when the culture medium is switched to RPMI, with
low d-glucose concentration (13.9 mmolÆL
)1
), or to a
DMEM medium with galactose gradually substituting
for glucose, HT-29 cells undergo a process of intesti-
nal-like differentiation [20]. On this basis, HT-29 cells
were cultured in RPMI (low d-glucose) or galactose-
containing medium, and their differentiation was
assessed by determining specific biochemical markers
[alkaline phosphatase (ALP) and sucrase-isomaltase]
and the rate of proliferation, and by electron-micro-
scopic examination. As shown in Fig. 1A, the levels of
ALP and sucrase-isomaltase in RPMI cells were not
significantly different from those in DMEM cells
(631 ± 32 versus 623 ± 25 mUÆmg
)1
protein for
ALP, and 80.7 ± 9.1 versus 77.8 ± 8.3 mUÆmg
)1
pro-
tein for sucrase-isomaltase, respectively), whereas
galactose-adapted cells showed a marked increase of
both ALP (830 ± 12 mUÆmg
)1
protein) and sucrase-
isomaltase (270 ± 20 mUÆmg
)1
protein). The prolifera-
tion rate (Fig. 1B) of cells cultured in RPMI and

galactose-adapted medium markedly decreased as com-
pared to DMEM cells, indicating a repression of their
tumoral condition. The cell morphology is shown in
Fig. 1C–E. DMEM cells appear to be completely
devoid of apical microvilli and junctional apparatus,
whereas RPMI cells present a well-developed brush
border, with microvilli on their apical side, together
with the presence of adherent junctions and desmo-
somes, and galactose-adapted cells display all the
features observed in RPMI cells, with, in addition,
characteristic intracellular laminae surrounded by
numerous and well-developed small microvilli. All of
these findings indicate that HT-29 cells grown in
RPMI or galactose-containing medium undergo a
remarkable process of intestinal-like differentiation,
confirming previous data [19,20]. From both the
quantitative and qualitative points of view, both
CPP activity and supramolecular structure C. Gravaghi et al.
5000 FEBS Journal 274 (2007) 4999–5011 ª 2007 The Authors Journal compilation ª 2007 FEBS
RPMI and galactose-adapted cells responded equally
to CPP administration, with an increase of [Ca
2+
]
i
.
Notably, undifferentiated HT-29 cells did not exhibit
the CPP effect (unpublished results). On this basis
and for purposes of simplicity, all further experi-
ments were performed by culturing cells in RPMI
medium.

To investigate the effect of CPP in increasing the
extracellular free Ca
2+
concentration ([Ca
2+
]
o
) in the
buffer solution, while avoiding the possible precipita-
tion of insoluble calcium phosphate salts, which would
affect biological and laser light scattering measure-
ments, we first explored whether the presence of phos-
phate was necessary for the biological effect of CPP.
To this end, a first dose–response set of experiments at
[Ca
2+
]
o
higher than 2 mmolÆL
)1
was performed using
cells grown in RPMI. As shown in Fig. 2A,B the
[Ca
2+
]
i
peaks of increase in HT-29 cells elicited by
b-CN(1–25)4P CPP at two different concentrations (50
and 100 lmolÆL
)1

) and in the presence of 2 or 4
mmolÆL
)1
[Ca
2+
]
o
, expressed as percentage of the basal
values, were the same regardless of the presence or
absence of phosphate. In more detail (Fig. 2C,D), for
2 mmolÆL
)1
[Ca
2+
]
o
and 50 lmolÆL
)1
b-CN(1–25)4P,
the basal Ca
2+
concentration was 100 nmolÆL
)1
in the
presence of phosphate (trace a) and 70 nmolÆL
)1
in the
absence of phosphate (trace b), whereas the increments
due to CPP were 25 nmolÆL
)1

and 22 nmolÆL
)1
, respec-
tively; for 2 mmolÆL
)1
[Ca
2+
]
o
and 100 lmolÆL
)1
b-CN(1–25)4P, the basal Ca
2+
concentration was
72 nmolÆL
)1
(trace c) and 84 nmolÆ L
)1
(trace d),
whereas the increments due to CPP were 48 nmolÆL
)1
and 47 nmolÆL
)1
, respectively, i.e. the same regardless
of the presence or absence of phosphate in the buffer.
Similar results were obtained with the CPP DMV
mixture and a
s1
-CN(59–79)5P, indicating that free
phosphate is not involved in the biological effect of

CPP. More details on the dose–response relationship
(in the absence of phosphate) are presented in Fig. 3,
where [Ca
2+
]
o
was raised to 6 mmolÆL
)1
and the three
different preparations of CPP, each at different con-
centrations, were employed. With b-CN(1–25)4P and
0
250
500
750
1000
A
B
GALACTOSERPMIDMEM
GALACTOSERPMIDMEM
*
*
ALP
sucrase
0
50
100
BrdU incorporation (%)
mU/mg protein
*

*
C
D
E
Fig. 1. HT-29 cell differentiation. (A) ALP
(white bars) and sucrase (black bars)
enzyme activities of DMEM (undifferenti-
ated), RPMI and galactose-adapted (differen-
tiated) cells. (B) Proliferation rate as
determined by BrdU incorporation in the
three cell populations expressed as percent-
age with respect to DMEM cells. (C,D,E)
Transmission electronmicrographs of araldite
ultrathin sections of DMEM cells (C), RPMI
cells (D) and galactose-adapted cells (E),
respectively. Starting from the apical side,
arrows in (D) indicate adherent junctions
and desmosomes. Original magnification:
(C,D) ·10 000; (E) ·14 000. Data reported in
(A) and (B) represent mean value ± SD
(n ¼ 5–6 experiments for each bar). Aster-
isks indicate significantly different values
(P<0.05) from DMEM.
C. Gravaghi et al. CPP activity and supramolecular structure
FEBS Journal 274 (2007) 4999–5011 ª 2007 The Authors Journal compilation ª 2007 FEBS 5001
CPP DMV mixture, the highest biological effects were
observed at 4 mmolÆL
)1
[Ca
2+

]
o
(Fig. 3A,B), the opti-
mal effect being obtained at 200 lmolÆL
)1
b-CN(1–
25)4P and 1280 lmolÆL
)1
CPP DMV, respectively. The
differences between CPP DMV and b-CN(1–25)4P
doses may be explained by considering that, whereas
b-CN(1–25)4P is a synthetic, pure peptide, CPP DMV
is a mixture of peptides with different primary
sequences, and possibly different biological efficacies.
The behavior of a
s1
-CN(59–79)5P, reported in Fig. 3C,
appears to be completely different. First of all, the
extent of the measured effect is much more limited,
over the whole CPP and [Ca
2+
]
o
concentration range
explored. Second, the highest activity, within the inves-
tigated range of Ca
2+
concentration (2–6 mmolÆL
)1
),

was recorded at 6 mmolÆL
)1
[Ca
2+
]
o
. Third, no signifi-
cant change in [Ca
2+
]
i
was observed when the CPP
concentration was increased. Notably, the absence of
phosphate in the culture media did not modify cell
morphology and viability. Also surprising was the find-
ing that when CPP was added to the cell-containing
mixtures before the addition of Ca
2+
, no [Ca
2+
]
i
rise
was recorded in HT-29 cells due to the presence of
CPP. It should be remembered that the ‘cluster
sequence’ is completely unable to elicit the increase in
[Ca
2+
]
i

[18].
Preliminary laser light scattering experiments showed
that an aqueous solution of b-CN(1–25)4P, as well as
0
50
100
AB
CD
peak calcium increase
(% on basal value)
[
β
-CN(1-25)4P]
50
μ
mol/L
200 00
600
600
100
50
[
β
-CN(1-25)4P]
μ
mol/L
*
*
[Ca
2+

]
o
2mmol/L
[
β
-CN(1-25)4P]
μ
mol/L
50 100
0
50
100
Ionomycin
Ionomycin
100 100
a
b
c
d
[
β
-CN(1-25)4P]
100
μ
mol/L
Fluorescence arbitrary units
Time (s)
200
200
200

400
400
*
*
[Ca
2+
]
o
4mmol/L
KRH (containing phosphate)
phosphate-free KRH
Fig. 2. Intracellullar Ca
2+
increases in response to administration of b-CN(1–25)4P peptide in KRH or in phosphate-free KRH. The data were
collected on fura-2-loaded HT-29 cell populations grown in RPMI and treated with two CPP concentrations (50 and 100 lmolÆL
)1
) and at two
different extracellular Ca
2+
concentrations, 2 mmolÆL
)1
(A) and 4 mmolÆL
)1
(B). HT-29 cells were resuspended, just before the experiment, in
KRH (black bars) or phosphate-free KRH (white bars). The data collected were expressed as the mean value of [Ca
2+
]
i
peak rise (calculated
as percentage on basal value) ± SD (n ¼ 3–4 experiments for each bar). Asterisks indicate significantly different values (P<0.05) from the

minimal CPP dose. In (C) and (D), the representative traces relative to 50 lmolÆL
)1
b-CN(1–25)4P (arrow) in KRH (trace a), and in phosphate-
free KRH (trace b), and the representative traces relative to 100 lmolÆ L
)1
b-CN(1–25)4P (arrow) in KRH (trace c), in phosphate-free KRH
(trace d) at 2 mmolÆL
)1
extracellular Ca
2+
concentration, are shown. The vertical scale indicates fluorescent intensity at 485 nm emission
wavelength after excitation at 343 nm.
CPP activity and supramolecular structure C. Gravaghi et al.
5002 FEBS Journal 274 (2007) 4999–5011 ª 2007 The Authors Journal compilation ª 2007 FEBS
of CPP DMV, a
s1
-CN(59–79)5P, b-CN(1–25)0P and
the ‘cluster sequence’, at the used concentrations, gave
a very low scattered intensity, similar to that of pure
solvent, indicating a condition where aggregation is
absent. Therefore, the CPP solution in water can be
considered a full monomer solution of CPP. In con-
trast, the solution of the same CPP in phosphate-free
or phosphate-containing KRH with no Ca
2+
showed a
remarkable increase of scattered light, of the order
of 10 times that of the pure solvent, indicating the
occurrence of some aggregation. An additional four-
fold increase of the scattered light occurred when the

solvent contained 4 mmolÆL
)1
[Ca
2+
]
o
, whereas addi-
tion of Ca
2+
to a pre-existing Ca
2+
-free CPP solution
did not induce any increase of scattered light (data are
shown in Fig. 4). The time needed for the occurrence
of aggregation corresponded to the duration of the
experimental manipulations (pipetting, mixing, etc.),
i.e. a few seconds. This indicates that the process of
aggregation, when it occurs, is very rapid. b-CN(1–
25)0P, the dephosphorylated form of b-CN(1–25)4P,
dissolved in phosphate-free KRH gave rise to a higher
scattered intensity with respect to the corresponding
b-CN(1–25)4P solution, but no significant influence of
Ca
2+
was observed (Fig. 4), suggesting the occurrence
of an aggregation process different from that of
b-CN(1–25)4P. Finally, the ‘cluster sequence’ did not
exhibit any aggregation in solution, regardless of the
presence of Ca
2+

, as its scattered intensity was not dis-
similar to that of pure solvent.
Dynamic light scattering experiments showed
(Table 1) that the three CPPs, b-CN(1–25)4P, a
s1
-
CN(59–79)5P and CPP DMV, dissolved in 4 mmolÆL
)1
Ca
2+
phosphate-free KRH, formed aggregated struc-
tures with the same hydrodynamic radius (R
H
¼
60 ± 2 nm). An identical hydrodynamic radius was
detected for the aggregates of b-CN(1–25)4P dissolved
in phosphate-free KRH in the the absence of Ca
2+
.
Instead, b-CN(1–25)0P formed much bigger aggregates
(R
H
¼ 150 ± 4 nm), regardless of the presence or
absence of Ca
2+
.
Concerning the three CPPs with the same hydro-
dynamic radius in solution, the recorded differences in
the intensity of the scattered light do reflect differences
in the concentration of the aggregates in solution.

Assuming as 100% reference value the concentration
of the aggregates of b-CN(1–25)4P in the presence of
0
150
300
A
B
C
0
150
300
26
0
150
300
[β-CN(1-25)4P]
200 μmol/L
1280 μmol/L
960 μmol/L
640 μmol/L
320 μmol/L
200 μmol/L
150 μmol/L
100 μmol/L
50 μmol/L
150 μmol/L
100 μmol/L
50 μmol/L
peak [Ca
2+

]
i
increase
(% on basal value)
[CPP DMV]
[Ca
2+
]
o
mmol/L
[α
s1
-CN(59-79)5P]
4
Fig. 3. CPP bioactivity is related to extracellular Ca
2+
and peptide
concentration. The data were collected after administering to fura-
2-loaded HT-29 cell populations various amounts of individual CPPs,
b-CN(1–25)4P and a
s1
-CN(59–79)5P, (A) and (C), respectively, and
of a mixture of CPPs (CPP DMV) (B). Each point on the graphs cor-
responds to the mean value of the [Ca
2+
]
i
peak rise ± SD, obtained
from three or four experiments, and expressed as a percentage of
the basal value; in all cases, a CPP single dose was provided to the

cells at a fixed extracellular Ca
2+
concentration. All values are signi-
ficantly different from each other (P<0.05).
0
25
50
β-CN(1-25)0Pβ-CN(1-25)4Pβ
I-1
r
1
2
3
4
5
Fig. 4. Excess scattered intensity relative to the solvent, I
r
) 1, for:
1, b-CN(1–25)4P in phosphate-free KRH containing 4 mmolÆL
)1
Ca
2+
2, b-CN(1–25)4P in phosphate-free KRH without Ca
2+
;3,
b-CN(1–25)4P prepared in phosphate-free KRH without Ca
2+
fol-
lowed by addition of 4 mmolÆL
)1

Ca
2+
;4,b-CN(1–25)0P in phos-
phate-free KRH containing 4 mmolÆL
)1
Ca
2+
;5,b-CN(1–25)0P in
phosphate-free KRH without Ca
2+
(for all solvents, phosphate-free
KRH with or without 4 mmolÆL
)1
Ca
2+
, the same very small scat-
tered intensity was measured).
C. Gravaghi et al. CPP activity and supramolecular structure
FEBS Journal 274 (2007) 4999–5011 ª 2007 The Authors Journal compilation ª 2007 FEBS 5003
4 mmolÆL
)1
Ca
2+
, which provides the highest scattered
intensity (Table 1), the relative concentration of
CPP DMV aggregates in the same solvent was 35%,
although with a solute concentration six times higher
than that of b-CN(1–25)4P, and that of a
s1
-CN(59–

79)5P was only 4.5%, with the same total solute con-
centration. The absence of Ca
2+
caused a reduction in
aggregation of b-CN(1–25)4P to only 25%, whereas
no significant change in the relative percentage of
aggregation was induced in b-CN(1–25)0P by the pres-
ence of Ca
2+
(2.5% versus 2.4%). Of course, in each
sample, aggregates are expected to coexist with dis-
aggregated molecules, in a mole fraction depending on
the physicochemical characteristics of the peptide.
However, the disaggregated fraction was shown to
make a negligible contribution to the scattered inten-
sity, less than 0.1%.
Laser light scattering measurements were also
performed on CPP DMV (1280 lmolÆL
)1
), a
s1
-CN(59–
79)5P (200 lmolÆL
)1
) and b-CN(1–25)4P (200
lmolÆL
)1
) as a function of Ca
2+
concentration, in the

same range of the Ca
2+
uptake experiments reported
in Fig. 3, and the results are shown in Fig. 5. As the
three CPPs form aggregated particles with the same
hydrodynamic radius, as already reported, the differ-
ences in excess scattered intensity relative to the sol-
vent, I
r
) 1, reflect the differences in the number of
aggregates in solution. The scattering intensity curves
of b-CN(1–25)4P (Fig. 5A) and CPP DMV (Fig. 5B)
present the same convex behavior, with a maximum
at 4 mmolÆL
)1
Ca
2+
. It is surprising that the shapes
closely correspond to those of the dose–biological
response (Fig. 3), showing that at 4 mmolÆL
)1
Ca
2+
,
where the maximal biological activity is reached, there
is the highest concentration of CPP aggregates. In the
case of a
s1
-CN(59–79)5P (Fig. 5C), the scattered inten-
sity is always very low (as low as the biological effect)

and shows a smooth increase of the number of aggre-
gates with increasing Ca
2+
content, again paralleling
the similar small increase of the biological effect.
Discussion
This work provides novel information regarding the
ability of CPPs to enhance Ca
2+
uptake by HT-29
cells, which have been shown to undergo differentia-
tion in culture, and demonstrates that this biological
effect depends on a particular type of CPP aggrega-
tion and the concentration of aggregates in solution.
For the first time, the supramolecular structural archi-
tecture of CPPs has been studied under experimental
conditions that allow the viability in culture of cells
such as differentiated HT-29 cells, and permit the
expression by these cells of an enhanced uptake of
extracellular Ca
2+
. Remarkably, the absence of phos-
phate ions (as KH
2
PO
4
) in the cell culture medium
did not affect this biological effect, or cell viability,
enabling us to explore the process of CPP aggrega-
tion (in the absence of any possible precipitation of

calcium phosphate salts) by a laser light scattering
technique.
Regarding the CPP-mediated enhancement of Ca
2+
uptake, a relevant observation is the existence of
an optimal CPP ⁄ Ca
2+
ratio for the effect [4 mmolÆL
)1
Ca
2+
⁄ 200 lmolÆL
)1
b-CN(1–25)4P]. This result,
obtained in an experimental model consisting of
in vitro cells, is in agreement with results obtained
using animals or everted intestinal tissue [21–24]. It is
0
246
150
300
Scattered Intensity
(relative units)
[Ca
2+
] mmol/L
β
-CN(1-25)4P
CPP DMV
α

S1
-CN(59-79)5P
Fig. 5. Scattered intensity of CPPs as a function of Ca
2+
concentra-
tion. Scattered intensity curve for b-CN(1–25)4P (200 lmolÆL
)1
), for
CPP DMV (1280 lmolÆL
)1
) and for a
s1
-CN(59–79)5P (200 lmolÆL
)1
).
Each value of scattered intensity is calculated in relative units,
i.e. with respect to the intensity scattered by the same amount of
peptide as a full monomer solution.
Table 1. Aggregative properties of CPPs.The data reported refer to
experiments where the concentration of CPP was 1280 lmolÆL
)1
for CPP DMV and 200 lmolÆL
)1
for each other peptide in 4 and
0 mmolÆL
)1
Ca
2+
in phosphate-free KRH. All data are referred to
those for b-CN(1–25)4P, which provided the highest intensity of

light scattering, assumed as 100%.
Hydrodynamic radius
of aggregates (nm)
Relative concentration
of aggregates (%)
[Ca
2+
]
o
4 mmolÆL
)1
b-CN(1–25)4P 60 100
CPP DMV 60 35
a
s1
-CN(59–79)5P 60 4.5
b-CN(1–25)0P 150 2.5
Cluster sequence 0 0
[Ca
2+
]
o
0 mmolÆL
)1
b-CN(1–25)4P 60 25
b-CN(1–25)0P 150 2.4
CPP activity and supramolecular structure C. Gravaghi et al.
5004 FEBS Journal 274 (2007) 4999–5011 ª 2007 The Authors Journal compilation ª 2007 FEBS
noteworthy that the conditions we used, with Ca
2+

concentrations up to 6 mmolÆL
)1
, are close to those
occurring in the intestinal lumen after a proper meal,
where Ca
2+
may reach a concentration of 3–4
mmolÆL
)1
in rats and 7–8 mmolÆL
)1
in humans [25].
The modest Ca
2+
uptake effect exerted on HT-29 cells
by a
S1
-CN(59–79)5P as compared to the much more
pronounced effect exerted by b-CN(1–25)4P is in line
with the differences in the Fe
2+ ⁄ 3+
absorption
mediated by the two CPPs [26,27], possibly associated
with different structural changes induced in the two
CPPs by Fe
2+ ⁄ 3+
(as well as Ca
2+
) binding [16,28,29].
The set of laser light scattering experiments clearly

demonstrated the occurrence of CPP self-aggregation
in solution, with precise features (very rapid occur-
rence; 60 nm hydrodynamic radius; absolute need for
concomitant presence of Ca
2+
and CPP for optimal
aggregation). At the same time, they also demonstrated
that the ability to aggregate, in terms of dimension
and concentration of aggregates, relied on the chemical
structure of CPP, as the ‘cluster sequence’ pentapep-
tides do not aggregate at all. An explanation of these
features can be given following a model of self-aggre-
gation similar to that proposed by Horne for b-casein
micelles [30,31], where the single monomers possess
hydrophilic and hydrophobic regions, and hydrophobic
interactions between monomers are important for the
aggregation. As CPPs are negatively charged, due to
the presence of phosphorylated serine and glutamic
acid residues, the repulsive interactions between mono-
mers prevent their aggregation when they are dissolved
in pure water, as we observed. At higher ionic
strengths, as in phosphate-free KRH, the effect of elec-
trostatic repulsion is screened, and aggregation can
take place, as we also observed. In addition, calcium
divalent counterions may facilitate the organization of
peptides in the aggregates, as they can coordinate two
charges belonging to different molecules, explaining
the marked increase that we observed in the relative
number of aggregates of b-CN(1–25)4P due just to
the presence of Ca

2+
. The scarce propensity of
a
s1
-CN(59–79)5P to aggregate, in term of aggregate
concentration, is most probably due to the additional
phosphorylated serines present, providing more nega-
tive charges, and fewer hydrophobic residues [9]
(Table 2). A strong contribution of repulsive interac-
tions among monomers results in a higher proportion
of monomeric forms, as compared to b-CN(1–25)4P.
The differences in the aggregation features and ability
to elicit the [Ca
2+
]
i
rise effect of b-CN(1–25)4P and
a
s1
-CN(59–79)5P probably reflect the different and
already described conformations of these CPPs [32,33].
With regard to the different aggregation properties of
b-CN(1–25)4P and b-CN(1–25)0P, b-CN(1–25)0P has
a much lower number of negative charges than
b-CN(1–25)4P, and an almost null coordination role
due to Ca
2+
. Furthermore, b-CN(1–25)0P is known to
assume a much more flexible and dynamic conforma-
tion in solution than b-CN(1–25)4P [32], which proba-

bly facilitates aggregation into bigger complexes. In
fact, the hydrodynamic radius of b-CN(1–25)0P is
150 nm versus the 60 nm of b-CN(1–25)4P. However,
the relative concentration of aggregates is about 2.5%
that of b-CN(1–25)4P, regardless of the presence or
absence of Ca
2+
. The two molecules do aggregate but
in a completely different manner, in terms of both size
and concentration of aggregates, b-CN(1–25)4P aggre-
gates exhibiting the Ca
2+
uptake effect and b-CN(1–
25)0P not at all. The absence of aggregation by the
‘cluster sequence’ is not surprising, as the presence of
three phosphorylated serines and two glutamic acids
accounts for such a strong negative charge that repul-
sive interactions prevail and prevent aggregation.
The most intriguing evidence provided by this inves-
tigation is the relationship between CPP aggregation
and the biological effect on differentiated HT-29 cells.
As shown in Fig. 5, the scattered intensity curves of
b-CN(1–25)4P and CPP DMV at different Ca
2+
concen-
trations exhibit the same convex behavior, with a maxi-
mum at 4 mmolÆL
)1
Ca
2+

, mimicking the profiles of
the Ca
2+
uptake effect (Fig. 3). As the Ca
2+
concen-
tration increases from 2 to 4 mmolÆL
)1
, the concentra-
tion of aggregates increases, owing to the complexing
power of Ca
2+
, but at higher contents, 6 mmolÆL
)1
,
the abundance of counterions leads to a higher number
of phosphopeptide monomers undergoing direct inter-
actions, preventing them from being involved in exten-
sive aggregation. In parallel, the biological effect rises
from 2 to 4 mmolÆL
)1
Ca
2+
, but decreases from 4 to
6 mmolÆL
)1
Ca
2+
, indicating that it follows the concen-
tration of aggregates. This evidence suggests the notion

that the aggregated forms are the active forms of a bio-
active CPP such as b-CN(1–25)4P. Further support for
this notion comes from the finding that for formation
Table 2. Synthetic CPP primary structures. The ‘cluster sequence’
characteristic of all CPPs is underlined and indicated in bold charac-
ters. S corresponds to phosphorylated serine. (For additional
details, see Ferraretto et al. [18].)
CPP Primary structure
a
s1
-CN(59–79)5P QMEAESISSSEEIVPNSVEQK(59–79)
b-CN(1–25)4P RELEELNVPGEIVESL
SSSEESITR(1–25)
b-CN(1–25)0P RELEELNVPGEIVESLSSSEESITR(1–25)
‘Cluster sequence’
pentapeptide
SSSEE
C. Gravaghi et al. CPP activity and supramolecular structure
FEBS Journal 274 (2007) 4999–5011 ª 2007 The Authors Journal compilation ª 2007 FEBS 5005
of the biologically active aggregates, the simultaneous
presence of CPP and Ca
2+
is needed while complexes
are forming. Presumably, the CPP aggregates formed
in the absence of Ca
2+
, although exhibiting a hydrody-
namic radius equal or similar to that of the Ca
2+
-con-

taining aggregates (60 nm), are different from those
formed in the presence of Ca
2+
. An additional relevant
point concerns the role of phosphate in the CPP-medi-
ated [Ca
2+
]
i
rise effect. The removal of phosphate (as
KH
2
PO
4
) from the buffer does not affect the biological
effect, whereas the removal of phosphate from the
serines totally abrogates it, emphasizing the fact that
the role of serine-linked phosphate is essential to:
(a) bind Ca
2+
; (b) induce correct aggregation of CPP;
and (c) elicit the biological effect.
A final matter of discussion regards the possible rel-
evance of our findings to the controversial issue [34,35]
of whether CPPs enhance Ca
2+
absorption at the
intestinal level, thus improving Ca
2+
bioavailability.

Investigations of this, performed on animals (rats,
chicks, chickens) and humans, were based on the evi-
dence that, in models of absorption such as everted
sacs [21,36] and ligated segments of rat ileum
[24,37,38], CPPs favor Ca
2+
absorption, particularly in
the presence of substances such as phosphate [36] that
are capable of forming insoluble calcium salts. This
effect was attributed to the ability of CPPs to form
complexes carrying ‘soluble’ calcium. Our studies refer
to a cell model, HT-29 cells differentiated in vitro.
Therefore, any extension to physiological situations in
animals has to be done with extremely caution. If we
take this model as valid, the flux of Ca
2+
from the
extracellular milieu into the cytosol of HT-29 cells may
mimic the Ca
2+
flux from the intestinal lumen to the
interior of enterocytes, particularly at the ileum level
(passive absorption). The overall Ca
2+
flux during
intestinal absorption is in the mmolÆ L
)1
order of mag-
nitude, whereas the observed increment of [Ca
2+

]
i
in
HT-29 cells due to the CPP effect is in the range of
about 50 nmolÆL
)1
. Whether this relatively small,
although rapid, increase of [Ca
2+
]
i
is responsible for
and sufficient to enable the passage of Ca
2+
along the
intestinal absorption route under physiological condi-
tions is a difficult question that, at present, cannot be
answered. What can be said is that the [Ca
2+
]
i
rise
effect does match the CPP-mediated enhanced Ca
2+
absorption observed in the rat ileum sacs or ligated
segments [21,24,36–38], substantiates the reports show-
ing a positive role of CPP treatment on Ca
2+
bioavail-
ability in animals [22,24,39–44], and suggests the

notion that CPPs not only maintain Ca
2+
in an
absorbable form but also interact with the plasma
membranes of certain cells, facilitating Ca
2+
uptake.
Concerning the conflicting results of human studies,
some in favor of the efficacy of CPP treatment [45–48]
and some not [49–51], it should be remembered that,
according to our findings, the ability of CPPs to elicit
the optimal biological effect relies on two critical con-
ditions, the presence of Ca
2
–CPP aggregates in the cor-
rect conformation and concentration, and a suitable
ratio between Ca
2+
and CPP, this latter condition
being in agreement with data determined in intestinal
model studies [21,51]. Examining the experimental
protocols of the above-cited papers [45–51] it is hard
to evaluate when (or whether) these critical conditions
were fulfilled. It is worth mentioning that we had evi-
dence (unpublished results) that the ability of CPPs to
elicit a transient rise in [Ca
2+
]
i
is acquired by HT-29

cells, as well as Caco-2 cells, upon differentiation (in
other words, it is peculiar to the differentiated state of
these intestinal-related cells), and is also exhibited by
human osteoblasts in culture, suggesting that the CPP
effect may be of more general significance in the mod-
ulation of Ca
2+
uptake by cells.
It is the purpose of our current and future research
to explore the molecular mechanism by which CPPs
elicit a transient rise in [Ca
2+
]
i
in sensitive cells, as well
as to set up and apply proper conditions to evaluate
the use of CPPs as possible functional foods enhancing
Ca
2+
bioavailability.
Experimental procedures
Cell culture media and all other reagents were purchased
from Sigma (St Louis, MO, USA). Fetal bovine serum was
from EuroClone Ltd (Wetherby, UK). Fura-2 acetoxy-
methyl ester, fura-2 pentasodium salt, and ionomycin (the
last two compounds used only for calibration purposes)
were obtained from Calbiochem (La Jolla, CA, USA).
Casein phosphopeptides
The CPP DMV preparation employed is a casein-derived
hydrolysate (CE 90 CPP III; DMV International, Veghel,

the Netherlands), comprising several components, each con-
taining the characteristic CPP ‘cluster sequence’, with the
following composition: 93.8% as dry matter; 96% purity;
10.8% total nitrogen content; 3.7% phosphorus content;
nitrogen ⁄ phosphorus ratio 3.1; P ⁄ Ser ratio 0.85 mol ⁄ mol;
average relative molecular mass 2500. This CPP mixture was
determined to be Ca
2+
-free as already reported [17]. The
individual CPPs a
s1
-CN(59–79)5P and b-CN(1–25)4P, the
dephosphorylated form of b-CN(1–25)4P [b-CN(1–25)0P]
and the ‘cluster sequence’ were synthetically produced by
Primm (Milan, Italy), and characterized for purity as already
reported [18]. The primary structure of all the used synthetic
CPP activity and supramolecular structure C. Gravaghi et al.
5006 FEBS Journal 274 (2007) 4999–5011 ª 2007 The Authors Journal compilation ª 2007 FEBS
peptides is shown in Table 2. All CPPs and CPP derivatives
were stored at ) 20 °C until use, when they were dissolved in
double-distilled water in stock solutions (1000· concen-
trated, with respect to the final concentration) and eventu-
ally brought to neutrality with 1 mmolÆL
)1
NaOH.
Cell culture
The colon carcinoma cell line HT-29 was obtained from the
Istituto Zooprofilattico Sperimentale di Brescia (Brescia,
Italy). In order to differentiate HT-29 cells, we used two
different approaches: (a) to change the medium from high-

d-glucose DMEM to low-d-glucose RPMI supplemented
with 2 mmolÆL
)1
l-glutamine, 0.1 mgÆL
)1
streptomycin,
1 · 10
5
UÆL
)1
penicillin and 0.25 mgÆL
)1
amphotericin B,
cells were cultured in RPMI medium until confluence, when
they were subcultured for at least 10 passages; (b) substitu-
tion of glucose with galactose in DMEM medium ) these
culture conditions guarantee a high degree of cell differenti-
ation [52,53], as assessed by (i) measurement of the activity
of ALP and sucrase-isomaltase, two well-known biochemi-
cal markers of intestinal cell differentiation, present on the
brush border cell fraction (P2) isolated from the cell homo-
genates, (ii) measurement of their proliferation rate, and
(iii) their fine morphology as analyzed by transmission elec-
tron microscopy.
Cell cultures were periodically checked for the presence
of mycoplasma and were found to be free of contamina-
tion. Cell viability, assessed by the Trypan blue exclusion
test, and cell morphology, examined by optical microscopy,
remained unaffected by treatment with each one of the used
CPPs or CPP derivatives up to 40 mmolÆL

)1
.
Electron microscopy
Cells grown in DMEM, RPMI and in galactose-adapted
DMEM were plated in 35 mm Petri dishes and allowed to
grow until about 80% confluence, when they were fixed for
60 min at room temperature with 2% glutaraldehyde in
0.1 m Sorensen phosphate buffer (pH 7.4), thoroughly
rinsed with the same buffer, postfixed in 1% osmium tetra-
oxide (OsO
4
) in 0.1 m Sorensen phosphate buffer, dehy-
drated through an ascending series of ethanol, and
embedded in araldite (Durcupan; Fluka, Milan, Italy).
Ultrathin sections were obtained with an Ultracut ultra-
microtome (Reichert Ultracut R-Ultramicrotome; Leika,
Wien, Austria), and stained with uranyl acetate and lead
citrate before examination using a Jeol CX100 electron
microscope (Jeol, Tokyo, Japan).
Cell proliferation assay
Cells (1 · 10
4
cells per well), cultured in their medium in
a Microtiter plate (96-well, Greiner bio-one; Cellstar,
Frickenhausen, Germany), were submitted to a 2 h pulse
with bromodeoxyuridine (BrdU), and BrdU incorporation
into DNA was quantified by the chemiluminescent immu-
noassay (Roche Applied Science, Milan, Italy), following
the manufacturer’s instructions.
Isolation of brush border fraction and enzyme

assays
For the determination of ALP and sucrase-isomaltase activ-
ities, cells were seeded in 75 cm
2
flasks and, after reaching
80–90% confluence, were harvested in ice-cold physiological
saline, washed three times, pelleted by centrifugation at
105 000 g using a Beckman TL-100 (Beckman Coulter,
Fullerton, CA, USA) rotor type TLA-100.3, and stored at
) 80 °C. Cell subfractions, particularly the P2 subfraction
enriched in brush borders, were prepared as described pre-
viously [54,55]. The ALP assay was performed as previously
described [56] on samples of 20–50 lg of P2 subfractions
resuspended to a final volume of 50 lL. The sucrase-iso-
maltase assay was performed following the one-step ultra-
micromethod [57] on P2 subfractions (about 20 lgof
protein) resuspended to a final volume of 20 lL. Results
are expressed as mUÆ(mg protein)
)1
, 1 unit being defined as
the enzyme activity that hydrolyzes 1 lmole of substrate
per minute. The protein content was measured following
the method of Lowry et al. [58].
[Ca
2+
]
i
measurement in cell populations
The procedure described in our previous work [17] was
employed. Briefly, cells grown as a monolayer in a 25 cm

2
flask in RPMI culture medium were detached with tryp-
sin ⁄ EDTA, washed several times with KRH [containing
(mmolÆL
)1
): NaCl 125.0, KCl 5.0, KH
2
PO
4
1.2, CaCl
2
2.0,
MgSO
4
1.2, glucose 6.0, and Hepes 25.0, pH 7.4), and
loaded for 30 min at 37 °C with 5 lmolÆ L
)1
fura-2 acet-
oxymethyl ester and 2.5 lmolÆL
)1
Pluronic F-127 in KRH.
The loaded cell suspension was rinsed extensively with
KRH, and divided into aliquots comprising 0.5 · 10
6
cells.
Each aliquot was gently pelleted and resuspended in 2 mL
of KRH, and then transferred to a 37 °C thermostated
cuvette in a Perkin-Elmer LS-50B spectrofluorimeter
(Perkin-Elmer, Beaconsfield, UK). This fura-2-loaded cell
suspension was continuously stirred, and concomitantly sub-

mitted to excitation at 343 nm, the fluorescence intensity
being recorded at 485 nm. As fura-2 fluorescence increases
with increasing [Ca
2+
]
i
at these wavelength settings, the
changes in fluorescence intensity reflected the changes in
[Ca
2+
]
i
concentration [59]. CPP was administered to cell sus-
pension at the final chosen concentration, and at the end of
each experiment a calibration was performed [17]. The peak
of [Ca
2+
]
i
increase was calculated as the difference between
the [Ca
2+
]
i
values recorded after and before (basal value)
C. Gravaghi et al. CPP activity and supramolecular structure
FEBS Journal 274 (2007) 4999–5011 ª 2007 The Authors Journal compilation ª 2007 FEBS 5007
CPP administration, and was expressed as percentage of the
basal value or in absolute terms (nmolÆL
)1

). Under these
conditions, the duration of the experiments was less than
10 min, including a 1–2 min interval between the addition of
Ca
2+
and that of CPP. In this period of time, cells main-
tained full viability. As a control, [Ca
2+
]
i
concentrations
after ionomycin treatment [59] were also measured.
Experiments with increasing extracellular Ca
2+
concentrations
In the experiments performed in the presence of [Ca
2+
]
o
higher than 2 mmolÆL
)1
, fura-2-loaded cell pellets were
suspended, immediately before starting the experiment, in
phosphate-free KRH containing (mmolÆL
)1
) 140.0 NaCl,
5.0 KCl, 0.55 MgCl
2
, 6.0 glucose and 10.0 Hepes,
adjusted to pH 7.4, to prevent any possible precipitation

of calcium phosphate. Then, CaCl
2
was added in order
to obtain the desired final Ca
2+
concentrations.
Characterization of CPP aggregation by laser
light scattering
The aggregative properties of CPPs were studied by laser
light scattering. Aliquots of different CPPs, dissolved in
pure water as concentrated stocks, were diluted to the final
concentrations in KRH or in phosphate-free KRH, con-
taining the appropriate Ca
2+
concentration, matching the
experimental conditions used to follow the CPP biological
effect. The absence of any calcium phosphate precipitation
was a prerequisite for light scattering measurement. The
samples were transferred in an appropriate measuring cell,
and quasielastic laser light scattering measurements were
carried out on a standard apparatus equipped with a BI9K
Digital correlator (Brookhaven Instruments Co., Holtsville,
NY, USA) [60]. The light source was an argon ion laser
operating on the 514 nm green line (Lexel, Fremont, CA,
USA). Both independent static and dynamic laser light
scattering measurements were performed on the same sam-
ples at room temperature. If molecules undergo aggregation
in solution, laser light scattering immediately reveals the
presence of aggregates, recognizing both the dimension
(hydrodynamic radius) and the concentration of the aggre-

gated particles. Static measurements provide combined
information about the average molecular mass and the con-
centration of macromolecules in solution. The measured
quantity is the average light intensity scattered by the solu-
tion relative to that scattered by the solvent. All of the sol-
vents used in our experiments (water, phosphate-free KRH
and KRH) showed the same extremely low scattered inten-
sity within experimental errors. The excess of scattered
light due to the presence of CPP, I
r
) 1 ¼ (I
CPPsolution
)
I
solvent
) ⁄ I
solvent
is proportional to both the average molecu-
lar mass and the concentration of CPP particles in solution
according to the equation:
I
r
À 1 ¼ A
dn
dc

2
c
X
c

n
c
M
n
¼ A
dn
dc

2
c < M > ð1Þ
where A is a calibration constant, dn ⁄ dc is the refractive
index increment of the solution, c is the CPP concentra-
tion (gÆmL
)1
), c
n
is the concentration of CPP forming
particles of molecular mass M
n
, and <M> is the aver-
age molecular mass of the CPP particles in solution.
Independently, dynamic measurements yield information
about the diffusion coefficient D of particles in solution,
and hence their hydrodynamic radius, R
H
, via the
Stokes–Einstein relation:
D ¼
k
B

T
6pgR
H
ð2Þ
where k
B
is Boltzmann’s constant, T is the absolute temper-
ature, and g is the viscosity of the solvent [60,61]. If parti-
cles of different dimensions are present in solution, they
can be resolved, as their contribution to the measured cor-
relation function has a characteristic decay time propor-
tional to their dimension. Therefore, the availability of
both static and dynamic laser light scattering measurements
enables us to decouple information about the average mass
and relative concentration of CPP aggregates in solution.
Statistical analysis
The data reported in Figs 1 and 2 are expressed as mean
values ± SD. Statistically significant differences between
two mean values were established by Student’s t-test,
and two independent population t-tests, performed with
origin 6.0 (Origin Lab Corporation, Northampton, MA,
USA) (a P-value < 0.05 was considered significant).
Acknowledgements
This work was supported in part by the EU FAIR
Programme Project CT98-3077 [Casein phosphopep-
tide (CPP): Nutraceutical ⁄ functional food ingredients
for food and pharmaceutical applications] and by
Fondazione Romeo ad Enrica Invernizzi (CPP: role in
the calcium intestinal absorption and its utilization. A
perspective study on their possible usage as nutraceuti-

cals or functional food to favour calcium bioavailabil-
ity). We thank Professor Mario Corti for helpful
reading and discussing the manuscript.
References
1 Allen LH (1982) Calcium bioavailability and absorp-
tion: a review. Am J Clin Nutr 35, 783–808.
2 Bosscher D, Van Caillie-Bertrand M, Van Cawwen-
bergh R & Deelsdra H (2003) Availabilities of calcium,
iron, and zinc from dairy infant formulas is affected by
CPP activity and supramolecular structure C. Gravaghi et al.
5008 FEBS Journal 274 (2007) 4999–5011 ª 2007 The Authors Journal compilation ª 2007 FEBS
soluble dietary fibers and modified starch fractions.
Nutrition 19, 641–645.
3 Shah NP (2000) Effects of milk-derived bioactives: an
overview. Br J Nutr 84, S3–S10.
4 Recker RR, Bammi A, Barger-Lux MJ & Heaney RP
(1988) Calcium absorbability from milk products, an
imitation milk, and calcium carbonate. Am J Clin Nutr
47, 93–95.
5 Holt C & Sawyer L (1988) Primary and predicted sec-
ondary structures of the caseins in relation to their bio-
logical functions. Protein Eng 2, 251–259.
6 Holt C (1992) Structure and stability of bovine casein
micelles. Adv Protein Chem 43, 63–151.
7 Holt C, de Kruiff CG, Tuinier R & Timmins PA (2003)
Substructure of bovine casein micelles by small-angle
X-ray and neutron scattering. Colloids Surfaces 213 ,
275–284.
8 Little FM & Holt C (2004) An equilibrium thermody-
namic model of the sequestration of calcium phosphate

by casein phosphopeptides. Eur Biophys J 33, 435–447.
9 Cross KJ, Huq LN, Palamara JE, Perich JW & Rey-
nolds EC (2005) Physicochemical characterization of
casein phosphopeptide–amorphous calcium phosphate
nanocomplexes. J Biol Chem 280, 15362–15369.
10 Rollema HS (1992) Casein association and micelle for-
mation. In Advanced Dairy Chemistry, Vol. 1: Proteins
(Fox PF, ed.), pp. 111–140. Elsevier Applied Science,
London.
11 Meisel H (1997) Biochemical properties of bioactive
peptide derived from milk proteins: potential nutraceuti-
cals for food and pharmaceutical applications. Livest
Prod Sci 50, 125–138.
12 Meisel H & Bockelmann W (1999) Bioactive peptides
encrypted in milk proteins: proteolytic activation and
thropho-functional properties. Antonie Van Leeuwen-
hoek 76, 207–215.
13 Meisel H & FitzGerald RJ (2003) Biofunctional pep-
tides from milk proteins: mineral binding and cytomod-
ulatory effects. Curr Pharm Des 9, 1289–1295.
14 Sato R, Shindo M, Gunshin H, Noguchi T & Naito H
(1991) Characterization of phosphopeptide derived from
bovine beta-casein: an inhibitor to intra-intestinal pre-
cipitation of calcium phosphate. Biochim Biophys Acta
1077, 413–415.
15 Holt C, Wahlgren MN & Drakenberg T (1996) Ability
of a b-casein phosphopeptide to modulate the precipita-
tion of calcium phosphate by forming amorphous dical-
cium phosphate nanoclusters. Biochem J 314, 1035–1039.
16 Holt C, Timmins PA, Errington N & Leaver J (1998) A

core-shell model of calcium phosphate nanoclusters
stabilized by b-casein phosphopeptides, derived from
sedimentation equilibrium and small-angle X-ray and
neutron-scattering measurements. Eur J Biochem 252,
73–78.
17 Ferraretto A, Signorile A, Gravaghi C, Fiorilli A & Tet-
tamanti G (2001) Casein phosphopeptides influence cal-
cium uptake by cultured human intestinal HT-29 tumor
cells. J Nutr 131, 1655–1661.
18 Ferraretto A, Gravaghi C, Fiorilli A & Tettamanti G
(2003) Casein-derived bioactive phosphopeptides. Role
of phosphorylation and primary structure in promoting
calcium uptake by HT-29 tumor cells. FEBS Lett 551,
92–98.
19 Polak-Charcon S, Hekmati M & Ben-Shaul Y (1989)
The effect of modifying the culture medium on cell
polarity in a human colon carcinoma cell line. Cell Diff
Dev 26, 119–129.
20 Pinto M, Appay MD, Simon-Assmann P, Chevalier G,
Dracopoli N, Fogh J & Zweibaum A (1982) Enterocytic
differentiation of cultured human colon cancer cells by
replacement of glucose by galactose in the medium. Biol
Cell 44, 193–196.
21 Erba D, Ciappellano S & Testolin G (2002) Effect of
the ratio of casein phosphopeptides to calcium (w ⁄ w)
on passive calcium transport in the distal small intestine
of rats. Nutrition 18, 743–746.
22 Saito Y, Lee YS & Kimura S (1998) Minimum effective
dose of casein phosphopeptides (CPP) for enhancement
of calcium absorption in growing rats. Int J Vit Nutr

Res 68, 335–340.
23 Tsuchita H, Suzuki T & Kuwata T (2001) The effect
of casein phosphopeptides on calcium absorption from
calcium-fortified milk in growing rats. Br J Nutr 85,
5–10.
24 Mykka
¨
nen HM & Wasserman RH (1980) Enhanced
absorption of calcium by casein phosphopeptides in
rachitis and normal chicks. J Nutr 119, 2141–2148.
25 Bronner F, Pansu D & Stein WD (1986) Analysis of
intestinal calcium transport across the intestine. Am J
Physiol 250, G561–G569.
26 Bouhallab S, Cinga V, Aı
´
t-Oukhatar N, Bureau F,
Neuville D, Arhan P, Maubois JL & Bougle
´
D (2002)
Influence of various phosphopeptides of caseins on iron
absorption. J Agric Food Chem 50, 7127–7130.
27 Kibangou IB, Bouhallab S, Henry G, Bureau F, Allou-
che S, Blais A, Gue
´
rin P, Arhan P & Bougle
´
D (2005)
Milk proteins and iron absorption: contrasting effects
of different caseinophosphopeptides. Pedriatr Res 58,
731–734.

28 Gaucheron F, Le Graet Y, Boyaval E & Piot M (1997)
Binding of cations to casein molecules: importance of
physico-chemical conditions. Milchwissenschaft 52, 322–
327.
29 Huq NL, Keith JC & Reynolds EC (1995) A 1H-NMR
study of the casein phosphopeptide alpha s1-casein(59–
79). Biochem Biophys Acta 1247, 201–208.
30 Horne DS (2002) Casein structure, self-assembly and
gelation. Curr Opin Colloid Interface Sci 7, 456–461.
C. Gravaghi et al. CPP activity and supramolecular structure
FEBS Journal 274 (2007) 4999–5011 ª 2007 The Authors Journal compilation ª 2007 FEBS 5009
31 Horne DS (2003) Substructure of bovine casein micelles
by small-angle X-ray and neutron scattering. Colloids
Surfaces A: Physicochem Eng Aspects 213, 255–263.
32 Farrell HM Jr, Qi PX, Wickham ED & Unruh JJ
(2002) Secondary structural studies of bovine caseins:
structure and temperature dependence of beta-casein
phosphopeptide(1–25) as analyzed by circular dichroism,
FTIR spectroscopy, and analytical ultracentrifugation.
J Protein Chem 21, 307–321.
33 Cross KJ, Huq NL, Bicknell W & Reynolds EC (2001)
Cation-dependent structural features of b-casein-(1–25).
Biochem J 356, 277–286.
34 Kitts DD & Yuan YV (1992) Casein phosphopeptides
and calcium bioavailability. Trends Food Sci Technol 3,
31–35.
35 FitzGerald RJ (1998) Potential uses of caseinophospho-
peptides. Int Dairy J 8, 451–457.
36 Erba D, Ciappellano S & Testolin G (2001) Effect of
caseinphosphopeptides on inhibition of calcium intesti-

nal absorption due to phosphate. Nutr Res 21, 649–656.
37 Lee YS, Noguchi T & Naito H (1983) Intestinal absorp-
tion of calcium in rats given diets containing casein or
aminoacid mixture: the role of casein phosphopeptides.
Br J Nutr 49, 67–76.
38 Sato R, Noguchi T & Naito H (1986) Casein phopsho-
peptides (CPP) enhance calcium absorption from the
ligated segment of rat small intestine. J Nutr Sci Vitami-
nol 32, 67–76.
39 Recker RR & Heaney RP (1985) The effect of milk sup-
plements on calcium metabolism, bone metabolism and
calcium balance. Am J Clin Nutr 41, 254–263.
40 Kitts DD, Yuan YV, Nagasawa T & Moriyama Y
(1992) Effect of casein, caseinphosphopeptides and cal-
cium intake on ileal
45
Ca disappearence and temporal
systolic blood pressure in spontaneously hypertensive
rats. Br J Nutr 69, 765–781.
41 Brommage R, Juillerat MA & Jost R (1991) Influence
of casein phosphopeptides and lactulose on intestinal
calcium absorption in adult female rats. Lait 71, 173–
180.
42 Kopra N, Scholz-Ahrens KE & Barth CA (1992) Effect
of casein phosphopeptides on utilisation of calcium in
vitamin D-deficient rats. Milchwissenschaft 47, 488–492.
43 Parkinson GB & Gransberg PH (2004) Effect of casein
phosphopeptides and 25-hydroxycholecalciferol on tibial
dyschondroplasia in growing broiler chicken. Br Poultry
Sci 45, 802–806.

44 Bennet T, Desmond A, Harrington M, McDonagh D,
FitzGerald R, Flyn A & Cashman KD (2000) The effect
of high intakes of casein and casein phosphopeptide on
calcium absorption in the rat. Br J Nutr 83, 673–680.
45 Mellander O (1950) The physiological importance of the
casein phosphopeptide calcium salts. II. Peroral calcium
dosage of infants. Acta Soc Med Ups 55, 247–255.
46 Heaney RP, Saito Y & Orimo H (1994) Effect of casein
phosphopeptide on absorbability of Co-ingested calcium
in normal post menopausal women. J Bon Miner Metab
12, 77–81.
47 Hansen M, Sandstro
¨
m B, Jensen M & Sorensen SS
(1997) Casein phosphopeptides improve zinc and cal-
cium absorption from rice-based but not from whole-
grain infant cereal. J Pediatr Gastroenterol Nutr 24,
56–62.
48 Hansen M, Sandstro
¨
m B, Jensen M & Sorensen SS
(1997) Effect of casein phosphopeptides on zinc and cal-
cium absorption from bread meals. J Trace Elements
Med Biol 11, 143–149.
49 Lopez-Huertas E, Teucher B, Boza JJ, Martinez-Ferez
A, Majsak-Newman G, Baro L, Carrero JJ, Santiago
MG, Fonolla
`
J & Fairweather-Tait S (2006) Absorp-
tion of calcium from milks enriched with fructo-oligo-

saccharides, caseinophosphopeptides, tricalcium
phosphate and milk solids. Am J Clin Nutr 83,
310–316.
50 Narva M, Ka
¨
rkka
¨
inen M, Poussa T, Lamberg-Allardt
C & Korpela R (2003) Casein phosphopeptides in milk
and fermented milk do not affect calcium metabolism
acutely in post menopausal women. J Am Coll Nutr 22,
88–93.
51 Teucher B, Majsak-Newman G, Dainty JR, McDonagh
D, FitzGerald RJ & Fairweather-Tait S (2006) Calcium
absorption is not increased by caseinophosphopeptides.
Am J Clin 84, 162–166.
52 Huet C, Sahuquillo-Merino C, Coudrier E & Louvard
D (1987) Absorptive and mucus-secreting subclones
isolated from multipotent intestinal cell line (HT29)
provide new models for cell polarity and terminal differ-
entiation. J Cell Biol 105, 345–357.
53 Zweibaum A, Laburthe M, Grasset E & Louvard D
(1991) The gastrointestinal system IV. In Handbook of
Physiology, 4th edn (Rauner BB, Field M, Frizzel RA
& Schultz SG, eds), pp. 223–255. American Physiologi-
cal Society, Bethesda, MD.
54 Schmitz J, Preise H, Maestracci D, Ghosh BK, Cerda
JJ & Crane RK (1973) Purification of the human intesti-
nal brush border membrane. Biochim Biophys Acta 323,
98–112.

55 Zweibaum A, Pinto M, Chevalier G, Dussaulx E, Tria-
dou N, Lacroix B, Haffen K, Brun JL & Rousset M
(1985) Enterocytic differentiation of a subpopulation of
the human colon tumor cell line HT-29 selected for
growth in sugar-free medium and its inhibition by glu-
cose. J Cell Physiol 122, 21–29.
56 Buras RR, Shabahang M, Davoodi F, Schumaker LM,
Cullen KJ, Byers S, Nauta RJ & Evans SRT (1995) The
effect of extracellular calcium on colonocytes: evidence
for differential responsiveness based upon degree of cell
differentiation. Cell Prolif 28, 245–262.
CPP activity and supramolecular structure C. Gravaghi et al.
5010 FEBS Journal 274 (2007) 4999–5011 ª 2007 The Authors Journal compilation ª 2007 FEBS
57 Messer M & Dahlquist A (1966) A one-step ultramicro
method for the assay of intestinal disaccharidases. Anal
Biochem 14, 376–392.
58 Lowry OH, Rosebrough N, Farr A & Randall R (1951)
Protein measurement with the folin phenol reagent.
J Biol Chem 193, 265–275.
59 McCormack JG & Cobbold PH (1991) Cellular Cal-
cium: a Practical Approach. The Practical Approach
Series. Oxford University Press, New York, NY.
60 Cantu` L, Corti M, Lago P & Musolino M (1991) Char-
acterization of a vesicle distribution in equilibrium with
larger aggregates by accurate static and dynamic laser
light scattering measurements. SPIE 1430, 144–159.
61 Corti M (1985) The laser light scattering technique and
its application to micellar solutions. In Physics of
Amphiphiles: Micelles, Vesicles and Microemulsions
(Degiorgio V & Corti M, eds), pp. 122–151. North-

Holland, Amsterdam.
C. Gravaghi et al. CPP activity and supramolecular structure
FEBS Journal 274 (2007) 4999–5011 ª 2007 The Authors Journal compilation ª 2007 FEBS 5011