Milk Lacking a-Casein Leads to Permanent Reduction in
Body Size in Mice
Andreas F. Kolb1,2*, Reinhard C. Huber3,4,6, Simon G. Lillico4, Ailsa Carlisle4, Claire J. Robinson1,
Claire Neil4, Linda Petrie2, Dorte B. Sorensen5, I. Anna S. Olsson3,6, C. Bruce A. Whitelaw4
1Molecular Recognition Group, Hannah Research Institute, Ayr, United Kingdom, 2Nutrition and Epigenetics Group, Vascular Health Division, Rowett Institute of Nutrition
and Health, University of Aberdeen, Aberdeen, United Kingdom, 3Laboratory Animal Science Group, Instituto de Biologia Molecular e Celular, Universidade do Porto,
Porto, Portugal, 4Division of Developmental Biology, The Roslin Institute and Royal (Dick) School of Veterinary Studies, University of Edinburgh, Edinburgh, United
Kingdom, 5Department of Large Animal Sciences, University of Copenhagen, Copenhagen, Denmark, 6Danish Centre for Bioethics and Risk Assessment, Faculty of Life
Sciences, University of Copenhagen, Copenhagen, Denmark
The major physiological function of milk is the transport of amino acids, carbohydrates, lipids and minerals to mammalian
offspring. Caseins, the major milk proteins, are secreted in the form of a micelle consisting of protein and calcium-
phosphate. We have analysed the role of the milk protein a-casein by inactivating the corresponding gene in mice.
Absence of a-casein protein significantly curtails secretion of other milk proteins and calcium-phosphate, suggesting a role
for a-casein in the establishment of casein micelles. In contrast, secretion of albumin, which is not synthesized in the
mammary epithelium, into milk is not reduced. The absence of a-casein also significantly inhibits transcription of the other
casein genes. a-Casein deficiency severely delays pup growth during lactation and results in a life-long body size reduction
compared to control animals, but has only transient effects on physical and behavioural development of the pups. The data
support a critical role for a-casein in casein micelle assembly. The results also confirm lactation as a critical window of
metabolic programming and suggest milk protein concentration as a decisive factor in determining adult body weight.
Citation: Kolb AF, Huber RC, Lillico SG, Carlisle A, Robinson CJ, et al. (2011) Milk Lacking a-Casein Leads to Permanent Reduction in Body Size in Mice. PLoS
ONE 6(7): e21775. doi:10.1371/journal.pone.0021775
Editor: Marc Tjwa, University of Frankfurt, Germany
Received September 7, 2010; Accepted June 11, 2011; Published July 18, 2011
Copyright: ? 2011 Kolb et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted
use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported by the Scottish Executive Environment and Rural Affairs Department (ROAME 31190), the BBSRC through ISPG support and
Gene Technologies underpinning Health Care project grant #12599, the Hannah Development Fund and the Genomia Seed Fund, the Fundac ¸a ˜o para a Cie ˆncia e
a Tecnologia (PhD grant to RH SFRH/BD/36682/2007), and the European Commission NEST029025 project INTEGRA. The funders had no role in study design, data
collection and analysis, decision to publish, or preparation of the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: email@example.com
Milk is a hallmark of mammals, providing the primary source of
nutrition for their young until weaning. It is an emulsion of fat
globules in a water-based fluid. The major proteins of this fluid are
the caseins. The caseins are serine rich phosphoproteins which are
almost exclusively expressed in the lactating mammary gland
[1,2]. In cows the casein proteins together constitute up to 80% of
total milk protein and also 80% of total mRNA in the lactating
mammary gland . The caseins are members of a large family of
serine rich phospho-proteins, which are clustered on chromosome
5 in the mouse (chromosome 6 in cattle, chromosome 4 in
humans). There are 5 functional casein genes in the mouse
whereas most other species (including ruminants) only express 3 or
4 functional genes . The casein proteins show little homology
with each other outside their signal peptide domain [2,4].
Inactivation of the b-casein gene has shown to have little effect
on milk secretion and growth of the offspring in the mouse . In
contrast, deletion of the k-casein gene in mice completely
abrogates milk production . Exactly why this happens is
unclear as neither the structure of the individual casein proteins
nor the structure of the casein micelles is currently known in detail
[7,8,9]. The available results, however, support a model in which
k-casein (the casein protein which is present in milk in least
abundance) plays a critical role in ensuring transport and solubility of
casein micelles . The sequences of a- (aS1 in ruminants) b-, and c-
casein (aS2-casein in ruminants) contain one or more clusters of
phosphorylated residues known as phosphate centres. Phosphate
centres can sequester amorphous calcium phosphate, probably in the
Golgi vesicles of mammary secretory cells, to form thermodynam-
ically stable complexes of defined chemical composition which are
then secreted through the apical membrane of the mammary
epithelial cells [10,11,12,13,14]. Milk, like many other biological
fluids, is supersaturated with respect to the crystalline mineral of
bones and teeth (apatite) but due to the sequestration reaction, is
under-saturated with respect to amorphous calcium phosphate. Since
apatite only forms by maturation of the amorphous phase of calcium
the majority of the total calcium is sequestered within the casein
micelles. In milks of different species, the total calcium and total
casein concentrations are highly correlated  and provided this
balance is maintained, the milks remain stable. If there is insufficient
casein to sequester the secreted calcium and orthophosphate then the
milk becomes unstable . In mouse milk, 3 of the phosphate
centres arein present a-casein and onlyone in b-casein. Thus the loss
of a-casein is likely to have more serious consequences for milk
stability than the loss of b-casein.
PLoS ONE | www.plosone.org1 July 2011 | Volume 6 | Issue 7 | e21775
Deficiencies for b-casein and a-casein exist as naturally
occurring genotypes in goats  but have no apparent
detrimental effect on milk production. However, the aS1-Cn0
allele has been reported to decrease the efficiency with which other
casein proteins are secreted . We set out to define the role of a-
casein in milk secretion and its nutritional role by inactivating the
corresponding gene in mice. Our results demonstrate that a-casein
deficiency has a significant effect on milk protein secretion and
growth of the offspring.
Materials and Methods
PCR amplifications were done using Taq Polymerase from
various suppliers. Oligonucleotides were purchased from MWG,
Sigma-Genosys or Invitrogen. Primer sequences, amplicon size
and annealing temperatures are given in table 1. Template DNA
for PCR analyses was isolated as described .
The targeting construct for the a-casein gene was generated
using a short arm of homology of 679 bp corresponding to
nucleotides 208 to 887 downstream of the transcriptional start site
(which was isolated as a EcoRV/BssHII fragment from the
bacterial artificial chromosome BAC 490H23 of a Research
Genetics BeloBAC11 library derived from 129SV mouse DNA)
and a long arm of homology of 6696 bp corresponding to
nucleotides 1422 to 8117 downstream of the transcriptional start
site (isolated as a StuI/EcoRV fragment). The targeting construct
carries a hygromycin-thymidine kinase fusion gene under the
control of the murine phospho-glycerol kinase (PGK) promoter as
Transgenic mice were generated at Genoway (Lyon, France) as
a SV129xC57BL/6 mixed background. After transfer to the
transgenic mouse facility at the Roslin Institute the mice were
maintained there in accordance with Home Office guidelines. This
Table 1. Primer combinations used for PCR analysis.
name sequence annealing temp.amplicon size
acas659 GTT CCA GTA CGC TAA GAG CTT AAA ACT GC 39
acas759 TAG GGG GCA AAA ATG TGT ATT ATC CTA GC 39
53uC 1566 bp
acas659 GTT CCA GTA CGC TAA GAG CTT AAA ACT GC 39
pBKpA59 GCT ATT GCT TTA TTT GTA ACC ATT A 39
53uC 848 bp
acas459 AAG TTT CCC CAG CAC AGC AAT CT 39
acas559 CCA AAG GGG AAA GGC ATC ATA CT 39
56uC 248 bp
bcas2159 GAT GCC CCT CCT TAA CTC TGA AA 39
bcas2259 TTG TGG AAG GAA GGG TGC TAC T 39
56uC 226 bp
gcas1059 ATA ACA CGC CCA CCC AGG AAT C 39
gcas1159 GGG AAA CCA CGA AGA AAC CAA T 39
54uC 272 bp
kcas159 TCG ACC CCA TTA CTC CCA TTG TGT 39
kcas259 TGT AAA AGG TAA GGG AAG ACG AGA AAG AT 39
53uC 289 bp
mGAPDH159 GCT TTC CAG AGG GGC CAT CCA CA 39
mGAPDH259 ACG GCA AAT TCA ACG GCA CAG TCA A 39
61uC 426 bp
acas159 ATG AAA CTC CTC ATC CTC ACC TGC 39
acas759 TAG GGG GCA AAA ATG TGT ATT ATC CTA GC 39
PGK559 AAG CGC ATG CTC CAG ACT GCC TTG GGA AA 39
acas759 TAG GGG GCA AAA ATG TGT ATT ATC CTA GC 39
Nol3-159 CTC CGG ACC ACA AGC CCG ACT C 39
Nol3-259 CTG GGT GCT TCT GGC GTC CAG 39
61uC 452 bp
Traf1-559 GCA AAC CCT GGC TCA AAA AG 39
Traf1-659 ACT CTG TGG CCG CTG GAA GG 39
57uC 409 bp
Birc5-15 CTG GAG GAC TGC GCC TGC AC9 39
Birc5-259 TGC TAG GAG GCC CTG GCT GG 39
65uC 432 bp
BiP-359 TCA CGT CCA ACC CCG AGA ACA C 39
BiP-459 GCC GCC ACC CAG GTC AAA CAC A 39
58uC 434 bp
grp94-159 GGG TGT TGG GCC TCT GCT GTG 39
grp94-259 ACC CGT GTC TGT GAC ATG CAG C 39
54uC 440 bp
PDIA6-159 GGC AGC CAT CAA TGC ACG CAA 39
PDIA6-259 TGG GCC GTG GCT CAC AAC TCA T 39
58uC 258 bp
REDD1-159 TGG TGC CCA CCT TTC AGT TG 39
REDD1-259 GTC AGG GAC TGG CTG TAA CC 39
53uC 121 bp
a-Casein Deficient Mice
PLoS ONE | www.plosone.org 2 July 2011 | Volume 6 | Issue 7 | e21775
study was approved by the Roslin Institute Animal Ethics
Committee and was performed under Home Office Licence 60/
Housing and maintenance of mice
Before establishing the mating couples, mice were kept in same-
sex groups of 2 to 6 littermates. The animals were kept in opaque
M3 type cages with dimensions of 48(15(13(cm from North Kent
Plastics, UK equipped with bedding (‘Eco-bedding’, B&K
Universal Ltd, UK) and nesting material (‘Nestlet’, Datesand
Ltd, UK). Cages were changed once per week and in case of the
nursing females the nesting material was transferred to the new
cage. The environment was kept at 21+/(2(C and 55+/(15%
relative air humidity with a 12(hour light 12(hour dark cycle (lights
off 8:00 pm) Food (BeeKay Transgenic Rodent Diet BK021E,
B&K Universal Ltd, UK) and tap water were available ad libitum,
water bottles being routinely changed twice per week.
Breeding of mice
Apparently pregnant females were separated from the male and
checked daily in the morning for pups in the cage. The first day
pups were found was considered the day of birth (day 1). On this
day the pups were marked individually by foot pad tattooing and
checked for the presence of a milk spot and signs of lack of
maternal care. The offspring of 3(wt females were fostered onto 3
(-casein deficient females and vice versa on day 2 (Table 2), the
cross fostered litters being matched for day of birth. The pups were
weaned on day 23 by separating them from the nursing female and
placing them in cages as described above with same-sex litter
mates. From weaning until the age of 6 weeks, water-soaked diet
was provided ad libitum in Petri dishes on the cage to weaned pups
in all groups to facilitate solid food intake by growth impaired
Pre-weaning mortality was low in all groups and unaffected by
dam genotype. One wild type female had 1 pup stillborn and 2
born alive but lost between day 1 and cross fostering. Two a-
casein deficient females lost 2 pups and one pup respectively
between day 1 and cross fostering. No pups were lost from cross
fostering until day 21, but on day 22 two pups were found dead in
the largest litter (11 pups before day 22) nursed by an a-casein
All monitoring took place in a testing room adjacent to the
housing room and with similar environmental conditions. The
entire housing cage was moved from the housing room to the
monitoring room. Physical and social aspects of mouse housing are
described in detail in table 3.
Behavioural analysis before weaning (early phenotyping)
The pups were individually weighed and monitored for general
behaviour and reactivity, health, sensory and muscle development,
and reflexes on day 1, 3, 7, 14 and 21 of age following a protocol
based on pre-weaning development stages and adapted from 
and . Additionally, from day 13 onwards pups were weighed
every second day and screened daily to detect the day when both
eyes were opened. During scoring the adult female was removed
from the home cage and placed into an empty opaque cage, food
and water being provided. Specifically for the respective
monitoring days, the following records were taken:
Day 3: presence/absence of a milk spot and the position of the
animals regarding the nest (inside/outside). Righting reflex: the pup
was positioned in dorsal recumbence and the time it took to turn
to a ventral position was taken with a cut off time of 60 sec.
Day 7: presence/absence of a milk spot and the position of the
animals regarding the nest (inside/outside). Presence/absence of
the first fur was recorded.
Day 14: Pups were inspected for the presence of a fully
developed fur and the opening of the eyelids, auditory canal and
the presence/absence of the incisors. Grip reflex: The animals were
stimulated at their 4 paws with a cotton swab and the reflex
reaction (closing of paws) recorded.
Day 21: At the originally scheduled weaning day (day 21), the
home cage was placed into the testing room for at least 10 minutes.
After this the pups were weighed and placed individually into an
empty opaque cage (observation cage; similar to the housing cage as
described above) for close inspection. There each animal was
screened for apparent deviations from fluid gait, horizontally
extended tail position, normal body position (e.g. hunched back),
normal body proportion (e.g. lumps) and normal respiration while
moving freely in the cage. While being restrained in a dorsal position,
signs of dehydration (lifting a skin flap between index finger and
thumb), general fur condition (smoothness), the presence of wounds,
trimmed whiskers, presence of discharge in nose and the anogenital
region, general eye condition and skin colour were recorded. Muscle
tonus in fore/hind legs and abdomen was recorded by applying slight
pressure with the index finger of the free hand.
After the assessment under restraint, the following tests were
performed (adapted from  and ): Provoked biting: To assess
reflexive biting reaction, the animal was stimulated with a cotton
swab at the muzzle followed by a check for conspicuous teeth and
mucosa condition. Grip strength: To assess the ability of the animal
to hold its own bodyweight, the individual was placed on a cage lid
which was then turned upside down. Animals being able to hold
on for at least 15 sec. were recorded as successful. Vertical pole test:
As a test of motor coordination and balance the animal was placed
on a horizontally positioned pole (80 cm long, 1.55 cm diameter,
taped) which was erected slowly to vertical position. Animals
falling off before reaching vertical position were recorded. Postural
reflex: The animal was placed in an empty observation cage which
was then gently shaken 3 times vertically and 3 times horizontally.
The presence/absence of the natural reaction to this treatment
(extending all 4 limbs) was recorded.
Thereafter, while still being in the observation cage, the senses
of touch and of hearing were measured. First a sudden airflow was
Table 2. Experimental groups.
Group Females Pups nursed n littersn pups n pups/litter min-max on day 21
G1 Wt +/+
Wt +/+ (own)
Wt +/+ (from G3)
Het +/2 (from G2)
G2 Null 2/2
3 25 6–11
a-Casein Deficient Mice
PLoS ONE | www.plosone.org3July 2011 | Volume 6 | Issue 7 | e21775
Table 3. Details and scores of the modified SHIRPA protocol.
Respiration rate: Observe the respiration rate in viewing jar.0 Gasping, irregular
1 Slow, shallow
Tremor: See if the mouse tremors in the viewing jar.0 None
Piloerection: Observe body hair and piloerection.0 None
1 Coat stood on end
Palpebral Closure: Observe both eyes and their closure.0 Eyes wide open
1 Eyes 1/2 closed
2 Eyes closed
Gait: Observe the gait of the mouse.0 Normal
1 Fluid but abnormal
2 Limited movement only
Pelvic Elevation: Visually measure the pelvis height of mouse0 Markedly flattened
in gait. Keep the eye level of the observer at the height of the1 Barely touches
mouse. Make the observation from the side of the arena.2 3 mm elevation Normal
3 Elevated (.3 mm elevation)
Tail Elevation: See if the mouse elevates the tail.0 Dragging
1 Horizontally extended
2 Elevated/Straub Tail (.45u)
Touch Escape: Approach the mouse with a finger sideways.0 No response
Observe how close the finger is when the mouse escapes.1 Escape response to touch
2 Escape response to approach
Positional Passivity: Hold up the mouse by the tail on the0 Struggles when held by tail
arena to see if the mouse resists. If the mouse does not resist, hold the mouse
in restraint by the neck, keep it on the
1 Struggles when held by neck (finger grip, not
back, and then hold it by the hind legs. See if the mouse resists at each step
and stop when the mouse resists.
2 Struggles when laid supine (on back)
3 Struggles when held by hind legs
4 No struggle
Trunk curl: See if the mouse brings the upper body up by stooping the ventral side
and shows a sit-up movement curl
when held up by the tail. Twisting the upper body (Trunk sideways is not trunk curl).1 Present
Limb grasping: See if the mouse holds the forelimbs and hind0 Absent
legs together (Limb grasping) when held up by the tail.1 Present
Grip Strength: Hold the mouse by the tail and drag it to the0 None
fringe of the metal net on the arena. Evaluate the grip1 Slight grip, semi-effective
strength felt by the hand of the observer.2 Moderate grip, effective
3 Active grip, effective
4 Unusually effective
Body Tone: Pinch the mouse with the thumb and forefinger of the observer on the
dorsal sides while allowing the mouse to
0 Flaccid, no return of cavity to normal
hold onto the metal net in the arena. Evaluate the resistance.1 Slight resistance
2 Extreme resistance, board like
Corneal Reflex: Stimulate the cornea of the mouse with the0 None
body (not the tip) of the wire attached to a dowel and see if 1 Active single eye blink
the mouse closes the eyelids.2 Multiple eye blink
a-Casein Deficient Mice
PLoS ONE | www.plosone.org4July 2011 | Volume 6 | Issue 7 | e21775
applied to the back of the animal from about 3 cm distance using a
rubber bulb air blower and the presence of reaction (ear flick,
shaking) was recorded. To test hearing, a ballpoint pen was
‘clicked’ about 5 cm behind the animal and the presence of
reaction (ear flick, shaking, jumping, fleeing) was recorded. During
the procedures mentioned above the number of faeces was
counted. The scale pan and the observation cage were cleaned
with surface cleaning wipes after each animal.
Behavioural analysis after weaning
The animals were weighed weekly after weaning and at 8 weeks
of age the animals were individually screened using a protocol
adapted from , including parameters on health, general
behaviour and reactivity, reflexes and muscle strength.
First the animal was placed into a transparent Perspex cylinder
(height 18 cm, Ø 15 cm) and observed regarding general activity,
respiration and occurrence of tremor. Then the animal was
transferred into a Makrolon IV cage (42630627 cm, Allentown)
and deviations from normal palpebral closure, piloerection, gait,
pelvic elevation, tail elevation, and positional passivity were
recorded. Touch escape response was assessed by slowly
approaching the mouse with a finger sideways and attempting to
In tail suspension, trunk curl and limb grasping were recorded
with the mouse hanging and body tone, corneal/pinna/toe pinch
reflex were observed while the animal was placed on a cage grid.
Efficiency of grip was assessed by gently pulling the animal back by
the tail. A wire manoeuvre was performed by letting the mouse
hold on to a horizontal wire (length 45 cm, Ø 3.1 mm) with the
forelimbs while being held by the tail and releasing it from a
Under supine restraint, skin colour, deviations from normal
heart rate, limb/abdominal tone, lacrimation, salivation and signs
of fear, irritability and aggression were recorded. Provoked biting
response was tested as described above. A negative geotaxis test
was performed by placing the animal on a horizontal grid and
turning the grid to vertical position with the head facing down.
Milk was isolated from mammary tissue at peak lactation (day
10 of lactation) after cervical dislocation of the mice. The milk was
harvested using a Pasteur pipette which was put onto the nipple
while pressure was applied to the tissue. Milk was drawn up into
the pipette transferred into a microcentrifuge tube and stored at
After thawing at room temperature, milk samples were well
suspended, diluted 1 in 4 in distilled water, and centrifuged for
5 sec. at 200 g in a table top centrifuge. The lower phase
containing the defatted whole milk was removed from the upper
lipid layer into a new tube and diluted with 2 volumes of distilled
water. Then one volume of a 46 concentrated protein sample
buffer was added. To separate the whey and casein fractions,
defatted whole milk was spun for 15 min. at 200 g. The
supernatant containing the whey fraction was diluted with 2
volumes of distilled water and one volume of a 46concentrated
protein sample buffer. The pellet containing the casein fraction
was re-suspended in an equivalent of 3 volumes (of starting
volume) of distilled water and then one volume of a 46
Toe pinch: Use forceps with the tips bent to stimulate the hind0 None
legs while allowing the mouse to hold onto the metal net on1 Slight withdrawal
the arena. Observe the response.2 Moderate withdrawal, not brisk
3 Brisk, rapid withdrawal
4 Very brisk repeated extension and flexion
Wire manoeuvre: Let the mouse hold on to a horizontal wire0 active grip with hindlegs
with the forelimbs, hold it by the tail and bring it to horizontal position before letting go.1 Difficulty to grasp with hindlimbs
2 Unable to grasp with hindlimbs
3 Unable to lift hindlegs, falls within seconds
4 Falls immediately
Skin color: Observe the color of the ventral sides of limbs0 Blanched
(palms and soles).1 Pink
2 Bright, deep red flush
3 Dark footpad, pigmentation
Limb tone: Press the rear side of the hind legs of the mouse0 No resistance
with the forefinger and the middle finger of the observer to1 Slight resistance
evaluate how violently the mouse kicks back.2 Moderate resistance
3 Marked resistance
4 Extreme resistance
Negative Geotaxis: Animal placed on horizontal grid, lifted to vertical with animal facing the floor -
0 animal stays in head down position
1 animal turns head up
2 animal turns head halfway up
Table 3. Cont.
a-Casein Deficient Mice
PLoS ONE | www.plosone.org5 July 2011 | Volume 6 | Issue 7 | e21775
concentrated protein sample buffer was added. The samples were
separated on a 10% polyacrylamide gel and stained with
Coomassie Blue. Milk protein expression visible in the stained
gel was then analysed on a Kodak Imaging Station using the
Kodak 1D imaging software.
Total protein extracts from tissues were isolated as described
. Briefly tissue fragments were dounced on ice at 100 mg/ml
in a buffer containing 1% SDS, 1% NP-40 and 0.5% deoxycholic
acid in 16PBS supplemented with a broad spectrum proteinase
inhibitor mix (SIGMA P-8340). The lysates were incubated on ice
for 30 min. and then centrifuged at 4uC for 10 min. at 10000 g.
The supernatants were then aliquoted and stored at 280uC.
Cytoplasmic protein extracts for the analysis of caspase activity
were prepared as described . Briefly, tissue fragments were
dounced on ice in a buffer containing 25 mM KPO4pH 7.8,
8 mM MgCl2, 1 mM EDTA, 1% Triton X-100 and 15% glycerol.
The extracts were incubated on ice for 5 min. and then
centrifuged at 4uC for 1 min. at 10000 g. The supernatants were
then aliquoted and stored at 280uC.
Western blots were done after semi-dry transfer of the proteins
to a nitrocellulose membrane as described . Mouse a-casein
protein was detected using a rabbit-anti a-casein antiserum (Santa
Cruz Biotechnology sc-98699) and a horse-radish-peroxidase
linked goat anti-rabbit serum (Cell Signalling Technologies
#7074). Mouse b-casein protein was detected using a goat-anti
b-casein antiserum (Santa Cruz Biotechnology sc-17969) and a
horse-radish-peroxidase linked rabbit anti-goat serum (Jackson
Immuno-Research). Mouse grp78/BiP protein was detected using
a goat-anti grp78/BiP antiserum (Santa Cruz Biotechnology sc-
1051) and a horse-radish-peroxidase linked rabbit anti-goat serum.
All antibodies were used at a dilution of 1:1000.
Milk proteins were separated on a 15% polyacrylamide gel and
bands were excised manually. The proteins represented by these
bands were trypsinised using a protocol of the Micromass
MassPrep Station (Micromass Ltd, Manchester, UK) and analysed
by electrospray LC-MS methods as described previously .
Calcium and phosphate concentration were measured in milk
derived from wild-type, a-casein deficient [2/2] and heterozy-
gous [+/2] mice using a Konelab 30 Clinical analyser (Thermo
Scientific) using the Konelab calcium (catalogue number 981367)
and phosphorus (catalogue number: 981386) analysis kits as
recommended by the supplier.
RNA and quantitative PCR
RNA was isolated from tissues using the Ambion RNAwiz
reagent following the supplier’s protocol (1 ml of RNAwiz per
100 mg of tissue). Reverse transcription of RNA was done with
MLV RNAse(-) reverse transcriptase (Promega) following the
suppliers recommendations. 2 mg of total RNA was used as
template for the cDNA synthesis reaction. A one in 10 dilution of
the reaction was subsequently used as template for quantitative
PCR (Applied Biosystems 7500 Fast System).
Quantitative PCR amplifications were done with a final primer
concentration of 0.5 mM. Primer design was done using the Primer
Select program of the DNA Star software suite. The sequences,
annealing temperatures and amplicon sizes of the oligonucleotides
used in this study are provided in Table 1. The PCR products
obtained by quantitative PCR were evaluated by melting point
analysis and agarose gel electrophoresis. Amplifications were done
at 40 cycles of 15 sec. at 95uC, 15 sec. at the indicated annealing
temperature and 30 sec. at 72uC. Data were collected at the end of
each PCR cycle. Standard curves for all genes were generated
from serial dilutions of a plasmid containing the cDNA for each
gene. The crossing points obtained from the sample analysis was
then correlated with the standard curves to provide a concentra-
tion of the individual PCR product. Expression of the casein genes
was then correlated with expression of the reference genes (b-actin
or GAPDH) in the same sample (expressed as pg of gene per pg of
PCR arrays were carried out using a mouse apoptosis array
(SABiosystems: catalogue number PAMM-012) following the
recommendations of the supplier.
Tissues were collected from mice at day 10 of lactation,after
schedule 1 cervical dislocation. The left lower mammary gland
was dissected out and placed in 4% para-formaldehyde overnight
at 4uC on a rocking platform. The para-formaldehyde was then
removed and the tissues were washed, first, in PBS for 30 minutes
at 4uC, then in 30% ethanol for 15 min. at room temperature.
This was followed by two 30 min. washes in 70% ethanol at room
temperature. The tissues were then transferred to fresh 70%
ethanol until ready for processing for histology. The tissues were
dehydrated through a series of alcohol dilutions followed by
penetration by wax under vacuum (in a Shandon Hypercentre).
The tissues were then embedded in wax blocks for sectioning.
Each sample was sectioned at three intervals at least 100 mm apart
and a 5 mm section mounted for immuno-histochemistry analysis
at each interval.
Sections were de-waxed using Histo-clear (Lamb, Inc.) for
5 min. and then re-hydrated in a graded series of ethanol (100%,
95%, 70%; 2 min. per incubation). The slides were then washed in
deionised water and 16 PBS (two 5 min. incubations). The
endogenous peroxidase activity was blocked by a 5 min.
incubation in 3% hydrogen peroxide. After two 5 min. washes
in PBS the slides were blocked in 1% BSA in 16PBS-Tween for
20 min. After addition of the primary antibody (rabbit-anti a-
casein antiserum; 1:100 dilution in 16PBS-Tween) the slides were
incubated at 4uC overnight. The slides were then washed twice for
5 min. in 16PBS and the secondary antibody was added (horse-
radish-peroxidase linked goat anti-rabbit serum; 1:1000 dilution in
16PBS-Tween). After a 30 min. incubation at room temperature
the slides were washed twice for 5 min. in 16PBS-Tween.
Horseradish peroxidise activity was detected using a diamino-
benzidine dye-kit (Vector-Laboratories). The slides were incubated
with substrate for 5 min. The reaction was stopped by washing the
slides with distilled water.
The slides were then counterstained with Haematoxylin (Cell-
Path) for 2 min. and washed in running tap water. The slides were
then de-hydrated in a graded series of ethanol (70%, 95%, 100%;
2 min. per incubation), cleared in Xylene (Fisher Biochemicals)
and coverslips were mounted in Pertex (Cell-Path). Following
staining the sections were photographed using a Leica Stereomi-
croscope at a 20 fold fold magnification.
Mouse RAW264 cells (ECACC catalogue number 85062803)
were grown in DMEM (4.5 g/l glucose) supplemented with 10%
foetal calf serum, 2 mM glutamine and penecillin/streptomycin.
To induce apoptosis the cells were treated with 10 mM
staurosporin for 6 h. Cytoplasmic extracts for the analysis of
caspase activity were prepared as described in section ‘‘milk
analysis’’ above. Protein concentrations were measured using a
Bradford assay (Sigma). Caspase activity of cell extracts was
a-Casein Deficient Mice
PLoS ONE | www.plosone.org6July 2011 | Volume 6 | Issue 7 | e21775
measured using the Caspase-3/7-Glo assay (Promega) as recom-
mended by the supplier. The assay provides a pro-luminescent
caspase-3/7 substrate, which contains the amino acid sequence
DEVD in a reagent optimised for caspase activity, luciferase
activity and cell lysis. Presence of caspase 3 or 7 in the cell extract
leads to cleavage of the substrate and light emission which is
detected in a luminometer (Berthold). Three or more parallel
samples were analysed to measure caspase activity. 50 ml of
protein extract (from cells and tissues) were mixed with 25 ml of the
Caspase Glo reagent and incubated at room temperature for
20 min. after which the readings were taken. Apoptosis in RAW
cells was also confirmed by measuring cell viability using the Cell-
Titre Blu reagent (Promega) and DNA laddering on a 2% agarose
Categorical data measured only once were analyzed using the
Fisher’s exact test (2-sided). Count data were analyzed using the
Mann Whitney U test. For comparison of body weight
development we fitted a linear model for weight using time
and group as covariates. Litter size was also tested as a covariate
but no significant effect was observed and it was removed from
the model. The interaction between time and treatment was
included in the model so each treatment would have its own
growth rate. We also considered a spline at day 23 (weaning
threshold). The spline allows the growth rate to change after the
day 23 for each treatment. This way we are able to test
differences of the growth rate between treatments before and
after weaning. The model was fitted using generalized
Figure 1. Targeting of the a-casein gene. Panel A: Schematic representation of the murine casein locus. Casein genes are represented as solid
arrows. Other predicted genes are shown as open arrows. Panel B: Schematic representation of the unmodified a-casein gene and the targeted a-
casein gene. Exons of the a-casein gene are indicated as solid boxes, the hytk selection marker gene is indicated as striped box. The PGK promoter
element directing expression of the selection marker gene is indicated as arrowhead. The relative positions of the EcoRI restriction sites (EI), the
Southern blot probe (probe), sizes of hybridising DNA fragments and the primer binding sites used for genotyping (horizontal arrows) are indicated.
Panel C: PCR analysis of genomic DNA isolated from the three representative ES cell clones using the primer combination acas6, acas7 and pBKpA2
(analysing the 59 end of the homologous recombination event). A 1566 bp band is detected in all samples and represents the unmodified a-casein
allele [U: unmodified]. The second clone carries a targeted a-casein allele as indicated by the occurrence of a 848 bp PCR product [T: targeted].
Marker: phage l digested with HindIII and EcoRI (l6H/E). Panel D: Southern blot analysis of EcoRI digested DNA derived from tail clips of a wild-type
(acas [+/+]), and a-casein mutant mice (heterozygous: acas [+/2], and homozygous a-casein [2/2]). The probe indicated in panel B detects a 7.5 kb
DNA fragment representative of the unmodified a-casein allele [U] and a 4.3 kb band representative of the targeted a-casein allele [T]. Panel E: PCR
analysis of genomic DNA isolated from two ES cell clones using the primer combination acas1, acas7 and PGK5 (analysing the 39 end of the
homologous recombination event). A 688 bp band is detected in both samples and represents the unmodified a-casein allele [U]. The second clone
carries a targeted a-casein allele as indicated by the occurrence of a 450 bp PCR product [T]. Marker: NEB PCR marker.
a-Casein Deficient Mice
PLoS ONE | www.plosone.org7 July 2011 | Volume 6 | Issue 7 | e21775
estimation equations (GEE) with compound symmetry for the
working correlation matrix because of the repeated measure-
ments for weight. The comparisons of the average growth rates
between the three groups were computed using the covariance
matrix of the model coefficients. The data was analyzed using
SPSS version 16.0 (SPSS for Windows, Release 16.0.1. 2007.
Chicago: SPSS Inc.). The GEE model was fitted using R (R
Development Core Team. R: A Language and Environment for
Statistical Computing. Vienna, Austria. 2009). The significance
level was set at 0.05.
Generation of a-casein deficient mice
In order to inactivate the a-casein gene (Fig. 1a) a targeting
construct was established based on previous experience with
constructs used to inactivate the b- and c-casein genes [5,27]. The
targeting event removes the complete second exon which includes
the translational start codon and the 15 amino acid signal peptide
(Fig. 1b). In 3 independent transfections into ES cells the targeting
frequency was found to be around 3%. Successfully modified ES
Figure 2. Milk protein analysis. Panel A: SDS-polyacrylamide gel analysis of milk derived from wild-type [+/+], heterozygous [+/2] and
homozygous [2/2] a-casein deficient mice. Milk was purified as indicated in the methods section and defatted whole milk, whey and the casein
fraction were separated on a 10% gel and stained with Coomassie Blue. The sizes of the protein molecular weight markers (New England Biolabs,
broad range protein marker) are indicated. Panel B: SDS-polyacrylamide gel analysis of milk derived from wild-type [+/+], heterozygous [+/2] and
homozygous [2/2] a-casein deficient mice. Defatted whole milk was separated on a 15% gel and stained with Coomassie Blue. The sizes of the
protein molecular weight markers are indicated as is the position of WAP (whey acidic protein). Panel C: SDS-polyacrylamide gel analysis of milk
derived from wild-type [+/+] mice. The proteins identified by mass-spectrometry analysis (shown in table 4) are indicated (bands #1 to 8). Panel D:
SDS-polyacrylamide gel analysis of milk derived from wild-type [+/+] and homozygous [2/2] a-casein deficient mice. Proteins specific to milk from a-
casein deficient mice identified by mass-spectrometry analysis (shown in table 4) are indicated (bands #9 to 13).
a-Casein Deficient Mice
PLoS ONE | www.plosone.org8 July 2011 | Volume 6 | Issue 7 | e21775
cells clones were identified using a PCR analysing the 59 end of the
targeted gene (Fig. 1c). The targeting event was then confirmed by
Southern blot analysis assessing the 59 end of the targeted a-casein
gene (Fig. 1d). The integration event was also confirmed using
PCR analysis and Southern blotting with primer combinations
and probes specific for the 39 end (Fig. 1e). Two targeted cell
clones were grown up and injected into blastocysts. Animals
carrying a homozygous deletion of the a-casein gene are
phenotypically normal prior to lactation.
Altered milk composition in a-casein deficient mice
Milk was isolated from lactating wild-type [+/+], heterozygous [+/
2] and null [2/2] animals, separated on a denaturing protein gel
and analysed by Coomassie Blue staining. Two representative
samples of milk derived from heterozygous [+/2] and null [2/2]
mice are shown are in Fig. 2a alongside a wild-type control milk
previously observed for a-, b- and c-casein (with apparent molecular
weights of 42, 30 and 25 kDa, respectively) . The identity of the
protein bands was confirmed using mass-spectrometry. Table 4
details the data for the 8 most prominent protein bands indicated in
Fig. 2c. These findings were also supported using Western blot
analyses for a-casein and b-casein (Fig. 3a and 3b). Milk from a-
casein deficient mice shows some characteristic protein bands which
are absent from milk of wild-type or heterozygous mice (Fig. 2d).
Mass-spectrometry analysis was used to identify some of these
proteins. The major additional band with an apparent molecular
weight of 78 kDa was identified as grp78/BiP. BiP is an ER resident
protein which is involved in the assembly of multi-protein complexes
like antibodies [29,30]. The identity of the grp78/BiP protein was
confirmed by Western blot analysis (Fig. 3c). In addition two further
ER-resident proteins, grp94 and PDIA6 (protein disulfide-isomerase
associated protein 6), were detected in milk of the a-casein deficient
mice (Fig. 2d, table 4).
Milk protein abundance was then analysed quantitatively using
a Kodak imaging system (Fig. 3d and e). In heterozygous mice the
amount of a-casein protein is reduced by around 50% when
compared to albumin (Fig. 3e). In contrast, the relative amounts of
b-casein and c-casein compared to albumin remain constant. In
the milk derived from homozygous a-casein deficient mice, no a-
casein protein can be detected (Fig. 2a, 3a and 3d). In addition, the
amounts of b- and c-casein are also significantly reduced (Fig. 2a,
3d and e). Whereas the ratio of b-casein to albumin in control milk
is 2:1, the ratio in milk from homozygous a-casein deficient mice is
1:1 indicating at least a 50% reduction in b-casein protein
concentration in milk (Fig. 3d and e). This reduction is also
detected by Western blot analysis (Fig. 3b) Analysis of the data
using one-way ANOVA demonstrated that the changes in the
concentration of b- and c-casein in the milk of a-casein deficient
mice are highly significant (p.0.001; Table 5). The concentration
of milk proteins (Fig. 3d) is consistent with published data .
In a-casein deficient milk the reduction in protein secretion also
affects the mouse WAP protein, which is secreted into the whey
fraction of milk (Fig. 2b). This suggests that all milk proteins
secreted from mammary epithelial cells are affected by the absence
of a-casein. In contrast, secretion of albumin, which is derived
from serum, is not affected by the absence of a-casein (Fig. 2a).
Furthermore, the overall amount of milk secreted by a-casein
deficient dams appears to be reduced (observation during
harvesting). Thus, a-casein deficiency causes a generalised
reduction in milk secretion and a reduction in milk protein
The presence of the ER proteins grp78/BiP, grp94 and PDIA6
in milk of a-casein deficient mice is unexpected. We therefore
analysed whether the grp78/BiP, grp94 and PDIA6 proteins are
up-regulated in lactating mammary tissue of a-casein null mice
using protein gel analysis (Fig. 4a) and Western blotting (Fig. 4c).
In total protein extracts derived from mammary tissue it is evident
that grp78/BiP and grp94 expression is indeed up-regulated in
response to the absence of a-casein. In contrast, the PDIA6
protein can not discerned in Coomassie stained gels of mammary
protein extracts, as it co-localises with other proteins of similar
molecular mass. Grp78, grp94 and PDIA6 are all ER resident
Table 4. Identification of milk proteins using mass-spectrometry.
band protein aaexpected MW reported MWobserved MW MS scorecoveragenumber of sequences
#1 transferrin 69778.8 kDa80 kDa2040 59% 35
#2 albumin576 67 KDa 67 kDa 139162%28
#3 L-AAO-1 523 58.5 kDa59 kDa771 42%16
a-casein 289 33.6 kDa43 kDa 42 kDa1069 38%7
k-casein16017.7 kDa 31 kDa 32 kDa 1349%2
b-casein216 23.6 kDa26 kDa28 kDa512 14%3
c-casein 16919.5 kDa 25 kDa26 kDa 19527%4
d-casein 14415.3 kDa 22 kDa189 46%4
#9grp94 80290.1 kDa 94 kDa100 kDa 59637% 23
#10 grp78/BiP65672.5 kDa78 kDa 76 kDa 2139 47% 29
#11PDIA6 445 48.7 kDa49 kDa 501 31% 10
b-casein14423.6 kDa 26 kDa28/27 kDa441 12%2
c-casein 14415.3 kDa 25 kDa23 kDa 435 46%4
The number of amino acids (aa) and the expected molecular weight (MW) are given for the mature proteins (i.e. without the signal peptide) where appropriate. In
addition, the reported molecular weights for the mouse caseins  and the molecular weights observed in the SDS-PAGE analysis are shown. The MS score is a
measure of confidence of the detected protein species. A score higher than 39 is significant. The number of peptide sequences and the fraction of the total protein
covered are also indicated. Protein bands #1-8 are the predominant protein species in the milk of wild-type mice. Protein bands #9-13 are detected in the milk of a-
casein deficient mice but not in the milk of wild-type or heterozygous mice. L-AAO-1: L-amino acid-oxidase1, PDIA6: protein disulfide isomerase associated 6.
a-Casein Deficient Mice
PLoS ONE | www.plosone.org9 July 2011 | Volume 6 | Issue 7 | e21775
proteins which are not normally secreted from the cells but act as
chaperones, aiding protein folding. Their up-regulation is typically
associated with ER stress [31,32,33]. This suggests that the lack of
a-casein protein in mammary epithelial cells induces ER stress.
Analysis of gene expression rates of grp78, grp94 and PDIA6
confirms that all three genes are significantly up-regulated in tissue
samples of a-casein deficient mice with respect to control mice and
heterozygous mice (Fig. 4d). In contrast, another protein often up-
regulated under conditions of ER stress, REDD1  is not
significantly regulated by the absence of a-casein (Fig. 4d),
Several other protein bands with apparent molecular weights of
28, 27 and 22 kDa were identified in milk from a-casein deficient
mice (Fig. 2d). Mass-spectrometry analysis confirmed that these
proteins are breakdown products of b- and c-casein (Table 4). The
lower molecular weight b-casein products are also detected by
Western blot analysis of milk samples derived from a-casein
deficient mice (Fig. 3b). Analysis of protein extracts from
mammary gland tissue suggests that the breakdown products are
the predominant intracellular forms of b and c-casein (Fig. 4a and
b) but are not secreted as efficiently as the full length protein.
Table 5. Significance of changes in protein ratios.
protein ratio[+/+] vs [+/2][+/2] vs [2/2][+/+] vs [2/2]
bcas/gcas 0.0430 0.65360.8148
Milk protein expression was analysed in wild-type mice [+/+] (n=3),
heterozygous mice [+/2] (n=6) and a-casein deficient mice [2/2] (n=5).
Expression was quantified by densitometric scanning of protein gels and
correlation of the net intensities with a molecular weight marker of known
concentration. The protein ratios of a-casein to albumin (acas/alb.), a-casein to
b-casein (acas/bcas), b-casein to c-casein (bcas/gcas), b-casein to albumin (bcas/
alb.) and c-casein to albumin (gcas/alb.) were determined. The data were
analysed by ANOVA for one way comparisons. The p values obtained for the
different comparisons are shown. P values in bold print are below the cut off
points of 0.01 or 0.001 (as indicated).
Figure 3. Analysis of milk protein expression. Panel A: Western blot analysis of milk derived from wild-type [+/+], heterozygous [+/2] and
homozygous [2/2] a-casein deficient mice. The a-casein protein was detected using a rabbit-anti a-casein antiserum. Panel B: Western blot analysis of
milk derivedfrom wild-type[+/+], heterozygous [+/2] andhomozygous[2/2] a-caseindeficient mice. Theb-casein protein was detectedusinga goat-anti
b-casein antiserum. Panel C: Western blot analysis of milk derived from wild-type [+/+], heterozygous [+/2] and homozygous [2/2] a-casein deficient
mice. The grp78/BiP protein was detected using a goat-anti grp78 antiserum. Panel D: Densitometric analysis of milk protein abundance as detected by
mice(n=5)werescannedinaKodakdensitometerandthenetintensity ofthemilk proteins was comparedwith theintensitiesofmolecularweight marker
proteins of know concentration. Protein concentrations detected in the three genotypes for albumin, a-casein (acas), b-casein (bcas) and c-casein protein
(gcas) are presented. Comparisons of [2/2] vs [+/2] as analysed by one-way ANOVA are significant with p,0.001 (***); Comparisons of [+/2] vs [+/+] are
significant with p,0.05 (*)where indicated. Panel E: The relative amounts of milk protein abundance were calculated for theratios of a-casein to albumin
(acas/alb.), a-casein to b-casein (acas/bcas), b-casein to c-casein (bcas/gcas), b-casein to albumin (bcas/alb.) and c-casein to albumin (gcas/alb.). For
comparisons against wild-type mice in a one-way ANOVA p,0.01 is indicated by **, p,0.001 by ***. Exact P values are presented in table 5.
a-Casein Deficient Mice
PLoS ONE | www.plosone.org 10July 2011 | Volume 6 | Issue 7 | e21775
The caseins are important for ion transport, mainly calcium and
phosphate [35,36]. Not surprisingly we therefore find that the
lower casein protein levels in a-casein deficient mice are
accompanied by a significantly lower concentration of calcium
and phosphate in milk (reduced by around 90% compared to milk
derived from wild-type animals; Fig. 5a).
a-Casein deficiency affects the expression of other casein
We analysed the expression of casein specific mRNAs in the
lactating mammary gland from control and a-casein deficient mice
(Fig. 5b and c). Expression of the casein genes was correlated with
expression of the reference genes GAPDH (shown in Fig. 5b and c)
and b-actin (Fig. 5d). Identical results were obtained for both
reference genes. As expected the amount of a-casein specific
mRNA in heterozygous mice is at 50% of control levels and absent
in homozygous knock-out mice. Expression of the corresponding
b-casein and c-casein casein genes is significantly reduced in a-
casein deficient mice (to around 10% of the expression levels
detected in wild-type and heterozygous mice; Fig. 5b, c and d).
This suggests that the absence of the a-casein protein not only
interferes with protein secretion from the mammary epithelial cells
but that absence of both alleles of the a-casein gene also reduces
the expression of the other casein genes. The expression of the k-
casein gene is also affected but not to the same extent. Its
expression is reduced to around 20% of control levels in a-casein
deficient mice (Fig. 5c and d). No consistent reduction of b-casein,
c-casein and k-casein expression is detected in the heterozygous
animals. Fig. 5d presents the expression of the casein genes as
percentages of wild-type expression. The consistent reduction of b-
, c- and k-casein gene expression in the a-casein deficient
mammary tissue is highly significant (p,0.001; table 6) whereas
expression changes in heterozygous vs wild-type mice are
marginally or not significant (all p values.0.01; table 6).
Figure 4. Analysis of cellular proteins in mammary tissue. Panel A: SDS-polyacrylamide gel analysis of total protein extracts derived from
lactating mammary tissue of wild-type [+/+], heterozygous [+/2] and homozygous [2/2] a-casein deficient mice. Proteins were separated on a 10%
gel and stained with Coomassie Blue. The sizes of the protein molecular weight markers (New England Biolabs, broad range protein marker) are
indicated as are the positions of the b-casein and c-casein proteins (arrows). The grp78/BiP and grp94 proteins and the breakdown products of the b-
casein and c-casein proteins are marked by arrowheads. Panel B: Western blot analysis of milk derived from wild-type [+/+], heterozygous [+/2] and
homozygous [2/2] a-casein deficient mice. The b-casein protein was detected using a goat-anti b-casein antiserum. Panel C: Western blot analysis
of milk derived from wild-type [+/+], heterozygous [+/2] and homozygous [2/2] a-casein deficient mice (9, 31 and 55). The grp78/BiP protein was
detected using a goat-anti grp78 antiserum. Panel D: Correlation of gene expression in wild type [+/+], heterozygous [+/2] and a-casein deficient
mice [2/2] using quantitative PCR. The results for the genes encoding the ER proteins BiP/grp78, PDIA6, grp94 and REDD1 were correlated with the
expression of the reference gene b-actin. Quantification was done in 3 [+/+], 5 [+/2] and 5 [2/2] mice. Statistical analysis using one-way ANOVA
demonstrates that the expression increases for BiP, grp94 and PDIA6 observed in a-casein deficient mice with respect to both wild-type and
heterozygous mice occur with p,0.05. For comparisons against wild-type mice in a one-way ANOVA p,0.05 is indicated by *, p,0.01 by **, and
p,0.001 by ***.
a-Casein Deficient Mice
PLoS ONE | www.plosone.org 11July 2011 | Volume 6 | Issue 7 | e21775
Histological analysis of mammary tissue in a-casein
In order to assess whether the reduction in protein and RNA
expression is correlated with overall morphological changes,
sections of mammary tissue were obtained from animals at peak
lactation and analysed by haematoxylin/eosin (H&E) staining and
immuno-histochemistry (Fig. 6). As expected a-casein expression
was clearly detectable in tissue from wild-type control mice and
heterozygous mice (Fig. 6a). In contrast no a-casein protein could
be detected in sections of a-casein deficient mice (Fig. 6a). As
expected sections analysed with pre-immune serum did not show
any a-casein specific staining (Fig. 6a). No gross morphological
alterations were detected in the sections of the different genotypes
(Fig. 6b). This suggests that the absence of a-casein, although
critical for overall milk composition, does not significantly impact
on the survival of the mammary gland. To address this question
further we measured the expression and activity of capase proteins
in the mammary gland. Firstly we assessed the presence of cleaved
caspase 3 in the protein extracts derived from control mice,
heterozygous mice and a-casein deficient mice. No cleaved
caspase 3 could be detected (Fig. 7a). In contrast cleaved caspase
3 protein with the expected molecular weight of 19 kDa was
readily detected in the extracts of RAW264 cells treated with
Figure 5. Analysis of milk calcium and phosphate levels and milk protein gene expression. Panel A: Calcium and phosphate content of
mouse milk was determined as indicated in the methods section. Concentrations are given in nM. Panel B: Quantitative PCR analysis of a-casein and
b-casein gene expression. cDNA derived from representative wild-type, heterozygous [+/2] and homozygous [2/2] a-casein deficient mice was
analysed using primer pairs specific for a-casein, b-casein and the reference gene GAPDH. Expression of the casein genes was correlated with the
reference gene and is expressed as pg casein/pg GAPDH. Panel C: Quantitative PCR analysis of c-casein and k-casein gene expression. Expression of
the c and k-casein genes was correlated with the reference gene and is expressed as pg casein/pg GAPDH. Panel D: Correlation of casein gene
expression in wild type [+/+], heterozygous [+/2] and a-casein deficient mice [2/2] using quantitative PCR. Casein gene expression was correlated
with the expression of the reference gene b-actin. Quantification of a-casein was done in 3 [+/+], 7 [+/2] and 5 [2/2] mice. Quantification of b-casein
was done in 3 [+/+], 8 [+/2] and 4 [2/2] mice. Quantification of c- and k-casein was done in 3 [+/+], 3 [+/2] and 3 [2/2] mice. Expression in
heterozygous and a-casein deficient mice is presented as percentage of median casein gene expression in wild-type control mice [+/+] (set to 100%).
Error bars represent standard deviations. For comparisons against wild-type mice in a one-way ANOVA p,0.05 is indicated by *, p,0.01 by **, and
p,0.001 by ***. Exact p values are presented in table 6.
Table 6. Significance of changes in casein gene expression.
RNA ratio[+/+] vs [+/2][+/2] vs [2/2][+/+] vs [2/2]
Casein gene expression was measured in wild-type mice [+/+], heterozygous
mice [+/2] and a-casein deficient mice [2/2] using quantitative PCR. The
results were correlated with the expression of the reference gene b-actin.
Quantification of a-casein was done in 3 [+/+], 7 [+/2] and 5 [2/2] mice.
Quantification of b-casein was done in 3 [+/+], 8 [+/2] and 4 [2/2] mice.
Quantification of c- and k-casein was done in 3 [+/+], 3 [+/2] and 3 [2/2] mice.
The data were analysed by ANOVA for one way comparisons. The p values
obtained for the different comparisons are shown. P values in bold print are
below the cut off points of 0.01 or 0.001 (as indicated).
a-Casein Deficient Mice
PLoS ONE | www.plosone.org12July 2011 | Volume 6 | Issue 7 | e21775
10 mM staurosporin (Fig. 7a). Similarly, no specific signals for
cleaved caspase 3 could be detected in immuno-histochemical
analyses of tissue sections. In addition, caspase 3 and caspase 7
activities was measured in cytoplasmic protein extracts derived
from mammary tissue of control, heterozygous and a-casein
deficient mice. No significant differences in caspase 3 and 7 were
detected (Fig. 7b). In contrast a significant increase in caspase
activity could be detected in RAW cells treated with 10 mM
staurosporin (Fig. 7b).
cDNAs derived from representative samples of control and a-
casein deficient mice were analysed for characteristic gene
expression changes related to apoptosis. Potential changes in
candidate genes were assessed using a PCR array containing 84
apoptosis related genes. The expression of 16 genes was changed
significantly (11 down-regulated and 5 up-regulated in a-casein
deficient mice vs. control). Five genes, which were changed by
more than 7 fold (1 down-regulated and 4 up-regulated), were
further analysed in cDNAs derived from 3 control, 5 heterozygous
and 5 a-casein deficient animals (Fig. 7c). Only 3 genes showed
consistent and significant expression differences between control
and a-casein deficient animals. Expression of the anti-apoptotic
genes survivin (baculoviral inhibitor of apoptosis repeat-contain-
Figure 6. Immuno-histochemistry analysis of mammary tissue. Panel A: Paraffin embedded sections of mammary tissue (day 10 lactation)
were analysed using a rabbit-anti a-casein antiserum. The slides were subsequently counterstained with haematoxylin. Representative sections
derived from wild-type [+/+], heterozygous [+/2] and homozygous [2/2] a-casein deficient mice are presented. The lower panels are control
sections incubated with a rabbit pre-immune serum in place of the a-casein specific antiserum. Panel B: Representative sections derived from wild-
type [+/+], heterozygous [+/2] and homozygous [2/2] a-casein deficient mice at day 10 of lactation stained with haematoxylin.
a-Casein Deficient Mice
PLoS ONE | www.plosone.org13 July 2011 | Volume 6 | Issue 7 | e21775
ing-5; Birc5) and nucleolar protein 3 (Nol3/ARC: nuclear
apoptosis repressor with caspase recruitment domain) was
increased by 22 and 2.5 fold respectively. In contrast expression
of another anti-apoptotic gene, TNF receptor-associated factor 1
(Traf1), was reduced to 5% of control levels in a-casein deficient
a-Casein deficient milk restricts the growth of offspring
In order to distinguish between the effect of the genetic
alteration in the offspring and the effect of the modified milk, the
mice were put into 3 groups and cross-fostered such that wild-type
offspring was nursed by a-casein deficient dams (3 litters with a
total of 25 pups; 6 to 11 pups per litter) and heterozygous a-casein
[+/2] offspring was nursed by wild-type dams (3 litters with a total
of 22 pups; 4 to 10 pups per litter). Wild-type offspring nursed by
wild-type dams were used as additional controls (3 litters with a
total of 34 pups; 10 to 13 pups per litter; Table 2).
Weight gain during lactation was significantly reduced in
offspring nursed by a-casein deficient dams (p,0.001 for all
comparisons of offspring nursed by a-casein deficient dams with
the other two groups after day 3). This effect was seen when a-
casein deficient dams nursed their own pups and if the offspring of
wild-type mice was nursed by deficient dams. This demonstrates
that the effect is mediated by the genotype of the nursing female.
This effect was also seen in two further litters after breeding of the
a-casein deficient mice onto a CD1 background (Kolb et al.,
unpublished). All females were able to nurse their pups, as
indicated by the presence of milk in the stomach (‘milk spot’) in all
groups of pups (Table 3). However, by day 6, pups nursed by a-
casein deficient dams were visibly emaciated (Fig. 8a) in
Figure 7. Analysis of markers of apoptosis in mammary tissue from a-casein deficient mice. Panel A: Western blot analysis of samples
derived from two a-casein deficient mice and one heterozygous mouse (all taken at mid-lactation). The protein extracts were separated on a 10%
(upper panel) and 15% (lower panel) polyacrylamide-gel blotted to nitrocellulose and detected using antisera against b-actin (upper panel) and the
cleavage product of caspase 3 (lower panel). Extracts from RAW264 cells treated with 10 mM staurosporin (STS) for 6 h were used as positive control.
The sizes of the protein molecular weight markers (Cell Signaling Technologies, biotinylated protein marker) are indicated as are the positions of the
b-actin and caspase 3 proteins (arrows) Panel B: Analysis of caspase 3 and caspase 7 activity in cytoplasmic extracts of mammary gland tissue of
control [+/+], heterozygous [+/2] and a-casein deficient mice [2/2] using a Caspase-Glo assay (Promega). Extracts derived from RAW264 cells treated
with staurosporin were used as positive control. Panel C: Correlation of gene expression in wild type [+/+], heterozygous [+/2] and a-casein
deficient mice [2/2] using quantitative PCR. Expression of the genes encoding the apoptosis related proteins nucleolar protein 3 (Nol3; up-
regulated), Birc5 (up-regulated) and Traf1 (down-regulated) were correlated with the expression of the reference gene b-actin. Quantification was
done in 3 [+/+], 6 [+/2] and 5 [2/2] mice. Statistical analysis using one-way ANOVA demonstrates that the expression changes for all three genes
observed in a-casein deficient mice with respect to both wild-type and heterozygous mice occur with p,0.05. For comparisons against wild-type
mice in a one-way ANOVA p,0.05 is indicated by *, and p,0.001 by ***.
a-Casein Deficient Mice
PLoS ONE | www.plosone.org14 July 2011 | Volume 6 | Issue 7 | e21775
comparison with pups nursed by control dams (Fig. 8b). By mid-
lactation offspring nursed by a-casein deficient mice are
significantly smaller than offspring nursed by control dams
(Fig. 8c). Pre-weaning mortality was low in all groups and
unaffected by dam genotype. Weighing of individual pups
throughout lactation demonstrates that offspring nursed by a-
casein deficient dams only reach a weight of around 3 g by the end
of lactation, whereas pups nursed by control dams weigh around
12 g (Fig. 8d and Fig. 9a). After weaning mice were maintained on
a standard chow diet ad libitum. Offspring nursed by a-casein
deficient dams display a significantly accelerated growth with
respect to the control group between weaning and day 50 of life
(Fig. 9b and 10). However, this brief growth spurt is insufficient to
bring the weight of pups nursed by a-casein deficient dams up to
the weight of control animals in the long term (Fig. 9c). The
difference in weight is displayed in both sexes (Fig. 9d).
Before the weaning day (day 23), wild-type pups nursed by a-
casein deficient dams had a significantly slower growth rate than
pups (wild-type and heterozygous) nursed by wild-type dams
(p,0.001 for both comparisons). The growth rate of wild-type and
heterozygous pups nursed by wild-type dams 1 and 3 was not
significantly different (p=0.729). After weaning, wild-type pups
nursed by a-casein deficient dams had a significantly higher
growth rate than pups (wild-type and heterozygous) nursed by
wild-type dams (p=0.039 and p=0.014 respectively). The growth
rate of wild-type and heterozygous pups nursed by wild-type dams
was not significantly different (p=0.449) For wild-type pups
nursed by a-casein deficient dams the average growth rate was
0.14 g/day for days 1–23 and 0.52 g/day for days 24–60 (cf.
Fig. 9b). For pups (wild-type and heterozygous) nursed by wild-
type dams the growth rates were for days 1–23 0.64 g/day and
0.63 g/day, respectively, and for days 24–60 0.47 g/day and
0.45 g/day, respectively (Fig. 10).
Physical development of mice nursed by a-casein
In order to assess whether the deficient nutrition during
lactation has an impact on general development, mice in the
three experimental groups were exposed to an early phenotyping
protocol including an adapted version of the general screen of the
SHIRPA assay . Pre-weaning observations indicated physical
impairment in the ‘small’ mice. Measures of maturation (e.g. eye
opening; tables 3 and 7) occur with a slight delay with respect to
control animals. 60% and 50% of the pups nursed by wild-type
dams had both eyes opened on day 14, all of the pups nursed by a-
casein deficient females still had their eyes closed at that age,
opening them only around 3 days later (Table 7). Physical abilities
which rely on physical strength and coordination when tested on
day 21 are significantly less developed in pups reared by a-casein
deficient dams (Table 8). Significantly fewer of the growth
Figure 8. Photographs of experimental animals. Panel A: Photograph of wild-type mice nursed by a-casein deficient [2/2] dams (at age of 6
days). Panel B: Photograph of heterozygous offspring nursed by wild-type dams (at age of 6 days). Panel C: Photograph of mice nursed by wild-
type and a-casein deficient dams at 11 days of age. Panel D: Photograph of mice nursed by wild-type and a-casein deficient dams at 21 days of age.
a-Casein Deficient Mice
PLoS ONE | www.plosone.org15 July 2011 | Volume 6 | Issue 7 | e21775
Figure 9. Impact of a-casein deficient milk on pup growth. Panel A: Growth curve of three different groups of mice during lactation (G1: wild-
type pups nursed by wild-type dams n=34; G2: wild-type pups nursed by a-casein deficient [2/2] dams, n=25; and G3: heterozygous pups nursed
by wild-type dams, n=22). Values shown are +/2 standard deviation. All weight differences between group G2 vs. G1 and G3 were significant from
day 7 (p,0.001) as assessed by ANOVA. Panel B: Percentage weight gain throughout different stages of life for the three experimental groups. The
weight of individual mice was compared on two days (as indicated: e.g. 1/3 corresponds to the interval between day 3 and day 1 of life) and the
percent weight increase was recorded. The average for all mice in the three experimental groups is shown for consecutive time periods. Panel C:
Growth curve of mice in the three groups over the first 6 months of life. Mice nursed by a-casein deficient dams show a consistent growth deficiency.
Panel D: Growth curve of mice in the three groups of mice over the first 6 months of life separated by gender. Error bars represent standard
Figure 10. Model estimates for weight increase across experimental groups. Average growth rates were determined for wild-type pups
nursed by wild-type dams (group 1), wild-type pups nursed by a-casein deficient dams (group 2) and heterozygous pups nursed by wild-type dams
(group 3). The estimated increases were compared to the observed values over the first 60 days of life.
a-Casein Deficient Mice
PLoS ONE | www.plosone.org16 July 2011 | Volume 6 | Issue 7 | e21775
impaired animals (36%, P-value,0.001 Fisher’s exact test) were
able to hold on to the pole in the vertical pole test, whereas none of
the pups nursed by wild-type dams fell off before the pole reached
a vertical position. Some animals nursed by a-casein deficient
dams presented a rough condition of the fur (84%, P-value,0.001
Fisher’s exact test) whereas all control animals had normal fur
condition. Grip strength in growth impaired pups was significantly
lower (P=0.001), as 36% of the animals could not hold on to the
cage lid for 15 seconds when it was turned upside down, compared
to 3% and 0% in the two control groups. Animals nursed by
knockout females defecated less during handling and scoring on
day 21 (P-value,0.001, Mann Whitney U). Also significantly
fewer (76%, P=0.004) pups nursed by a-casein deficient females
showed a provoked biting response at day 21 when compared to
pups nursed by wild-type dams (100%). In summary, at pre-
weaning the a-casein deficient milk suckled pups were smaller and
physically weaker than those reared on wild-type milk.
Subsequently, post-weaning observations clearly indicated that
the ‘small’ mice had, in-effect, caught up with respect to their
relative physical characteristics. Specifically, when the animals
were assessed again at 8 weeks of age, there were no significant
differences between groups for most of the parameters observed
(Table 9). This suggests that in offspring nursed by a-casein
deficient mice, general development of pups is less impaired than
what would be expected from their significant growth retardation
which has persisted throughout their life so far (currently the mice
are almost 2 years old).
The role of caseins in milk protein secretion
Deficiency for a-casein severely impairs milk protein secretion
in mice. In contrast b-casein deficiency does not result in a similar
decline . This suggests that, in contrast to b-casein, a-casein
plays a more critical role in the establishment of a functional casein
micelle thereby affecting secretion of all casein proteins.
Alternatively or additionally due to the higher number of
phosphate centres present in a-casein vs. b-casein (3 vs. 1),
deficiency for a-casein (but not for b-casein) may decrease the
stability of the biofluid resulting in a precipitation of calcium-
phosphate in the Golgi vesicles. These results argue against a
functional redundancy of the calcium-sensitive caseins (a-, b-, c-
and d-casein). Whereas overall secretion of milk proteins is not
adversely affected in heterozygous mice (apart from a 50%
reduction in a-casein protein secretion), milk protein secretion
from mammary epithelial cells is severely curtailed in homozygous
a-casein deficient mice. This affects both, caseins and whey
proteins like WAP. In contrast, secretion of albumin which is not
secreted by the mammary epithelial cells is not affected by the
absence of a-casein. These results suggest a functional role for a-
casein in casein micelle formation and/or stabilisation (similar to
that of k-casein) . We find that the glucose regulated protein 78
(grp78 or BiP) is significantly up-regulated in response to the
absence of a-casein. Grp78/BiP is an endoplasmic reticulum (ER)-
resident protein whose expression is enhanced under conditions of
ER stress  [31,33]. Surprisingly the protein is also found in
milk of a-casein deficient dams. This suggests that grp78/BiP may
be critically involved in the assembly of the casein micelle. One
Table 7. Eye-opening in pups.
group (pups –
dams)day 14 day 15 day 16day 17 day 18 median
G1 (wt – wt) 1814101 day 14
G2 (wt – null)00 10 141 day 17
G3 (het – wt)11 11000 day 14.5
The number of pups opening their eyes on the indicated day are presented.
Table 8. Behavioural differences between pups nursed by
wild-type dams and pups nursed by a-casein deficient dams
during the lactation period; days of assessment: 1, 3, 7, 14 and
Parameterday of assessment change
milk-spot y/n1 ns
inside/outside nest3 ns
milk-spot y/n3 ns
inside/outside nest7 ns
milk-spot y/n7 ns
teeth y/n 14ns
fur y/n 14 ns
grip-reflex 2/4 palms14 ns
ears open y/n14 ns
eyes open y/n14p,0.05
body-posture norm/consp21 ns
respiration norm/consp 21 ns
wounds n/y21 ns
eye condition norm/consp 21 ns
whiskers trimmed n/y21 ns
discharge from nose n/y 21 ns
condition of teeth and mucosa norm/consp 21 ns
colour of skin/ears/paws norm/consp21 ns
tonus of abdomen norm/consp21 ns
tonus in forelegs norm/consp 21ns
tonus in hindlegs norm/consp21 ns
position-reflex empty cage y/n21ns
sense of touch y/n 21ns
vision y/n 21ns
gait norm/consp 21ns
hearing y/n 21 ns
body conformation norm/consp 21ns
tail position norm/consp 21p,0.05
provoked biting y/n 21p,0.05
hanging on grid y/n21p,0.05
fur norm/consp 21p,0.05
faeces weighing cage 21p,0.05
faeces novel environment 21p,0.05
vertical pole test21p,0.05
Significance level p,0.05 for Fisher’s exact test. Y=yes, n=no, ns=not
significant, norm=normal, consp=conspicuous.
a-Casein Deficient Mice
PLoS ONE | www.plosone.org17July 2011 | Volume 6 | Issue 7 | e21775
can speculate that grp78/BiP is associated with the maturing
casein micelle and is co-secreted with the immature micelle into
milk. Two further ER resident proteins, grp94 and PDIA6 are also
secreted into milk of a-casein deficient animals indicating that the
lack of a-casein has significant impacts on protein processing and
transport in the mammary epithelium. Interestingly, secretion of
grp78/BiP and grp94 is also observed in mice over-expressing
human protein C [37,38]. This may suggest that perturbation of
the balance of milk proteins invokes a common ER stress response
resulting in the co-secretion of ER resident proteins into milk.
Despite the clear evidence for changes to the expression and
localisation of ER resident proteins there is no clear evidence for
significant increases in mammary apoptosis in the absence of a-
casein. Firstly, lactation is not terminated prematurely in the
deficient animals suggesting that at least a significant portion of the
mammary epithelial cells is viable. Under physiological conditions
mammary gland involution after weaning leads to a complete
cessation of milk secretion, apoptosis and regression of the
epithelial tissue [39,40]. In homozygous protein C transgenic
mice no successful lactation is established, suggesting that in that
case ER stress is followed by loss of mammary epithelium function
. Secondly, we cannot detect any gross morphological
alteration of the secretory mammary epithelium in a-casein
deficient mice. Thirdly, we also failed to detect increases in caspase
activity in a-casein deficient mammary tissue. Finally, the
apoptosis related genes which showed consistent and significant
increases in mRNA expression in mammary tissue of a-casein
deficient mice include the anti-apoptotic genes survivin (Birc5) and
ARC/Nol3, suggesting that the mammary cells are actively
avoiding apoptosis. On the other hand a reduction in the
expression of the Traf1 gene, encoding an anti-apoptotic gene in
the context of the mammary gland  suggests a balance of pro-
apoptotic and anti-apoptotic responses to a-casein deficiency.
Taken together these findings may suggest that the ER-stress
response in mammary epithelial cells of a-casein deficient mice is
sufficient to rescue protein secretion, thereby preventing large
scale apoptosis and loss of tissue function.
Interestingly, we find that steady state levels of mRNAs
encoding b-casein, c-casein and k-casein are also severely reduced
in the mammary gland of a-casein deficient mice (cf. Fig. 5d). In
contrast no such reduction is observed in heterozygous mice.
There could be (at least) two explanations for this. The deletion of
cis-acting DNA elements in the a-casein gene may affect the
expression of the entire casein gene locus. This genetic explanation
would imply that the presence of one functional a-casein allele is
sufficient to mediate the activation of both copies of the casein
gene locus. Alternatively there may be physiological explanations.
ER stress has been associated with an attenuation of protein
synthesis . One can speculate that in the mammary gland this
leads to a decrease in the transcription of milk protein genes which
account for most of the transcripts during lactation to prevent
overloading of a compromised endoplasmic reticulum . A
decrease in the abundance of milk protein gene specific mRNAs
was also observed in the transgenic mice over expressing human
protein C . Alternatively, the viability of the secretory
mammary epithelial cells may be impaired to some degree by
the ER stress response, leading to a lower relative number of
casein producing cells in the mammary gland.
Not surprisingly a-casein deficiency and the reduced concen-
tration of casein proteins is accompanied by a reduction in the
concentration of both calcium and phosphate in milk. The effect of
a-casein deficiency in mice is similar to that reported for a aS1-
casein deficient goat breed (Cn0) in that total casein secretion is
reduced by around 75% . While there are no data available
which suggest a critical growth delay in the offspring of aS1-Cn0
goats, it is unlikely that an allele that would have a nutritional
impact similar to the inactivation of the a-casein in mice could
have persisted in the natural environment. In addition, the
Table 9. Comparison regarding general health and behaviour between pups nursed by wild-type dams and pups nursed by a-
casein deficient dams.
ParameterP-value G1 vs G2P-value G1 vs G3ParameterP-value G1 vs G2 P-value G1 vs G3
Respiration rate constantconstant Irritability 0.407constant
Piloerection constantconstant Fearconstant 0.407
Palpebral closure constantconstant Gait
Trunk curl constantconstant Grip strength
Visual placingconstant constant Heartrate
Positional passivityconstantconstant Limb tone
Aggression constant constant Tail elevation
Abdominal toneconstant constantUrination jar 0.3830.573
Corneal reflexconstantconstant Urination Makrolon cage0.103 0.103
Skin colour constantconstant Defecation Makrolon cage0.8920.178
Lacrimation constantconstantTouch escape 0.654 0.406
Salivation constantconstantToe pinch 0.0840.623
Negative Geotaxisconstantconstant Wire manoeuvre 0.2640.258
Tremor0.393constantDefecation jar* 0.2920.070
P-values of parameters without significant differences between group 1 (G1: wild-type pups nursed by wild-type dams) and 2 (G2: wild-type pups nursed by a-casein
deficient dams; P-value G1 vs. G2) and group 1 and 3 (G3: heterozygous pups nursed by wild-type dams; P-value G1 vs. G3) at 8 weeks of age (Fisher’s exact test and
*Mann-Whitney-U test). Constant values indicate that all animals in the groups compared had the same, normal, score.
a-Casein Deficient Mice
PLoS ONE | www.plosone.org18 July 2011 | Volume 6 | Issue 7 | e21775
absence of a-casein does not appear to have a critical effect on
calcium secretion in goats . This indicates that there may be
species (and maybe also strain-) specific differences in the assembly
of casein micelles.
The consequences of a-casein deficiency for the
The most critical consequence of a-casein deficiency for the
offspring is a sustained shortfall in body size. Pups nursed by a-
casein deficient dams only reach 25% of body weight of control
animals at weaning and even at later stages in life remain around
33% lighter than control animals. The experiments described in
this paper have been carried out with 3 litters of each genotype on
a C57BL/6 background. However we have observed the same
phenotype of reduced weight gain during lactation with two more
litters on a CD1 background (Kolb et al., unpublished data),
suggesting that the phenotype is similar in different genetic
backgrounds. The observed changes in calcium concentration in
the milk of a-casein deficient animals are unlikely to play a role in
this permanent reduction in body size as calcium restriction during
lactation only exerts a transient growth delay in rodents .
In contrast, protein deficient diets have been implicated as
important regulators of metabolic programming and later health
outcomes in mice and men [43,44,45,46]. Gestational protein
deficiency in the mother leads to intra-uterine growth delay in the
offspring. However, when such offspring is then cross-fostered
onto dams on a control diet, they show a phase of rapid catch up
growth during lactation . The combination of poor foetal and
rapid post-natal growth leads to increased susceptibility for
metabolic disease and reduces life-span [48,49]. In contrast,
nutritional restriction during lactation leads to a growth delay in
rodent pups which is maintained throughout life .
Milk supply in rodents can also be modulated by a number of
approaches [48,50,51]. Variations in litter size lead to differences in
adult body weight, although generally the effects in mice are less
drastic (around 10% in adulthood) than those observed in offspring
nursed by a-casein deficient dams. Alternatively, milk protein
concentration can be manipulated by changing the protein content
of diet of the nursing dams . Specific alterations in milk
composition have been more difficult to achieve technically and have
relied on rearing rodent pups by gastrostomal nutrition away from
their mothers . However, this results in a radically different early
environment as these pups are not exposed to the normal maternal
behaviour . Therefore the effect of specific protein restrictions
during lactation has not been addressed in detail.
The molecular mechanisms by which early nutrition alters body
size permanently are not fully understood at the moment. But the
growth hormone/IGF axis and programming of appetite control
have been implicated as key mediators. Attenuated growth rates in
early post-natal life provide a significant protection against metabolic
disease and extend health- and lifespan [45,51,54,55,56]. Many
genetic mutations of the growth hormone/IGF axis result in both,
small body size and an increased life- and health-span. Increased life-
span and increased resistance to metabolic disease is also observed in
mice nursed by dams on a low protein diet . This suggests that
permanent changes in body size induced by altered protein supply
during lactation are accompanied by an improvement in metabolic
health. In humans formula feeding has been associated with both, a
more rapid increase in body weight during lactation  and an
increased susceptibility to metabolic disorders [46,58,59].
Mice reared on a-casein deficient milk show a marked delay in
the development of abilities which are related to physical
capabilities. However, the deficiencies displayed at weaning had
disappeared at 8 weeks of age. This indicates that these
impairments are transient, whereas the changes in body weight
Thus the phenotype of a-casein deficient dams and their
offspring confirm the critical role of lactation in determining life-
long body size and underline the decisive role of protein supply
during this developmental window. These mice represent a
genetically defined model system of nutritional restriction with a
significant impact on whole animal metabolism. This model can
be exploited to study mechanistic aspects of metabolic program-
ming of body size, resistance to obesity and long-term health
We would like to thank Armando Teixeira Pinto (Department of
Biostatistics and Medical Informatics, Faculty of Medicine, University of
Porto) for statistical advice, and the Roslin Institute Small Animal Unit for
animal care. We also want to thank Norrie Russel (Roslin Institute) for
taking the photographs shown in Figs. 8a–d. We also thank Dr. Carl Holt
(University of Glasgow, UK) for critical reading of the manuscript.
Conceived and designed the experiments: AFK IASO CBAW RCH SGL.
Performed the experiments: AFK RCH SGL AC CJR CN LP. Analyzed
the data: AFK RCH IASO DBS SGL CBAW. Contributed reagents/
materials/analysis tools: AFK. Wrote the paper: AFK RCH IASO CBAW.
1. Rijnkels M (2002) Multispecies Comparison of the Casein Gene Loci and
Evolution of Casein Gene Family. Journal of Mammary Gland Biology and
Neoplasia 7: 327–345.
2. Kawasaki K, Weiss KM (2003) Mineralized tissue and vertebrate evolution: The
secretory calcium-binding phosphoprotein gene cluster. Proc Natl Acad Sci USA
3. Rijnkels M, Kooiman PM, Krimpenfort PJ, de Boer HA, Pieper FR (1995)
Expression analysis of the individual bovine beta-, alpha s2- and kappa-casein
genes in transgenic mice. Biochem J 311: 929–937.
4. Ginger MR, Grigor MR (1999) Comparative aspects of milk caseins. Comp
Biochem Physiol B Biochem Mol Biol 124: 133–145.
5. Kumar S, Clarke AR, Hooper ML, Horne DS, Law AJ, et al. (1994) Milk
composition and lactation of beta-casein-deficient mice. Proc Natl Acad Sci
USA 91: 6138–6142.
6. Shekar PC, Goel S, Rani SD, Sarathi DP, Alex JL, et al. (2006) kappa-casein-
deficient mice fail to lactate. Proc Natl Acad Sci U S A 103: 8000–8005.
7. Holt C (1992) Structure and stability of bovine casein micelles. Adv Protein
Chem 43: 63–151.
8. Holt C, Hasnain SS, Hukins DW (1982) Structure of bovine milk calcium
phosphate determined by X-ray absorption spectroscopy. Biochim Biophys Acta
9. McMahon DJ, Oommen BS (2008) Supramolecular structure of the casein
micelle. J Dairy Sci 91: 1709–1721.
10. Neville MC, Peaker M (1979) The secretion of calcium and phosphorus into
milk. J Physiol 290: 59–67.
11. Boisgard R, Chanat E (2000) Phospholipase D-dependent and -independent
mechanisms are involved in milk protein secretion in rabbit mammary epithelial
cells. Biochim Biophys Acta 1495: 281–296.
12. Pauloin A, Delpal S, Chanat E, Lavialle F, Aubourg A, et al. (1997) Brefeldin A
differently affects basal and prolactin-stimulated milk protein secretion in
lactating rabbit mammary epithelial cells. Eur J Cell Biol 72: 324–336.
of calcium phosphate by casein phosphopeptides. Eur Biophys J 33: 435–447.
14. Holt C (2004) An equilibrium thermodynamic model of the sequestration of
calcium phosphate by casein micelles and its application to the calculation of the
partition of salts in milk. Eur Biophys J 33: 421–434.
15. Holt C, Sorensen ES, Clegg RA (2009) Role of calcium phosphate nanoclusters
in the control of calcification. Febs J 276: 2308–2323.
16. Jenness R, Holt C (1987) Casein and lactose concentrations in milk of 31 species
are negatively correlated. Experientia 43: 1015–1018.
17. Tziboula-Clarke A, Hubert R (2002) goat milk. Encyclopedia of Dairy Sciences.
Oxford: Elsevier. 1270 p.
a-Casein Deficient Mice
PLoS ONE | www.plosone.org19 July 2011 | Volume 6 | Issue 7 | e21775
18. Chanat E, Martin P, Ollivier-Bousquet M (1999) Alpha(S1)-casein is required for Download full-text
the efficient transport of beta- and kappa-casein from the endoplasmic reticulum
to the Golgi apparatus of mammary epithelial cells. J Cell Sci 112: 3399–3412.
19. Kolb AF, Ansell R, McWhir J, Siddell SG (1999) Insertion of a foreign gene into
the beta-casein locus by Cre-mediated site-specific recombination. Gene 227:
20. Marques JM Swedish University of Agricultural Sciences, Uppsala; 2007.
21. Crawley JN (2007) What’s Wrong With My Mouse? Behavioral Phenotyping of
Transgenic and Knockout Mice: 2ndEdn. HobokenNJ: Wiley-Liss. 544 p.
22. Rogers DC, Fisher EM, Brown SD, Peters J, Hunter AJ, et al. (1997) Behavioral
and functional analysis of mouse phenotype: SHIRPA, a proposed protocol for
comprehensive phenotype assessment. Mamm Genome 8: 711–713.
23. Matsukura S, Jones PA, Takai D (2003) Establishment of conditional vectors for
hairpin siRNA knockdowns. Nucleic Acids Res 31: e77.
24. Kolb AF, Maile J, Heister A, Siddell SG (1996) Characterization of functional
domains in the human coronavirus HCV 229E receptor. J Gen Virol 77:
25. Szymanowska M, Hendry KA, Robinson C, Kolb AF (2009) EMMPRIN
(basigin/CD147) expression is not correlated with MMP activity during adult
mouse mammary gland development. J Cell Biochem 106: 52–62.
26. de Roos B, Geelen A, Ross K, Rucklidge G, Reid M, et al. (2008) Identification
of potential serum biomarkers of inflammation and lipid modulation that are
altered by fish oil supplementation in healthy volunteers. PROTEOMICS 8:
27. Robinson C, Kolb AF (2009) Analysis of mammary specific gene locus
regulation in differentiated cells derived by somatic cell fusion. Exp Cell Res 315:
28. Stevenson EM, Leaver J (1994) Chromatographic separation of the proteins of
mouse milk. International Dairy Journal 4: 205.
29. Kozutsumi Y, Normington K, Press E, Slaughter C, Sambrook J, et al. (1989)
Identification of immunoglobulin heavy chain binding protein as glucose-
regulated protein 78 on the basis of amino acid sequence, immunological cross-
reactivity, and functional activity. J Cell Sci Suppl 11: 115–137.
30. Kozutsumi Y, Segal M, Normington K, Gething MJ, Sambrook J (1988) The
presence of malfolded proteins in the endoplasmic reticulum signals the
induction of glucose-regulated proteins. Nature 332: 462–464.
31. Melnick J, Aviel S, Argon Y (1992) The endoplasmic reticulum stress protein
GRP94, in addition to BiP, associates with unassembled immunoglobulin chains.
Journal of Biological Chemistry 267: 21303–21306.
32. Kantor L, Pinchasi D, Mintz M, Hathout Y, Vanderver A, et al. (2008) A Point
Mutation in Translation Initiation Factor 2B Leads to a Continuous Hyper
Stress State in Oligodendroglial-Derived Cells. PLoS ONE 3: e3783.
33. Ron D, Walter P (2007) Signal integration in the endoplasmic reticulum
unfolded protein response. Nat Rev Mol Cell Biol 8: 519.
34. Whitney ML, Jefferson LS, Kimball SR (2009) ATF4 is necessary and sufficient
for ER stress-induced upregulation of REDD1 expression. Biochemical and
Biophysical Research Communications 379: 451.
35. Sawyer L, Holt C (1993) The secondary structure of milk proteins and their
biological function. J Dairy Sci 76: 3062–3078.
36. Holt C, Sawyer L (1988) Primary and predicted secondary structures of the
caseins in relation to their biological functions. Protein Eng 2: 251–259.
37. McManaman JL, Palmer CA, Anderson S, Schwertfeger K, Neville MC (2004)
Regulation of milk lipid formation and secretion in the mouse mammary gland.
Adv Exp Med Biol 554: 263–279.
38. Palmer CA, Lubon H, McManaman JL (2003) Transgenic mice expressing
recombinant human protein C exhibit defects in lactation and impaired
mammary gland development. Transgenic Res 12: 283–292.
39. Flint DJ, Boutinaud M, Whitelaw CB, Allan GJ, Kolb AF (2006) Prolactin
inhibits cell loss and decreases matrix metalloproteinase expression in the
involuting mouse mammary gland but fails to prevent cell loss in the mammary
glands of mice expressing IGFBP-5 as a mammary transgene. J Mol Endocrinol
40. Sorrell DA, Szymanowska M, Boutinaud M, Robinson C, Clarkson RW, et al.
(2005) Regulation of genes encoding proteolytic enzymes during mammary
gland development. J Dairy Res 72: 433–441.
41. Cao Y, Karin M (2003) NF-kappaB in mammary gland development and breast
cancer. J Mammary Gland Biol Neoplasia 8: 215–223.
42. Krukowski M (1987) Calcium deficiency during lactation and in the first two
weeks after weaning: decreased ash and increased magnesium in bone of rat
pups. Bone 8: 251–257.
43. Bellinger L, Lilley C, Langley-Evans SC (2004) Prenatal exposure to a maternal
low-protein diet programmes a preference for high-fat foods in the young adult
rat. Br J Nutr 92: 513–520.
44. Brawley L, Itoh S, Torrens C, Barker A, Bertram C, et al. (2003) Dietary protein
restriction in pregnancy induces hypertension and vascular defects in rat male
offspring. Pediatr Res 54: 83–90.
45. Ozanne SE, Lewis R, Jennings BJ, Hales CN (2004) Early programming of
weight gain in mice prevents the induction of obesity by a highly palatable diet.
Clin Sci (Lond) 106: 141–145.
46. Koletzko B, von Kries R, Closa R, Escribano J, Scaglioni S, et al. (2009) Lower
protein in infant formula is associated with lower weight up to age 2 y: a
randomized clinical trial. Am J Clin Nutr 89: 1836–1845.
47. Hales CN, Ozanne SE (2003) The dangerous road of catch-up growth. J Physiol
48. Levin BE (2006) Metabolic imprinting: critical impact of the perinatal
environment on the regulation of energy homeostasis. Philosophical Transac-
tions of the Royal Society B: Biological Sciences 361: 1107–1121.
49. McMillen IC, Robinson JS (2005) Developmental Origins of the Metabolic
Syndrome: Prediction, Plasticity, and Programming. Physiol Rev 85: 571–633.
50. Velkoska E, Cole TJ, Dean RG, Burrell LM, Morris MJ (2008) Early
undernutrition leads to long-lasting reductions in body weight and adiposity
whereas increased intake increases cardiac fibrosis in male rats. J Nutr 138:
51. Kappeler L, De Magalhaes Filho C, Leneuve P, Xu J, Brunel N, et al. (2009)
Early postnatal nutrition determines somatotropic function in mice. Endocri-
nology 150: 314–323.
52. Beierle EA, Chen MK, Hartwich JE, Iyengar M, Dai W, et al. (2004) Artificial
rearing of mouse pups: development of a mouse pup in a cup model. Pediatr Res
53. Chatterjee D, Chatterjee-Chakraborty M, Rees S, Cauchi J, de Medeiros CB,
et al. (2007) Maternal isolation alters the expression of neural proteins during
development: ‘Stroking’ stimulation reverses these effects. Brain Res 1158:
54. Berryman DE, Christiansen JS, Johannsson G, Thorner MO, Kopchick JJ
(2008) Role of the GH/IGF-1 axis in lifespan and healthspan: lessons from
animal models. Growth Horm IGF Res 18: 455–471.
55. Cripps RL, Martin-Gronert MS, Archer ZA, Hales CN, Mercer JG, et al. (2009)
Programming of hypothalamic neuropeptide gene expression in rats by maternal
dietary protein content during pregnancy and lactation. Clin Sci (Lond) 117:
56. Kappeler L, De Magalhaes Filho CM, Dupont J, Leneuve P, Cervera P, et al.
(2008) Brain IGF-1 receptors control mammalian growth and lifespan through a
neuroendocrine mechanism. PLoS Biol 6: e254.
57. Ziegler EE (2006) Growth of breast-fed and formula-fed infants. Nestle Nutr
Workshop Ser Pediatr Program 58: 51–59; discussion 59–63.
58. Koletzko B (2006) Long-term consequences of early feeding on later obesity risk.
Nestle Nutr Workshop Ser Pediatr Program 58: 1–18.
59. Koletzko B, von Kries R, Monasterolo RC, Subias JE, Scaglioni S, et al. (2009)
Can infant feeding choices modulate later obesity risk? Am J Clin Nutr 89:
a-Casein Deficient Mice
PLoS ONE | www.plosone.org20 July 2011 | Volume 6 | Issue 7 | e21775