Laboratory of Genetics and
Institute of Diabetes and
Digestive and Kidney
Institutes of Health,
Bethesda, Maryland 20892,
Correspondence to L.H.
A secretory organ in mammals
that produces milk during
lactation to feed the young. The
gland is composed of alveoli,
ducts and a stromal
Lateral aggregates of cholesterol
and sphingomyelin that are
thought to occur in the plasma
(permeant to molecules up to
1 kDa) between adjacent cells,
which is composed of
12 connexin protein subunits,
six of which form a connexon
or hemichannel contributed by
each of the coupled cells.
INFORMATION NETWORKS IN THE
Lothar Hennighausen and Gertraud W. Robinson
Abstract | Unique developmental features during puberty, pregnancy, lactation and post-
lactation make the mammary gland a prime object to explore genetic circuits that control the
specification, proliferation, differentiation, survival and death of cells. Steroids and simple
peptide hormones initiate and carry out complex developmental programmes, and reverse
genetics has been used to define the underlying mechanistic connections.
Organogenesis requires a sequence of cellular processes
that involve commitment to a specific cell fate, prolifer-
ation of committed progenitor cells, the initiation and
implementation of differentiation programmes and the
maintenance of tissue homeostasis by the controlled
turnover of cells. In the MAMMARY GLAND, these events
are strictly regulated by steroid and peptide hormones.
A number of reviews cover this subject in depth1–4.
Over the past two decades, studies with numerous
strains of genetically altered mice have resulted in the
characterization of an integrated framework of signal-
ling networks that control the normal development
BOX 1 and neoplastic conversion of mammary tissue.
Concepts have emerged as to how mammary cells have
acquired ancient signalling pathways and tailored them
to satisfy the diverse needs of the developing tissue
during puberty, pregnancy, lactation and regression of
the gland in involution. In particular, it has become
clear that various cytokine receptors signal through a
limited set of tyrosine kinases and transcription factors.
Recent studies have also identified ‘brakes’ that stop the
flow of information emanating from the cell-surface
receptors. These include molecules from the suppres-
sor of cytokine signalling (SOCS) family and, possibly
more unexpectedly, LIPID RAFTS. The role of structural
components such as GAP JUNCTIONS has long been under-
appreciated but, as it turns out, such components are
also crucial in mammary cell differentiation.
Tissue culture cells, primary cells and genetically
altered mice have been used over the past two decades
to dissect pathways that control mammary develop-
ment and cellular differentiation. Although the use of
primary cultures and tissue culture cell lines has shed
light on the mechanistic cues that underlie signalling
pathways, this approach could not be expected to reca-
pitulate the developing architecture and physiological
intricacies that occur during pregnancy, and final
confirmation still relies on the use of animal models.
More than 100 genes have been shown to control vari-
ous aspects of mammary physiology, from the develop-
ment of the mammary ANLAGE to remodelling of the
gland during involution. Whereas some of these genes
fit into pathways that can mechanistically explain their
function, the position of others in signalling networks
remains to be determined. Owing to space restrictions,
this review focuses on the signalling pathways that
function during pregnancy. This developmental win-
dow has been studied most intensively and a picture
is emerging that links hormones to specific molecules
that carry out their function.
Two genetic approaches have been used to explore
proteins that control mammary development: the
ectopic expression of genes in transgenic animals and
the inactivation of genes from the mouse genome.
In the first approach researchers ask which cellular
events can be modulated by a given protein, whereas
the second approach directly addresses the function of
a protein in vivo. This review was written according to
the motto, ‘don’t ask what a protein can do for a cell but
ask what a cell can or cannot do without it’.
Mammary structure and cellular composition
Two tissue compartments constitute the mammary
gland: the epithelium, which consists of DUCTS and
NATURE REVIEWS | MOLECULAR CELL BIOLOGY
VOLUME 6 | SEPTEMBER 2005 | 715
The embryonic primordium of
An epithelial structure
Small ducts that branch off a
major duct during ductal
Peptide hormones produced by
the placenta during pregnancy.
These hormones bind to the
prolactin receptor and are
involved in the induction of
proliferation and functional
differentiation of luminal cells.
Connective tissue supporting
the epithelial compartment of
the mammary gland. The
prevalent cell types are fat cells
(adipocytes) in addition to
fibroblasts, vasculature and
composed of two cell types
surrounding a central lumen.
The luminal cells synthesize and
secrete milk components and
the basal cells are contractile.
milk-producing alveolar cells; and the STROMA, or con-
nective tissue, which is also called the mammary fat
pad (FIG. 1; BOX 1. In general, the epithelial cells form
ducts and ALVEOLI with a central lumen that opens
to the body surface through the nipple. By contrast,
monotremes have mammary glands without a nip-
ple or central lumen and the ducts open directly to
a confined area known as the milk patch. Most epi-
thelial cells are LUMINAL, SECRETORY CELLS, which undergo
functional differentiation in pregnancy to produce
milk. They are encased by a mesh-like system of BASAL,
MYOEPITHELIAL CELLS, which are contractile and partici-
pate in the delivery of milk. The extensive system of
ducts and alveoli is embedded in the stroma, the main
components of which are ADIPOCYTES, but fibroblasts,
cells of the haematopoietic system, blood vessels and
neurons are also present.
The existence of distinct cell lineages that are
derived from an elusive mammary STEM CELL has been
proposed and models of how these lineages develop
have been defined5–7. It has been proposed that the cells
forming the epithelial compartment of the mammary
gland are derived from mammary stem cells (MSCs),
which have the capacity to self-renew and give rise to
committed epithelial precursor cells (EPCs; FIG. 2). The
progeny of EPCs then becomes restricted to a ductal
or alveolar fate. The ductal precursor cells (DPs) form
basal cells and luminal cells, the two cell types that con-
stitute ducts. During pregnancy, alveoli are generated
from alveolar precursors (APs), which give rise to basal
and luminal cells — the differentiated, milk-producing
cells. The alveolar epithelium expands during preg-
nancy, secretes milk during lactation and undergoes
apoptosis and remodelling during involution. Loss
of the prolactin signal after suckling is stopped leads
to massive death of luminal cells in a process called
involution. This restores a ductal system that contains
multipotent and committed ductal and luminal precur-
sor cells (FIG. 2). The presence of stem cells is the basis
of the profound capacity for alveolar renewal in each
subsequent pregnancy. Stem cells have the capacity to
renew themselves and also give rise to progenitor cells,
which are destined for either a basal or a luminal fate.
The ability of progenitor cells to function as a source
for alveolar development was highlighted by experi-
ments in which small sections of a duct generated an
entire ductal tree when transplanted into a cleared fat
pad8–10 BOX 2.
Growing ducts in pubescent animals have conspicuous,
club-shaped structures at their tips, which are known as
terminal end buds (TEB; FIG. 1a). These are sites at which
cells divide at a high rate to advance progression of the
ducts into the fat pad11. They disappear once the entire fat
pad has been filled with ducts (FIG. 1b). Two morphologi-
cally distinct cell types can be found in the TEBs — an
outer layer of cap cells and the more centrally located
body cells. The former gives rise to basal cells whereas
luminal cells are derived from body cells.
In postnatal mammary tissue, most epithelial cells
express receptors for oestrogen and progesterone to
enable these hormones to stimulate ductal outgrowth
and branching. Part of the action of oestrogen derives
from the induction of progesterone receptors (PRs).
In the pubescent gland, PRs are expressed in pro-
liferating cells12, and in virgin mice (female mice
that have not been mated) that are treated with high
doses of oestrogen and progesterone (to mimic preg-
nancy), proliferating cells are preferentially localized
adjacent to PR-positive cells13. Both oestrogen and
progesterone have pleiotropic actions in the uterus,
ovaries and the hypothalamic–pituitary axis in regu-
lating sexual development. In the mammary gland,
oestrogen and progesterone control ductal outgrowth
and alveolar expansion, respectively, namely by regu-
lating cell proliferation and cellular turnover in the
Oestrogen. Oestrogen binds to two distinct receptors,
oestrogen receptor (ER) α and ERβ, which are encoded
by two genes. Like other steroid receptors, these are
members of the large family of nuclear receptors,
which function as transcription factors when bound to
the steroid hormone. Whereas deletion of ERβ has no
adverse effects on ductal and alveolar development14,
both stromal and epithelial ERα are required for nor-
mal ductal elongation and outgrowth during puberty15.
By contrast, ERα is dispensable for pregnancy-mediated
Progesterone. There are also two isoforms — A and
B — of the PR, which display differential activities
as transcription factors. These isoforms are encoded
by two transcripts derived from the same gene by
Box 1 | Life and death of a mammary gland
The mammary gland forms as an appendage of the skin and has its evolutionary
origin in skin glands. The number and location of glands vary among different
classes of mammals. In mice, five pairs of glands develop along a line that runs
slightly ventral to the limb buds, whereas only one pair develops in the thoracic
region in humans. Development of the mammary gland commences in the foetus.
The initial cues that induce the formation of small buds on the ventral surface of
foetuses are not known. Sequential and reciprocal signals between the epithelium
and surrounding mesenchyme (the embryonic stroma) direct the outgrowth of a
small duct into deeper layers of the dermal mesenchyme and formation of the nipple,
the opening for milk removal. Through further elongation and bifurcation a small
ductal system forms that associates with the subdermal fat pad. During puberty the
cyclical production of ovarian oestrogen and progesterone accelerates ductal
outgrowth and branching. In the mature animal, the entire fat pad is filled with a
regularly spaced system of primary and secondary ducts that are decorated with
SIDE BRANCHES, which form and disappear during each oestrous cycle. Proliferation
and maturation of the alveolar compartment occurs during pregnancy and is
controlled mainly by prolactin and PLACENTAL LACTOGENS. At term, the mammary
gland reaches maturity and produces and secretes milk to support the young. In
mice, mammary tissue produces milk equivalent to 20% of the body weight of the
dam. At the end of lactation, the loss of suckling stimuli and the pressure build-up on
cessation of milk removal initiates a remodelling programme called involution. This
causes massive cell death, the collapse of the alveoli and the remodelling of the
epithelial compartment to restore a simple ductal structure again. A new round of
alveolar expansion, maturation and lactation is initiated with the next pregnancy.
716 | SEPTEMBER 2005 | VOLUME 6
b Mature virgin
ERBB4 ligands, RANK-L
LUMINAL, SECRETORY CELLS
Cells lining the lumen of alveoli.
They synthesize milk proteins
and secrete milk.
BASAL, MYOEPITHELIAL CELLS
Cells that surround the luminal
cells as a layer. They contain
smooth muscle actin and
contract in response to oxytocin
to mediate milk let-down
through the main ducts to the
A fat cell.
Cells that have the capacity for
self renewal and generation of
differentiating daughter cells.
differential initiation of transcription. Alveolar
development is perturbed in the combined absence
of both isoforms17, but there is no adverse effect on
development if only PR-A is deleted18, indicating that
the PR-B form is required to carry out the prolifera-
tive effects of progesterone on mammary epithelial
cells19. It should be noted that the PR is essential for
the expansion of the alveolar compartment, but its
contribution to ductal elongation and branching is
only minor. PR-expressing cells are evenly spaced
in the ducts of young mice and their distribution
becomes mosaic in mature virgin mice and during
early pregnancy20,21. In early pregnancy, PR-positive
cells are found closely apposed to proliferating cells,
which implies that the proliferative effect of PR is
mediated in part through PARACRINE activities, as shown
by mixing experiments. PR-negative cells, which by
themselves cannot develop into functional alveoli, can
participate in the development of alveoli when they are
in close proximity to wild-type cells22. Progesterone
seems to induce the production of a signal that induces
the proliferation of neighbouring cells. One candidate
for this activity is receptor activator of nuclear factor
κB (NF-κB)-ligand (RANK-L, formerly called osteo-
protegerin)19, a molecule that belongs to the tumour
necrosis factor (TNF) family and is an important
regulator of OSTEOCLAST development23.
Processing information during pregnancy
Three characteristic and temporally coordinated cel-
lular events, which are regulated by hormones, occur
during pregnancy. These are the proliferation of alveo-
lar epithelium, its differentiation and its survival. This
developmental programme can be initiated and car-
ried out by a single peptide hormone, prolactin (PRL),
which is produced mainly by the LACTOTROPHS in the
anterior pituitary gland. Over the past decade, several
other CYTOKINES, including RANK-L24 and ligands of the
epidermal growth factor (EGF) family25, have joined
the ranks of ‘inducers of mammary development’. The
transcription factors signal transducer and activator
of transcription-5a (STAT5A) and STAT5B (referred
to collectively as STAT5) are shared downstream
mediators of peptide hormones that signal through the
prolactin receptor (PRLR) and ERBB4, and they are
the central switch controlling proliferation, differentia-
tion and survival of mammary cells (see below). The
activation and nuclear localization of STAT5 and the
expression of genes that encode milk proteins define
the differentiation programme during pregnancy
Prolactin and placental lactogens. In 1929, Gruyter and
Strijker injected pituitary extract from lactating rab-
bits into virgin rabbits and showed that this undefined
mixture could induce mammary development and lac-
tation (for historical references, see REF. 3). Four years
later, Riddle and his colleagues identified the active
component and named it ‘prolactin’ after its function
in promoting lactation. Prolactin has two essential
roles in reproduction — the maintenance of the corpus
luteum during early pregnancy and the induction of
mammary development. Through the maintenance
of the CORPUS LUTEUM, PRL ensures the secretion of
oestrogen and progesterone, which themselves are
required for ductal and alveolar development, respec-
tively (as outlined above). After mid-pregnancy, both
functions of PRL are replaced by PLACENTAL LACTOGENS,
but PRL takes over again after birth. PRL is essential
for maintaining lactation, as evidenced by lactational
failure after blocking PRL secretion with dopamine
antagonists such as bromocriptine. Although the main
source of PRL is the lactotrophs, local production of
PRL by mammary epithelium has been reported26,
where it functions as a paracrine mediator of mam-
mary epithelial development27.
PRL mediates its action through PRLR, a trans-
membrane protein of the class I cytokine receptor
family28. PRLR dimers undergo a conformational
change following ligand binding, and the associated
Janus kinase-2 (JAK2) phosphorylates specific tyro-
sine residues in the PRLR, which allows docking and
Figure 1 | Mouse mammary gland development during puberty, pregnancy and
lactation. Schematic (Aa–d) and wholemount (Ba–d) presentation of the different stages and
the principal hormones that control development. A rudimentary ductal design within the
mammary fat pad is visible at birth, which grows at the same rate as the animal until the onset
of puberty. During puberty, the cyclical production of ovarian oestrogen and progesterone
promotes and accelerates ductal outgrowth (Aa, Ba). At this stage, conspicuous club-shaped
structures (terminal end buds (TEB)), where the highest levels of cell division occur, appear at
the ductal tips. In the mature virgin, the entire fat pad is filled with a regularly spaced system of
primary and secondary ducts, with side branches that form and disappear in each oestrous
cycle (Ab, Bb). Hormonal changes that occur when pregnancy begins (the release of
prolactin, placental lactogens and progesterone) increase cell proliferation and the formation
of alveolar buds (Ac, Bc), which grow and differentiate into milk-secreting alveoli at the end of
pregnancy (Ad, Bd). During lactation, alveoli are fully matured and the luminal cells synthesize
and secrete milk components into the lumina. RANK-L, receptor activator of nuclear factor κB
NATURE REVIEWS | MOLECULAR CELL BIOLOGY
VOLUME 6 | SEPTEMBER 2005 | 717
Describing, or relating to, a
regulatory cell that secretes an
agonist into intercellular spaces
from which it diffuses to a
target cell other than the one
that produces it.
A mesenchymal cell that can
differentiate into a bone-
Prolactin-producing cells in the
anterior lobe of the pituitary
Initially identified as regulatory
polypeptides secreted by
immune cells, this family also
molecules such as growth
hormone and prolactin.
A temporary endocrine gland
in the ovaries that produces
A state in which loss of only one
of two alleles of a gene
detectably disables its function.
activation of STAT5. As well as STAT5, PRLR can
signal through the mitogen-activated protein kinase
(MAPK) pathway and others that are dependent on
JAK2. In addition to the full-length form (the long
form), other isoforms of the PRLR have been detected.
These include a short form that lacks the cytoplasmic
domain except for 23 amino acids (see below).
Inactivation of the genes encoding PRL and its
receptor, and the expression of transgenes encoding
mutant PRLR, have helped to elucidate the role of this
cytokine — at least during early stages of pregnancy.
Female mice with inactive PrlR alleles had non-func-
tional corpora lutea and so could not maintain preg-
nancies29. This, however, could be partially overcome
with progesterone treatment. Although ductal out-
growth during puberty was overtly normal in these
mice, no functional alveolar compartment was formed
during pregnancy, as a result of a lack of luminal cell
proliferation29. Transplantation experiments of mutant
mammary tissue into wild-type hosts BOX 2 estab-
lished that the proliferation and differentiation defect
was autonomous to the epithelial compartment and
not the result of altered systemic hormone levels30,31.
A threshold level of PRLR is required for nor-
mal development, as evidenced by a genetic HAPLO
INSUFFICIENCY29. In the presence of only one PrlR gene,
alveolar proliferation and differentiation was stalled in
the second half of pregnancy. This developmental block
was partially alleviated after several pregnancies29,31.
Kelly and colleagues explored the molecular basis of
this by studying the expression of a transgene encod-
ing a short version of PRLR that lacks the cytoplasmic
domain except for 23 amino acids32. Expression of this
transgene restored mammary gland architecture during
pregnancy, as well as STAT5 activation, the expression
of milk proteins and the ability to nurse pups success-
fully. This short form of the PRLR can bind to JAK2
and activate MAPK activity, but does not contain any
PRLR docking sites. So the restoration of proliferation
by this short form of PRLR during pregnancy triggered
differentiation even when only half of the STAT5 dock-
ing sites were present. This concept is supported by
the fact that mice with only one functional Stat5 allele
but with two functional PrlR alleles develop functional
mammary tissue and can lactate33.
Successful lactation depends on a pulsatile release
of PRL from the pituitary gland. Galanin, a 29-amino-
acid peptide, is a mitogen for lactotrophic cells, and
mice with an inactivated galanin gene have reduced
PRL levels during pregnancy, which causes an inabil-
ity to nurse pups34. Lactation in these mice could be
restored by PRL application35, which further supports
the hierarchy of these two hormones. In addition to
regulating PRL concentration by controlling lac-
totrophic cells, galanin also influences the mammary
epithelium directly. The pregnancy-induced alveolar
architecture and milk-protein gene expression was
not completely rescued in mice carrying two inactive
galanin alleles35. However, similar to PRL, galanin
induced the differentiation of mammary epithelium
in vitro, as evidenced by the activation of STAT5.
It was originally proposed that PRL instructs the
proliferation and differentiation of mammary epithe-
lium through mechanisms that are specific to the PRLR
in inducing luminal-cell-specific genes. However,
experiments with hybrid receptors that contain the
ligand-binding domain of the PRLR and the intracel-
lular domain of the erythropoietin receptor (EPOR),
which can recruit STAT5, have led to a revision of this
model36. This hybrid receptor is activated by PRL and
the signal, which is conveyed through the cytoplasmic
domain of the EPOR and STAT5, is sufficient to restore
the phenotype of PRLR-null mammary epithelium.
Current evidence indicates that PRL — and prob-
ably other cytokines — represents a generic cue that
activates transcriptional programmes that are shared
between several cytokine receptors. Although these
programmes might contain some cell-specific compo-
nents, they seem to be of a general nature, mediating
responses such as proliferation and cell survival. As
discussed later, STAT5 and JAK2 are shared among
many cytokine receptors and it seems that they convert
the signal from a cell-specific receptor into a generic
Distinct roles of ERBB tyrosine kinase family members.
Neuregulins, heregulins and other polypeptides that
are related to EGF control vital cellular functions. Their
effects are mediated by four distinct receptors from
the ERBB family. The EGF receptor (ERBB1), ERBB2,
ERBB3 and ERBB4 are unique receptor tyrosine kinases
Figure 2 | Cell lineages in mammary epithelium. The mammary gland is a derivative of the
ectoderm, which also gives rise to the skin and other appendages as well as the
neuroectoderm. Models, supported by experimental evidence, have been developed that
propose the existence of a mammary stem cell (MSC) and distinct cell lineages that lead to the
formation of different cell types in mammary tissue5. It has been suggested that the multipotent
MSCs give rise to epithelial precursor cells (EPCs), the progeny of which develop into either
ductal or alveolar cells. Myoepithelial cells and luminal cells are formed from ductal precursors
(DP) as the ducts grow out postnatally, particularly during puberty. On initiation of pregnancy,
alveolar precursor cells (AP) give rise to myoepithelial and luminal cells, the latter of which
synthesize and secrete milk. After lactation, the alveolar cells are subject to programmed cell
death during the process of involution. A simple ductal system containing multipotent (yellow)
and committed ductal (green) and luminal (orange) precursor cells persists that will develop into
a fully functional epithelium in subsequent pregnancies.
718 | SEPTEMBER 2005 | VOLUME 6
a Transplant into cleared fat pad
b Transplant without clearing
A defective protein that retains
interaction capabilities and so
competes with normal proteins,
thereby impairing protein
that can undergo homo- and heterodimerization, and
activate several signalling pathways. An essential role
for ERBB2 in a subset of human breast cancers has
been recognized for more than 15 years37, and the
inhibition of ERBB2 signalling by the monoclonal
antibody Herceptin to treat certain cases of metastatic
breast cancers has highlighted its significance in cel-
lular transformation38. A role for this family of tyrosine
kinase receptors in normal mammary development
and function has only recently emerged.
Both ERBB1 and ERBB4 are necessary for mam-
mary development during pregnancy, although they
exert their effects in different compartments. The first
indication for a contribution of ERBB1 came from
waved-2 mice, in which ErbB1 is mutated. These mice
show impaired alveolar development39. For normal
mammary development, the presence of ERBB1 within
the stroma is required — an absence of ERBB1 from
the epithelium does not abrogate pregnancy-medi-
ated alveologenesis31,40. As ERBB1 can be activated by
several ligands, the ablation of individual EGF-related
hormones leads to no, or only partial, developmental
defects, implying that there is considerable functional
redundancy between these ligands41.
Additional evidence that members of the ERBB
family contribute to mammary function has come
from transgenic mice expressing a DOMINANTNEGATIVE
form of ERBB2 REF. 42. These mice fail to develop a
functional alveolar compartment during pregnancy. As
this protein lacks specificity and can form dimers with
all ERBB-family members, the identity of the relevant
ERBB2-dimerizing partner that was responsible for the
phenotype could not be identified. Two lines of genetic
evidence now point to ERBB4 as the molecule in ques-
tion. Two research teams used different strategies to
bypass the embryonic lethality of ERBB4-null mice25,43
and derived similar conclusions. ERBB4-null foetuses
die at embryonic day 11 (E11) owing to heart defects,
and Golding and colleagues rescued this lethality with
a transgene that targeted ERBB4 expression specifi-
cally to cardiac myocytes43. Rescued ERBB4-null mice
were fertile but failed to nurse their litters. Histological
Box 2 | Mammary epithelial transplants
Frequently, the deletion of a gene in the mouse influences the development and function of more than one organ.
In particular, the absence of hormone receptors also affects ovarian function, which makes it difficult to distinguish
between direct and indirect effects on mammary gland development. Furthermore, some mutants are not viable and
so mammary epithelial development cannot be studied directly. These problems can be circumvented by
transplanting mammary epithelial cells into a wild-type host8,80. In a 3-week-old mouse the epithelial ducts are still
confined to the most proximal part of the fat pad near the nipple (see figure, part a). This area can be removed
surgically to leave a ‘cleared fat pad’ into which epithelial cells from another (gene-deleted) animal can be
transplanted. The transplanted epithelium (indicated in dark pink) can then develop in a wild-type stroma where it is
exposed to the hormonal milieu of a normal animal. If the endogenous epithelium is left in place (see figure, part b),
the mutant (dark pink) and wild-type (purple) epithelia will develop in the same fat pad under identical conditions,
allowing side-by-side comparisons of both epithelia by ‘in situ’ analyses such as histology, immunohistochemistry or
in situ hybridization.
NATURE REVIEWS | MOLECULAR CELL BIOLOGY
VOLUME 6 | SEPTEMBER 2005 | 719
Mid-pregnancy Start of lactation
A tool for cell-specific gene
deletion. It uses Cre
bacteriophage P1, which
mediates intra- and
recombination between loxP
The functional unit in the
mammary gland that produces
analysis of mammary tissue at delivery revealed that
the alveolar epithelium had undergone proliferation
and expansion, and that its differentiation status,
as measured by the expression of milk proteins, was
within normal limits. However, in contrast to control
tissue, lipid droplets accumulated within the luminal
cells, which implied that these cells had failed to secrete
milk efficiently, the hallmark of a fully differentiated
cell. These studies could not distinguish between a
direct effect of the absence of ERBB4 in mammary
epithelial cells and an indirect, systemic effect. By
deleting the ERBB4 gene specifically in mammary
epithelium using CRELOXPMEDIATED RECOMBINATION,
Jones and colleagues resolved this issue25. Similar to
the germline ERBB4 deletion, alveolar units in these
mice failed to undergo complete functional differen-
tiation and some reduction in alveolar proliferation
and expansion was also observed. So the absence of
ERBB4 in mammary epithelium has a direct effect
on the cells. Like the PRLR, stimulation of ERBB4
activates STAT5 to convey growth and differentiation
signals in the alveolar luminal cells. However, neither
study could detect activated STAT5 in ERBB4-null
alveolar epithelium at birth. This could indicate that
ERBB4, through the activation of STAT5, has a more
prominent role in the functional luminal cell during
lactation than PRL does.
Interpreting multiple languages
JAK2 and STAT5. In evolutionary terms, STAT5 is an
ancient transcription factor that responds to biochemi-
cal cues from a plethora of diverse cytokines, including
growth hormone, erythropoietin, interleukins, PRL
and ligands of the EGF family. Originally identified as
a ‘mammary gland factor’ (MGF) that is activated by
PRL and binds to promoter sequences in milk-protein
genes44, STAT5 has now taken centre stage in many
cytokine signalling pathways45. STAT5A and STAT5B
show 96% similarity at the amino-acid level and are
encoded by two juxtaposed genes46,47. Mice from which
Stat5a or Stat5b or both genes have been inactivated
show biochemical alterations in many cell types. These
alterations are consistent with those arising from
inactivations or mutations of the respective activating
cytokines and their receptors45.
Inactivation of Stat5a in mice caused no overt phe-
notype, except for the failure to lactate48, suggesting a
partial compensation through STAT5B. Proliferation
and expansion of the alveolar compartment was only
slightly reduced but the luminal cells failed to undergo
functional differentiation to produce milk. Analysis of
mammary tissue devoid of both STAT5A and STAT5B49
has confirmed that these two transcription factors dis-
play partially redundant functions in mammary devel-
opment50. As mice in which both STAT5 genes had been
targeted were infertile owing to non-functional corpora
lutea49 — similar to the PRLR-null mice29 — pregnancy-
mediated mammary development therefore had to be
investigated in tissue from STAT5A/B-null mice that
was transplanted into immunocompromised, but oth-
erwise wild-type, mice BOX 2. The complete absence
of lobuloalveolar development, probably as a result of
the lack of proliferation, was comparable to that seen
in the absence of PRLR or JAK2 REFS 13,29,31,50,51. The
use of mice in which the entire Stat5 locus was flanked
by loxP sites provided further insight into the multiple
roles of STAT5 throughout pregnancy33. As expected,
Cre-mediated inactivation of Stat5 in MAMMARY ALVEO
LAR progenitor cells resulted in the same physiological
consequences that were observed when the gene was
ablated in the germline. By contrast, loss of Stat5 late
in pregnancy, after the epithelium has initiated the
maturation programme, led to premature cell death,
indicating a role for STAT5 in cell survival. STAT5
has now emerged as a crucial switch not only for the
proliferation and differentiation of mammary luminal
Figure 3 | Progression of differentiation during
pregnancy. Several prognostic markers of mammary
epithelial cell differentiation are acquired gradually in the
course of pregnancy. At mid-pregnancy (left images) the
gland consists of immature alveoli (arrows in part a) with a
small lumen that is surrounded by cuboidal luminal cells.
Moderate levels of STAT5A, the transcription factor that
mediates signalling from the prolactin receptor (PRLR), are
found in the majority of luminal cells (red staining in part c). In
some of the alveoli, small amounts of the milk protein whey
acidic protein (WAP) are made and secreted into the lumen
(red staining in part e). At term, the alveoli are large (arrows in
part b) and filled with milk, indicated by the pink staining
material and lipid droplets (arrowhead in part b). Strong
nuclear STAT5A staining (red stain in part d, arrowheads)
reflects extensive signalling through the PRLR and ERBB4,
and can be used as an indicator of active gene transcription.
Large amounts of WAP (and other milk proteins) are
synthesized and secreted into the lumen at the apical
membrane (red stain in f). The cell adhesion molecule
E-cadherin (green staining) outlines the luminal epithelial cells
in parts c–f. STAT, signal transducer and activator of
720 | SEPTEMBER 2005 | VOLUME 6
Specialized domains in the
plasma membrane that form
small invaginations and are
involved in vesicular trafficking
and cell signalling.
cells, but also as a regulator of cell survival and func-
tion during lactation.
Although the individual Stat5 knockouts show dis-
tinct phenotypes, this does not mean that STAT5A and
STAT5B have inherently different properties — instead
it reflects differences in the expression patterns of the
two genes. For example, STAT5A is the predominant
isoform in mammary tissue and STAT5A-, but not
STAT5B-, null mice show blunted mammary develop-
ment. By contrast, STAT5B is the predominant isoform
in liver and so STAT5B-null mice show defects in this
tissue52. STAT5 can be activated in the mammary
epithelium through the phosphorylation of a specific
tyrosine residue by JAK2 and ERBB4. JAK2 seems to be
the crucial activator of STAT5, at least during the early
stages of pregnancy, as evidenced by the mammary
phenotype of JAK2-null mammary epithelium51,53,54.
The concept that PRL and placental lactogens are
the key hormones that activate STAT5 and thereby
induce a developmental programme leading to the
production of milk-secreting cells was firmly estab-
lished until the finding that ERBB4 also conveys its
information through STAT5. An intriguing interpre-
tation would be that both PRLR and ERBB4 control
distinct and possibly overlapping features of alveolar
development. Clearly, the PRLR is absolutely essential
for alveolar proliferation in early pregnancy, whereas
the loss of ERBB4 affects the functional differentia-
tion of luminal cells after the proliferative phase. Yet
only the specific inactivation of the PRLR in differen-
tiated luminal cells will address its relevance during
Modulating the flow of information
The almost identical nature of mammary defects after
deletion of PrlR, Jak2 or Stat5 illustrates the impor-
tance of this cascade in luminal cell proliferation and
differentiation. Although it is known how hormones
activate this pathway, the nature of the ‘brakes’ that
allow modulation and cessation of these signals has so
far been an enigma. Two main components and distinct
mechanisms to avoid either precocious or sustained
JAK–STAT signalling during pregnancy have now been
discovered. These are mediated by members of the
SOCS family of proteins and, unexpectedly, caveolin-1
(CAV1), a structural component of lipid rafts.
SOCS members as brakes. The eight members of
the SOCS family (SOCS1–7 and cytokine-inducible
Src-homology-2 (SH2) protein (CIS)) interact with
JAKs and cytokine receptors to curtail the activation
of STAT proteins55. SOCS1, which is linked to the
modulation of interferon signalling in the immune
system, has also been shown genetically to be an
essential regulator of PRL signalling in the mammary
epithelium during pregnancy56. Socs1-null mice have
immunological defects, but these can be overcome by
simultaneously deleting Ifn1 genes. Such mice showed
excessive alveolar proliferation during pregnancy and
elevated STAT5 activity during lactation, implying that
SOCS1 inhibits PRL signalling. As mentioned above,
luminal cell proliferation and differentiation during
the first pregnancy is incomplete in the presence of
only one PrlR allele. An equivalent loss of one SOCS1
allele in PRLR hemizygous mice restored functional
alveolar development and STAT5 activity, which
elegantly shows the role of SOCS1 in modulating PRL
In vitro studies have shown that SOCS molecules
bind to tyrosine residues in cytokine receptors and
block the binding and activation of STAT molecules.
Expression of SOCS3 is controlled by STAT5 and
its levels are induced during pregnancy57, which
indicates that it is part of a negative-feedback loop.
SOCS3 binds to tyrosine residues in the gp130 recep-
tor, which is the shared subunit of receptors for
cytokines such as interleukin-6 (IL-6) and leukaemia
inhibitory factor (LIF). Activation of gp130 signalling
in the mammary epithelium occurs during involution
and leads to the activation of STAT3, the presence of
which is required for involution58–61. Therefore, it can
be predicted that SOCS3 has a role in modulating
gp130-controlled remodelling during involution.
Socs3-null embryos die because of a placental defect
at mid-pregnancy (~12 days of embryonic develop-
ment)62,63, at which point it is too early to rescue the
mammary epithelium by transplantation. But inacti-
vation of Socs3 specifically in mammary epithelium
using Cre-loxP-mediated recombination established
a role for SOCS3 in the remodelling of mammary
tissue during involution (M. Pacher-Zavisin, G.W.R.
and L.H., unpublished observations), indicating that
it is part of the cytokine signalling pathway that is
activated by the gp130 receptor60.
CAV1. The discovery that CAV1 is required for con-
trolling STAT5 activity provided an unexpected twist
to cytokine signalling64. CAV1 is an essential structural
component of CAVEOLAE. The loss of both Cav1 alleles
results in precocious mammary gland development
during pregnancy and concomitant precocious acti-
vation of STAT5. Molecular analyses have established
that CAV1 abrogates PRL-induced gene expression by
sequestering JAK2, which therefore cannot activate
STAT5. These studies highlight that the compartmen-
talization of components within the cell controls their
availability on cytokine stimulation.
Executing STAT5 signals
Inactivation of the genes encoding PRLR, ERBB4, JAK2
and STAT5 has highlighted a role for this pathway in
the proliferation, survival and differentiation of mam-
mary epithelium. Implementing these programmes is
expected to involve several proteins, and the removal
of a single STAT5 target gene will not recapitulate the
lesions resulting from the absence of STAT5. Several
types of STAT5 target gene have been identified so far
(FIG. 4): genes that encode proteins that merely reflect
the differentiation status of the cell; those encoding
proteins that promote cell proliferation and control
cell survival; a gene encoding a growth factor and one
encoding a transcription factor.
NATURE REVIEWS | MOLECULAR CELL BIOLOGY
VOLUME 6 | SEPTEMBER 2005 | 721
A small evagination from a
main duct that forms during the
oestrous cycle and at the
initiation of pregnancy.
A conserved structural motif in
kinase domains that needs to be
phosphorylated for full
activation of the kinase.
RANK-L, RANK, cyclin D1 and IGF-2. The RANK-L
gene, which encodes an osteoclast differentiation fac-
tor from the TNF family, is under the control of PRL
itself65 and is a bona fide STAT5 target. This indicates
that PRL, a growth factor by itself, can activate pro-
duction of another growth factor. A role for RANK-L
and its receptor RANK in mammary gland develop-
ment was established in mice from which either of
the two genes was deleted24. These mice were unable
to nurse their young because of an inhibition of
alveolar development during pregnancy. In mam-
mary epithelial cells from these mice, activation of
the anti-apoptotic molecule AKT/protein kinase
B (PKB) was reduced, alveolar cell death during
pregnancy was increased and cell proliferation was
reduced. However, ALVEOLAR BUDS still formed, which
implied that RANK-L is required for later steps of
alveolar development, such as differentiation. One
of the downstream events in RANK signalling is the
activation of NF-κB. The existence of this link in the
mammary gland was established by mutating IκB
kinase-α (IKKα), a subunit of inhibitor of NF-κB
(IκB)66. Mutations of serine residues in the ACTIVATION
LOOP rendered IKKα inactive and severely impaired
NF-κB activity. Mammary epithelium in these mice
expanded during pregnancy but was unable to func-
tionally differentiate53,66. The concomitant reduction
in cyclin D1 expression observed during pregnancy
in this mutant identified this cell-cycle regulator as a
crucial downstream target of NF-κB signalling. This
was supported by evidence that expression of a cyclin
D1 transgene rescued epithelial differentiation and
the lactation defect in these mice53,66.
The cyclin D1 gene is activated not only by the
RANK–NF-κB pathway but also by STAT5 directly
through a γ-IFN-activated site (GAS) within the pro-
moter. Loss of cyclin D1 leads to a paucity of alveolar
cells, which also fail to functionally differentiate67,68.
Insulin-like growth factor-2 (IGF-2) has been
proposed as another mediator of PRL signalling69.
Its expression can be induced in primary mammary
cells by PRL, and ectopic expression of IGF-2 can
partially rescue pregnancy-mediated expansion of
PRLR-null mammary epithelial cells69. IGF-2-null
mice can nurse their pups, which shows that this
growth factor is not essential for normal mammary
development per se. However, a transient mammary
epithelial growth defect in mid-pregnancy has been
observed, which was alleviated at term, indicating
that IGF-2 might have a modulatory role during a
specific time window.
Junctional integrity. The establishment of functional
alveoli depends on the polarization of the luminal
cells and the formation of junctions between them.
The gene encoding connexin-26 (Cx26), an essential
integral component of gap junctions, is a direct target
gene of STAT5 REF. 70. It is expressed during preg-
nancy and its loss is not compensated for by other
connexins. Its essential role in mammary develop-
ment has been shown in cell-specific knockout
mice71. Specific deletion of Cx26 in the mammary
epithelium led to a high level of cell death, causing
a failure to nurse pups. Like Cx26, the Cx32 gene
is also controlled by STAT5, but its expression is
confined to the lactation stage. Loss of Cx32 did not
alter mammary function, which can be explained by
a compensatory function of connexin-26 REF. 70.
Milk proteins. The genes for milk proteins, the
expression of which defines a mature and differenti-
ated mammary luminal cell, are the classic targets of
STAT5 in the mammary epithelium, and their tran-
scriptional stimulation during pregnancy is mediated
by GAS sequences within their promoter regions72,73.
In the absence of both STAT5A and STAT5B, none
of the milk proteins, including whey acidic protein
(WAP) and β-casein, is expressed50. A considerable
reduction can be observed in the absence of only
Figure 4 | Processing of cytokine information during pregnancy. Cytokine signalling
through the prolactin receptor (PRLR) and ERBB4 activates signal transducer and activator of
transcription-5 (STAT5). On PRL binding, the conformation of the PRLR dimer changes and the
associated Janus kinase JAK2 phosphorylates PRLR on specific tyrosine residues. When
STAT5 binds to these residues, JAK2 phosphorylates one tyrosine residue on STAT5, inducing
its dimerization and nuclear translocation. Suppressor of cytokine signalling-1 (SOCS1) binds
to PRLR and negatively regulates STAT5 signals. STAT5 dimers bind to specific sites in the
promoters of target genes and induce their transcription. Bona fide target genes include those
encoding SOCS3, RANK-L (receptor activator of nuclear factor κB (NF-κB)-ligand), connexin-26,
milk proteins, cyclin D1 and the transcription factor ELF5. SOCS3 negatively regulates cytokine
signalling through gp130-containing receptors, which are active mainly during involution60.
RANK-L is secreted and activates the RANK–NF-κB pathway (right-hand cell); connexin-26 is
a component of gap junctions; and cyclin D1 promotes mammary epithelium proliferation.
ELF5 is a transcription factor that, together with STAT5, induces transcription of whey acidic
protein, a prominent milk component. STAT5 itself activates the transcription of several milk-
protein genes. Milk components, including micelles, are assembled in the Golgi apparatus, and
vesicles (blue circles) are secreted into the alveolar lumen. JAK2 can also associate with
caveolin-1 (CAV1). IKKα, inhibitor of NF-κB (IκB) kinase-α; P, phosphate.
722 | SEPTEMBER 2005 | VOLUME 6
Survival pathways. Little is known about the path-
ways that convey luminal cell survival. Among the
members of the BCL2 family, Bcl-X is by far the most
prominently expressed in the mammary epithelium
and its gene is induced during pregnancy74. In sev-
eral in vitro systems, Bcl-X is controlled by STAT5
through a GAS motif in the promoter75. However,
inactivation of Bcl-X specifically in mammary epi-
thelium did not abrogate luminal cell survival dur-
ing pregnancy, but instead resulted in accelerated
remodelling after cessation of lactation74.
ELF5. DNA microarray analyses have identified ELF5
as a protein that is preferentially expressed in mam-
mary tissue and the expression of which is strongly
induced during pregnancy76 — this implies that it
could be under the control of PRL signalling70. ELF5
is a member of the ETS FAMILY of transcription factors
and its expression is specific to epithelial cells of EXO
CRINE GLANDS35,77. Elf5-null foetuses die at embryonic
day 7.5, which points to a crucial function during
embryogenesis. Mice that have only one functional
Elf5 allele showed abrogated pregnancy-mediated
alveologenesis78. Notably, the lesions were similar
to those observed in the absence of PRLR29, STAT5
REFS 33,50 or JAK2 REFS 51,54. Alveolar buds formed
but failed to proliferate and mature, as indicated by
the absence of milk proteins. Although circumstantial,
this points to the Elf5 gene being a bona fide STAT5
target. It is intriguing that STAT5 as a transcription
factor not only activates genes that control specific fea-
tures of epithelial cells, such as cell proliferation and
differentiation, but also a gene that encodes another
transcription factor, which itself could control dif-
ferentiation programmes. ETS transcription factors
modulate the expression of the milk protein gene
WAP during pregnancy, but not during lactation, as
shown in transgenic mice79, and it can be postulated
that they activate other genes in mammary epithelium
in synergy with STAT5.
A picture is emerging of how peptide hormones
control the genesis of a functional mammary gland
during pregnancy (FIG. 4; BOX 3. It is now firmly
established that, on ligand binding, the receptors
convey their signals through tyrosine kinases and
the transcription factor STAT5. The timing, intensity
Box 3 | Key proteins during pregnancy
The following proteins control the proliferation, differentiation, survival and function of mammary secretory
epithelial cells during pregnancy.
• Progesterone: steroid hormone synthesized by the corpus luteum.
• Prolactin (PRL): peptide hormone produced preferentially in lactotrophic cells in the pituitary gland but also in
mammary epithelial cells.
• EGF family: a class of growth factors structurally related to epidermal growth factor.
• Receptor activator of nuclear factor κ κB (NF-κ κB)-ligand (RANK-L): osteoclast differentiation factor from the
tumour necrosis factor (TNF) family.
• Prolactin receptor (PRLR): the PRLR is activated by the peptide hormones prolactin (PRL) and placental lactogens.
• ERBB4: member of the EGF (ERBB) receptor family. ERBB4 is a tyrosine kinase and forms homodimers and
heterodimers with other members of the ERBB family.
• RANK: receptor activator of NF-κ κB.
• JAK2: Janus kinase (JAK) 2 is a tyrosine kinase associated with type 1 cytokine receptors, including the PRLR,
growth hormone receptor and the erythropoietin receptor. It phosphorylates specific tyrosine residues on itself,
the receptors, and on STAT3 and STAT5.
• STAT5: member of the family of signal transducers and activators of transcription.
• ELF5: member of the ETS family.
• Cyclin D1.
• Connexin-26 and connexin-32: members of the connexin family of proteins that form gap junctions.
• WAP: the whey acidic protein (WAP) is a prominent protein in mice and its transcription is controlled to a large
extent by STAT5, which binds in the promoter region. In addition, an ETS site (possibly recognized by ELF5) in the
WAP gene promoter is necessary for the temporal regulation of WAP expression during pregnancy.
An anti-apoptotic protein. It is
the founding member of the
BCL2 family of pro- and anti-
Proto-oncogene family related
to v-ets, one of the oncogenes of
the acutely transforming avian
erythroblastosis virus E26.
A gland of epithelial origin that
secretes (directly or through a
duct) onto an epithelial surface.
A genetic analysis that proceeds
from phenotype to genotype by
positional cloning or
NATURE REVIEWS | MOLECULAR CELL BIOLOGY
VOLUME 6 | SEPTEMBER 2005 | 723
and duration of these signals are modulated by SOCS
proteins and caveolin. The genes activated by STAT5
encode proteins that control cell proliferation, gap
junction components, growth factors, members
of the SOCS family, milk proteins and at least one
other transcription factor, which, in conjunction
with STAT5, can activate the transcription of milk-
protein genes. Experiments from many laboratories
have identified additional genes that are either acti-
vated or suppressed during pregnancy. The chal-
lenge ahead is to place these genes and the proteins
they encode into the larger picture of mammary
development and function. Although essential,
traditional mouse genetics will not be the only pil-
lar for the understanding of the physiological sig-
nificance of each protein in the puzzle of mammary
gland development. It will be necessary to introduce
in vitro approaches that use primary cells or cell
lines combined with FORWARD GENETICS. Only when
it is possible to rescue the absence of a gene in the
mammary epithelium by introducing a gene encod-
ing a downstream target will it be possible to define
their position in the signalling network. Towards
this goal, it will be necessary to develop primary
cells that can form a functional epithelium in vivo
after being genetically manipulated in vitro.
1. Imagawa, W., Bandyopadhyay, G. K. & Nandi, S.
Regulation of mammary epithelial cell growth in mice and
rats. Endocr. Rev. 11, 494–523 (1990).
Topper, Y. J. & Freeman, C. S. Multiple hormone
interactions in the developmental biology of the mammary
gland. Physiol. Rev. 60, 1049–1106 (1980).
Hennighausen, L. & Robinson, G. W. Signaling pathways
in mammary gland development. Dev. Cell 1, 467–475
Hennighausen, L. & Robinson, G. W. Think globally, act
locally: the making of a mouse mammary gland. Genes
Dev. 12, 449–455 (1998).
Smalley, M. & Ashworth, A. Stem cells and breast cancer:
a field in transit. Nature Rev. Cancer 3, 832–844 (2003).
Provides an accurate and up-to-date introduction
into the field of mammary stem cells.
Wagner, K. U. et al. An adjunct mammary epithelial cell
population in parous females: its role in functional
adaptation and tissue renewal. Development 129,
Smith, G. H. & Boulanger, C. A. Mammary epithelial stem
cells: transplantation and self-renewal analysis. Cell Prolif.
36, 3–15 (2003).
Deome, K. B., Faulkin, L. J., Bern, H. A. & Blair, P. B.
Development of mammary tumors from hyperplastic
alveolar nodules transplanted into gland-free mammary fat
pads of female C3H mice. Cancer Res. 19, 515–520 (1959).
Daniel, C. W., De Ome, K. B., Young, J. T., Blair, P. B. &
Faulkin, L. J., The in vivo life span of normal and
preneoplastic mouse mammary glands: a serial
transplantation study. Proc. Natl Acad. Sci. USA 61,
10. Kordon, E. C. & Smith, G. H. An entire functional
mammary gland may comprise the progeny from a single
cell. Development 125, 1921–1930 (1998).
Provides evidence for the existence of lineage-
committed cells in mammary epithelium.
11. Humphreys, R. C. et al. Apoptosis in the terminal endbud
of the murine mammary gland: a mechanism of ductal
morphogenesis. Development 122, 4013–4022 (1996).
12. Zeps, N., Bentel, J. M., Papadimitriou, J. M. &
Dawkins, H. J. Murine progesterone receptor expression in
proliferating mammary epithelial cells during normal
pubertal development and adult estrous cycle. Association
with ERα and ERβ status. J. Histochem. Cytochem. 47,
13. Grimm, S. L. et al. Disruption of steroid and prolactin
receptor patterning in the mammary gland correlates with
a block in lobuloalveolar development. Mol. Endocrinol.
16, 2675–2691 (2002).
14. Forster, C. et al. Involvement of estrogen receptor β in
terminal differentiation of mammary gland epithelium. Proc.
Natl Acad. Sci. USA 99, 15578–15583 (2002).
15. Bocchinfuso, W. P. et al. Induction of mammary gland
development in estrogen receptor-α knockout mice.
Endocrinology 141, 2982–2994 (2000).
16. Mueller, S. O., Clark, J. A., Myers, P. H. & Korach, K. S.
Mammary gland development in adult mice requires
epithelial and stromal estrogen receptor α. Endocrinology
143, 2357–2365 (2002).
17. Lydon, J. P. et al. Mice lacking progesterone receptor
exhibit pleiotropic reproductive abnormalities. Genes Dev.
9, 2266–78 (1995).
18. Mulac-Jericevic, B., Mullinax, R. A., DeMayo, F. J., Lydon,
J. P. & Conneely, O. M. Subgroup of reproductive functions
of progesterone mediated by progesterone receptor-B
isoform. Science 289, 1751–1754 (2000).
19. Mulac-Jericevic, B., Lydon, J. P., DeMayo, F. J. &
Conneely, O. M. Defective mammary gland
morphogenesis in mice lacking the progesterone receptor
B isoform. Proc. Natl Acad. Sci. USA 100, 9744–9749
20. Silberstein, G. B., Van Horn, K., Shyamala, G. &
Daniel, C. W. Progesterone receptors in the mouse
mammary duct: distribution and developmental regulation.
Cell Growth Differ. 7, 945–952 (1996).
21. Ismail, P. M., Li, J., DeMayo, F. J., O’Malley, B. W. & Lydon,
J. P. A novel LacZ reporter mouse reveals complex
regulation of the progesterone receptor promoter during
mammary gland development. Mol. Endocrinol. 16,
Shows that the introduction of a lacZ gene into the
endogenous progesterone gene locus allows the
analysis of the expression of the progesterone
22. Brisken, C. et al. A paracrine role for the epithelial
progesterone receptor in mammary gland
development. Proc. Natl Acad. Sci. USA 95, 5076–5081
23. Theill, L. E., Boyle, W. J. & Penninger, J. M. RANK-L and
RANK: T cells, bone loss, and mammalian evolution.
Annu. Rev. Immunol. 20, 795–823 (2002).
24. Fata, J. E. et al. The osteoclast differentiation factor
osteoprotegerin-ligand is essential for mammary gland
development. Cell 103, 41–50 (2000).
25. Long, W. et al. Impaired differentiation and lactational
failure of Erbb4-deficient mammary glands identify ERBB4
as an obligate mediator of STAT5. Development 130,
26. Vonderhaar, B. K. Prolactin involvement in breast cancer.
Endocr. Relat. Cancer 6, 389–404 (1999).
27. Naylor, M. J., Lockefeer, J. A., Horseman, N. D. &
Ormandy, C. J. Prolactin regulates mammary epithelial cell
proliferation via autocrine/paracrine mechanism. Endocrine
20, 111–114 (2003).
28. Boutin, J. M. et al. Cloning and expression of the rat
prolactin receptor, a member of the growth hormone/
prolactin receptor gene family. Cell 53, 69–77 (1988).
29. Ormandy, C. J. et al. Null mutation of the prolactin receptor
gene produces multiple reproductive defects in the mouse.
Genes Dev. 11, 167–178 (1997).
30. Brisken, C. et al. Prolactin controls mammary gland
development via direct and indirect mechanisms.
Dev. Biol. 210, 96–106 (1999).
31. Gallego, M. I. et al. Prolactin, growth hormone, and
epidermal growth factor activate Stat5 in different
compartments of mammary tissue and exert different and
overlapping developmental effects. Dev. Biol. 229,
32. Binart, N., Imbert-Bollore, P., Baran, N., Viglietta, C. &
Kelly, P. A. A short form of the prolactin (PRL) receptor is
able to rescue mammopoiesis in heterozygous
PRL receptor mice. Mol. Endocrinol. 17, 1066–1074
33. Cui, Y. et al. Inactivation of Stat5 in mouse mammary
epithelium during pregnancy reveals distinct functions in
cell proliferation, survival, and differentiation. Mol. Cell Biol.
24, 8037–8047 (2004).
34. Wynick, D. et al. Galanin regulates prolactin release and
lactotroph proliferation. Proc. Natl Acad. Sci. USA 95,
35. Naylor, M. J. et al. The neuropeptide galanin augments
lobuloalveolar development. J. Biol. Chem. 278,
36. Brisken, C., Socolovsky, M., Lodish, H. F. & Weinberg, R.
The signaling domain of the erythropoietin receptor
rescues prolactin receptor-mutant mammary epithelium.
Proc. Natl Acad. Sci. USA 99, 14241–14245 (2002).
Shows that the intracellular domains of the EPO and
PRL receptors display equivalent functions, and that
the EPO- and PRL-specific responses depend on the
37. Slamon, D. J. et al. Studies of the HER-2/neu proto-
oncogene in human breast and ovarian cancer. Science
244, 707–712 (1989).
38. Fischer, O. M., Streit, S., Hart, S. & Ullrich, A. Beyond
Herceptin and Gleevec. Curr. Opin. Chem. Biol. 7,
39. Fowler, K. J. et al. A mutation in the epidermal growth
factor receptor in waved-2 mice has a profound effect on
receptor biochemistry that results in impaired lactation.
Proc. Natl Acad. Sci. USA 92, 1465–1469 (1995).
40. Wiesen, J. F., Young, P., Werb, Z. & Cunha, G. R. Signaling
through the stromal epidermal growth factor receptor is
necessary for mammary ductal development.
Development 126, 335–344 (1999).
41. Luetteke, N. C. et al. Targeted inactivation of the EGF and
amphiregulin genes reveals distinct roles for EGF receptor
ligands in mouse mammary gland development.
Development 126, 2739–2750 (1999).
42. Jones, F. E. & Stern, D. F. Expression of dominant-negative
ErbB2 in the mammary gland of transgenic mice reveals a
role in lobuloalveolar development and lactation.
Oncogene 18, 3481–3490 (1999).
43. Tidcombe, H. et al. Neural and mammary gland defects in
ErbB4 knockout mice genetically rescued from embryonic
lethality. Proc. Natl Acad. Sci. USA 100, 8281–8286
44. Wakao, H., Gouilleux, F. & Groner, B. Mammary gland
factor (MGF) is a novel member of the cytokine regulated
transcription factor gene family and confers the prolactin
response. EMBO J. 13, 2182–2191 (1994).
45. Ihle, J. N. The Stat family in cytokine signaling. Curr. Opin.
Cell Biol. 13, 211–217 (2001).
46. Liu, X., Robinson, G. W., Gouilleux, F., Groner, B. &
Hennighausen, L. Cloning and expression of Stat5 and an
additional homologue (Stat5b) involved in prolactin signal
transduction in mouse mammary tissue. Proc. Natl Acad.
Sci. USA 92, 8831–8835 (1995).
47. Cui, Y. et al. The Stat3/5 locus encodes novel
endoplasmic reticulum and helicase-like proteins that are
preferentially expressed in normal and neoplastic
mammary tissue. Genomics 78, 129–134 (2001).
48. Liu, X. et al. Stat5a is mandatory for adult mammary gland
development and lactogenesis. Genes Dev. 11, 179–186
49. Teglund, S. et al. Stat5a and Stat5b proteins have
essential and nonessential, or redundant, roles in cytokine
responses. Cell 93, 841–850 (1998).
50. Miyoshi, K. et al. Signal transducer and activator of
transcription (Stat) 5 controls the proliferation and
differentiation of mammary alveolar epithelium. J. Cell Biol.
155, 531–542 (2001).
51. Shillingford, J. M. et al. Jak2 is an essential tyrosine kinase
involved in pregnancy-mediated development of
mammary secretory epithelium. Mol. Endocrinol. 16,
52. Udy, G. B. et al. Requirement of STAT5b for sexual
dimorphism of body growth rates and liver gene
expression. Proc. Natl Acad. Sci. USA 94, 7239–7244
724 | SEPTEMBER 2005 | VOLUME 6
53. Shillingford, J. M. et al. Proteotyping of mammary tissue Download full-text
from transgenic and gene knockout mice with
immunohistochemical markers: a tool to define
developmental lesions. J. Histochem. Cytochem. 51,
54. Wagner, K. U. et al. Impaired alveologenesis and
maintenance of secretory mammary epithelial cells in Jak2
conditional knockout mice. Mol. Cell Biol. 24, 5510–5520
References 29, 33, 48–51 and 54 describe the role of
the PRL receptor, JAK2 and STAT5 in the
development and differentiation of mammary
epithelium during pregnancy and lactation.
55. Kubo, M., Hanada, T. & Yoshimura, A. Suppressors of
cytokine signaling and immunity. Nature Immunol. 4,
56. Lindeman, G. J. et al. SOCS1 deficiency results in
accelerated mammary gland development and rescues
lactation in prolactin receptor-deficient mice. Genes Dev.
15, 1631–1636 (2001).
Shows for the first time that SOCS1 can modulate
PRL signalling during pregnancy.
57. Tonko-Geymayer, S. et al. Regulation and function of the
cytokine-inducible SH-2 domain proteins, CIS and
SOCS3, in mammary epithelial cells. Mol. Endocrinol. 16,
58. Kritikou, E. A. et al. A dual, non-redundant, role for LIF as a
regulator of development and STAT3-mediated cell death
in mammary gland. Development 130, 3459–3468 (2003).
59. Humphreys, R. C. et al. Deletion of Stat3 blocks mammary
gland involution and extends functional competence of the
secretory epithelium in the absence of lactogenic stimuli.
Endocrinology 143, 3641–3650 (2002).
60. Zhao, L. et al. Mammary gland remodeling depends on
gp130 signaling through Stat3 and MAPK. J. Biol. Chem.
279, 44093–44100 (2004).
61. Zhao, L., Melenhorst, J. J. & Hennighausen, L. Loss of
interleukin 6 results in delayed mammary gland involution:
a possible role for mitogen-activated protein kinase and
not signal transducer and activator of transcription 3.
Mol. Endocrinol. 16, 2902–2912 (2002).
62. Roberts, A. W. et al. Placental defects and embryonic
lethality in mice lacking suppressor of cytokine signaling 3.
Proc. Natl Acad. Sci. USA 98, 9324–9329 (2001).
63. Takahashi, Y. et al. SOCS3: an essential regulator of LIF
receptor signaling in trophoblast giant cell differentiation.
EMBO J. 22, 372–384 (2003).
64. Park, D. S. et al. Caveolin-1-deficient mice show
accelerated mammary gland development during
pregnancy, premature lactation, and hyperactivation of the
Jak-2/STAT5a signaling cascade. Mol. Biol. Cell 13,
65. Srivastava, S. et al. Receptor activator of NF-κB ligand
induction via Jak2 and Stat5a in mammary epithelial cells.
J. Biol. Chem. 278, 46171–46178 (2003).
66. Cao, Y. et al. IKKα provides an essential link between
RANK signaling and cyclin D1 expression during
mammary gland development. Cell 107, 763–775 (2001).
67. Fantl, V., Stamp, G., Andrews, A., Rosewell, I. &
Dickson, C. Mice lacking cyclin D1 are small and show
defects in eye and mammary gland development.
Genes Dev. 9, 2364–2372 (1995).
68. Sicinski, P. et al. Cyclin D1 provides a link between
development and oncogenesis in the retina and breast.
Cell 82, 621–630 (1995).
69. Brisken, C. et al. IGF-2 is a mediator of prolactin-induced
morphogenesis in the breast. Dev. Cell 3, 877–887 (2002).
70. Ormandy, C. J. et al. Investigation of the transcriptional
changes underlying functional defects in the mammary
glands of prolactin receptor knockout mice. Recent Prog.
Horm. Res 58, 297–323 (2003).
71. Bry, C. et al. Loss of connexin 26 in mammary epithelium
during early but not during late pregnancy results in
unscheduled apoptosis and impaired development.
Dev. Biol. 267, 418–429 (2004).
72. Pittius, C. W., Sankaran, L., Topper, Y. J. & Hennighausen, L.
Comparison of the regulation of the whey acidic protein
gene with that of a hybrid gene containing the whey acidic
protein gene promoter in transgenic mice. Mol. Endocrinol.
2, 1027–1032 (1988).
73. Li, S. & Rosen, J. M. Distal regulatory elements required
for rat whey acidic protein gene expression in transgenic
mice. J. Biol. Chem. 269, 14235–14243 (1994).
74. Walton, K. D. et al. Conditional deletion of the bcl-x gene
from mouse mammary epithelium results in accelerated
apoptosis during involution but does not compromise cell
function during lactation. Mech. Dev. 109, 281–293
75. Socolovsky, M., Fallon, A. E., Wang, S., Brugnara, C. &
Lodish, H. F. Fetal anemia and apoptosis of red cell
progenitors in Stat5a-/-5b-/- mice: a direct role for
Stat5 in Bcl-X(L) induction. Cell 98, 181–191
76. Renou, J. P. et al. Identification of genes differentially
expressed in mouse mammary epithelium transformed
by an activated β-catenin. Oncogene 22, 4594–4610
77. Lapinskas, E. J. et al. A major site of expression of the ets
transcription factor Elf5 is epithelia of exocrine glands.
Histochem. Cell Biol. 122, 521–526 (2004).
78. Zhou, J. et al. Elf5 is essential for early embryogenesis and
mammary gland development during pregnancy and
lactation. EMBO J. 24, 635–644 (2005).
79. McKnight, R. A. et al. An Ets site in the whey acidic
protein gene promoter mediates transcriptional activation
in the mammary gland of pregnant mice but is
dispensable during lactation. Mol. Endocrinol. 9, 717–724
80. Robinson, G. W., Accili, M. & Hennighausen, L. in
Methods in Mammary Gland Biology and Breast Cancer
Research. (eds Ip, M. M. & Asch, B. B.) 307–316 (Kluwer
Academic/Plenum Publishers, 2000).
L.H. and G.W.R. are supported by the Intramural Research
Program of the National Institutes of Health (National Institute of
Diabetes and Digestive and Kidney Diseases).
Competing interests statement
The authors declare no competing financial interests.
The following terms in this article are linked online to:
Bcl-X | CAV1 | Cx26 | Cx32 | ELF5 | ERBB1 | ERBB2 | ERBB4 |
gp130 | JAK2 | PRL | PRLR | RANK | RANK-L | SOCS1 |
SOCS3 | STAT3 | STAT5A | STAT5B
Lothar Hennighausen’s laboratory:
Access to this interactive links box is free online.
NATURE REVIEWS | MOLECULAR CELL BIOLOGY
VOLUME 6 | SEPTEMBER 2005 | 725