© 2006 Nature Publishing Group
Generation of a functional mammary gland from a
single stem cell
Mark Shackleton1,2, Franc ¸ois Vaillant1,2, Kaylene J. Simpson3†, John Stingl4,5, Gordon K. Smyth1,
Marie-Liesse Asselin-Labat1,2, Li Wu1, Geoffrey J. Lindeman1,2& Jane E. Visvader1,2
from evidence that the mammary gland can be regenerated by
transplantation of epithelial fragments in mice1–3. Interest in
MaSCs has been further stimulated by their potential role in
breast tumorigenesis4. However, the identity and purification of
isolated discrete populations of mouse mammary cells on the
basis of cell-surface markers and identified a subpopulation
(Lin2CD29hiCD241) that is highly enriched for MaSCs by trans-
plantation. Here we show that a single cell, marked with a LacZ
transgene, can reconstitute a complete mammary gland in vivo.
The transplanted cell contributed to both the luminal and myo-
epithelial lineages and generated functional lobuloalveolar units
during pregnancy. The self-renewing capacity of these cells was
demonstrated by serial transplantation of clonal outgrowths. In
support of a potential role for MaSCs in breast cancer, the stem-
cell-enriched subpopulationwas expanded in premalignant mam-
mary tissue from MMTV-wnt-1 mice and contained a higher
number of MaSCs. Our data establish that single cells within
the Lin2CD29hiCD241population are multipotent and self-
renewing, properties that define them as MaSCs.
and lobuloalveolar structures, the latter of which arise during
pregnancy5,6. Several features of the developing gland appear to
require MaSCs. There are two primary epithelial cell lineages—
myoepithelial and luminal (comprising ductal and alveolar sub-
types)—which are presumed to arise from a common progenitor
cell. The massive expansion of mammary epithelium during puberty
and pregnancy, together with the remarkable regenerative capacity
apparent during successive reproductive cycles, also implicate stem-
like cells. Furthermore, serial transplantation of retrovirally tagged
epithelial fragments has suggested that a single progenitor cell may
repopulate an entire mammary gland3. Progenitor-enriched popu-
lations have been reported in the mouse mammary gland7,8, and in
human breast tissue, bi-potential cells capable of generating both
luminal and myoepithelial cells in vitro have been identified9–11.
Here we describe the prospective isolation of mouse MaSCs using
specific cell-surface markers and performing transplantations into
the mammary stroma. Because the mammary gland comprises a
heterogeneous mix of cell types, we used antibodies against well-
characterized endothelial (CD31) and haematopoietic (CD45 and
TER119) antigens to deplete these cells using fluorescence-activated
cell sorting (FACS). The substantial CD45þand CD31þpopulations
were excluded from freshly isolated mammary cell suspensions by
gating on the CD452CD312TER1192(Lin2) population (Fig. 1a,
in the identification of haematopoietic stem cells, to determine the
frequency of mammary repopulating ‘units’ (MRUs) in subpopu-
lations of cells. Lin2cells were isolated by FACS and transplanted in
decreasing numbers into cleared mammary fat pads (MFPs) of
recipient mice. The percentage of epithelial outgrowths was estab-
lished for each injected cell number, and the frequency of MRUs in
the Lin2population was calculated to be 1/4,900 (Supplementary
Table 1). An example of an outgrowth arising from 5,000 trans-
planted Lin2cells is shown in Fig. 1b. In contrast, 22 transplants of
3,000 Linþcells (Fig. 1a, R1) produced no outgrowths (Fig. 1b),
indicating that MRUs are not enriched in this subset.
We defined four distinct Lin2subpopulations based on the
expression of CD29 (b1-integrin), a stem-cell marker in skin13, and
CD24 (heat-stable antigen), which has been used to enrich neural
frequency of MRUs in these four populations was determined
following isolation by FACS and transplantation in numbers pro-
portional to their frequency in the Lin2population15. MRUs
were enriched approximately eightfold in the Lin2CD29hiCD24þ
population, whereas no significant enrichment was found in
the other three subsets (Supplementary Table 2). Notably, the
Lin2CD29hiCD24þsubpopulation was found to be enriched for
long-term label-retaining cells, consistent with the presence of
quiescent or asymmetrically dividing cells16(Supplementary Fig. 1).
In agreement with ref. 17, analysis of CD49f (a6-integrin) co-
expression revealed significant enrichment of CD49fþþcells in the
Lin2CD29hiCD24þgate (data not shown). However, neither high
Sca-1 (stem cell antigen) expression nor Hoechst33342dye exclusion
(which are previously reported characteristics of mammary stem/
progenitor cells7,8) were enriched in the Lin2CD29hiCD24þsub-
population (Fig. 1d, e), and these observations were corroborated by
transplantation of these subpopulations (Supplementary Tables 3
counting and determining cell viability before transplantation. In
addition, we transplanted cells from Rosa-26 mice, which carry a
ubiquitously expressed LacZ transgene18, into wild-type recipients to
allow verification of donor origin by staining for LacZ (b-galactosi-
dase) activity in the harvested gland. Using this more quantitative
method, the calculated MRU frequency in the Lin2CD29hiCD24þ
population was increased to 1/64 without being significantly altered
for the other populations (Supplementary Table 5). A LacZ-positive
(LacZþ) epithelial outgrowth obtained from one of these transplants
is depicted in Fig. 1f. Given that cells are inevitably lost during
transplantation, the actual MRU frequency is likely to be higher than
1The Walter and Eliza Hall Institute of Medical Research, 1G Royal Parade, Parkville, Victoria 3050,2Bone Marrow Research Laboratories, Royal Melbourne Hospital, Parkville,
Victoria 3050,3Department of Biochemistry and Molecular Biology, The University of Melbourne, Victoria 3010, Australia.4Terry Fox Laboratory, British Columbia Cancer
Agency, Vancouver, British Columbia V5Z 1L3, Canada.5Stem Cell Technologies Inc., 570 West 7th Avenue, Suite 400, Vancouver, British Columbia V5Z 1B3, Canada.
†Present address: Department of Cell Biology, Harvard Medical School, Boston, Massachusetts 02115, USA.
Vol 439|5 January 2006|doi:10.1038/nature04372
© 2006 Nature Publishing Group
To characterize epithelial cell types within the different subpopu-
lations, flow cytometric analysis was performed using antibodies
against CD24, CD29 and either the luminal marker cytokeratin 18
(K18) or the myoepithelial marker cytokeratin 14 (K14) (Fig. 2a).
Whereas almost all Lin2CD29loCD24þcells expressed K18, the
majority of K14hicells resided within the Lin2CD29hiCD24þsub-
population and these presumably correspond to mature myoepithe-
lial cells. Interestingly, a distinct K14lopopulation was also revealed
in the Lin2CD29loCD24þgate, indicating that many K18þcells in
the mammary gland express low levels of K14.
Epithelial cell culture assays9provided further evidence that the
cells. Only the two Lin2CD24þpopulations yielded significant
colonies and the CD29hisubset exhibited a two- to three-fold higher
colony frequency andsubstantially largercolonies (Fig.2b). Toassess
the differentiation capacity of the cells, we compared the growth of
Figure 2 | Characterization of cytokeratin expression in Lin2cells and
increased progenitor capacity of Lin2CD29hiCD241mammary cells
in vitro. a, Expression of CD29 and CD24 in subpopulations of Lin2cells
defined by K18 and K14 expression. b, Colony-forming ability of the four
Lin2cell populations defined by CD29 and CD24 expression (histogram
shows mean ^ s.e.m., n ¼ 5). c, Representative structures produced by
cells (lower row); bright-field views of gels (left; scale bars, 100mm),
H&E-stained sections (middle; scale bars, 10mm), and labelling with
anti-milk antibody are shown (right; arrowheads indicate milk-producing
structures, arrow indicates a non-milk-producing structure). Insets in right
panels show isotype-labelled control sections: red, milk; blue, DAPI; scale
bars, top 40mm, bottom 20mm. d, Expression of CD29 and CD24 in a
terminal endbud (left column; scale bar, 40mm) and a mature duct (right
column; scale bar, 16mm). The merged confocal images are shown below.
Insets show isotype-labelled control sections: green, CD29; red, CD24; blue,
Figure 1 | Enrichment of mammary repopulating units in the
Lin2CD29hiCD241population. a, Expression of haematopoietic and
endothelial cell-surface markers in mammary cell suspensions (left); gating
strategy (right) used to select Lin2(R2 gate) and Linþ(R1 gate) cells.
of CD24 and CD29 in the Lin2population (left); gating strategy used to
the Lin2CD29hiCD24þpopulation (left, dotted line shows isotype
labelling); gating strategy used to purify cells according to Sca-1 expression
and size (right, R3–5 gates), as defined by forward scatter (FSC). Sixty-three
per cent of the Lin2population expressed high levels of Sca-1. e, Depletion
of Hoechst side-population (SP) cells in the Lin2CD29hiCD24þ
subpopulation (left) compared to the overall Lin2population (middle);
gating strategy used to purify cells according to Hoechst staining (middle);
loss of SP cells in the presence of 100mM verapamil (right). MP, main
population. f, A LacZþoutgrowth arising from transplantation of 13
visualized, double-sorted Lin2CD29hiCD24þcells. Scale bar in f, 250mm.
NATURE|Vol 439|5 January 2006
© 2006 Nature Publishing Group
Lin2CD29hiCD24þand Lin2CD29loCD24þcells in Matrigel under
lactogenicconditions. Cells fromthe Lin2CD29loCD24þpopulation
predominantly (85%) formed single-cell layered, alveolar-like
structures that produced milk protein (Fig. 2c and Supplementary
that have a luminal cell fate under lactogenic conditions. In contrast,
Lin2CD29hiCD24þcells formed a heterogeneous mix of morpho-
logically distinguishable structures, including ductal forms and
multicellular spheroid bodies, as well as occasional (3%) alveolar-
like structuresakin to thosefromthe Lin2CD29loCD24þpopulation
(Fig. 2c and Supplementary Fig. 2a). Co-staining of the Matrigel-
derived structures with antibodies to K14 and K18 revealed marked
differences between Lin2CD29loCD24þand Lin2CD29hiCD24þcell-
derived structures with respect to their pattern of cytokeratin
expression (Supplementary Fig. 2b). The expanded differentiative
cells suggest that this population is enriched for mammary progeni-
tors. Compatible with these findings, high levels of CD29 expression
were apparent in the cap cell region of terminal end buds, presumed
to be rich in stem cells19,20, relative to mature ducts in which high
expression was predominantly basolateral, as previously shown21
To test the ‘common-progenitor model’ of lineage development in
MRU constituted a single cell. Double-sorted Lin2CD29hiCD24þ
cells from Rosa-26 mice were resuspended at a concentration of one
cell per injection volume, with or without supporting cells (5,000
cells from a wild-type population depleted of Lin2CD29hiCD24þ
cells). Eight LacZþepithelial outgrowths were produced from 68
affect the likelihood of an outgrowth or its size. Although the eight
outgrowths could have resulted from more than one distinct pro-
genitor, calculations showed this to be extremely unlikely (see
Supplementary Methods). To prove definitively that a single cell
could repopulate a cleared MFP, we transplanted individual, double-
sorted Lin2CD29hiCD24þRosa-26 cells that had been viewed
microscopically in 10-ml Terasaki wells. Six LacZþoutgrowths were
produced from 102 transplants in three independent experiments
(Supplementary Table 6 and Fig. 3a) and, as previously observed,
supporting cells had no effect. Substantial engraftment of the fat pad
was evident and histological sectioning of the outgrowths revealed
normal ductal structures composed of both luminal and myoepithe-
lial cells (Fig. 3b). Furthermore, repopulated glands exhibited com-
plete functional differentiation at parturition (Fig. 3a) and sections
derived from pregnant recipients revealed lipid droplets and abun-
dant milk protein within alveoli and ductal lumens, respectively
(Fig. 3b and c). Thus, a single Lin2CD29hiCD24þcell can recon-
stitute a completely functional mammary gland, demonstrating its
high proliferative and multi-lineage differentiation capacity.
Further evidence for the clonality of outgrowths arising from
limiting numbers of Lin2CD29hiCD24þcells was derived from
‘mixing’ experiments in which cells from wild-type and Rosa-26
mice were co-injected. Seventy cells from each donor were mixed
before transplantation and outgrowths analysed in virgin recipients.
Pure wild-type or LacZþoutgrowths were observed in 95/97 cases
(Supplementary Fig. 3), indicating that combinatorial activity
between independent MaSCs is not essential to generate a mammary
To evaluate whether the Lin2CD29hiCD24þrepopulating cell can
self-renew, epithelial outgrowths derived from primary transplants
of Lin2CD29hiCD24þcells were analysed by flow cytometry and
serial transplantation. The primary outgrowths comprised the same
CD29 and CD24 profiles as donor mice (Fig. 3d, left panel), whereas
cell suspensions from cleared, untransplanted mammary fat
pads were CD242(Fig. 3d, right panel), demonstrating that the
CD24þcells were donor-derived. For secondary transplantation,
we used primary outgrowths that developed from up to 25
Lin2CD29hiCD24þRosa-26 cells and which were therefore likely to
be clonal (see Supplementary Methods). Cells from each primary
outgrowth generated LacZþoutgrowths in multiple secondary reci-
pients (Supplementary Table 7 and Supplementary Methods), indi-
cating that self-renewal had occurred in the primary outgrowths.
Figure 3 | A single, self-renewing Lin2CD29hiCD241cell can repopulate a
MFP. a, Wholemount analysis of epithelial outgrowths arising from
transplantation of a single LacZþLin2CD29hiCD24þcell; virgin recipients
harvested 10 and 8.5 weeks after transplantation (top left and middle; scale
High-magnification images of ductal structures (bottom left; scale bar,
100mm), terminal endbuds (bottom middle; scale bar, 50mm), and lobulo-
alveolarstructuresinafull-term pregnant recipient(bottomright; scalebar,
100mm). b, Sections of single-cell origin, LacZþoutgrowths stained with
nuclear fast red show ductal (arrowheads) and myoepithelial (arrows) cell
lineages (left, scale bar, 5mm) and a terminal endbud (middle; scale bar,
10mm) in a virgin recipient, and lobulo-alveolar epithelium in a mid-term
pregnant recipient (right, arrows indicate lipid droplets; scale bar, 10mm).
L, lumen.c, Immunofluorescencestaining with anti-milkantibodyof a duct
arising from a single LacZþLin2CD29hiCD24þcell in a pregnant recipient;
inset, isotype-labelled control section; green, milk; blue, DAPI. d, Flow
cytometric analysis of cell suspensions prepared from MFPs transplanted
with Lin2CD29hiCD24þcells (left) and untransplanted cleared MFPs
of cells derived from an outgrowth of single Lin2CD29hiCD24þcell origin;
scale bar, 100mm. f, A tertiary LacZþoutgrowth harvested at parturition
(left; scale bar, 250mm). Section of this outgrowth stained with nuclear fast
red (right; scale bar, 10mm).
NATURE|Vol 439|5 January 2006
© 2006 Nature Publishing Group
capacity was retained in these outgrowths (Fig. 3f). To definitively
confirm the self-renewing capability of the Lin2CD29hiCD24þ
repopulating cell, we transplanted a primary outgrowth of single-
cell origin and produced 15 secondary outgrowths (Supplementary
Table 7) that exhibited complete differentiation at parturition (Fig.
3e). Thus, the Lin2CD29hiCD24þmammary repopulating cell is
capable of self-renewal, a defining feature of stem cells22.
Recent evidence suggests the existence of a tumour stem cell for
in hyperplastic but premalignant mammary tissue from two strains
of mice prone to develop tumours. Despite an age-related expansion
of the Lin2CD29hiCD24þsubpopulation in wild-type glands,
significantly increased epithelial cellularity and percentage of
Lin2CD29hiCD24þcells were evident in MMTV-wnt-1 mice
(Fig. 4a and b). In addition, transplantation studies showed the
relative MaSC frequencies in Lin2CD29hiCD24þsubpopulations
from MMTV-wnt-1 and control mice to be 1/57 (1/37 to 1/86) and
1/86 (1/51 to 1/147), respectively. Collectively, these data indicated a
6.4-fold (^1.2s.e.m.) increase in the absolute number of MaSCs in
premalignant MMTV-wnt-1 transgenic glands, implying a role for
Wnt signalling in the self-renewal of MaSCs analogous to its role in
haematopoietic stem cells23. In addition, the epithelial outgrowths
arising from transplantation of Lin2CD29hiCD24þMMTV-wnt-1
mammary cells were profoundly hyperplastic at 5weeks post-
transplantation (Fig. 4c). Our findings are compatible with the
proposal that the wnt-1 oncogene gives rise to heterogeneous
tumours because it perturbs the MaSC pool24,25. Interestingly,
pre-neoplastic mammary tissue from MMTV-neu mice26, which
develop luminal epithelial tumours, showed no expansion of the
Lin2CD29hiCD24þpopulation (Fig. 4a and b). Taken together, our
data suggest that different epithelial cell-types are the targets of
transformation in the wnt-1 and neu mammary tumorigenesis
This study provides the first description, to our knowledge, of the
has implications for the isolation of stem cells from other epithelial
tissues. It is not known whether there is a hierarchy of mammary
stem and progenitor cells, analogous to the haematopoietic sys-
tem22,27. However, there is evidence suggesting that distinct cellular
progenitors for ductal and lobular structures exist in the mammary
gland3,28. We have established that a single stem cell is capable of
reconstituting the entire mammary tree. Presumably this cell under-
goes division in its stromal milieu to yield progenitors as well as
mammary epithelium into a branching ductal network. It seems
likely that b1-integrin and a6-integrin (ref. 17) participate in
mediating interactions between MaSCs and the mammary stroma.
Delineation of genes that are expressed in stem and progenitor cells
should allow the identification of further markers of the MaSC and
the putative breast cancer stem cell.
Mice. FVB/NJ, C57BL/6, BALB/c, Rosa-26 (ref. 18) (C57BL/6), MMTV-wnt-1
(BALB/c), and MMTV-neu (FVB/NJ) mice were bred and maintained in our
animal facility according to institutional guidelines. All experiments were
approved by the Animal Research Ethics Committee of the Melbourne Health
Mammary cell preparation. Mammary glands were dissected from 8-week-old
female mice. After mechanical dissociation with a McIlwain tissue chopper
(Mickle Laboratory Engineering), the tissue was placed in culture medium
(DME HAM with 1mM glutamine, 5mgml21insulin, 500ngml21hydrocorti-
sone, 10ngml21epidermal growth factor and 20ngml21cholera toxin)
supplemented with 5% bovine calf serum and containing 300Uml21collagen-
The resultant organoid suspension was sequentially resuspended in 0.25%
trypsin-EGTA for 1–2min, 5mgml21dispase (Roche Diagnostics) and
0.1mgml21DNase (Worthington) for 5min, and 0.64% NH4Cl for 3min
before filtration through a 40-mm mesh and labelling.
Antibodies. Antibodies against mouse antigens were purchased from BD
Pharmingen unless otherwise specified, and included CD24-PE, biotinylated
milk (Nordic Immunological Laboratories), anti-cytokeratin 14 (Covance), and
anti-cytokeratin 18 (Progen Biotechnik). Streptavidin-APC was purchased from
BD Pharmingen. Fluorochrome-conjugated secondary antibodies included anti-
Celllabelling,flowcytometry andsorting.Hoechst stainingwasperformedfor
1h at 378C with 6mgml21Hoechst33342(Sigma). For labelling of intracellular
epitopes, cells were fixed in chilled acetone for 1min and permeablized in 0.1%
Tween/PBS on ice for 5 min before blocking. Blocking was performed in rat
g-globulin (Jackson Laboratories) and anti-CD16/CD32 Fcg III/II receptor
antibody (BD Pharmingen) for 10min. Antibody incubations were performed
at 48C for 25min (45min for intracellular epitope labelling). Cells were
resuspended in 0.5mgml21propidium iodide (Sigma) before analysis. Data
analysis was performed on the single, live cell gate using WEASEL software
(http://www.wehi.edu.au/cytometry/WEASELv2.html). Cell sorting was carried
out on a FACSDiVa, FACStar or FACS Vantage cell sorter (Becton Dickinson).
The purity of sorted populations was routinely more than 95%.
Mammary fat pad transplantation and analysis. Sorted cells were resuspended
in PBS with 0.04% trypan blue (Sigma) and 50% fetal calf serum (FCS), and
had been cleared of endogenous epithelium1. Visualization of cells before
transplantation was performed in 10-ml Terasaki wells. Recipient glands were
removed for evaluation after 5–10weeks unless otherwise stated. Wholemounts
of wild-type mammary outgrowths were stained with haematoxylin. LacZþ
outgrowthswere detectedbyX-Galstaining29,and/orpolymerase chain reaction
analysis for the LacZ gene18. An outgrowthwas defined as an epithelial structure
Figure 4 | The Lin2CD29hiCD241population is expanded in MMTV-wnt-1
in cell suspensions from preneoplastic MMTV-wnt-1 and MMTV-neu
transgenic mammary glands. Tissue was taken from parous MMTV-wnt-1
mice (n ¼ 3) at 4 months of age and virgin MMTV-neu mice (n ¼ 3) at 6
same premalignant glands (scale bars, 40mm). b, Histogram depicting the
percentages (mean ^ s.e.m.) ofCD29hicells in the Lin2CD24þpopulations
of MMTV-wnt-1 (n ¼ 3; 74%) and MMTV-neu (n ¼ 3; 43%) mammary
glands compared with age- and parity-matched controls (n ¼ 2; 38% and
40%, respectively). A similar fold expansion of the Lin2CD29hiCD24þ
subpopulation was observedin nulliparous and parous MMTV-wnt-1 mice.
However, this population was found to increase with age in wild-type mice
(1.5- to twofold larger in 4–6-month-old mice versus 8-week-old mice).
transplantationof 25 Lin2CD29hiCD24þcells derived from control (left) or
MMTV-wnt-1 (central) donor mice; scale bars, 250mm. Right panel shows
an H&E-stained section prepared from the structure depicted in the middle
panel. Scale bars, 20mm.
NATURE|Vol 439|5 January 2006
© 2006 Nature Publishing Group
comprising ducts arising from a central point, with lobules and/or terminal end
In vitro assays. For colony assays, cells were sorted directly into the wells of
24-well plates containing culture medium with 1% FCS in the presence of
14,000cm22irradiated NIH-3T3 cells. After 24h the media was replaced with
serum-free culture medium containing 0.1% bovine serum albumin, and 5days
later the colonies were fixed with acetone:methanol (1:1), stained with Giemsa,
and counted. For three-dimensional assays, cells were resuspended in chilled
100% Matrigel and the gels allowed to set before covering with culture medium
to DME-HAM containing 1mM glutamine, 5mgml21insulin, 500ngml21
hydrocortisone, 5mgml21prolactin and 1% FCS after 1week. The cells were
cultured for a total of 2weeks before fixation in 4% paraformaldehyde and
embedding in paraffin for sectioning and staining with haematoxylin and eosin
(H&E), or fixation in chilled acetone:methanol (1:1) for immunostaining.
Immunostaining. Frozen sections were prepared from tissues embedded in
OCT compound. After fixation in 100% acetone, sections were rehydrated and
blocked with 5% bovine calf serum in PBS. Paraffin-embedded sections were
dewaxed, washed in PBS, and subjected to antigen retrieval by boiling in 10mM
citrate buffer for 20min and treatment with 150mM glycine for 15min, before
blocking as above. Primary antibody staining was performed overnight at 48C,
while secondary antibody staining was performed for 30min and DAPI staining
for 5min at room temperature. Matrigel cultures were prepared and immuno-
stained as described30. Sections were imaged on a Leica TCS4 SP2 spectral
confocal scanner linked to a Leica DMIRE2 inverted microscope.
Received 1 August; accepted 13 October 2005.
1.DeOme, K. B., Faulkin, L. J. Jr, 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., DeOme, K. B., Young, J. T., Blair, P. B. & Faulkin, L. J. Jr. The
in vivo life span of normal and preneoplastic mouse mammary glands: a serial
transplantation study. Proc. Natl Acad. Sci. USA 61, 53– -60 (1968).
Kordon, E. C. & Smith, G. H. An entire functional mammary gland may
comprise the progeny from a single cell. Development 125, 1921– -1930 (1998).
Al-Hajj, M., Wicha, M. S., Benito-Hernandez, A., Morrison, S. J. & Clarke, M. F.
Prospective identification of tumorigenic breast cancer cells. Proc. Natl Acad.
Sci. USA 100, 3983– -3988 (2003).
Hennighausen, L. & Robinson, G. W. Think globally, act locally: the making of a
mouse mammary gland. Genes Dev. 12, 449– -455 (1998).
Daniel, C. W. & Smith, G. H. The mammary gland: a model for development.
J. Mammary Gland Biol. Neoplasia 4, 3– -8 (1999).
Welm, B. E. et al. Sca-1poscells in the mouse mammary gland represent an
enriched progenitor cell population. Dev. Biol. 245, 42– -56 (2002).
Alvi, A. J. et al. Functional and molecular characterisation of mammary side
population cells. Breast Cancer Res. 5, R1– -R8 (2003).
Stingl, J., Eaves, C. J., Zandieh, I. & Emerman, J. T. Characterization of bipotent
mammary epithelial progenitor cells in normal adult human breast tissue.
Breast Cancer Res. Treat. 67, 93– -109 (2001).
10. Gudjonsson, T. et al. Isolation, immortalization, and characterization of a
human breast epithelial cell line with stem cell properties. Genes Dev. 16,
693– -706 (2002).
11.Dontu, G. et al. In vitro propagation and transcriptional profiling of human
mammary stem/progenitor cells. Genes Dev. 17, 1253– -1270 (2003).
12. Bonnefoix, T., Bonnefoix, P., Verdiel, P. & Sotto, J. J. Fitting limiting dilution
experiments with generalized linear models results in a test of the single-hit
Poisson assumption. J. Immunol. Methods 194, 113– -119 (1996).
13. Jones, P. H., Harper, S. & Watt, F. M. Stem cell patterning and fate in human
epidermis. Cell 80, 83– -93 (1995).
14. Rietze, R. L. et al. Purification of a pluripotent neural stem cell from the adult
mouse brain. Nature 412, 736– -739 (2001).
15. Uchida, N. & Weissman, I. L. Searching for hematopoietic stem cells: evidence
that Thy-1.1loLin2Sca-1þcells are the only stem cells in C57BL/Ka-Thy-1.1
bone marrow. J. Exp. Med. 175, 175– -184 (1992).
16. Smith, G. H. Label-retaining epithelial cells in mouse mammary gland divide
asymmetrically and retain their template DNA strands. Development 132,
681– -687 (2005).
17. Stingl, J. et al. Purification and unique properties of mammary epithelial stem
cells. Nature advance online publication, 4 January 2006 (doi:10.1038/
18. Friedrich, G. & Soriano, P. Promoter traps in embryonic stem cells: a genetic
screen to identify and mutate developmental genes in mice. Genes Dev. 5,
1513– -1523 (1991).
19. Kenney, N. J., Smith, G. H., Lawrence, E., Barrett, J. C. & Salomon, D. S.
Identification of stem cell units in the terminal end bud and duct of the mouse
mammary gland. J. Biomed. Biotechnol. 1, 133– -143 (2001).
20. Williams, J. M. & Daniel, C. W. Mammary ductal elongation: differentiation of
myoepithelium and basal lamina during branching morphogenesis. Dev. Biol.
97, 274– -290 (1983).
21. White, D. E. et al. Targeted disruption of b1-integrin in a transgenic mouse
model of human breast cancer reveals an essential role in mammary tumour
induction. Cancer Cell 6, 159– -170 (2004).
22. Weissman, I. L. Stem cells: units of development, units of regeneration, and
units in evolution. Cell 100, 157– -168 (2000).
23. Reya, T. et al. A role for Wnt signalling in self-renewal of haematopoietic stem
cells. Nature 423, 409– -414 (2003).
24. Li, Y. et al. Evidence that transgenes encoding components of the Wnt
signalling pathway preferentially induce mammary cancers from progenitor
cells. Proc. Natl Acad. Sci. USA 100, 15853– -15858 (2003).
25. Liu, B. Y., McDermott, S. P., Khwaja, S. S. & Alexander, C. M. The transforming
activity of Wnt effectors correlates with their ability to induce the
accumulation of mammary progenitor cells. Proc. Natl Acad. Sci. USA 101,
4158– -4163 (2004).
26. Guy, C. T. et al. Expression of the neu protooncogene in the mammary
epithelium of transgenic mice induces metastatic disease. Proc. Natl Acad. Sci.
USA 89, 10578– -10582 (1992).
27. Dick, J. E. Stem cells: Self-renewal writ in blood. Nature 423, 231– -233 (2003).
28. Smith, G. H. Experimental mammary epithelial morphogenesis in an in vivo
model: evidence for distinct cellular progenitors of the ductal and lobular
phenotype. Breast Cancer Res. Treat. 39, 21– -31 (1996).
29. Sum, E. Y. M., O’Reilly, L. A., Jonas, N., Lindeman, G. J. & Visvader, J. E. The
LIM domain protein Lmo4 is highly expressed in proliferating mouse epithelial
tissues. J. Histochem. Cytochem. 53, 475– -486 (2005).
30. Debnath, J., Muthuswamy, S. K. & Brugge, J. S. Morphogenesis and
oncogenesis of MCF-10A mammary epithelial acini grown in three-dimensional
basement membrane cultures. Methods 30, 256– -268 (2003).
Supplementary Information is linked to the online version of the paper at
Acknowledgements We are grateful to N. Forrest for expert assistance,
J. Adams and A. Harris for critical review of the manuscript, S. Mihajlovic for
histology, and F. Battye and A. Holloway for FACS support. This work was
supported by the Victorian Breast Cancer Research Consortium and the
National Health and Medical Research Council (Australia). M.-L.A.-L. is
supported by a Fondation pour la Recherche Medicale Fellowship, K.J.S. by a
Peter Doherty Fellowship and J.S. by Fellowships from the Canadian Breast
Cancer Foundation and the Natural Sciences and Engineering Research Council
Author Contributions M.S. and F.V. contributed equally to this work. G.J.L. and
J.E.V. contributed equally to this work.
Author Information Reprints and permissions information is available at
npg.nature.com/reprintsandpermissions. The authors declare no competing
financial interests. Correspondence and requests for materials should be
addressed to J.E.V. (firstname.lastname@example.org) or G.J.L. (email@example.com).
NATURE|Vol 439|5 January 2006