© 2006 Nature Publishing Group
Pten dependence distinguishes
haematopoietic stem cells from
O¨mer H. Yilmaz1, Riccardo Valdez2, Brian K. Theisen2, Wei Guo3, David O. Ferguson2, Hong Wu3
& Sean J. Morrison1
Recent advances have highlighted extensive phenotypic and functional similarities between normal stem cells and cancer
stem cells. This raises the question of whether disease therapies can be developed that eliminate cancer stem cells
without eliminating normal stem cells. Here we address this issue by conditionally deleting the Pten tumour suppressor
gene in adult haematopoietic cells. This led to myeloproliferative disease within days and transplantable leukaemias
within weeks. Pten deletion also promoted haematopoietic stem cell (HSC) proliferation. However, this led to HSC
depletion via a cell-autonomous mechanism, preventing these cells from stably reconstituting irradiated mice. In contrast
to leukaemia-initiating cells, HSCs were therefore unable to maintain themselves without Pten. These effects were
mostly mediated by mTOR as they were inhibited by rapamycin. Rapamycin not only depleted leukaemia-initiating cells
but also restored normal HSC function. Mechanistic differences between normal stem cells and cancer stem cells can
thus be targeted to deplete cancer stem cells without damaging normal stem cells.
Cancer stem cells have notable phenotypic and mechanistic simi-
larities to normal stem cells in the same tissues1–4. Acute myeloid
leukaemia (AML) is sustained by leukaemic stem cells that are also
called leukaemia-initiating cells because they are defined by their
ability to transfer disease on transplantation into irradiated mice5–7.
Leukaemia-initiating cells expressmarkers similar to normal HSCs5,6
and dependon similar mechanismsto self-renew8,9. Braincancerstem
cells also express markers of normal neural stem cells and depend on
similar pathways for their proliferation4,10. The Hedgehog, Wnt and
Notch pathways that often promote cancer cell proliferation also
promote normal stem cell self-renewal1,2,11,12. Conversely, tumour
suppressors that inhibit cancer cell proliferation—such as p53,
p16INK4aand p19ARF—also inhibit stem cell self-renewal11,13,14.
Whether cancer stem cells arise from normal stem cells or other
cells, their similarity to normal stem cells indicates that they inherit
or acquire stem cell properties. This raises the question of whether it
will be possible to identify therapies that eliminate cancer stem cells
without eliminating normal stem cells in the same tissues.
on leukaemia-initiating cells and normal HSCs. PTEN (for phospha-
tase and tensin homologue) is a phosphatase that negatively regulates
signalling through the phosphatidylinositol-3-OH kinase (PI(3)K)
pathway, inhibiting proliferation and survival15,16. Pten is commonly
topoietic malignancies18–21. Here we report that whereas Pten deletion
causes the generation of transplantable leukaemia-initiating cells, it
also causes the depletion of normal HSCs, thus identifying a
mechanistic difference between the maintenance of normal stem
cells and cancer stem cells.
Pten deletion leads to leukaemogenesis
Pten was conditionally deleted from 6-to-8-week-old Ptenfl/fl;Mx-1-
Cre mice by administering seven doses of polyinosine-polycytidine
(pIpC) over 14days to induce Cre expression22,23. After 14days, Pten
seemed to be completely deleted from HSCs and other haemato-
poietic cells (Supplementary Fig. 1). We analysed Ptenfl/fl;Mx-1-Cre
mice, as well as Ptenfl/þ;Mx-1-Cre littermate control mice, five days
developed myeloproliferative disease marked by a tenfold increase in
spleen cellularity (Fig. 1c), complete histological effacement of the
splenic architecture (Fig. 1b), reduced bone marrow cellularity
(Fig. 1c), and increased blast cell frequency (Fig. 1d). The increased
spleen cellularity was largely attributable to extramedullary haemato-
poiesis (Supplementary Fig. 2c, d) with a prominent expansion in the
number of immature myeloid cells (Supplementary Fig. 2e–g; Sup-
plementary Table 1). None out of 20 Ptenfl/þ;Mx-1-Cre littermates
showed these changes after pIpC treatment (Fig. 1c, a, d).
Within 4 to 6 weeks after pIpC treatment, most Ptenfl/fl;Mx-1-Cre
mice progressed to frank leukaemia24, including AML and acute
lymphoblastic leukaemia (ALL), and died (for the criteria used to
diagnose leukaemias, see Supplementary Table 2). AMLs were
characterized by large numbers of chloroacetate-esterase-positive
myeloid blasts in the spleen (Fig. 1e), and ALLs were characterized
by large numbers of terminal deoxynucleotidyl transferase (TdT)-
positive lymphoid blasts throughout the thymus, which was also
enlarged and effaced (Fig. 1f). The bone marrow contained
Mac-1þGr-1lowCD42myeloid blasts and CD4þCD8þCD3þMac-
12lymphoid blasts (Fig. 1h, i; data not shown). Karyotypic analysis
of myeloid blasts from four Pten-deleted mice with AML revealed
1Howard Hughes Medical Institute, Life Sciences Institute, Department of Internal Medicine, and Center for Stem Cell Biology, University of Michigan, Ann Arbor, Michigan
48109-2216, USA.2Department of Pathology, University of Michigan, Ann Arbor, Michigan 48109-2216, USA.3Molecular and Medical Pharmacology, UCLA School of Medicine,
Los Angeles, California 90095-1735, USA.
Vol 441|25 May 2006|doi:10.1038/nature04703
© 2006 Nature Publishing Group
marked aneuploidy and/or chromosomal translocations, suggesting
leukaemogenesis was associatedwithadditional mutations afterPten
deletion (Supplementary Fig. 3). Common abnormalities across
multiple karyotypes from each mouse suggested that the AMLs
were clonal or oligoclonal (Supplementary Fig. 3).
HSCs proliferate after Pten deletion but become depleted
We examined the cell cycle status of whole bone marrow cells and
Flk-22Sca-1þLin2c-KitþCD482cells five days after completing
pIpC administration. Flk-22Sca-1þLin2c-KitþCD482cells are
highly enriched for HSCs25–27, representing only 0.005% of bone
marrow cells. Nineteen out of 20 mice transplanted with 15 Flk-
22Sca-1þLin2c-KitþCD482cells from control mice showed long-
term multilineage reconstitution by the donor cells (Supplementary
Table 3). There was no significant effect of Pten deletion on the cell
there was a three-to-four-fold increase in the percentage of dividing
Flk-22Sca-1þLin2c-KitþCD482cells, and in the percentage that
incorporated the nucleotide analogue BrdU (5-bromodeoxyuridine)
overa19-h period (Fig.2c–f).These datasuggest that Ptenpromotes
quiescence in HSCs, and that in the absence of Pten HSCs are driven
Consistent with this, the absolute number of Flk-22Sca-1þLin2c-
KitþCD482cells per Ptenfl/fl;Mx-1-Cre mouse increased byapproxi-
mately threefold within five days of pIpC treatment (Fig. 2g; see
Methods for details). Notably, when Ptenfl/fl;Mx-1-Cre mice were
Figure 1 | Pten deletion from adult haematopoietic cells leads to
myeloproliferative disease that progresses to AML and ALL. a–c, Within
five days of Pten deletion, Ptenfl/fl;Mx-1-Cre mice developed
myeloproliferative disease marked by increased spleen cellularity (c),
reduced bonemarrowcellularity (c),andcomplete effacementofthesplenic
architecture (compare b with a) owing to myeloid-predominant
extramedullary haematopoiesis (Supplementary Fig. 2). Values are
mean ^ s.d.*,P , 0.05.d,Blastcellfrequencywassignificantlyincreasedin
Ptenfl/fl;Mx-1-Cre mice, but five days after pIpC treatment only a minority
of these mice (2 out of 6) showed more than 20% blasts. Bars indicate
mean values. *, P , 0.01. e, f, By six weeks after pIpC treatment, most
were acutely ill, and showed the pathological features of both AML
lymphoid blasts in the thymus; white arrows). g–i, The bone marrow of
Ptenfl/fl;Mx-1-Cre mice contained a blast cell population (box in h), with
myeloid (i; Mac-1-positive) and lymphoid (i; CD4-positive) cells, not
evident in control mice (g).
NATURE|Vol 441|25 May 2006
© 2006 Nature Publishing Group
examined 24 to 39 days after pIpC treatment, the number of
Flk-22Sca-1þLin2c-KitþCD482cells declined significantly below
that found in control mice (Fig. 2g). This suggested that HSCs
transiently expanded in number after Pten deletion, but were unable
to maintain themselves and subsequently became depleted.
To test the function of Pten-deficient HSCs, we transplanted 15
Flk-22Sca-1þLin2c-KitþCD482donor cells from Ptenfl/fl;Mx-1-Cre
mice or littermate controls (five days after pIpC) into irradiated
recipient mice along with 200,000 recipient bone marrow cells
(Fig. 2h). Whereas control cells gave high levels of long-term multi-
lineage reconstitution in all five recipients, Pten-deficient cells
initially gave multilineage reconstitution, but by eight weeks after
transplantation none of the seven recipients remained multilineage-
numbers of Pten-deficient HSCs were thus capable of efficiently
engrafting and undergoing multilineage differentiation but became
depleted over time and gave only transient, rather than long-term,
of 23 mice that were injected with 15 Flk-22Sca-1þLin2c-
KitþCD482cells from Ptenfl/fl;Mx-1-Cre donors did not develop
leukaemia. Eighteen of these mice showed transient multilineage
reconstitution whereas only two remained long-term multilineage-
reconstituted, and levels of reconstitution declined continuously in
both of these mice (Supplementary Table 3). The fact that donor
developleukaemia,demonstrates that thedepletion ofPten-deficient
HSCs is not caused by leukaemogenesis.
We performed similar experiments with whole bone marrow cells.
Whereas all recipients of control bone marrow cells showed long-
term multilineage reconstitution (Fig. 2i) with high levels of donor
cells (Fig. 2j), the percentage of recipients that were multilineage-
reconstituted by Ptenfl/fl;Mx-1-Cre cells, and the levels of reconstitu-
tion in these recipients, declined over time. The results with whole
changed their surface-marker phenotype. Moreover, only 3 out of 10
ALL, further demonstrating that the loss of HSC activity was not
secondary to leukaemogenesis.
Pten acts cell-autonomously to maintain HSCs
The possibility remained that a few days of exposure to the myelo-
proliferative disease in donor mice might have irreversibly damaged
the Pten-deficient HSCs before transplantation. To test whether the
depletion of HSCs in Ptenfl/fl;Mx-1-Cre mice was an indirect con-
sequence of neoplasms/altered haematopoietic environment, or
whether Pten is required cell-autonomously for the maintenance of
Figure 2 | HSCs proliferate after Pten deletion, transiently expanding in
cycle status of whole bone marrow (WBM) cells from Ptenfl/fl;Mx-1-Cre
mice (b) and littermate controls (a) was observed five days after pIpC
treatment. BrdU was administered for 19h to mark cells that entered S
are mean ^ s.d. of three experiments. c, d, Flk-22Sca-1þLin2c-KitþCD482
HSCs from Ptenfl/fl;Mx-1-Cre mice (d) included significantly
(*, P , 0.05) more dividing cells relative to controls (c). Values are
mean ^ s.d. of three experiments. e, f, Significantly (*, P , 0.01) fewer
Flk-22Sca-1þLin2c-KitþCD482cells from Ptenfl/fl;Mx-1-Cre mice (f) were
of DNA/RNAcontent. Values are mean ^ s.d. of threeexperiments. g, Total
number of Flk-22Sca-1þLin2c-KitþCD482cells in Ptenfl/fl;Mx-1-Cre mice
wassignificantly increasedrelativetolittermate controlsfivedaysafterpIpC
treatment, but significantly decreased relative to controls 24 to 39 days after
pIpC administration. Values are mean ^ s.d. of six experiments. h, Fifteen
controls were transplanted into irradiated recipients along with 200,000
recipient bone marrow cells. Whereas control cells gave long-term
multilineage reconstitution in all recipients (n ¼ 5), Pten-deficient cells
gave only transient multilineage reconstitution (n ¼ 7; *, P , 0.05).
i, j, Similar results were observed when 300,000 bone marrow cells from
Ptenfl/fl;Mx-1-Cre mice or controls were transplanted into irradiated
recipients (n ¼ 10 and 4, respectively; *, P , 0.05) along with 300,000
recipient bone marrow cells. In h and j, error bars represent standard
NATURE|Vol 441|25 May 2006
© 2006 Nature Publishing Group
HSCs, we transplanted Ptenfl/fl;Mx-1-Cre bone marrow cells or
control bone marrow cells (both CD45.2þ) into recipient mice
(CD45.1þ) along with half as many recipient bone marrow cells
(Fig. 3a). Six weeks after transplantation, when these mice showed
stable chimaerism, Pten was deleted and the relative frequencies of
donor and recipient HSCs were monitored over time. As expected,
thirds of HSCs in the bone marrow irrespective of whether they were
from control (Fig. 3b) or Pten-deficient (Fig. 3c) mice.
Consistent with a cell-autonomous requirement for Pten in HSCs,
the number of Pten-deficient HSCs declined over time and the
number of wild-type recipient HSCs in the same mice increased. By
5 to 20 weeks after pIpC administration, control Ptenfl/þ;Mx-1-Cre
donor cells still accounted for 64% of bone marrow Flk-22Sca-
1þLin2c-KitþCD482cells (Fig. 3d), but Ptenfl/fl;Mx-1-Cre donor
cells accounted for only 15% of bone marrow Flk-22Sca-1þLin2c-
KitþCD482cells (Fig. 3e). The total number of control Ptenfl/þ;Mx-
Mx-1-Cre donor HSCs initially dominated the HSC pool, but by5 to
20 weeks after pIpC treatment recipient HSCs outnumbered Pten-
deficient donor HSCs by 6.4-fold (Fig. 3g). To confirm functionally
the depletion of donor HSCs, we transplanted 1 £ 106bone marrow
cells, or 100 Flk-22Sca-1þLin2c-KitþCD482cells, into irradiated
mice. All recipients of bone marrow cells from control mice (from
(Supplementary Table 4). However, recipients of Pten-deficient
donor (from Fig. 3e) bone marrow cells or Flk-22Sca-1þLin2c-
KitþCD482cells never achieved multilineage donor cell reconstitu-
tion (Supplementary Table 4). As Pten-deficient HSCs were depleted
and control HSCs were expanded within the same mice, these data
confirmthatPtenisrequiredcell-autonomously for themaintenance
Pten deletion seemed to deplete HSCs by inhibiting self-renewal
cell death, as assessed by staining for annexin Vor activated caspase-3,
ineither whole bonemarrowcells orFlk-22Sca-1þLin2c-KitþCD482
cells from Ptenfl/fl;Mx-1-Cre mice four weeks after pIpC treatment
(Supplementary Fig. 4). Moreover, approximately 90% of single
Flk-22Sca-1þLin2c-KitþCD482cells from either Pten-deleted
mice or control mice formed colonies in methylcellulose irrespective
of whether they were isolated five days or four weeks after pIpC
treatment (Supplementary Fig. 5). If HSCs were destined to undergo
cell death or rapid terminal differentiation after Pten deletion, then
have formed fewer colonies in methylcellulose. Together with the
ability of Pten-deficient HSCs to efficiently engraft and transiently
reconstitute irradiated mice (Fig. 2), these data suggest that HSCs
show less self-renewal potential after Pten deletion.
Pten-deficient leukaemias are transplantable
Leukaemia-initiating cells are defined by their ability to transfer
disease upon transplantation into irradiated mice5–7. These cells are
rare among unfractionated leukaemia cells but are highly enriched
among cells that express HSC markers5–7. To test whether the
neoplasms in Ptenfl/fl;Mx-1-Cre mice were transplantable, we
transplanted bone marrow cells, splenocytes, Flk-22Sca-1þLin2c-
Mac-1þB2202CD32myeloid cells, CD4þMac-12CD45hilymphoid
blasts, or CD3þMac-12/B220þMac-12lymphoid cells from five
independent Ptenfl/fl;Mx-1-Cre donors (which were euthanized
owing to illness) into irradiated recipients. Virtually every recipient
of 5 £ 105to 2 £ 106Ptenfl/fl;Mx-1-Cre donor bone marrow cells
(Fig. 4a) or splenocytes (Fig. 4c) died within four weeks of trans-
plantation with ALL and/or AML. Only a minority of the recipients
of 3 £ 105Ptenfl/fl;Mx-1-Cre donor bone marrow cells died, with 2
out of 14 developing AML and 2 out of 14 developing ALL (Fig. 4b).
of every 600,000 bone marrow cells (0.00017%) used in these
experiments were capable of initiating AML or ALL. Recipients of
control bone marrow cells (n ¼ 20), splenocytes (n ¼ 7) or HSCs
(n ¼ 17) from Ptenfl/þ;Mx-1-Cre mice never developed leukaemia
(data not shown).
To test whether leukaemia-initiating cellsco-purify with HSCs, we
transplanted 10 to 15 Flk-22Sca-1þLin2c-KitþCD482cells from
Ptenfl/fl;Mx-1-Cre donors into 33 irradiated recipients, five of which
died from AMLwithin four weeks (Fig. 4d). This suggests that 1 out
Figure 3 | Pten is required cell-autonomously for HSC maintenance.
(CD45.2) Ptenfl/fl;Mx-1-Cre bone marrow cells (experimental treatment),
or control bone marrow cells (control treatment), to recipient bone
marrow cells. Six weeks after transplantation, Pten was deleted.
b, c, Two days after pIpC treatment, donor cells accounted for 69 to 71% of
Flk-22Sca-1þLin2c-KitþCD482HSCs in recipient mice as expected,
whether the mice had been transplanted with control (Ptenfl/þ;Mx-1-Cre)
donor cells (b) or Ptenfl/fl;Mx-1-Cre donor cells (c). Values are mean ^ s.d.
of three mice per treatment. d, e, Subsequently, 5 to 20 weeks after pIpC
treatment, control cells still accounted for 64% of bone marrow
Flk-22Sca-1þLin2c-KitþCD482HSCs (d), but Ptenfl/fl;Mx-1-Cre donor
cells accounted for only 15% of Flk-22Sca-1þLin2c-KitþCD482HSCs (e).
Values are mean ^ s.d. of ten mice per treatment. *, P , 0.05. f, g, Control
HSCs were stable over time (f), whereas Ptenfl/fl;Mx-1-Cre donor HSCs
Values are mean ^ s.d.
NATURE|Vol 441|25 May 2006
© 2006 Nature Publishing Group
of every 81 cells in this population (1.2%) were capable of initiating
AML28—a considerable enrichment compared with whole bone
marrow, although the vast majority of cells that co-purified with
HSCs did not transfer disease.
Half of the recipients of 15,000 to 25,000 myeloid blasts died from
out of every 36,000 (0.003%) myeloid blast cells initiated AML.
compared with whole bone marrow, but not nearly as enriched as
among cells that expressed HSC markers. Recipients of bulk myeloid
cells, lymphoid blasts and bulk lymphoid cells also developed ALL
and/or AML, although leukaemia-initiating cells were not as
enriched within these populations (Fig. 4f–h). Thus, a variety of
cell populations contained leukaemia-initiating cells.
Rapamycin depletes leukaemia-initiating cells
The observation that Pten deletion leads to the depletion of normal
HSCs but promotes the generation of leukaemia-initiating cells
provided a rare distinction between the mechanisms that regulate
the maintenance of normal stem cells compared with leukaemia-
initiating cells. The PI(3)K pathway is highly branched, but activates
the mammalian target of rapamycin (mTOR) among other down-
stream effectors29,30. mTOR kinase activity is inhibited by the drug
rapamycin31,32, and human AMLs and ALLs have been shown to
respond to rapamycin33–35. Therefore, we administered rapamycin
to Ptenfl/fl;Mx-1-Cre mice to test whether it depleted leukaemia-
initiating cells or rescued normal HSC function.
Ptenfl/fl;Mx-1-Cre mice became overtly ill after pIpC treatment as
they developed leukaemias, exhibiting lethargy, ruffling of fur, and
hunched posture (Fig. 5a). All three such mice in this experiment
died within 3 to 4 weeks of pIpC treatment from AML and ALL
(Fig. 5c). In contrast, three mice that were maintained on daily
injections of rapamycin (4mg per kg of body weight) remained
healthy and active four weeks after pIpC treatment (Fig. 5b). These
rapamycin-treated mice did not show any histological evidence of
neoplasm, as the spleens had normal architecture with only focal
areas of erythroid-predominant haematopoiesis (Fig. 5c). Daily
injections of rapamycin for seven days after pIpC treatment also
prevented the decrease in bone marrow cellularity (Fig. 5d) and the
increase in spleen cellularity (Fig. 5e) observed in Ptenfl/fl;Mx-1-Cre
mice, without significantly affecting these parameters in control
mice. Hence, mice maintained on rapamycin immediately after
Pten deletion did not develop signs of haematopoietic malignancy.
To determine whether rapamycin eliminated leukaemia-initiating
cells, we treated Pten-deleted mice with vehicle or rapamycin for six
weeks and then transplanted graded doses of whole bone marrow
cells into irradiated mice (which no longer received rapamycin).
Recipients of bone marrow cells from vehicle-treated mice all died in
a dose-dependent manner within 20 to 31 days of transplantation
treated mice remained healthy and never showed signs of leukaemia,
irrespective of the dose of cells transplanted (Fig. 5f). This demon-
strates that rapamycin inhibits the generation or maintenance of
To test whether rapamycin was effective against established leu-
kaemias, mice that had been transplanted with Ptenfl/fl;Mx-1-Cre
bone marrow cells were treated with daily injections of vehicle or
rapamycin, beginning 15weeks after pIpC administration. Although
all three vehicle-treated mice died from ALL and/or AMLwithin five
weeks, all three rapamycin-treated mice remained overtly healthy
(Supplementary Table 5). Almost all recipients of bone marrow cells
from a vehicle-treated mouse died in a dose-dependent manner
(Fig. 5g). In contrast, most recipients of bone marrow cells from
rapamycin-treated micesurvived(Fig.5g).Thus, rapamycinreduced
the frequency of leukaemia-initiating cells even when treatment was
initiated after the onset of frank leukaemia. The two rapamycin-
treated mice that were not sacrificed to provide a source of cells for
transplantation were treated with daily injections of rapamycin for
15weeks. Although these mice seemed overtly healthy, with normal-
sized spleens and thymuses, one showed histological evidence of
myeloproliferative disease and the other showed signs of AML and
Rapamycin treatment was also initiated after the transplantation
of 2 £ 106bone marrowcellsfrom a Pten-deficient mouse with AML
and ALL into irradiated recipients. Vehicle-treated recipients all died
Figure 4 | AML- and ALL-initiating cells are rare in Pten-deleted mice,
but are transplantable, are contained within multiple distinct populations,
and are highly enriched among cells that express HSC markers.
a–h, Whole bone marrow (WBM) cells (a, b), whole splenocytes (c),
12lymphoid blasts (g), or CD3þMac-12/B220þMac-12lymphoid cells (h)
were transplanted into irradiated recipients from 3 to 5 independent
Ptenfl/fl;Mx-1-Cre donor mice with leukaemia. Donor (CD45.2þ) cells were
competed against 200,000 recipient (CD45.1 þ ) bone marrow cells for
radioprotection. Bars indicate the proportion of recipients that died with
AML, ALL, or AML and ALL, or which survived with no signs of neoplasm
(‘Negative’). Recipients of Ptenfl/þ;Mx-1-Cre control bone marrow cells
(n ¼ 20), splenocytes (n ¼ 7) or HSCs (n ¼ 17) never developed leukaemia
(data not shown).
NATURE|Vol 441|25 May 2006
© 2006 Nature Publishing Group
within 25days of transplantation (Fig. 5h). In contrast, rapamycin-
treated recipients died 40 to 60 days after transplantation (Fig. 5h).
When initiated after the onset of leukaemia, rapamycin was effective
in prolonging the life of mice, but not in curing the leukaemias.
Rapamycin also inhibited the survival and proliferation of clono-
genic leukaemia cells in culture. Freshly isolated or cultured myeloid
blast cells from Ptenfl/fl;Mx-1-Cre mice with AML were plated into
methylcellulose. Rapamycin significantly reduced the percentage of
blast cells that formed colonies, as well as colony size, in a dose-
dependent manner (Supplementary Fig. 6a–e). Rapamycin also
significantly reduced the percentage of myeloid blasts in S phase of
the cell cycle, and increased the percentage of cells expressing
activated caspase-3 (Supplementary Fig. 6f, g).
Rapamycin rescues Pten-deficient HSCs
Rapamycin also restored the capacity of Pten-deficient HSCs to
provide long-term multilineage reconstitution to irradiated mice.
Daily injections of rapamycin for seven days after pIpC adminis-
but did normalize the cell cycle distribution of Flk-22Sca-1þLin2c-
KitþCD482cells in Ptenfl/fl;Mx-1-Cre mice without affecting the
proliferation of HSCs from control littermates (Fig. 6a). Rapamycin
also eliminated the HSC expansion observed seven days after pIpC
treatment (Fig. 6b), and the HSC depletion observed after four
weeks in Ptenfl/fl;Mx-1-Cre mice (Fig. 6c), without affecting HSC
numbers in control mice. Most notably, rapamycin restored the
potential of Flk-22Sca-1þLin2c-KitþCD482cells isolated from
Figure 5 | Rapamycin depletes leukaemia-initiating cells. a, b, Four weeks
after pIpC treatment, Ptenfl/fl;Mx-1-Cremice exhibitedlethargy,ruffled fur,
hunched posture, and required euthanization (a; n ¼ 3), but if treated daily
with 4mg per kg rapamycin remained healthy and active (b; n ¼ 3).
c, Rapamycin-treated mice had normal spleen architecture (right panel),
whereas vehicle-treated littermates developed AML and ALL (left panel).
d, e, Rapamycin prevented the decrease in bone marrow cellularity (d) and
the increase in spleen cellularity (e) in Ptenfl/fl;Mx-1-Cre mice, but did not
affecttheseparametersincontrols.Valuesaremean ^ s.d.of6to7miceper
treatment. f, Mice transplanted with 1 £ 106Ptenfl/fl;Mx-1-Cre donor
(CD45.2) bone marrow cells along with 0.5 £ 106control (CD45.1) bone
marrow cells were treated with vehicle or rapamycin (4mg per kg per day)
for six weeks, starting immediately after pIpC treatment. Graded doses of
whole bone marrow (WBM) cells from these mice were transplanted into
lethally irradiated recipients along with 2 £ 105recipient (CD45.1) bone
marrow cells. Recipients of bone marrow from vehicle-treated mice died
with leukaemia,whereas recipients of marrow from rapamycin-treated mice
remained healthy. g, Similar donor mice were treated for five weeks with
vehicle or rapamycin (4mg per kg per day) beginning 15 weeks after pIpC
treatment. Nearly all recipients of bone marrow from vehicle-treated mice
died of leukaemia in a dose-dependent manner, whereas recipients of cells
from rapamycin-treated mice usually survived. h, Finally, 2 £ 106donor
ALLwere transplanted into sublethally irradiated recipients. The recipients
were treated with vehicle or rapamycin (0.4mg per kg per day). Panels f and
h show one representative experiment out of three.
NATURE|Vol 441|25 May 2006
© 2006 Nature Publishing Group
Ptenfl/fl;Mx-1-Cre mice to provide long-term multilineage reconsti-
tution seven days after pIpC treatment (Fig. 6d, e). This confirms
that Flk-22Sca-1þLin2c-KitþCD482cells from Ptenfl/fl;Mx-1-Cre
The mechanism responsible for the depletion of Pten-deficient
HSCs remains to be elucidated. One possibility is that persistent
activation of the PI(3)K pathway following the loss of Pten leads to
reduced HSC self-renewal via a gradual increase in the rate at which
HSCs exit the stem cell pool. Another possibility is that Pten
deficiency induces the gradual senescence of HSCs. Conditional
deletion of Pten leads to a p53-dependent senescence of prostate
cells36. Leukaemias might acquire secondary mutations that inactivate
the senescence response.
Deletion of the cyclin-dependent kinase inhibitor p21Cip1also
compounds that promote stem cell quiescence might consistently
have different effects on normal stem cells and cancer stem cells.
Conditional Pten deletion in the fetal central nervous system
increases the self-renewal and frequency of neural stem cells38. This
is the opposite of what we observed, and may represent a general
difference between fetal and adult stem cells rather than a difference
between tissues. The balance of proto-oncogenes and tumour sup-
pressor genes that regulate stem cell self-renewal changes between
embryonic, fetal and adult life as the organogenic demand decreases
and the risk of cancer increases11,13,39.
These data demonstrate that it is possible to identify—and to
target therapeutically—pathways that have distinct effects on normal
stem cells and cancer stem cells within the same tissue. This is an
important finding, because it has been proposed that oncogenic
mutations confer self-renewal potential by activating pathways used
by normal stem cells, irrespective of whether the mutations occur in
stem cells or other cells1,2,12. By comparing the mechanisms that
regulate the maintenance of normal stem cells versus cancer stem
cells, it should be possible to design new therapies and to improve
Mice. Mice were housed in the Unit for Laboratory Animal Medicine at the
University of Michigan. Ptenfl/þand Mx-1-Cre mice were backcrossed for eight
and six generations, respectively, onto a C57BL/Ka-CD45.2:Thy-1.1 background.
Recipients in reconstitution assays were adult C57BL/Ka-CD45.1:Thy-1.2 mice.
Flowcytometry andisolationofHSCs.Bone marrowcellswereflushedfromthe
or magnesium, supplemented with 2% heat-inactivated calf serum (HBSSþ;
GIBCO). Cells were triturated and filtered through nylon screen (45mm; Sefar
America) to obtain a single-cell suspension. Flk-22Sca-1þLin2c-KitþCD482
HSCs and Thy-1.1lowSca-1þMac-1lowCD4lowB2202multipotent progenitors
(MPPs) were isolated as previously described25,26,40. For isolation of Flk-22Sca-
1þLin2c-KitþCD482HSCs, whole bone marrow cells were incubated with
unconjugated monoclonal antibodies to lineage (Lin) markers including B220
(6B2), CD3 (KT31.1), CD4 (GK1.5), CD5 (53-7.3), CD8 (53-6.7), Gr-1 (8C5),
Mac-1 (M1/70) and Ter119. After washing, cells were resuspended in anti-rat
IgG conjugated to phycoerythrin (PE; Jackson ImmunoResearch). Cells were
then stained with directly conjugated antibodies to Sca-1 (Ly6A/E-APC), c-Kit
(2B8-biotin), Flk-2 (A2F10-PE; eBioscience) and CD48 (HM48-1-FITC; BD
Pharmingen). To identify CD45.2þHSCs, antibodies to CD45.2 (104-FITC; BD
Pharmingen)and CD48(HM48-1-PE;eBioscience)were used. HSCs were often
pre-enriched by selecting c-Kitþcells using paramagnetic microbeads and
autoMACS (Miltenyi Biotec). To identify leukaemic blast cells, anti-CD45
(30-F11-APC; eBioscience) was used.
The total number of Flk-22Sca-1þLin2c-KitþCD482cells per mouse was
calculated based on the frequency of this population in the bone marrow and
spleen,the cellularity of the spleen andlong bones, andbyassumingthat 15%of
all bone marrow is within the long bones41. The blood and other tissues do not
contribute significantly to the overall size of the HSC pool.
Long-term competitive reconstitution assays. Adult recipient mice were
irradiated with an Orthovoltage X-ray source delivering 300radmin21. Reci-
pient mice received two doses of 540rads each, delivered 3h apart. Forsublethal
irradiation, mice were administered one dose of 800rads. Donor (CD45.2þ)
HSCs were sorted and then re-sorted (for purity) into individual wells of a 96-
well plate containing 200,000 CD45.1þwhole bone marrowcells in HBSSþ. The
contents of individual wells were injected into the retro-orbital venous sinus of
irradiated CD45.1þrecipients. For at least 16weeks after transplantation, blood
was obtained from the tail veins of recipient mice, subjected to ammonium
chloride/potassium bicarbonate red-cell lysis, and stained with directly con-
jugated antibodies to CD45.2 (104-FITC), B220 (6B2), Mac-1 (M1/70), CD3
conducted as described in Supplementary Methods.
Administration of pIpC and rapamycin. As described previously42, polyinosine-
polycytidine (pIpC; Sigma) was resuspended in Dulbecco’s-PBS at 2mgml21
and passed through a 0.22-mm filter. Mice received 25mg of pIpC per gram of
body mass every other day for two weeks. Rapamycin (Calbiochem and LC
Laboratories) was administered by intraperitoneal injection at the indicated
Figure 6 | Rapamycin rescues normal HSC function after Pten deletion.
a, Seven daysof rapamycin (4mg per kg per day) did not affect proliferation
in whole bone marrow (WBM; 19-h pulse with BrdU), but did eliminate
the increase in Flk-22Sca-1þLin2c-KitþCD482HSC proliferation in
Ptenfl/fl;Mx-1-Cre mice without affecting proliferation in controls. Values
are mean ^ s.d. of three experiments. b, Rapamycin eliminated the HSC
expansion observed seven days after pIpC treatment in Ptenfl/fl;Mx-1-Cre
mice without affecting HSCs in Ptenfl/þ;Mx-1-Cre controls. Values are
mean ^ s.d. of 6 to 7 mice per treatment. c, Rapamycin eliminated the
depletion of HSCs observed 3 to 4 weeks after pIpC treatment in
Ptenfl/fl;Mx-1-Cre mice without affecting controls. Values are mean ^ s.d.
of three mice per treatment. d, e, Rapamycin restored the long-term
(.16weeks after transplantation) multilineage reconstitution potential of
Flk-22Sca-1þLin2c-KitþCD482cells from Ptenfl/fl;Mx-1-Cre mice (d).
Fifteen Flk-22Sca-1þLin2c-KitþCD482donor HSCsfrom Ptenfl/fl;Mx-1-Cre
or control mice that had been treated with rapamycin (4mg per kg per day)
or vehicle for seven days after pIpC treatment were transplanted along with
200,000 recipient bone marrow cells into irradiated recipients. Recipients
received vehicle or rapamycin (0.4mg per kg per day) starting within two
weeks of transplantation. The frequency of donor white blood (CD45.2þ)
cells, myeloid (Mac-1þ) cells, B (B220þ) cells and T (CD3þ) cells was
determined 16 to 18 weeks after transplantation (e). Values are mean ^ s.d.
of three experiments.
NATURE|Vol 441|25 May 2006
© 2006 Nature Publishing Group Download full-text
doses. It was reconstituted in absolute ethanol at 10mgml21or 1mgml21and
diluted in 5% Tween-80 (Sigma) and 5% PEG-400 (Hampton Research) before
injection. The final volume of all injections was 200ml.
Received 1 October 2005; accepted 1 March 2006.
Published online 5 April 2006.
1.Reya, T., Morrison, S. J., Clarke, M. F. & Weissman, I. L. Stem cells, cancer, and
cancer stem cells. Nature 414, 105– -111 (2001).
Pardal, R., Clarke, M. F. & Morrison, S. J. Applying the principles of stem-cell
biology to cancer. Nature Rev. Cancer 3, 895– -902 (2003).
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).
Singh, S. K. et al. Identification of a cancer stem cell in human brain tumors.
Cancer Res. 63, 5821– -5828 (2003).
Lapidot, T. et al. A cell initiating human acute myeloid leukemia after
transplantation into SCID mice. Nature 17, 645– -648 (1994).
Bonnet, D. & Dick, J. E. Human acute myeloid leukemia is organized as a
hierarchy that originates from a primitive hematopoietic cell. Nature Med. 3,
730– -737 (1997).
Warner, J. K., Wang, J. C., Hope, K. J., Jin, L. & Dick, J. E. Concepts of human
leukemic development. Oncogene 23, 7164– -7177 (2004).
Park, I.-K. et al. Bmi-1 is required for the maintenance of adult self-renewing
haematopoietic stem cells. Nature 423, 302– -305 (2003).
Lessard, J. & Sauvageau, G. Bmi-1 determines the proliferative capacity of
normal and leukaemic stem cells. Nature 423, 255– -260 (2003).
10. Hemmati, H. D. et al. Cancerous stem cells can arise from pediatric brain
tumors. Proc. Natl Acad. Sci. USA 100, 15178– -15183 (2003).
11.Molofsky, A. V., Pardal, R. & Morrison, S. J. Diverse mechanisms regulate stem
cell self-renewal. Curr. Opin. Cell Biol. 16, 700– -707 (2004).
12. Taipale, J. & Beachy, P. A. The Hedgehog and Wnt signalling pathways in
cancer. Nature 411, 349– -354 (2001).
13. Molofsky, A. V. et al. Bmi-1 promotes neural stem cell self-renewal and neural
development but not mouse growth and survival by repressing the p16Ink4aand
p19Arfsenescence pathways. Genes Dev. 19, 1432– -1437 (2005).
14. Lowe, S. W. & Sherr, C. J. Tumor suppression by Ink4a– -Arf: progress and
puzzles. Curr. Opin. Genet. Dev. 13, 77– -83 (2003).
15. Maehama, T. & Dixon, J. E. The tumor suppressor, PTEN/MMAC1,
dephosphorylates the lipid second messenger, phosphatidylinositol 3,4,5-
trisphosphate. J. Biol. Chem. 273, 13375– -13378 (1998).
16. Stiles, B., Groszer, M., Wang, S., Jiao, J. & Wu, H. PTENless means more. Dev.
Biol. 273, 175– -184 (2004).
17. Di Cristofano, A. & Pandolfi, P. P. The multiple roles of PTEN in tumor
suppression. Cell 100, 387– -390 (2000).
18. Aggerholm, A., Gronbaek, K., Guldberg, P. & Hokland, P. Mutational analysis of
the tumour suppressor gene MMAC1/PTEN in malignant myeloid disorders.
Eur. J. Haematol. 65, 109– -113 (2000).
19. Roman-Gomez, J. et al. Promoter hypermethylation of cancer-related genes: a
strong independent prognostic factor in acute lymphoblastic leukemia. Blood
104, 2492– -2498 (2004).
20. Dahia, P. L. et al. PTEN is inversely correlated with the cell survival factor
Akt/PKB and is inactivated via multiple mechanisms in haematological
malignancies. Hum. Mol. Genet. 8, 185– -193 (1999).
21. Cheong, J. W. et al. Phosphatase and tensin homologue phosphorylation in the
C-terminal regulatory domain is frequently observed in acute myeloid
leukaemia and associated with poor clinical outcome. Br. J. Haematol. 122,
454– -456 (2003).
22. Hock, H. et al. Tel/Etv6 is an essential and selective regulator of adult
hematopoietic stem cell survival. Genes Dev. 18, 2336– -2341 (2004).
23. Ku ¨hn, R., Schwenk, F., Aguet, M. & Rajewsky, K. Inducible gene targeting in
mice. Science 269, 1427– -1429 (1995).
24. Kogan, S. C. et al. Bethesda proposals for classification of nonlymphoid
hematopoietic neoplasms in mice. Blood 100, 238– -245 (2002).
25. Christensen, J. L. & Weissman, I. L. Flk-2 is a marker in hematopoietic stem
cell differentiation: a simple method to isolate long-term stem cells. Proc. Natl
Acad. Sci. USA 98, 14541– -14546 (2001).
26. Kiel, M. J., Yilmaz, O. H., Iwashita, T., Terhorst, C. & Morrison, S. J. SLAM
family receptors distinguish hematopoietic stem and progenitor cells and
reveal endothelial niches for stem cells. Cell 121, 1109– -1121 (2005).
27. Yilmaz, O. H., Kiel, M. J. & Morrison, S. J. SLAM family markers are conserved
among hematopoietic stem cells from old and reconstituted mice and
markedly increase their purity. Blood 107, 924– -930 (2005).
28. Smith, L. G., Weissman, I. L. & Heimfeld, S. Clonal analysis of hematopoietic
stem-cell differentiation in vivo. Proc. Natl Acad. Sci. USA 88, 2788– -2792
29. Majumder, P. K. et al. mTOR inhibition reverses Akt-dependent prostate
intraepithelial neoplasia through regulation of apoptotic and HIF-1-dependent
pathways. Nature Med. 10, 594– -601 (2004).
30. Inoki, K., Corradetti, M. N. & Guan, K. L. Dysregulation of the TSC– -mTOR
pathway in human disease. Nature Genet. 37, 19– -24 (2005).
31. Podsypanina, K. et al. An inhibitor of mTOR reduces neoplasia and normalizes
p70/S6 kinase activity in Ptenþ/2mice. Proc. Natl Acad. Sci. USA 98,
10320– -10325 (2001).
32. Neshat, M. S. et al. Enhanced sensitivity of PTEN-deficient tumors to inhibition
of FRAP/mTOR. Proc. Natl Acad. Sci. USA 98, 10314– -10319 (2001).
33. Recher, C. et al. Antileukemic activity of rapamycin in acute myeloid leukemia.
Blood 105, 2527– -2534 (2005).
34. Avellino, R. et al. Rapamycin stimulates apoptosis of childhood acute
lymphoblastic leukemia cells. Blood 106, 1400– -1406 (2005).
35. Teachey, D. T. et al. The mTOR inhibitor CCI-779 induces apoptosis and
inhibits growth in preclinical models of primary adult human ALL. Blood 107,
1149– -1155 (2006).
36. Chen, Z. et al. Crucial role of p53-dependent cellular senescence in suppression
of Pten-deficient tumorigenesis. Nature 436, 725– -730 (2005).
37. Cheng, T. et al. Hematopoietic stem cell quiescence maintained by
p21cip1/waf1. Science 287, 1804– -1808 (2000).
38. Groszer, M. et al. Negative regulation of neural stem/progenitor cell
proliferation by the Pten tumor suppressor gene in vivo. Science 294,
2186– -2189 (2001).
39. Molofsky, A. V. et al. Bmi-1 dependence distinguishes neural stem cell self-
renewal from progenitor proliferation. Nature 425, 962– -967 (2003).
40. Morrison, S. J. & Weissman, I. L. The long-term repopulating subset of
hematopoietic stem cells is deterministic and isolatable by phenotype.
Immunity 1, 661– -673 (1994).
41. Smith, L. H. & Clayton, M. L. Distribution of injected 59Fe in mice. Exp.
Hematol. 20, 82– -86 (1970).
42. Mikkola, H. K. et al. Haematopoietic stem cells retain long-term repopulating
activity and multipotency in the absence of stem-cell leukaemia SCL/tal-1
gene. Nature 421, 547– -551 (2003).
Supplementary Information is linked to the online version of the paper at
Acknowledgements This work was supported by the Howard Hughes Medical
Institute. O.H.Y. was supported by a predoctoral fellowship from the University
of Michigan (UM) Institute of Gerontology. We thank the UM Flow-cytometry
Core Facility, which was supported by the UM-Comprehensive Cancer Center.
We also thank E. Smith in the Hybridoma Core Facility for antibody production,
supported in part through the Rheumatic Disease Core Center; A. Burgess and
N. McAnsh of the UM Comprehensive Cancer Center Tissue Core; and
C. Mountford for excellent mouse colony management.
Author Contributions O.H.Y. performed all experiments and participated in the
design and interpretation of experiments. R.V. performed all pathology on the
mice with help from O.H.Y. B.K.T. and D.O.F. performed spectral karyotype
analysis with help from O.H.Y. W.G. and H.W. provided the Ptenfl/flmice and
discussed pre-publication results. S.J.M. participated in the design and
interpretation of experiments, and wrote the paper with O.H.Y.
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 S.J.M. (email@example.com).
NATURE|Vol 441|25 May 2006